BSON reader: recover from the document size checks made on reading the size

Since #5930, a size below 5, a size that exceeds the enclosing document, and
a size too small for the document's content are reported before the document
is read. When recovering, such a document is now read up to its terminator,
within the enclosing document, and its size is not reported a second time.
BJData now checks an ndarray's type before the object that would hold it is
begun, so that test expects null.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>

.github/CONTRIBUTING.md

@@ -111,6 +111,8 @@ context which existing file needs to be extended, and only very few cases requir
 When fixing a bug, edit `unit-regression3.cpp` and add a section referencing the fixed issue.
 `unit-regression2.cpp` holds the older tests; the two files exist because a single one grew large enough for the
 MinGW linker to fail relocating it, so please keep adding to the smaller file rather than growing the larger one.
+Tests that call `sax_parse` go into `unit-sax_parse.cpp` rather than into a large test file: every call instantiates
+the parser and the binary reader that recover from errors, which grows a test file considerably.
 
 #### Exceptions
 

.github/workflows/cancel_closed_pr_runs.yml

@@ -24,7 +24,7 @@ jobs:
     permissions:
       actions: write
 
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
 
     steps:
       - name: Harden Runner

.github/workflows/check_amalgamation.yml

@@ -19,7 +19,7 @@ permissions:
 jobs:
   save:
     if: github.event_name == 'pull_request'
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     steps:
       - name: Harden Runner
         uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2

.github/workflows/check_docs_links.yml

@@ -16,7 +16,7 @@ concurrency:
 jobs:
   check_docs_links:
     if: github.repository == 'nlohmann/json'
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     timeout-minutes: 30
     steps:
       - name: Harden Runner

.github/workflows/comment_check_amalgamation.yml

@@ -12,7 +12,7 @@ jobs:
   comment:
     # push runs on develop have no PR to comment on (and no "pr" artifact)
     if: ${{ github.event.workflow_run.conclusion == 'failure' && github.event.workflow_run.event == 'pull_request' }}
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     permissions:
       contents: read
       actions: read

.github/workflows/dependency-review.yml

@@ -18,7 +18,7 @@ permissions:
 
 jobs:
   dependency-review:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     steps:
       - name: Harden Runner
         uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2

.github/workflows/flawfinder.yml

@@ -24,7 +24,7 @@ on:
 jobs:
   flawfinder:
     name: Flawfinder
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     permissions:
       actions: read
       contents: read

.github/workflows/labeler.yml

@@ -19,7 +19,7 @@ jobs:
       contents: read
       pull-requests: write
 
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
 
     steps:
       - name: Harden Runner

.github/workflows/publish_documentation.yml

@@ -34,7 +34,7 @@ jobs:
       contents: write
 
     if: github.repository == 'nlohmann/json'
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     steps:
       - name: Harden Runner
         uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2

.github/workflows/scorecards.yml

@@ -24,7 +24,7 @@ permissions:
 jobs:
   analysis:
     name: Scorecard analysis
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     permissions:
       # Needed to upload the results to code-scanning dashboard.
       security-events: write

.github/workflows/stale.yml

@@ -8,7 +8,7 @@ permissions:
 
 jobs:
   stale:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
 
     permissions:
       issues: write

.github/workflows/ubuntu.yml

@@ -18,7 +18,7 @@ concurrency:
 
 jobs:
   ci_test_gcc:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     container: gcc:latest
     steps:
       - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
@@ -130,7 +130,7 @@ jobs:
         run: cmake --build build --target ${{ matrix.target }}
 
   ci_static_analysis_clang:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     container: silkeh/clang:22
     strategy:
       matrix:
@@ -149,7 +149,7 @@ jobs:
         run: cmake --build build --target ${{ matrix.target }}
 
   ci_cmake_options:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     container: ubuntu:24.04
     strategy:
       matrix:
@@ -209,7 +209,7 @@ jobs:
           fail-on-error: false
 
   ci_test_compilers_gcc_old:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     strategy:
       matrix:
         compiler: ['4.8', '4.9', '5', '6']
@@ -251,7 +251,7 @@ jobs:
         run: cmake --build build --target ci_test_compiler_g++-${{ matrix.compiler }}
 
   ci_test_compilers_gcc:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     strategy:
       matrix:
         # older GCC docker images (4, 5, 6) fail to check out code
@@ -269,7 +269,7 @@ jobs:
         run: cmake --build build --target ci_test_compiler_default
 
   ci_test_compilers_clang:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     strategy:
       matrix:
         compiler: ['3.4', '3.5', '3.6', '3.7', '3.8', '3.9', '4', '5', '6', '7', '8', '9', '10', '11', '12', '13', '14', '15-bullseye', '16', '17', '18', '19', '20', '21', '22', 'latest']
@@ -289,7 +289,7 @@ jobs:
         run: cmake --build build --target ci_test_compiler_default
 
   ci_test_standards_gcc:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     container: gcc:latest
     strategy:
       matrix:
@@ -306,7 +306,7 @@ jobs:
         run: cmake --build build --target ci_test_gcc_cxx${{ matrix.standard }}
 
   ci_test_standards_clang:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     container: silkeh/clang:latest
     strategy:
       matrix:
@@ -330,7 +330,7 @@ jobs:
         if: ${{ matrix.stdlib == 'libstdcxx' }}
 
   ci_cuda_example:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     strategy:
       fail-fast: false
       matrix:
@@ -356,7 +356,7 @@ jobs:
     strategy:
       matrix:
         container: ['gcc:latest', 'silkeh/clang:latest']
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     container: ${{ matrix.container }}
     steps:
       - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
@@ -383,7 +383,7 @@ jobs:
         run: cmake --build build --target ci_module_cpp20
 
   ci_icpc:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     # Intel discontinued the classic icc/icpc compiler in oneAPI 2024.0; this is
     # Intel's own last officially published image that still includes it.
     container: intel/oneapi-hpckit:2023.2.1-devel-ubuntu22.04
@@ -402,7 +402,7 @@ jobs:
         run: cmake --build build --target ci_icpc
 
   ci_icpx:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     container: intel/oneapi-hpckit:latest
     steps:
       - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
@@ -416,7 +416,7 @@ jobs:
         run: cmake --build build --target ci_icpx
 
   ci_nvhpc:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
     container: nvcr.io/nvidia/nvhpc:25.5-devel-cuda12.9-ubuntu22.04
     steps:
       - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1

README.md

@@ -495,7 +495,7 @@ bool key(string_t& val);
 bool parse_error(std::size_t position, const std::string& last_token, const detail::exception& ex);
 ```
 
-The return value of each function determines whether parsing should proceed.
+The return value of each function determines whether parsing should proceed. For `parse_error`, returning `true` [recovers from the error](https://json.nlohmann.me/features/parsing/error_recovery/): the parser repairs the input and continues.
 
 To implement your own SAX handler, proceed as follows:
 
@@ -503,7 +503,7 @@ To implement your own SAX handler, proceed as follows:
 2. Create an object of your SAX interface class, e.g. `my_sax`.
 3. Call `bool json::sax_parse(input, &my_sax)`; where the first parameter can be any input like a string or an input stream and the second parameter is a pointer to your SAX interface.
 
-Note the `sax_parse` function only returns a `bool` indicating the result of the last executed SAX event. It does not return a  `json` value - it is up to you to decide what to do with the SAX events. Furthermore, no exceptions are thrown in case of a parse error -- it is up to you what to do with the exception object passed to your `parse_error` implementation. Internally, the SAX interface is used for the DOM parser (class `json_sax_dom_parser`) as well as the acceptor (`json_sax_acceptor`), see file [`json_sax.hpp`](https://github.com/nlohmann/json/blob/develop/include/nlohmann/detail/input/json_sax.hpp).
+Note the `sax_parse` function only returns a `bool` indicating whether the input was parsed without errors and no SAX event returned `false`. It does not return a  `json` value - it is up to you to decide what to do with the SAX events. Furthermore, no exceptions are thrown in case of a parse error -- it is up to you what to do with the exception object passed to your `parse_error` implementation. Internally, the SAX interface is used for the DOM parser (class `json_sax_dom_parser`) as well as the acceptor (`json_sax_acceptor`), see file [`json_sax.hpp`](https://github.com/nlohmann/json/blob/develop/include/nlohmann/detail/input/json_sax.hpp).
 
 ### STL-like access
 

docs/mkdocs/docs/api/basic_json/sax_parse.md

@@ -104,7 +104,9 @@ The SAX event lister must follow the interface of [`json_sax`](../json_sax/index
 
 ## Return value
 
-return value of the last processed SAX event
+`#!cpp true` if the input was parsed without errors and no SAX event returned `#!cpp false`; `#!cpp false` otherwise.
+In particular, the result is `#!cpp false` for input with errors, even if the SAX parser recovered from all of them
+(see [error recovery](../../features/parsing/error_recovery.md)).
 
 ## Exception safety
 
@@ -163,6 +165,7 @@ use the same file with stdio functions, but it must not wait for another thread
 - Added `ignore_trailing_commas` in version 3.13.0.
 - Added `tag_handler` in version 3.13.0.
 - Extended container support (1) to include types with lvalue-only ADL `begin`/`end` (matching `std::begin`/`std::end` semantics) in version 3.13.0.
+- Recovering from parse errors (see [`parse_error`](../json_sax/parse_error.md)) added in version 3.13.0.
 - Extended overload (2) to accept heterogeneous iterator+sentinel pairs (C++20 ranges support) in version 3.13.0.
 - `JSON_PRECISE_STREAM_POSITION` added in version 3.13.0 to optionally leave a `#!cpp std::istream` positioned right
   after the parsed value when `strict` is `#!cpp false`.

docs/mkdocs/docs/api/basic_json_document/accept.md

@@ -13,9 +13,15 @@ function never throws an exception for invalid input, and the returned `#!cpp bo
 is returned.
 
 The result is the same as that of `BasicJsonType::accept()` by design; the difference is how the check is made. This
-function runs the document's index builder (and discards the index), while `BasicJsonType::accept()` runs the SAX
+function runs the document's builder, which checks the text as [`parse()`](parse.md) does, while
-parser with a handler that ignores every event. Use this one if you go on to parse the same input with
+`BasicJsonType::accept()` runs the SAX parser with a handler that ignores every event. Use this one if you go on to
-[`parse()`](parse.md) anyway, so that both use the same code; either is fine for a plain yes/no answer.
+parse the same input with [`parse()`](parse.md) anyway, so that both use the same code; either is fine for a plain
+yes/no answer.
+
+When both `ignore_comments` and `ignore_trailing_commas` are `#!cpp false` (the default), the builder writes no index.
+With either option set, it builds the index as [`parse()`](parse.md) does and discards it: the result is the same, only
+the time differs. (Neither option needs the index; the variant that skips it exists only for the default options to
+keep the code size down.)
 
 ## Template parameters
 

docs/mkdocs/docs/api/basic_json_view/at.md

@@ -68,9 +68,10 @@ Strong guarantee: there are no changes to the view or the document it refers to.
     - Throws [`out_of_range.404`](../../home/exceptions.md#jsonexceptionout_of_range404) if `ptr` cannot be resolved
       because a reference token is used on a primitive value.
 
-None of these exceptions carry a [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) path: the view has no
+As for `BasicJsonType`, with [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) these exceptions name the JSON
-`BasicJsonType` value to point at, so the exception is created without one, even if `BasicJsonType` was built with
+Pointer path of the value within the document, and with
-`JSON_DIAGNOSTICS` enabled.
+[`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md) its bytes in the source text (when
+[`materialize()`](materialize.md) sets them).
 
 ## Complexity
 

docs/mkdocs/docs/api/basic_json_view/back.md

@@ -20,9 +20,9 @@ Strong guarantee: there are no changes to the view or the document it refers to.
 Throws [`invalid_iterator.214`](../../home/exceptions.md#jsonexceptioninvalid_iterator214) if the view is
 [null](is_null.md) or [discarded](is_discarded.md), or if it is an empty array or object.
 
-This exception does not carry a [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) path: the view has no
+As for `BasicJsonType`, with [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) the exception names the JSON Pointer
-`BasicJsonType` value to point at, so the exception is created without one, even if `BasicJsonType` was built with
+path of the value within the document, and with [`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md)
-`JSON_DIAGNOSTICS` enabled.
+its bytes in the source text (when [`materialize()`](materialize.md) sets them).
 
 ## Complexity
 

docs/mkdocs/docs/api/basic_json_view/front.md

@@ -20,9 +20,9 @@ Strong guarantee: there are no changes to the view or the document it refers to.
 Throws [`invalid_iterator.214`](../../home/exceptions.md#jsonexceptioninvalid_iterator214) if the view is
 [null](is_null.md) or [discarded](is_discarded.md), or if it is an empty array or object.
 
-This exception does not carry a [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) path: the view has no
+As for `BasicJsonType`, with [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) the exception names the JSON Pointer
-`BasicJsonType` value to point at, so the exception is created without one, even if `BasicJsonType` was built with
+path of the value within the document, and with [`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md)
-`JSON_DIAGNOSTICS` enabled.
+its bytes in the source text (when [`materialize()`](materialize.md) sets them).
 
 ## Complexity
 

docs/mkdocs/docs/api/basic_json_view/get.md

@@ -80,14 +80,13 @@ Strong guarantee: there are no changes to the view or the document it refers to.
 - For every other `T`: whatever [`materialize().get<T>()`](../basic_json/get.md) throws — typically `type_error.302`,
   or whatever a user-provided `from_json()` throws.
 
-None of the exceptions thrown directly by this function or by a `from_json()` that reads the view (the first four
+The exceptions thrown directly by this function or by a `from_json()` that reads the view (the first four bullets
-bullets above) carry a
+above) carry the context of those of `BasicJsonType`: with [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) the
-[`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) path: the view has no `BasicJsonType` value to point at. An
+JSON Pointer path of the value within the document, and with
-exception thrown while converting through `materialize()` (the last bullet) is different: it is thrown by a real
+[`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md) its bytes in the source text. An exception
-`BasicJsonType` value, so it **does** carry a `JSON_DIAGNOSTICS` path if `BasicJsonType` was built with it enabled.
+thrown while converting through `materialize()` (the last bullet) is thrown by the materialized `BasicJsonType` value
-That path is relative to the root of the view's subtree, because `materialize()` builds only the subtree (with parent
+instead: its path is relative to the root of the view's subtree, because `materialize()` builds only the subtree
-pointers): it is empty if the failing value is the subtree's root itself, and it never is the path within the whole
+(with parent pointers), so it is empty if the failing value is the subtree's root itself.
-document.
 
 ## Complexity
 

docs/mkdocs/docs/api/basic_json_view/get_string.md

@@ -22,9 +22,9 @@ Strong guarantee: there are no changes to the view or the document it refers to.
 Throws [`type_error.302`](../../home/exceptions.md#jsonexceptiontype_error302) if the value is not a string; example:
 `"type must be string, but is array"`.
 
-This exception does not carry a [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) path: the view has no
+As for `BasicJsonType`, with [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) the exception names the JSON Pointer
-`BasicJsonType` value to point at, so the exception is created without one, even if `BasicJsonType` was built with
+path of the value within the document, and with [`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md)
-`JSON_DIAGNOSTICS` enabled.
+its bytes in the source text (when [`materialize()`](materialize.md) sets them).
 
 ## Complexity
 

docs/mkdocs/docs/api/basic_json_view/number_token.md

@@ -22,9 +22,9 @@ Strong guarantee: there are no changes to the view or the document it refers to.
 Throws [`type_error.302`](../../home/exceptions.md#jsonexceptiontype_error302) if the value is not a number; example:
 `"type must be number, but is string"`.
 
-This exception does not carry a [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) path: the view has no
+As for `BasicJsonType`, with [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) the exception names the JSON Pointer
-`BasicJsonType` value to point at, so the exception is created without one, even if `BasicJsonType` was built with
+path of the value within the document, and with [`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md)
-`JSON_DIAGNOSTICS` enabled.
+its bytes in the source text (when [`materialize()`](materialize.md) sets them).
 
 ## Complexity
 

docs/mkdocs/docs/api/basic_json_view/operator[].md

@@ -73,9 +73,10 @@ Strong guarantee: there are no changes to the view or the document it refers to.
     - [`parse_error.109`](../../home/exceptions.md#jsonexceptionparse_error109) if an array index in `ptr` is not a
       number.
 
-None of these exceptions carry a [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) path: the view has no
+As for `BasicJsonType`, with [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) these exceptions name the JSON
-`BasicJsonType` value to point at, so the exception is created without one, even if `BasicJsonType` was built with
+Pointer path of the value within the document, and with
-`JSON_DIAGNOSTICS` enabled.
+[`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md) its bytes in the source text (when
+[`materialize()`](materialize.md) sets them).
 
 ## Complexity
 

docs/mkdocs/docs/api/basic_json_view/value.md

@@ -61,9 +61,10 @@ Strong guarantee: there are no changes to the view or the document it refers to.
    `default_value` instead of throwing, exactly as [`BasicJsonType::value`](../basic_json/value.md) catches
    `out_of_range` and returns `default_value`.
 
-None of these exceptions carry a [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) path: the view has no
+As for `BasicJsonType`, with [`JSON_DIAGNOSTICS`](../macros/json_diagnostics.md) these exceptions name the JSON
-`BasicJsonType` value to point at, so the exception is created without one, even if `BasicJsonType` was built with
+Pointer path of the value within the document, and with
-`JSON_DIAGNOSTICS` enabled.
+[`JSON_DIAGNOSTIC_POSITIONS`](../macros/json_diagnostic_positions.md) its bytes in the source text (when
+[`materialize()`](materialize.md) sets them).
 
 ## Complexity
 

docs/mkdocs/docs/api/json_sax/index.md

@@ -7,7 +7,8 @@ struct json_sax;
 
 This class describes the SAX interface used by [sax_parse](../basic_json/sax_parse.md). Each function is called in
 different situations while the input is parsed. The boolean return value informs the parser whether to continue
-processing the input.
+processing the input; for [`parse_error`](parse_error.md), it decides whether to
+[recover from the error](../../features/parsing/error_recovery.md).
 
 For instance, parsing the JSON text `{"a": [1, true]}` triggers the following callbacks, in order:
 

docs/mkdocs/docs/api/json_sax/parse_error.md

@@ -21,11 +21,18 @@ A parse error occurred.
 
 ## Return value
 
-Whether parsing should proceed (**must return `#!cpp false`**).
+Whether to recover from the error:
+
+- `#!cpp false` stops parsing.
+- `#!cpp true` recovers from the error: the error is repaired and parsing continues. If that is not possible, which
+  happens in the binary formats when the end of the item with the error is unknown, the value read so far is completed
+  and parsing stops. See [error recovery](../../features/parsing/error_recovery.md) for how errors are repaired.
+
+Either way, [`sax_parse`](../basic_json/sax_parse.md) returns `#!cpp false`.
 
 ## Examples
 
-??? example
+??? example "Example: (1) the SAX interface"
 
     The example below shows how the SAX interface is used.
 
@@ -39,13 +46,30 @@ Whether parsing should proceed (**must return `#!cpp false`**).
     --8<-- "examples/sax_parse.output"
     ```
 
+??? example "Example: (2) recovering from errors"
+
+    The example below shows how a SAX parser recovers from errors.
+
+    ```cpp
+    --8<-- "examples/sax_parse__error_recovery.cpp"
+    ```
+    
+    Output:
+    
+    ```
+    --8<-- "examples/sax_parse__error_recovery.output"
+    ```
+
 ## See also
 
 - [`sax_parse`](../basic_json/sax_parse.md) - SAX parser
 - [Parsing and Exceptions](../../features/parsing/parse_exceptions.md) - the article on handling parse errors without
   exceptions
+- [Error Recovery](../../features/parsing/error_recovery.md) - the article on recovering from parse errors
 - [SAX Interface](../../features/parsing/sax_interface.md) - guide to event-based parsing
 
 ## Version history
 
 - Added in version 3.2.0.
+- Returning `#!cpp true` recovers from the error since version 3.13.0; before, parsing stopped, but the result of
+  [`sax_parse`](../basic_json/sax_parse.md) could be wrong.

docs/mkdocs/docs/examples/sax_parse__error_recovery.cpp

@@ -0,0 +1,43 @@
+#include <iostream>
+#include <iomanip>
+#include <nlohmann/json.hpp>
+
+using json = nlohmann::json;
+
+// a SAX parser that creates a JSON value like json::parse does, but that
+// recovers from parse errors instead of stopping at the first one
+class recovering_parser : public nlohmann::detail::json_sax_dom_parser<json>
+{
+  public:
+    explicit recovering_parser(json& result)
+        : nlohmann::detail::json_sax_dom_parser<json>(result, false)
+    {}
+
+    bool parse_error(std::size_t position,
+                     const std::string& /*last_token*/,
+                     const json::exception& ex)
+    {
+        std::cout << "byte " << position << ": " << ex.what() << '\n';
+
+        // repair the input and continue
+        return true;
+    }
+};
+
+int main()
+{
+    // JSON text with several mistakes that ends too early
+    const std::string text = R"({
+    "name": "Hello World",
+    "tags": ["a" "b",],
+    "valid": tru,
+    "size": 1.,
+    "nested": {"x": 1)";
+
+    json result;
+    recovering_parser sax(result);
+    const bool valid = json::sax_parse(text, &sax);
+
+    std::cout << "\nvalid JSON: " << std::boolalpha << valid << '\n'
+              << std::setw(4) << result << std::endl;
+}

docs/mkdocs/docs/examples/sax_parse__error_recovery.output

@@ -0,0 +1,19 @@
+byte 49: [json.exception.parse_error.101] parse error at line 3, column 20: syntax error while parsing array - unexpected string literal; expected ']'
+byte 51: [json.exception.parse_error.101] parse error at line 3, column 22: syntax error while parsing value - unexpected ']'; expected '[', '{', or a literal
+byte 70: [json.exception.parse_error.101] parse error at line 4, column 17: syntax error while parsing value - invalid literal; last read: '"valid": tru,'
+byte 86: [json.exception.parse_error.101] parse error at line 5, column 15: syntax error while parsing value - invalid number; expected digit after '.'; last read: '1.,'
+byte 109: [json.exception.parse_error.101] parse error at line 6, column 22: syntax error while parsing object - unexpected end of input; expected '}'
+
+valid JSON: false
+{
+    "name": "Hello World",
+    "nested": {
+        "x": 1
+    },
+    "size": 1,
+    "tags": [
+        "a",
+        "b"
+    ],
+    "valid": null
+}

docs/mkdocs/docs/features/json_view.md

@@ -134,9 +134,21 @@ document: `#!cpp auto v = json_document::parse(text).root();` does not compile.
 - **Macros.** [`JSON_STRICT_NUL_HANDLING`](../api/macros/json_strict_nul_handling.md) and
   [`JSON_NOEXCEPTION`](../api/macros/json_noexception.md)/[`JSON_THROW_USER`](../api/macros/json_throw_user.md)
   behave the same way they do for `<nlohmann/json.hpp>`. These are the macros that change what the parser accepts or
-  how it reports errors. The others either do not apply ([`JSON_DIAGNOSTICS`](../api/macros/json_diagnostics.md)
+  how it reports errors. Two more macros add context to the exceptions of views, as they do for those of
-  needs parent pointers in `basic_json` values) or only affect the values `materialize()` produces, through
+  `#!cpp basic_json`:
-  `BasicJsonType`.
+
+    - With [`JSON_DIAGNOSTICS`](../api/macros/json_diagnostics.md), an exception names the JSON Pointer path of the
+      value within the document.
+    - With [`JSON_DIAGNOSTIC_POSITIONS`](../api/macros/json_diagnostic_positions.md), an exception names the bytes of
+      the value in the source text.
+
+    A view has neither parent pointers nor stored positions, so both are found from the document's index when the
+    exception is thrown. This takes time linear in the size of the document. One exception has neither:
+    `key()` of an iterator over an array or a primitive value
+    ([`invalid_iterator.207`](../home/exceptions.md#jsonexceptioninvalid_iterator207)), because a view's iterator does
+    not know the value it iterates over.
+
+    The other macros only affect the values `materialize()` produces, through `BasicJsonType`.
 
 ## What is different
 
@@ -167,10 +179,6 @@ document: `#!cpp auto v = json_document::parse(text).root();` does not compile.
   [`at`](../api/basic_json_view/at.md) throws for every missing value.
 - **Duplicate keys: lookups find the first member, `parse()` keeps the last.** See [Duplicate keys](#duplicate-keys)
   below.
-- **No [`JSON_DIAGNOSTICS`](../api/macros/json_diagnostics.md) path.** Exceptions thrown by `basic_json_view`'s own
-  element access and lookup functions never carry the JSON Pointer path `JSON_DIAGNOSTICS` would otherwise add: the
-  view has no `basic_json` value to point at, so the exception is created without one, regardless of how
-  `BasicJsonType` was built.
 - **Ordering comparisons are not provided** by `basic_json_view` — there is no `#!cpp operator<`.
   [`operator==`](../api/basic_json_view/operator_eq.md) and [`operator!=`](../api/basic_json_view/operator_ne.md) are
   provided, though: two views, or a view and a `BasicJsonType` value, compare equal exactly when
@@ -253,6 +261,34 @@ Two conversions never copy at all:
 
 Both results are only valid as long as the view — and, for a string with no escapes, the borrowed source text — is.
 
+### `string_view` members
+
+A struct whose member is a `#!cpp std::string_view` (C++17) can be decoded from a view with the
+`NLOHMANN_DEFINE_TYPE_*` macros, and the member is not copied either: for a string without escape sequences it points
+into the document's source text, and for a string with escape sequences into the document's buffer of decoded strings.
+The member is valid as long as the document is alive and, for a document that
+[borrows its source text](#ownership-and-lifetime) (as [`parse()`](../api/basic_json_document/parse.md) does for
+an lvalue `#!cpp std::string`, a `#!cpp std::string_view`, or a C string), as long as that text is.
+
+```cpp
+struct record
+{
+    std::string_view name;
+    int count;
+    NLOHMANN_DEFINE_TYPE_INTRUSIVE(record, name, count)
+};
+
+const std::string text = R"({"name": "widget", "count": 3})";
+const json_document doc = json_document::parse(text);  // borrows text
+const record r = doc.root().get<record>();             // r.name points into text, valid while text and doc are
+```
+
+!!! warning "Decoding into a `string_view` from a `json` value dangles at once"
+
+    The same struct also compiles with `#!cpp json::parse(text).get<record>()`, but there `name` points into the
+    temporary `json` value, which is destroyed at the end of the statement. Use a named
+    [`json_document`](../api/json_document.md), as above, to decode such a struct.
+
 ## Writing generic code
 
 A view's type names its `BasicJsonType` and its [`view_mode`](../api/view_mode.md):
@@ -329,14 +365,10 @@ Its body may only use what views and `basic_json` have in common: the type check
 a view does not have; such a type is converted through `materialize()`. A `from_json()` for `json` only is
 converted through `materialize()` as well.
 
-Two things differ when a `from_json()` reads the view:
+One thing differs when a `from_json()` reads the view: members looked up by name (`at()`, `value()`, ... and thus
-
+the macros) find the **first** member with a repeated key, see [Duplicate keys](#duplicate-keys);
-- **Duplicate keys.** Members looked up by name (`at()`, `value()`, ... and thus the macros) find the **first** member
+`#!cpp materialize().get<T>()` would see the last one. Containers built from an object, such as `#!cpp std::map`,
-  with a repeated key, see [Duplicate keys](#duplicate-keys); `#!cpp materialize().get<T>()` would see the last one.
+keep the **last** value either way, as [`parse()`](../api/basic_json/parse.md) does.
-  Containers built from an object, such as `#!cpp std::map`, keep the **last** value either way, as
-  [`parse()`](../api/basic_json/parse.md) does.
-- **[`JSON_DIAGNOSTICS`](../api/macros/json_diagnostics.md).** Exceptions thrown by the view carry no path to the
-  value.
 
 `to_json()` is unaffected: it writes `basic_json` values, never views.
 

docs/mkdocs/docs/features/parsing/error_recovery.md

@@ -0,0 +1,122 @@
+# Error Recovery
+
+By default, parsing stops at the first error. With the [SAX interface](sax_interface.md), you can instead ask the
+parser to *recover*: to repair the error and continue, so that you get as much as possible out of malformed input, for
+instance a file that was cut off, JSON edited by hand, or the output of a language model.
+
+## Recovering from errors
+
+The SAX parser's [`parse_error`](../../api/json_sax/parse_error.md) function is called for every error. Its return value
+decides what happens next:
+
+- `#!cpp false` stops parsing. This is what the SAX parsers of the library do, so [`parse`](../../api/basic_json/parse.md)
+  and [`accept`](../../api/basic_json/accept.md) never recover.
+- `#!cpp true` repairs the error and continues parsing.
+
+When recovering, the SAX parser still receives well-formed events: every `start_object` or `start_array` is followed by
+the matching `end_object` or `end_array`, and every `key` is followed by exactly one value. A SAX parser that creates a
+JSON value, such as the one in the example below, therefore gets a complete value. Parsing always ends, and
+[`sax_parse`](../../api/basic_json/sax_parse.md) returns `#!cpp false` for input that is not valid JSON, even if every
+error was repaired. Each token is reported at most once, and the SAX parser can stop at any error by returning
+`#!cpp false`.
+
+!!! example
+
+    The example below derives a SAX parser from the library's parser for `json` values (`json_sax_dom_parser`),
+    and recovers from all errors.
+
+    ```cpp
+    --8<-- "examples/sax_parse__error_recovery.cpp"
+    ```
+    
+    Output:
+    
+    ```
+    --8<-- "examples/sax_parse__error_recovery.output"
+    ```
+
+## How errors are repaired
+
+Each error is repaired with the smallest local edit: a missing separator is inserted, a stray token is removed, what can
+be read of a broken string or number is kept, and a value that cannot be read at all becomes `#!json null`.
+
+| Mistake                   | Repair                                                                         | Example                                  | Result                     |
+|---------------------------|--------------------------------------------------------------------------------|------------------------------------------|----------------------------|
+| missing `,` or `:`        | inserted                                                                       | `#!json [1 2]`, `#!json {"a" 1}`         | `[1,2]`, `{"a":1}`         |
+| missing value             | `#!json null` for an object key or between commas in an array                  | `#!json {"a":}`, `#!json [1,,2]`         | `{"a":null}`, `[1,null,2]` |
+| trailing comma            | removed                                                                        | `#!json [1,2,]`                          | `[1,2]`                    |
+| broken string             | invalid escapes and bytes are replaced (see below); a line break ends the string | `#!json ["a\qb"]`                        | `["aqb"]`                  |
+| broken number             | the longest valid beginning is kept                                            | `#!json [1., 2e+]`                       | `[1,2]`                    |
+| unreadable value          | `#!json null`                                                                  | `#!json [1, NaN, tru]`                   | `[1,null,null]`            |
+| number too large          | passed as infinity, together with its text                                     | `#!json [1e999]`                         | infinity (see below)       |
+| stray `:`                 | removed                                                                        | `#!json ["a":1]`                         | `["a",1]`                  |
+| member without a key      | skipped up to the next `,` or `}`                                              | `#!json {1:2, "b":3}`                    | `{"b":3}`                  |
+| wrong closing bracket     | closes the innermost array or object                                           | `#!json {"a":[1,2}, "b":3}`              | `{"a":[1,2],"b":3}`        |
+| input ends too early      | all open arrays and objects are closed                                         | `#!json {"a":[1,2`                       | `{"a":[1,2]}`              |
+| text before the value     | skipped                                                                        | `#!json )]}'{"a":1}`                     | `{"a":1}`                  |
+
+In a string, an unknown escape like `\q` stands for the escaped character (`q`), as in JavaScript. An invalid `\u`
+escape, a lone surrogate, and ill-formed UTF-8 are each replaced by U+FFFD (REPLACEMENT CHARACTER), and control
+characters are kept. A string without its closing quote ends at the next line break or at the end of the input.
+
+The input after the top-level value is not repaired: as without recovery, it is reported as an error, and parsing stops.
+
+## Binary formats
+
+The binary formats ([BJData](../binary_formats/bjdata.md), [BON8](../binary_formats/bon8.md),
+[BSON](../binary_formats/bson.md), [CBOR](../binary_formats/cbor.md), [MessagePack](../binary_formats/messagepack.md),
+and [UBJSON](../binary_formats/ubjson.md)) have no delimiters to find the next value by. So what can be repaired depends
+on whether the end of the item with the error is known, a distinction that
+[RFC 8949, Section 5.3](https://www.rfc-editor.org/rfc/rfc8949.html#section-5.3) makes for CBOR, too.
+
+If the item is complete, but cannot be passed on as it is, it is replaced, and parsing continues after it:
+
+| Mistake                                                             | Formats                 | Repair                                                                  |
+|---------------------------------------------------------------------|-------------------------|-------------------------------------------------------------------------|
+| tag, if `tag_handler` is `cbor_tag_handler_t::error` (the default)  | CBOR                    | ignored                                                                 |
+| simple value other than `false`, `true`, and `null`, like undefined | CBOR                    | `#!json null`                                                           |
+| number too large for a custom `number_float_t`, like `float`        | all                     | infinity                                                                |
+| character (`C`) that is not ASCII                                   | BJData, UBJSON          | U+FFFD                                                                  |
+| invalid high-precision number (`H`)                                 | BJData, UBJSON          | the longest valid beginning is kept, as for JSON text, or `#!json null` |
+| high-precision number too large                                     | BJData, UBJSON          | passed as infinity, together with its text                              |
+| object key that is not a string                                     | BON8, CBOR, MessagePack | the member is skipped                                                   |
+| element of a type the library does not read, like ObjectId or date  | BSON                    | `#!json null`                                                           |
+| string without its terminator                                       | BSON                    | kept                                                                    |
+| document whose size does not match its content                      | BSON                    | kept                                                                    |
+
+CBOR tags and simple values are repaired as [RFC 8949, Section 6.1](https://www.rfc-editor.org/rfc/rfc8949.html#section-6.1)
+suggests for converting CBOR to JSON. By default, [`sax_parse`](../../api/basic_json/sax_parse.md) reports every tag as
+an error; when recovering, tags are then ignored like with
+[`cbor_tag_handler_t::ignore`](../../api/basic_json/cbor_tag_handler_t.md). Strings that are not valid UTF-8 are no
+error: like [`from_cbor`](../../api/basic_json/from_cbor.md) and the other functions by default, `sax_parse` passes
+them on as they are.
+
+After any other error, the end of the item is unknown: the input ended, a byte is not a valid type marker, or a size
+cannot be right. Parsing then stops, and the value read so far is completed: a key that waits for its value gets
+`#!json null`, and all open arrays and objects are closed. This keeps everything before the error of an input that was
+cut off. The exception is BSON, which stores the size of every document: an element whose end is unknown gets
+`#!json null`, the rest of its document is skipped, and parsing continues after the document.
+
+## Limitations
+
+- A repair is a guess. For example, `#!json {"a" "b": 1}` could be meant as `#!json {"a": "b"}` or as
+  `#!json {"a": null, "b": 1}`; it is repaired to the former. Treat recovered values as a best effort, and check the
+  reported errors.
+- A closing bracket always closes the innermost array or object. If a bracket is missing rather than wrong, the
+  repair differs from the intention: `#!json {"a": {"b": [1, 2}, "c": 3}` is repaired to
+  `#!json {"a": {"b": [1, 2], "c": 3}}`, although `#!json {"a": {"b": [1, 2]}, "c": 3}` may have been meant.
+- Keys without quotes, and strings in single quotes, are not supported; such members are skipped.
+- In the binary formats, a member that is skipped because its key is not a string is lost, and so are the elements of a
+  BSON document after one whose end is unknown.
+- A number that is too large for `number_float_t` is passed as positive or negative infinity. The SAX parser's
+  `number_float` also gets the number's text, but a JSON value cannot store it, and
+  [`dump`](../../api/basic_json/dump.md) serializes infinity as `#!json null`.
+- When parsing is not strict (see [`sax_parse`](../../api/basic_json/sax_parse.md)), a repair may read parts of the
+  input after the value, for instance of the next value in a stream of concatenated values.
+
+## See also
+
+- [SAX interface](sax_interface.md) - implement a custom SAX handler
+- [`parse_error`](../../api/json_sax/parse_error.md) - the SAX event for parse errors
+- [`sax_parse`](../../api/basic_json/sax_parse.md) - generate SAX events
+- [parsing and exceptions](parse_exceptions.md) - control error handling

docs/mkdocs/docs/features/parsing/index.md

@@ -141,7 +141,7 @@ See [Zero-copy JSON views](../json_view.md) for the concept and the trade-offs.
 When the input is not valid JSON, the `parse` function throws an exception by default. If exceptions are undesired or
 unavailable, the parser can instead return a discarded value, or [`accept`](../../api/basic_json/accept.md) can be used
 to only check whether an input is valid JSON. See [parsing and exceptions](parse_exceptions.md) for the available
-options.
+options. To get as much as possible out of malformed input, a SAX parser can [recover from errors](error_recovery.md).
 
 ## See also
 
@@ -155,6 +155,7 @@ options.
 - [SAX interface](sax_interface.md) - implement a custom SAX handler
 - [Zero-copy JSON views](../json_view.md) - parse into a read-only document and read it through views
 - [parsing and exceptions](parse_exceptions.md) - control error handling
+- [error recovery](error_recovery.md) - get as much as possible out of malformed input
 - [parsing untrusted input](untrusted_input.md) - what to consider when parsing input from untrusted sources
 - [`operator>>`](../../api/operator_gtgt.md) - deserialize from stream
 - [`json_sax`](../../api/json_sax/index.md) - documentation of the SAX interface

docs/mkdocs/docs/features/parsing/parse_exceptions.md

@@ -68,7 +68,8 @@ bool parse_error(std::size_t position,
                  const json::exception& ex);
 ```
 
-The return value indicates whether the parsing should continue, so the function should usually return `#!cpp false`.
+The return value decides whether to stop parsing (`#!cpp false`) or to repair the error and continue
+(`#!cpp true`); see [error recovery](error_recovery.md) for the latter.
 
 ??? example "Example: report parse errors without exceptions"
 

docs/mkdocs/docs/features/parsing/sax_interface.md

@@ -62,7 +62,8 @@ bool key(string_t& val);
 bool parse_error(std::size_t position, const std::string& last_token, const json::exception& ex);
 ```
 
-The return value of each function determines whether parsing should proceed.
+The return value of each function determines whether parsing should proceed. For `parse_error`, returning
+`#!cpp true` [recovers from the error](error_recovery.md).
 
 Object keys are passed to `key` decoded and validated: escape sequences are already resolved, so different spellings of
 the same name yield the same string. See [`key`](../../api/json_sax/key.md#notes) for the exact guarantees.
@@ -101,7 +102,7 @@ To implement your own SAX handler, proceed as follows:
 2. Create an object of your SAX interface class, e.g. `my_sax`.
 3. Call `#!cpp bool json::sax_parse(input, &my_sax);` where the first parameter can be any input like a string or an input stream and the second parameter is a pointer to your SAX interface.
 
-Note the `sax_parse` function only returns a `#!cpp bool` indicating the result of the last executed SAX event. It does not return `json` value - it is up to you to decide what to do with the SAX events. Furthermore, no exceptions are thrown in case of a parse error - it is up to you what to do with the exception object passed to your `parse_error` implementation. Internally, the SAX interface is used for the DOM parser (class `json_sax_dom_parser`) as well as the acceptor (`json_sax_acceptor`), see file `json_sax.hpp`.
+Note the `sax_parse` function only returns a `#!cpp bool` indicating whether the input was parsed without errors and no SAX event returned `#!cpp false`. It does not return `json` value - it is up to you to decide what to do with the SAX events. Furthermore, no exceptions are thrown in case of a parse error - it is up to you what to do with the exception object passed to your `parse_error` implementation. Internally, the SAX interface is used for the DOM parser (class `json_sax_dom_parser`) as well as the acceptor (`json_sax_acceptor`), see file `json_sax.hpp`.
 
 ## See also
 

docs/mkdocs/docs/home/debugging.md

@@ -10,6 +10,13 @@ at its root, a [Natvis](https://learn.microsoft.com/en-us/visualstudio/debugger/
 file that gives `json`/`ordered_json` values a friendly, key/value debugger view instead of showing raw internal
 fields, when debugging with the MSVC debug engine (`cppvsdbg`) in Visual Studio or VS Code.
 
+The file also has visualizers for `json_document` and `json_view` (and their `ordered_json`/editable variants). They
+show the value of the document or view, and its elements as an expandable list. Because a document is only a flat index
+into the source text, the strings of the direct elements are shown, but the elements of nested arrays and objects are
+shown as index nodes (their kind, size, and numbers; strings only by length). Arrays and objects that have been edited
+in a `json_editable_document` are not expanded. The visualizers read the internal layout of the MSVC STL (`std::array`,
+`std::unique_ptr`).
+
 The checked-in visualizers cover namespace variants with zero, one, or two ABI tags. Configurations with three or more
 ABI tags are not covered and may show raw internal fields instead. To generate a Natvis file for such a configuration:
 
@@ -29,7 +36,7 @@ ABI tags are not covered and may show raw internal fields instead. To generate a
 
 4. Use the generated file instead of `nlohmann_json.natvis`.
 
-`--abi-tags` can be repeated for several configurations, and `--all` generates every combination (about 60 MB for 12
+`--abi-tags` can be repeated for several configurations, and `--all` generates every combination (about 60 MB for 11
 ABI tags, doubling with each further tag). See
 [`tools/generate_natvis`](https://github.com/nlohmann/json/tree/develop/tools/generate_natvis) for all options.
 

docs/mkdocs/mkdocs.yml

@@ -142,6 +142,7 @@ nav:
       - features/performance.md
       - Parsing:
           - features/parsing/index.md
+          - features/parsing/error_recovery.md
           - features/parsing/json_lines.md
           - features/parsing/large_and_partial_input.md
           - features/parsing/parse_exceptions.md

include/nlohmann/detail/exceptions.hpp

@@ -431,6 +431,28 @@ std::nullptr_t exception_context(const View& /*unused*/) noexcept
     return nullptr;
 }
 
+namespace view
+{
+// the path and bytes of the value of a view (defined in <nlohmann/detail/view/errors.hpp>)
+template<typename View>
+std::string diagnostics_of(const View& v);
+}  // namespace view
+
+// what goes in front of the message of an exception about j: nothing for a
+// basic_json value (exception_context() passes the value itself), the context
+// basic_json would add (JSON_DIAGNOSTICS, JSON_DIAGNOSTIC_POSITIONS) for a view
+template<typename BasicJsonType, enable_if_t<is_basic_json<BasicJsonType>::value, int> = 0>
+const char* exception_prefix(const BasicJsonType& /*unused*/) noexcept
+{
+    return "";
+}
+
+template<typename View, enable_if_t<is_json_view<View>::value, int> = 0>
+std::string exception_prefix(const View& v)
+{
+    return view::diagnostics_of(v);
+}
+
 /*!
 @brief throws because @a j does not have the type a conversion expects
 @param[in] expected  the expected type(s), e.g. "array" or "binary or array"
@@ -442,7 +464,7 @@ JSON_HEDLEY_NO_RETURN JSON_HEDLEY_NEVER_INLINE void throw_type_must_be(const cha
 {
     static_cast<void>(expected); // unused when JSON_NOEXCEPTION is defined
     static_cast<void>(j);
-    JSON_THROW(type_error::create(exception_id::type_mismatch, concat("type must be ", expected, ", but is ", j.type_name()), exception_context(j)));
+    JSON_THROW(type_error::create(exception_id::type_mismatch, concat(exception_prefix(j), "type must be ", expected, ", but is ", j.type_name()), exception_context(j)));
 }
 
 /*!

include/nlohmann/detail/input/input_adapters.hpp

@@ -796,17 +796,22 @@ class iterator_input_adapter
         return get_consumed_count() + eof_reads;
     }
 
-    // number of newlines in the consumed range [0, last_index); @a line_start
+    // number of newlines in the consumed range [first_index, last_index); if
-    // is set to the index after the last of them (0 if there is none); only
+    // there is one, @a line_start is set to the index after the last of them,
-    // valid when supports_seek
+    // otherwise it is left unchanged; only valid when supports_seek
-    std::size_t count_consumed_lines(std::size_t last_index, std::size_t& line_start) const
+    std::size_t count_consumed_lines(std::size_t first_index, std::size_t last_index, std::size_t& line_start) const
     {
-        const auto to = std::next(begin, static_cast<typename std::iterator_traits<IteratorType>::difference_type>(last_index));
+        using difference_type = typename std::iterator_traits<IteratorType>::difference_type;
+        const auto from = std::next(begin, static_cast<difference_type>(first_index));
+        const auto to = std::next(from, static_cast<difference_type>(last_index - first_index));
         const auto newline = char_traits<char_type>::to_char_type('\n');
         using reverse_it = std::reverse_iterator<IteratorType>;
-        const auto last_newline = std::find(reverse_it(to), reverse_it(begin), newline);
+        const auto last_newline = std::find(reverse_it(to), reverse_it(from), newline);
-        line_start = static_cast<std::size_t>(std::distance(begin, last_newline.base()));
+        if (last_newline.base() != from)
-        return static_cast<std::size_t>(std::count(begin, last_newline.base(), newline));
+        {
+            line_start = static_cast<std::size_t>(std::distance(begin, last_newline.base()));
+        }
+        return static_cast<std::size_t>(std::count(from, last_newline.base(), newline));
     }
 
     // append the already-consumed characters in the half-open range

include/nlohmann/detail/input/json_sax.hpp

@@ -138,7 +138,9 @@ struct json_sax
     @param[in] position    the position in the input where the error occurs
     @param[in] last_token  the last read token
     @param[in] ex          an exception object describing the error
-    @return whether parsing should proceed (must return false)
+    @return whether to recover from the error: false stops parsing; true
+            repairs the error and continues, or, if that is not possible,
+            stops after completing the value read so far
     */
     virtual bool parse_error(std::size_t position,
                              const std::string& last_token,
@@ -485,6 +487,9 @@ stored in place as it arrives, so after a parse error the value holds what was
 read up to it.
 
 @tparam BasicJsonType  the JSON type
+@tparam InputAdapterType  the input adapter of the lexer that can be passed to
+                          the constructor to record diagnostic positions; it
+                          does not matter if no lexer is passed
 @tparam Collect  whether to collect the values of containers of unknown size on
                  a stack (only for a caller that does not use the value after a
                  parse error)
@@ -493,7 +498,7 @@ read up to it.
                    it, so any type with its three functions works (as
                    materialize() of json_view uses it)
 */
-template<typename BasicJsonType, typename InputAdapterType, bool Collect = false, typename LexerType = lexer<BasicJsonType, InputAdapterType>>
+template<typename BasicJsonType, typename InputAdapterType = string_input_adapter_type, bool Collect = false, typename LexerType = lexer<BasicJsonType, InputAdapterType>>
 class json_sax_dom_parser : public json_sax_dom_error_state
 {
   public:
@@ -799,6 +804,8 @@ class json_sax_dom_parser : public json_sax_dom_error_state
         JSON_HEDLEY_NEVER_INLINE
         void grow()
         {
+            // the stack starts with local_capacity frames and only grows
+            JSON_ASSERT(capacity != 0);
             const std::size_t new_capacity = capacity * 2;
             std::unique_ptr<frame[]> new_heap(new frame[new_capacity]); // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
             for (std::size_t i = 0; i < count; ++i)
@@ -1614,7 +1621,7 @@ class json_sax_dom_binary_parser : public json_sax_dom_error_state
     lexer_t* m_lexer_ref = nullptr;
 };
 
-template<typename BasicJsonType, typename InputAdapterType>
+template<typename BasicJsonType, typename InputAdapterType = string_input_adapter_type>
 class json_sax_dom_callback_parser : public json_sax_dom_error_state
 {
   public:

include/nlohmann/detail/input/parser.hpp

@@ -72,14 +72,14 @@ This class implements a parser for JSON text. Nested arrays and objects are trac
 stack instead of recursion, so deeply nested input does not exhaust the call stack, and what is read
 is reported as SAX events.
 */
-template<typename BasicJsonType, typename InputAdapterType>
+template<typename BasicJsonType, typename InputAdapterType, bool DiscardValues = false>
 class parser
 {
     using number_integer_t = typename BasicJsonType::number_integer_t;
     using number_unsigned_t = typename BasicJsonType::number_unsigned_t;
     using number_float_t = typename BasicJsonType::number_float_t;
     using string_t = typename BasicJsonType::string_t;
-    using lexer_t = lexer<BasicJsonType, InputAdapterType>;
+    using lexer_t = lexer<BasicJsonType, InputAdapterType, DiscardValues>;
     using token_type = typename lexer_t::token_type;
 
   public:
@@ -88,10 +88,9 @@ class parser
                     parser_callback_t<BasicJsonType> cb = nullptr,
                     const bool allow_exceptions_ = true,
                     const bool ignore_comments = false,
-                    const bool ignore_trailing_commas_ = false,
+                    const bool ignore_trailing_commas_ = false)
-                    const bool discard_number_values_ = false)
         : callback(std::move(cb))
-        , m_lexer(std::move(adapter), ignore_comments, discard_number_values_)
+        , m_lexer(std::move(adapter), ignore_comments)
         , allow_exceptions(allow_exceptions_)
         , ignore_trailing_commas(ignore_trailing_commas_)
     {
@@ -153,25 +152,72 @@ class parser
     bool accept(const bool strict = true)
     {
         json_sax_acceptor<BasicJsonType> sax_acceptor;
-        return sax_parse(&sax_acceptor, strict);
+        return sax_parse_impl<false>(&sax_acceptor, strict);
     }
 
+    /*!
+    @brief public SAX interface
+
+    If the SAX parser's parse_error() returns true, the parser recovers from
+    the error: it repairs the input and continues (see #3989).
+
+    @param[in] sax     the SAX parser
+    @param[in] strict  whether to expect the last token to be EOF
+    @return whether the input was parsed without errors and no SAX event
+            returned false
+    */
     template<typename SAX>
     JSON_HEDLEY_NON_NULL(2)
     bool sax_parse(SAX* sax, const bool strict = true)
     {
-        (void)detail::is_sax_static_asserts<SAX, BasicJsonType> {};
+        return sax_parse_impl<true>(sax, strict);
-        const bool result = sax_parse_internal(sax);
-
-        if (result)
-        {
-            return finish_input(*sax, strict);
-        }
-
-        return false;
     }
 
   private:
+    /// what sax_parse_internal() does after an object key was expected
+    enum class next_step : std::uint8_t
+    {
+        /// stop parsing
+        stop,
+        /// parse a value that begins with last_token
+        parse_value,
+        /// evaluate the state of the innermost container, which reads
+        /// last_token again
+        evaluate_state
+    };
+
+    template<bool AllowRecovery, typename SAX>
+    JSON_HEDLEY_NON_NULL(2)
+    bool sax_parse_impl(SAX* sax, const bool strict)
+    {
+        (void)detail::is_sax_static_asserts<SAX, BasicJsonType> {};
+        const bool result = sax_parse_internal<AllowRecovery>(sax);
+
+        if (result)
+        {
+            if (strict)
+            {
+                // strict mode: next byte must be EOF; after recovering from an
+                // error, the end of the input may already have been read
+                if (last_token != token_type::end_of_input && get_token() != token_type::end_of_input)
+                {
+                    // the value is complete, so there is nothing to recover
+                    static_cast<void>(report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(), exception_message(token_type::end_of_input, "value"), nullptr),
+                                                   std::integral_constant<bool, AllowRecovery> {}));
+                    return false;
+                }
+            }
+            else
+            {
+                // the caller keeps using the input: position it right after
+                // the value by leaving the character that terminated it
+                m_lexer.release_lookahead();
+            }
+        }
+
+        return result && !error_reported;
+    }
+
     /// the text of the current float token for @a sax (see sax_ignores_number_text)
     template<typename SAX>
     const string_t& number_float_text(SAX* /*sax*/)
@@ -239,17 +285,26 @@ class parser
     template<typename DomSax>
     bool parse_dom(DomSax& sdp, const bool strict)
     {
-        sax_parse_internal(&sdp);
+        sax_parse_internal<false>(&sdp);
 
         static_cast<void>(finish_input(sdp, strict));
 
         return !sdp.is_errored();
     }
 
-    template<typename SAX>
+    /*!
+    @brief parse a JSON value and pass it to a SAX parser
+
+    @tparam AllowRecovery  whether to recover from an error if the SAX parser's
+                           parse_error() returns true; false for the SAX parsers
+                           of parse() and accept(), which never do, so that no
+                           code for recovering is generated for them
+    */
+    template<bool AllowRecovery, typename SAX>
     JSON_HEDLEY_NON_NULL(2)
     bool sax_parse_internal(SAX* sax)
     {
+        const std::integral_constant<bool, AllowRecovery> allow_recovery{};
         // value to avoid a goto (see comment where set to true)
         bool skip_to_state_evaluation = false;
 
@@ -288,10 +343,18 @@ class parser
                             break;
                         }
 
-                        // parse key
+                        // remember we are now inside an object
+                        states.push_back(0);
+
+                        // parse key (the steps of parse_key(), which are
+                        // repeated here and below for speed)
                         if (JSON_HEDLEY_UNLIKELY(last_token != token_type::value_string))
                         {
-                            return syntax_error(*sax, exception_message(token_type::value_string, "object key"));
+                            if (!continue_after(key_error(sax, allow_recovery, false), skip_to_state_evaluation))
+                            {
+                                return false;
+                            }
+                            continue;
                         }
                         if (JSON_HEDLEY_UNLIKELY(!sax->key(m_lexer.get_string_token())))
                         {
@@ -301,12 +364,13 @@ class parser
                         // parse separator (:)
                         if (JSON_HEDLEY_UNLIKELY(!get_token_expecting(token_type::name_separator)))
                         {
-                            return syntax_error(*sax, exception_message(token_type::name_separator, "object separator"));
+                            if (!continue_after(key_error(sax, allow_recovery, true), skip_to_state_evaluation))
+                            {
+                                return false;
+                            }
+                            continue;
                         }
 
-                        // remember we are now inside an object
-                        states.push_back(0);
-
                         // parse values
                         get_token();
                         continue;
@@ -342,9 +406,11 @@ class parser
 
                         if (JSON_HEDLEY_UNLIKELY(!std::isfinite(res)))
                         {
-                            return sax->parse_error(m_lexer.get_position(),
+                            if (!overflow_error(sax, res, allow_recovery))
-                                                    m_lexer.get_token_string(),
+                            {
-                                                    out_of_range::create(exception_id::number_overflow, concat("number overflow parsing '", m_lexer.get_token_string(), '\''), nullptr));
+                                return false;
+                            }
+                            break;
                         }
 
                         if (JSON_HEDLEY_UNLIKELY(!sax->number_float(res, number_float_text(sax))))
@@ -412,16 +478,63 @@ class parser
                     case token_type::parse_error:
                     {
                         // using "uninitialized" to avoid an "expected" message
-                        return syntax_error(*sax, exception_message(token_type::uninitialized, "value"));
+                        if (!report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(), exception_message(token_type::uninitialized, "value"), nullptr), allow_recovery))
+                        {
+                            return false;
+                        }
+
+                        // recover: keep what can be read of the token
+                        recover_token(allow_recovery);
+                        if (last_token != token_type::uninitialized)
+                        {
+                            // a string or a number
+                            continue;
+                        }
+                        if (states.empty())
+                        {
+                            // look for the value after the garbage
+                            if (!skip_to_value(allow_recovery))
+                            {
+                                return false;
+                            }
+                            continue;
+                        }
+                        // nothing could be read
+                        if (JSON_HEDLEY_UNLIKELY(!sax->null()))
+                        {
+                            return false;
+                        }
+                        break;
                     }
                     case token_type::end_of_input:
                     {
                         if (JSON_HEDLEY_UNLIKELY(m_lexer.get_chars_read_total() == 1))
                         {
-                            return syntax_error(*sax, empty_input_message());
+                            // there is nothing to recover
+                            static_cast<void>(report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(),
+                                                           empty_input_message(), nullptr), allow_recovery));
+                            return false;
                         }
 
-                        return syntax_error(*sax, exception_message(token_type::literal_or_value, "value"));
+                        if (!report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(), exception_message(token_type::literal_or_value, "value"), nullptr), allow_recovery))
+                        {
+                            return false;
+                        }
+
+                        // recover: the input ends where a value is missing
+                        if (states.empty())
+                        {
+                            // there is no value
+                            return false;
+                        }
+                        if (!recover_missing_value(sax, states, allow_recovery))
+                        {
+                            return false;
+                        }
+                        // the state evaluation reads the token again
+                        unget_token(allow_recovery);
+                        skip_to_state_evaluation = true;
+                        continue;
                     }
                     case token_type::uninitialized:
                     case token_type::end_array:
@@ -431,7 +544,35 @@ class parser
                     case token_type::literal_or_value:
                     default: // the last token was unexpected
                     {
-                        return syntax_error(*sax, exception_message(token_type::literal_or_value, "value"));
+                        if (!report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(), exception_message(token_type::literal_or_value, "value"), nullptr), allow_recovery))
+                        {
+                            return false;
+                        }
+
+                        // recover
+                        if (states.empty())
+                        {
+                            // look for the value after the garbage
+                            if (!skip_to_value(allow_recovery))
+                            {
+                                return false;
+                            }
+                            continue;
+                        }
+                        if (last_token == token_type::name_separator)
+                        {
+                            // a stray ':'; the value may follow
+                            get_token();
+                            continue;
+                        }
+                        if (!recover_missing_value(sax, states, allow_recovery))
+                        {
+                            return false;
+                        }
+                        // the state evaluation reads the token again
+                        unget_token(allow_recovery);
+                        skip_to_state_evaluation = true;
+                        continue;
                     }
                 }
             }
@@ -482,7 +623,30 @@ class parser
                     continue;
                 }
 
-                return syntax_error(*sax, exception_message(token_type::end_array, "array"));
+                if (!report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(), exception_message(token_type::end_array, "array"), nullptr), allow_recovery))
+                {
+                    return false;
+                }
+
+                // recover
+                if (last_token == token_type::end_of_input)
+                {
+                    // the input ends inside the array
+                    return close_containers(sax, states, allow_recovery);
+                }
+                if (last_token == token_type::end_object)
+                {
+                    // a wrong closing bracket closes the innermost container
+                    if (JSON_HEDLEY_UNLIKELY(!sax->end_array()))
+                    {
+                        return false;
+                    }
+                    states.pop_back();
+                    skip_to_state_evaluation = true;
+                }
+                // otherwise, a missing ',' (or a stray ':', which value
+                // parsing drops): the next value begins here
+                continue;
             }
 
             // states.back() is false -> object
@@ -499,9 +663,12 @@ class parser
                     // parse key
                     if (JSON_HEDLEY_UNLIKELY(last_token != token_type::value_string))
                     {
-                        return syntax_error(*sax, exception_message(token_type::value_string, "object key"));
+                        if (!continue_after(key_error(sax, allow_recovery, false), skip_to_state_evaluation))
+                        {
+                            return false;
+                        }
+                        continue;
                     }
-
                     if (JSON_HEDLEY_UNLIKELY(!sax->key(m_lexer.get_string_token())))
                     {
                         return false;
@@ -510,7 +677,11 @@ class parser
                     // parse separator (:)
                     if (JSON_HEDLEY_UNLIKELY(!get_token_expecting(token_type::name_separator)))
                     {
-                        return syntax_error(*sax, exception_message(token_type::name_separator, "object separator"));
+                        if (!continue_after(key_error(sax, allow_recovery, true), skip_to_state_evaluation))
+                        {
+                            return false;
+                        }
+                        continue;
                     }
 
                     // parse values
@@ -538,10 +709,546 @@ class parser
                 continue;
             }
 
-            return syntax_error(*sax, exception_message(token_type::end_object, "object"));
+            if (!report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(), exception_message(token_type::end_object, "object"), nullptr), allow_recovery))
+            {
+                return false;
+            }
+
+            // recover
+            if (last_token == token_type::end_of_input)
+            {
+                // the input ends inside the object
+                return close_containers(sax, states, allow_recovery);
+            }
+            if (last_token == token_type::end_array)
+            {
+                // a wrong closing bracket closes the innermost container
+                if (JSON_HEDLEY_UNLIKELY(!sax->end_object()))
+                {
+                    return false;
+                }
+                states.pop_back();
+                skip_to_state_evaluation = true;
+                continue;
+            }
+            if (!continue_after(recover_member(sax, allow_recovery), skip_to_state_evaluation))
+            {
+                return false;
+            }
         }
     }
 
+    /*!
+    @brief continue sax_parse_internal() after a recovery
+    @return whether to continue parsing
+    */
+    bool continue_after(const next_step step, bool& skip_to_state_evaluation)
+    {
+        if (step == next_step::evaluate_state)
+        {
+            // the state evaluation reads the token again
+            m_lexer.unget_token();
+            skip_to_state_evaluation = true;
+        }
+        return step != next_step::stop;
+    }
+
+    /// the parser for parse() and accept() never recovers: stop parsing
+    JSON_INTERNAL_ALWAYS_INLINE
+    static std::false_type continue_after(std::false_type /*step*/, bool& /*skip_to_state_evaluation*/) noexcept
+    {
+        return {};
+    }
+
+    /*!
+    @brief parse an object key and the name separator (:) after it
+
+    last_token is the token where the key is expected. sax_parse_internal()
+    repeats these steps rather than calling this function, which is used
+    when recovering from an error.
+
+    @return next_step::parse_value if the value follows, with last_token its
+            first token; next_step::evaluate_state if the object's state is
+            to be evaluated after recovering from an error; next_step::stop
+            to stop parsing
+    */
+    template<typename SAX>
+    next_step parse_key(SAX* sax)
+    {
+        const std::true_type allow_recovery{};
+
+        if (JSON_HEDLEY_UNLIKELY(last_token != token_type::value_string))
+        {
+            return key_error(sax, allow_recovery, false);
+        }
+
+        if (JSON_HEDLEY_UNLIKELY(!sax->key(m_lexer.get_string_token())))
+        {
+            return next_step::stop;
+        }
+
+        // parse separator (:)
+        if (JSON_HEDLEY_UNLIKELY(get_token() != token_type::name_separator))
+        {
+            return key_error(sax, allow_recovery, true);
+        }
+
+        // the value begins with the next token
+        get_token();
+        return next_step::parse_value;
+    }
+
+    /*!
+    @brief report a number that is too large for number_float_t, and recover
+           from the error by passing the value on; the SAX parser gets the
+           number's text as well
+
+    This is a separate function, as reading other numbers is measurably
+    slower if the error is handled where they are read.
+
+    @param[in] sax    the SAX parser
+    @param[in] value  the value that is not finite
+    @return whether to continue parsing
+    */
+    template<typename SAX>
+    bool overflow_error(SAX* sax, const number_float_t value, std::true_type allow_recovery)
+    {
+        if (!report_error(sax, out_of_range::create(exception_id::number_overflow, concat("number overflow parsing '", m_lexer.get_token_string(), '\''), nullptr), allow_recovery))
+        {
+            return false;
+        }
+        return sax->number_float(value, number_float_text(sax));
+    }
+
+    /// @copydoc overflow_error
+    template<typename SAX>
+    JSON_INTERNAL_ALWAYS_INLINE
+    std::false_type overflow_error(SAX* sax, const number_float_t /*value*/, std::false_type allow_recovery)
+    {
+        return report_error(sax, out_of_range::create(exception_id::number_overflow, concat("number overflow parsing '", m_lexer.get_token_string(), '\''), nullptr), allow_recovery);
+    }
+
+    /*!
+    @brief report a missing key, or a missing name separator (:) after the
+           key; the parser for parse() and accept() never recovers
+
+    @param[in] key_read  whether the key was read, so that the name separator
+                         is missing
+    @return std::false_type, see report_error()
+    */
+    template<typename SAX>
+    JSON_INTERNAL_ALWAYS_INLINE
+    std::false_type key_error(SAX* sax, std::false_type allow_recovery, const bool key_read)
+    {
+        return report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(), key_read
+                            ? exception_message(token_type::name_separator, "object separator")
+                            : exception_message(token_type::value_string, "object key"), nullptr), allow_recovery);
+    }
+
+    /*!
+    @brief report a missing key, or a missing name separator (:) after the
+           key, and recover from it
+
+    @param[in] key_read  whether the key was read, so that the name separator
+                         is missing
+    */
+    template<typename SAX>
+    next_step key_error(SAX* sax, std::true_type allow_recovery, const bool key_read)
+    {
+        if (!key_read)
+        {
+            if (!report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(), exception_message(token_type::value_string, "object key"), nullptr), allow_recovery))
+            {
+                return next_step::stop;
+            }
+            return recover_key(sax);
+        }
+
+        if (!report_error(sax, parse_error::create(exception_id::syntax_error, m_lexer.get_position(), exception_message(token_type::name_separator, "object separator"), nullptr), allow_recovery))
+        {
+            return next_step::stop;
+        }
+        return recover_name_separator(sax);
+    }
+
+    /////////////////////
+    // error recovery
+    /////////////////////
+
+    /*
+    The functions below repair an error after the SAX parser's parse_error()
+    returned true (see #3989). Each mistake is repaired by the smallest local
+    edit: a missing ',' or ':' is inserted, a stray token is removed, what can
+    be read of an invalid string or number is kept (see
+    lexer::recover_token()), a missing value becomes null, a wrong closing
+    bracket closes the innermost container, and the end of the input closes
+    all of them. The events stay balanced, and every key() is followed by
+    exactly one value.
+
+    A repair hands a token to the state evaluation, by returning it to the
+    lexer (lexer::unget_token()) so that the state evaluation reads it again,
+    only if it is ',', ']', '}', or the end of the input. The state evaluation
+    hands a token to value or key parsing only if it is none of them, so a
+    token is never handed back and forth. Every other step reads a token or
+    closes a container, so parsing always ends.
+    */
+
+    /*!
+    @brief report an error to the SAX parser; the parser for parse() and
+           accept() never recovers
+
+    @return std::false_type rather than false: its value is known where the
+            function is called even if the call is not inlined, so the code
+            for recovering is not generated
+    */
+    template<typename SAX, typename Exception>
+    JSON_INTERNAL_ALWAYS_INLINE
+    std::false_type report_error(SAX* sax, const Exception& ex, std::false_type /*allow_recovery*/)
+    {
+        static_cast<void>(sax); // MSVC 2015 does not count calling a static parse_error() as using it
+        error_reported = true;
+        static_cast<void>(sax->parse_error(m_lexer.get_position(), m_lexer.get_token_string(), ex));
+        return {};
+    }
+
+    /*!
+    @brief report an error to the SAX parser
+    @return whether to recover from the error
+    */
+    template<typename SAX, typename Exception>
+    bool report_error(SAX* sax, const Exception& ex, std::true_type /*allow_recovery*/)
+    {
+        static_cast<void>(sax); // MSVC 2015 does not count calling a static parse_error() as using it
+        const std::size_t position = m_lexer.get_position().chars_read_total;
+        if (error_reported && position == last_error_position && last_token == last_error_token)
+        {
+            // a repair handed on the token of the error it repaired; the
+            // token was reported already, and the SAX parser asked to recover
+            return true;
+        }
+
+        error_reported = true;
+        last_error_position = position;
+        last_error_token = last_token;
+
+        if (!sax->parse_error(m_lexer.get_position(), m_lexer.get_token_string(), ex))
+        {
+            return false;
+        }
+
+        // the token string of the next error begins here
+        m_lexer.restart_token_string();
+        return true;
+    }
+
+    /*!
+    @brief keep what can be read of the token that the lexer rejected
+
+    The error was reported for the rejected token, so it is not reported again
+    for the token it is repaired to (see lexer::recover_token()).
+    */
+    token_type recover_token()
+    {
+        last_token = m_lexer.recover_token();
+        last_error_position = m_lexer.get_position().chars_read_total;
+        last_error_token = last_token;
+        return last_token;
+    }
+
+    // The functions below are called from sax_parse_internal() after an error
+    // was reported. The parser for parse() and accept() stopped there, so for
+    // it they are stubs: the code for recovering is then not even referenced,
+    // and unoptimized builds do not emit it either.
+
+    /// @copydoc recover_token()
+    JSON_INTERNAL_ALWAYS_INLINE
+    token_type recover_token(std::true_type /*allow_recovery*/)
+    {
+        return recover_token();
+    }
+
+    /// @copydoc recover_token()
+    JSON_INTERNAL_ALWAYS_INLINE
+    static std::false_type recover_token(std::false_type /*allow_recovery*/) noexcept
+    {
+        return {};
+    }
+
+    /// return the token that was read last to the lexer (see lexer::unget_token())
+    JSON_INTERNAL_ALWAYS_INLINE
+    void unget_token(std::true_type /*allow_recovery*/)
+    {
+        m_lexer.unget_token();
+    }
+
+    /// @copydoc unget_token(std::true_type)
+    JSON_INTERNAL_ALWAYS_INLINE
+    static void unget_token(std::false_type /*allow_recovery*/) noexcept {}
+
+    /// @copydoc skip_to_value
+    JSON_INTERNAL_ALWAYS_INLINE
+    static std::false_type skip_to_value(std::false_type /*allow_recovery*/) noexcept
+    {
+        return {};
+    }
+
+    /// @copydoc recover_missing_value
+    template<typename SAX>
+    JSON_INTERNAL_ALWAYS_INLINE
+    static std::false_type recover_missing_value(SAX* /*sax*/, const std::vector<std::uint8_t>& /*states*/, std::false_type /*allow_recovery*/) noexcept
+    {
+        return {};
+    }
+
+    /// pass the end events of all open containers
+    template<typename SAX>
+    bool close_containers(SAX* sax, std::vector<std::uint8_t>& states, std::true_type /*allow_recovery*/)
+    {
+        while (!states.empty())
+        {
+            const bool is_array = states.back() != 0;
+            states.pop_back();
+            if (JSON_HEDLEY_UNLIKELY(is_array ? !sax->end_array() : !sax->end_object()))
+            {
+                return false;
+            }
+        }
+        return true;
+    }
+
+    /// @copydoc close_containers
+    template<typename SAX>
+    JSON_INTERNAL_ALWAYS_INLINE
+    static std::false_type close_containers(SAX* /*sax*/, std::vector<std::uint8_t>& /*states*/, std::false_type /*allow_recovery*/) noexcept
+    {
+        return {};
+    }
+
+    /*!
+    @brief read tokens until one begins a value, skipping everything before
+           the top-level value
+    @return whether a value begins with last_token
+    */
+    bool skip_to_value(std::true_type /*allow_recovery*/)
+    {
+        while (true)
+        {
+            switch (get_token())
+            {
+                case token_type::begin_array:
+                case token_type::begin_object:
+                case token_type::literal_false:
+                case token_type::literal_null:
+                case token_type::literal_true:
+                case token_type::value_float:
+                case token_type::value_integer:
+                case token_type::value_string:
+                case token_type::value_unsigned:
+                    return true;
+
+                case token_type::end_of_input:
+                    return false;
+
+                case token_type::parse_error:
+                    recover_token();
+                    if (last_token != token_type::uninitialized)
+                    {
+                        return true;
+                    }
+                    break;
+
+                case token_type::uninitialized:
+                case token_type::end_array:
+                case token_type::end_object:
+                case token_type::name_separator:
+                case token_type::value_separator:
+                case token_type::literal_or_value:
+                default:
+                    break;
+            }
+        }
+    }
+
+    /*!
+    @brief skip the rest of an object member that cannot be read
+
+    Reads tokens, beginning with last_token, until a ',', '}', or ']' that is
+    not inside a container that begins in the skipped tokens, or the end of
+    the input.
+    */
+    void skip_member()
+    {
+        std::size_t depth = 0;
+        while (true)
+        {
+            switch (last_token)
+            {
+                case token_type::begin_array:
+                case token_type::begin_object:
+                    ++depth;
+                    break;
+
+                case token_type::end_array:
+                case token_type::end_object:
+                    if (depth == 0)
+                    {
+                        return;
+                    }
+                    --depth;
+                    break;
+
+                case token_type::value_separator:
+                    if (depth == 0)
+                    {
+                        return;
+                    }
+                    break;
+
+                case token_type::end_of_input:
+                    return;
+
+                case token_type::parse_error:
+                    recover_token();
+                    break;
+
+                case token_type::uninitialized:
+                case token_type::literal_true:
+                case token_type::literal_false:
+                case token_type::literal_null:
+                case token_type::value_string:
+                case token_type::value_unsigned:
+                case token_type::value_integer:
+                case token_type::value_float:
+                case token_type::name_separator:
+                case token_type::literal_or_value:
+                default:
+                    break;
+            }
+            get_token();
+        }
+    }
+
+    /*!
+    @brief pass a value where it is missing
+
+    last_token is ',', ']', '}', or the end of the input, where a value was
+    expected. In an object, the key gets null; in an array, a ',' where a
+    value is missing stands for null (as in JavaScript), while an array that
+    ends there just ends.
+    */
+    template<typename SAX>
+    bool recover_missing_value(SAX* sax, const std::vector<std::uint8_t>& states, std::true_type /*allow_recovery*/)
+    {
+        JSON_ASSERT(!states.empty());
+        if (states.back() == 0 || last_token == token_type::value_separator)
+        {
+            return sax->null();
+        }
+        return true;
+    }
+
+    /// recover from a missing key; last_token is where it was expected
+    template<typename SAX>
+    next_step recover_key(SAX* sax)
+    {
+        switch (last_token)
+        {
+            case token_type::value_separator:
+            case token_type::end_object:
+            case token_type::end_array:
+            case token_type::end_of_input:
+                // no member: the object's state handles the token
+                return next_step::evaluate_state;
+
+            case token_type::parse_error:
+                recover_token();
+                if (last_token == token_type::value_string)
+                {
+                    // a key that could be repaired
+                    return parse_key(sax);
+                }
+                skip_member();
+                return next_step::evaluate_state;
+
+            case token_type::uninitialized:
+            case token_type::literal_true:
+            case token_type::literal_false:
+            case token_type::literal_null:
+            case token_type::value_string:
+            case token_type::value_unsigned:
+            case token_type::value_integer:
+            case token_type::value_float:
+            case token_type::begin_array:
+            case token_type::begin_object:
+            case token_type::name_separator:
+            case token_type::literal_or_value:
+            default:
+                // a member without a key
+                skip_member();
+                return next_step::evaluate_state;
+        }
+    }
+
+    /// recover from a missing name separator (:) after the key; last_token
+    /// is where it was expected
+    template<typename SAX>
+    next_step recover_name_separator(SAX* sax)
+    {
+        switch (last_token)
+        {
+            case token_type::value_separator:
+            case token_type::end_object:
+            case token_type::end_array:
+            case token_type::end_of_input:
+                // the value is missing as well
+                return sax->null() ? next_step::evaluate_state : next_step::stop;
+
+            case token_type::uninitialized:
+            case token_type::literal_true:
+            case token_type::literal_false:
+            case token_type::literal_null:
+            case token_type::value_string:
+            case token_type::value_unsigned:
+            case token_type::value_integer:
+            case token_type::value_float:
+            case token_type::begin_array:
+            case token_type::begin_object:
+            case token_type::name_separator:
+            case token_type::parse_error:
+            case token_type::literal_or_value:
+            default:
+                // a missing ':'; the value begins here
+                return next_step::parse_value;
+        }
+    }
+
+    /// recover from a token after an object member that is neither ',' nor
+    /// '}' (nor ']' or the end of the input, which the caller handles)
+    template<typename SAX>
+    next_step recover_member(SAX* sax, std::true_type /*allow_recovery*/)
+    {
+        if (last_token == token_type::parse_error)
+        {
+            recover_token();
+        }
+        if (last_token == token_type::value_string)
+        {
+            // a missing ','; the next key begins here
+            return parse_key(sax);
+        }
+        skip_member();
+        return next_step::evaluate_state;
+    }
+
+    /// the parser for parse() and accept() never recovers (and does not come
+    /// here, as report_error() returned false)
+    template<typename SAX>
+    JSON_INTERNAL_ALWAYS_INLINE
+    std::false_type recover_member(SAX* /*sax*/, std::false_type /*allow_recovery*/) const noexcept
+    {
+        return {};
+    }
+
     /// get next token from lexer
     token_type get_token()
     {
@@ -612,6 +1319,12 @@ class parser
     const bool allow_exceptions = true;
     /// whether trailing commas in objects and arrays should be ignored (true) or signaled as errors (false)
     const bool ignore_trailing_commas = false;
+    /// whether an error was reported to the SAX parser
+    bool error_reported = false;
+    /// the position of the last reported error
+    std::size_t last_error_position = 0;
+    /// the token of the last reported error
+    token_type last_error_token = token_type::uninitialized;
 };
 
 }  // namespace detail

include/nlohmann/detail/input/simd_scan.hpp

@@ -322,6 +322,78 @@ JSON_SCAN_SSSE3_TARGET JSON_HEDLEY_NEVER_INLINE inline bool vector_utf8_run(cons
     block_out = block;
     return false;
 }
+
+/*!
+@brief skip 16-byte blocks of well-formed UTF-8 at p (a character boundary)
+
+Like vector_utf8_run, but quotes, backslashes, and control characters are
+ordinary characters. Returns the start of the first block in which the check
+found an error (which may belong to a sequence begun in the block before) or
+where fewer than 16 bytes are left; a sequence may be cut at that point, so the
+caller resumes at its start, one sequence at a time.
+*/
+JSON_SCAN_SSSE3_TARGET JSON_HEDLEY_NEVER_INLINE inline const unsigned char* vector_utf8_valid_blocks(const unsigned char* block, const unsigned char* e) noexcept
+{
+    using lookup = utf8_lookup4<>;
+#if JSON_SCAN_NEON
+    const uint8x16_t t1h = vld1q_u8(lookup::byte_1_high.data());
+    const uint8x16_t t1l = vld1q_u8(lookup::byte_1_low.data());
+    const uint8x16_t t2h = vld1q_u8(lookup::byte_2_high.data());
+    uint8x16_t prev = vdupq_n_u8(0);
+    while (e - block >= 16)
+    {
+        const uint8x16_t in = vld1q_u8(block);
+        if (vmaxvq_u8(vorrq_u8(in, prev)) < 0x80)
+        {
+            // ASCII here, and no sequence from the block before to finish
+            prev = in;
+            block += 16;
+            continue;
+        }
+        const uint8x16_t prev1 = vextq_u8(prev, in, 15);
+        const uint8x16_t sc = vandq_u8(vandq_u8(vqtbl1q_u8(t1h, vshrq_n_u8(prev1, 4)), vqtbl1q_u8(t1l, vandq_u8(prev1, vdupq_n_u8(0x0F)))), vqtbl1q_u8(t2h, vshrq_n_u8(in, 4)));
+        const uint8x16_t must23 = vorrq_u8(vqsubq_u8(vextq_u8(prev, in, 14), vdupq_n_u8(0xE0 - 0x80)), vqsubq_u8(vextq_u8(prev, in, 13), vdupq_n_u8(0xF0 - 0x80)));
+        const uint8x16_t err = veorq_u8(vandq_u8(must23, vdupq_n_u8(0x80)), sc);
+        if (vmaxvq_u8(err) != 0)
+        {
+            break;
+        }
+        prev = in;
+        block += 16;
+    }
+#else
+    const __m128i t1h = _mm_loadu_si128(static_cast<const __m128i*>(static_cast<const void*>(lookup::byte_1_high.data()))); // NOLINT(bugprone-casting-through-void)
+    const __m128i t1l = _mm_loadu_si128(static_cast<const __m128i*>(static_cast<const void*>(lookup::byte_1_low.data()))); // NOLINT(bugprone-casting-through-void)
+    const __m128i t2h = _mm_loadu_si128(static_cast<const __m128i*>(static_cast<const void*>(lookup::byte_2_high.data()))); // NOLINT(bugprone-casting-through-void)
+    const __m128i nibble = _mm_set1_epi8(0x0F);
+    const __m128i zero = _mm_setzero_si128();
+    __m128i prev = zero;
+    while (e - block >= 16)
+    {
+        const __m128i in = _mm_loadu_si128(static_cast<const __m128i*>(static_cast<const void*>(block))); // NOLINT(bugprone-casting-through-void)
+        if (_mm_movemask_epi8(_mm_or_si128(in, prev)) == 0)
+        {
+            prev = in;
+            block += 16;
+            continue;
+        }
+        const __m128i prev1 = _mm_alignr_epi8(in, prev, 15);
+        const __m128i sc = _mm_and_si128(_mm_and_si128(_mm_shuffle_epi8(t1h, _mm_and_si128(_mm_srli_epi16(prev1, 4), nibble)),
+                                         _mm_shuffle_epi8(t1l, _mm_and_si128(prev1, nibble))),
+                                         _mm_shuffle_epi8(t2h, _mm_and_si128(_mm_srli_epi16(in, 4), nibble)));
+        const __m128i must23 = _mm_or_si128(_mm_subs_epu8(_mm_alignr_epi8(in, prev, 14), _mm_set1_epi8(0xE0 - 0x80)),
+                                            _mm_subs_epu8(_mm_alignr_epi8(in, prev, 13), _mm_set1_epi8(0xF0 - 0x80)));
+        const __m128i err = _mm_xor_si128(_mm_and_si128(must23, _mm_set1_epi8(static_cast<char>(-128))), sc);
+        if (_mm_movemask_epi8(_mm_cmpeq_epi8(err, zero)) != 0xFFFF)
+        {
+            break;
+        }
+        prev = in;
+        block += 16;
+    }
+#endif
+    return block;
+}
 #endif
 
 }  // namespace detail

include/nlohmann/detail/input/string_scan.hpp

@@ -90,6 +90,34 @@ JSON_HEDLEY_ALWAYS_INLINE std::size_t find_string_special(const unsigned char* d
     return n;
 }
 
+// return the index of the first 0x00 byte in [data, data+n), or n if there is
+// none; scans 8 bytes at a time (BSON's C-style strings)
+JSON_HEDLEY_ALWAYS_INLINE std::size_t find_zero_byte(const unsigned char* data, std::size_t n) noexcept
+{
+    constexpr std::uint64_t ones = 0x0101010101010101ull;
+    constexpr std::uint64_t high = 0x8080808080808080ull;
+    std::size_t i = 0;
+    for (; i + 8 <= n; i += 8)
+    {
+        const std::uint64_t v = read_eight_bytes(data + i);
+        const std::uint64_t zero = (v - ones) & ~v & high;
+        if (zero != 0)
+        {
+            // the lowest flagged byte is the first zero byte (see
+            // find_string_special())
+            return i + (static_cast<std::size_t>(count_trailing_zeros(zero)) / 8);
+        }
+    }
+    for (; i < n; ++i)
+    {
+        if (data[i] == 0x00)
+        {
+            return i;
+        }
+    }
+    return n;
+}
+
 // classify a byte as one the serializer must NOT copy verbatim when
 // ensure_ascii is requested: the closing quote, an escape, a control character
 // (< 0x20), DEL (0x7F), or any non-ASCII byte (>= 0x80). Everything else -
@@ -236,6 +264,64 @@ inline std::size_t valid_utf8_prefix(const unsigned char* data, std::size_t n) n
     return pos;
 }
 
+/*!
+@brief whether [data, data+n) is valid UTF-8, for checking whole strings
+
+The same answer as valid_utf8_prefix(data, n) == n. Short ASCII strings, the
+most common keys and values, are decided in one test, and long strings are
+checked 16 bytes at a time where the vector check is available.
+*/
+inline bool valid_utf8_string(const unsigned char* data, std::size_t n) noexcept
+{
+    constexpr std::uint64_t high = 0x8080808080808080ull;
+    if (n <= 16)
+    {
+        // two overlapping words cover 8 to 16 bytes, one partial word fewer
+        std::uint64_t first = 0;
+        std::uint64_t last = 0;
+        if (n >= 8)
+        {
+            std::memcpy(&first, data, sizeof(first));
+            std::memcpy(&last, data + n - 8, sizeof(last));
+        }
+        else if (n != 0)
+        {
+            std::memcpy(&first, data, n);
+        }
+        if (((first | last) & high) == 0)
+        {
+            return true;
+        }
+    }
+#if JSON_SCAN_VECTOR_UTF8
+    else if (n >= 32
+#if JSON_SCAN_SSSE3_DISPATCH
+             && cpu_has_ssse3()
+#endif
+            )
+    {
+        const unsigned char* block = vector_utf8_valid_blocks(data, data + n);
+        // the rest from the start of a sequence that may be cut at block
+        for (int i = 1; i <= 3 && block - i >= data; ++i)
+        {
+            const unsigned char c = block[-i];
+            if (c < 0x80)
+            {
+                break;
+            }
+            if (c >= 0xC0)
+            {
+                block -= i;
+                break;
+            }
+        }
+        const auto checked = static_cast<std::size_t>(block - data);
+        return checked + valid_utf8_prefix(block, n - checked) == n;
+    }
+#endif
+    return valid_utf8_prefix(data, n) == n;
+}
+
 #if JSON_SCAN_VECTOR_UTF8
 /// the end of scan_string_vector from block, where the vector loop stopped
 /// (ill-formed UTF-8, or fewer than 16 bytes left): one byte or sequence at a

include/nlohmann/detail/macro_scope.hpp

@@ -215,6 +215,17 @@
     #define JSON_NO_UNIQUE_ADDRESS
 #endif
 
+// Inlines small functions even in unoptimized builds, so that they are not
+// emitted. The parsers for parse(), accept(), and from_*() use it for the
+// functions that stand in for the code recovering from errors (see #3989).
+// MSVC is left to decide, as it warns (C4714) where it does not inline a
+// __forceinline function.
+#if defined(_MSC_VER) && !defined(__clang__)
+    #define JSON_INTERNAL_ALWAYS_INLINE
+#else
+    #define JSON_INTERNAL_ALWAYS_INLINE JSON_HEDLEY_ALWAYS_INLINE
+#endif
+
 // Clang targeting MinGW does not survive the thread_local storage the copy
 // constructor uses to bound its descent: every test that copies a value
 // segfaults with clang 11.0.1 and clang 18.1.8, while the same tests pass with
@@ -347,7 +358,7 @@
             return ej_pair.second == j;                                                         \
         });                                                                                     \
         if (it != std::end(m)) e = it->first;                                                   \
-        else ::nlohmann::detail::templated_json_throw<nlohmann::detail::out_of_range>(nlohmann::detail::out_of_range::create(nlohmann::detail::exception_id::value_out_of_range, nlohmann::detail::concat("enum value out of range for " #ENUM_TYPE ": ", nlohmann::detail::from_json_dump(j)), nlohmann::detail::exception_context(j))); \
+        else ::nlohmann::detail::templated_json_throw<nlohmann::detail::out_of_range>(nlohmann::detail::out_of_range::create(nlohmann::detail::exception_id::value_out_of_range, nlohmann::detail::concat(nlohmann::detail::exception_prefix(j), "enum value out of range for " #ENUM_TYPE ": ", nlohmann::detail::from_json_dump(j)), nlohmann::detail::exception_context(j))); \
     }
 
 // Ugly macros to avoid uglier copy-paste when specializing basic_json. They

include/nlohmann/detail/macro_unscope.hpp

@@ -19,6 +19,7 @@
 #undef NLOHMANN_CAN_CALL_STD_FUNC_IMPL
 #undef JSON_INLINE_VARIABLE
 #undef JSON_NO_UNIQUE_ADDRESS
+#undef JSON_INTERNAL_ALWAYS_INLINE
 #undef JSON_DISABLE_ENUM_SERIALIZATION
 #undef JSON_DISABLE_TUPLE_REFERENCE_CONVERSION
 #undef JSON_UNLOCKED_FILE_IO

include/nlohmann/detail/output/binary_writer.hpp

@@ -2088,7 +2088,8 @@ class binary_writer
         /// arrays: the first index with one more digit; only maintained by
         /// @ref calc_bson_sizes
         std::uint64_t index_boundary = 10;
-        /// @ref calc_bson_sizes only: where its size goes in the table
+        /// @ref calc_bson_sizes: where its size goes in the table; @ref
+        /// write_bson_document_single_pass: the position of its size field
         std::size_t size_slot;
         /// @ref calc_bson_sizes only: the size of its entries seen so far
         std::size_t entries_size = 0;
@@ -2339,6 +2340,199 @@ class binary_writer
         }
     }
 
+    template<typename SinkType>
+    using sink_written_t = decltype(std::declval<const SinkType&>().written());
+
+    /// whether every integer fits the 64 bits BSON has, so that writing an
+    /// array of numbers needs no range check
+    static constexpr bool bson_integers_fit_64_bits =
+        std::numeric_limits<typename BasicJsonType::number_integer_t>::is_specialized
+        && std::numeric_limits<typename BasicJsonType::number_integer_t>::digits <= std::numeric_limits<std::int64_t>::digits
+        && std::numeric_limits<typename BasicJsonType::number_unsigned_t>::is_specialized
+        && std::numeric_limits<typename BasicJsonType::number_unsigned_t>::digits <= std::numeric_limits<std::uint64_t>::digits;
+
+    /*!
+    @brief whether BSON documents can be written in a single pass: the sink
+           can overwrite what it has written (output_vector_sink), and
+           @ref bson_integers_fit_64_bits
+    */
+    using bson_single_pass_sink = conjunction<is_detected<sink_written_t, OutputSinkType>, std::integral_constant<bool, bson_integers_fit_64_bits>>;
+
+    /// write a placeholder for the size of an object or array whose entries
+    /// follow; @return its position, for @ref close_bson_container
+    std::size_t write_bson_size_placeholder()
+    {
+        const std::size_t pos = oa.written();
+        // zero in any byte order
+        const std::array<CharType, sizeof(std::int32_t)> zero{};
+        oa.write_characters(zero.data(), zero.size());
+        return pos;
+    }
+
+    /// end the object or array whose size is at @a pos and write its size
+    /// there: everything from that field to the terminating null byte
+    /// @return false (and the size is not written) if it does not fit
+    bool close_bson_container(const std::size_t pos)
+    {
+        oa.write_character(to_char_type(0x00));
+        const std::size_t size = oa.written() - pos;
+        if (JSON_HEDLEY_UNLIKELY(!value_in_range_of<std::int32_t>(size)))
+        {
+            return false;
+        }
+        const auto le_size = static_cast<std::uint32_t>(size);
+        oa.overwrite(pos, is_little_endian ? le_size : byteswap(le_size));
+        return true;
+    }
+
+    /*!
+    @brief check the value @a j of a BSON entry as @ref calc_bson_value_size
+           does, for @ref write_bson_document_single_pass
+    @return at least the size of the value
+    A number, boolean, or null needs no check there (its integers have at most
+    64 bits) and has at most 8 characters; a discarded value is rejected with
+    the same exception when it is written.
+    */
+    std::size_t check_bson_value(const BasicJsonType& j)
+    {
+        return (j.is_string() || j.is_binary()) ? calc_bson_value_size(j) : 8;
+    }
+
+    /// whether an entry of @a entry_size characters keeps the document that
+    /// begins at @a start, closed by @a open_containers null bytes, within
+    /// the 32-bit size of BSON
+    bool bson_entry_fits(const std::size_t start, const std::size_t entry_size, const std::size_t open_containers) const noexcept
+    {
+        const auto limit = static_cast<std::size_t>((std::numeric_limits<std::int32_t>::max)());
+        const std::size_t written = oa.written() - start;
+        return written <= limit && entry_size <= limit - written && open_containers <= limit - written - entry_size;
+    }
+
+    /*!
+    @brief Serializes the JSON object @a document as a BSON document in a
+           single pass, writing the size of each object and array once it is
+           complete
+
+    Checks every key and value as @ref calc_bson_sizes does, in the same
+    order, so the same exception is thrown; the output written up to then is
+    discarded with the vector it was written to.
+
+    @return false as soon as the document or a string, binary value, object,
+            or array in it would be larger than the 32-bit BSON sizes allow;
+            nothing large has been written then, and the caller writes the
+            document in two passes, which reports that error
+    */
+    template<typename SinkType = OutputSinkType, typename std::enable_if<is_detected<sink_written_t, SinkType>::value, int>::type = 0>
+    bool write_bson_document_single_pass(const BasicJsonType& document)
+    {
+        const std::size_t start = oa.written();
+        // the object or array whose entries are being written (with the
+        // position of its size), and the ones it is in
+        bson_frame current(&document, write_bson_size_placeholder());
+        std::vector<bson_frame> parents;
+
+        while (true)
+        {
+            const BasicJsonType* nested = nullptr;
+            std::size_t nested_size_pos = 0;
+            if (current.value->is_object())
+            {
+                const auto& object = *current.value->m_data.m_value.object;
+                while (nested == nullptr && current.member != object.cend())
+                {
+                    const auto& el = *current.member;
+                    ++current.member;
+                    const std::size_t header_size = calc_bson_entry_header_size(el.first, el.second);
+                    const std::size_t value_size = el.second.is_structured() ? 0 : check_bson_value(el.second);
+                    if (JSON_HEDLEY_UNLIKELY(!bson_entry_fits(start, header_size + value_size, parents.size() + 1)))
+                    {
+                        return false;
+                    }
+                    if (el.second.is_structured())
+                    {
+                        write_bson_entry_header(el.first, el.second.is_object() ? 0x03 : 0x04);
+                        nested_size_pos = write_bson_size_placeholder();
+                        nested = &el.second;
+                    }
+                    else
+                    {
+                        write_bson_value(el.first, el.second);
+                    }
+                }
+            }
+            else
+            {
+                const auto& array = *current.value->m_data.m_value.array;
+                while (nested == nullptr && current.index < array.size())
+                {
+                    if (is_bson_fixed_size(array[current.index]))
+                    {
+                        write_bson_array_fixed_entries(current);
+                        if (current.index == array.size())
+                        {
+                            break;
+                        }
+                    }
+
+                    // a string, binary value, object or array
+                    const BasicJsonType& el = array[current.index];
+                    const string_t index_name(current.index_name.data(), current.index_name_size);
+                    current.next_index();
+                    const std::size_t value_size = el.is_structured() ? 0 : check_bson_value(el);
+                    if (JSON_HEDLEY_UNLIKELY(!bson_entry_fits(start, 2 + index_name.size() + value_size, parents.size() + 1)))
+                    {
+                        return false;
+                    }
+                    if (el.is_structured())
+                    {
+                        write_bson_entry_header(index_name, el.is_object() ? 0x03 : 0x04);
+                        nested_size_pos = write_bson_size_placeholder();
+                        nested = &el;
+                    }
+                    else
+                    {
+                        write_bson_value(index_name, el);
+                    }
+                }
+            }
+
+            if (nested != nullptr)
+            {
+                if (nested->is_array())
+                {
+                    // an array of booleans, numbers and null is written without a frame on the stack
+                    bson_frame leaf(nested, nested_size_pos);
+                    write_bson_array_fixed_entries(leaf);
+                    if (leaf.index == nested->m_data.m_value.array->size())
+                    {
+                        if (JSON_HEDLEY_UNLIKELY(!close_bson_container(nested_size_pos)))
+                        {
+                            return false;
+                        }
+                        continue;
+                    }
+                    parents.push_back(std::move(current));
+                    current = std::move(leaf);
+                    continue;
+                }
+                parents.push_back(std::move(current));
+                current = bson_frame(nested, nested_size_pos);
+                continue;
+            }
+
+            if (JSON_HEDLEY_UNLIKELY(!close_bson_container(current.size_slot)))
+            {
+                return false;
+            }
+            if (parents.empty())
+            {
+                return true;
+            }
+            current = std::move(parents.back());
+            parents.pop_back();
+        }
+    }
+
     /*!
     @brief Serializes the JSON object @a document as a BSON document
 
@@ -2350,6 +2544,39 @@ class binary_writer
     @pre       document.type() == value_t::object
     */
     void write_bson_document(const BasicJsonType& document)
+    {
+        write_bson_document(document, bson_single_pass_sink());
+    }
+
+    /// the sink can overwrite: one pass, unless the error handler replaces or
+    /// drops ill-formed UTF-8, which changes the sizes the checks compute
+    void write_bson_document(const BasicJsonType& document, std::true_type /*single_pass*/)
+    {
+        const std::size_t start = oa.written();
+        if ((error_handler == error_handler_t::keep || error_handler == error_handler_t::strict)
+                && write_bson_document_single_pass(document))
+        {
+            return;
+        }
+        // a length does not fit: start over, so that it is reported as the
+        // two passes report it, before anything is written
+        oa.rewind(start);
+        write_bson_document_two_pass(document);
+    }
+
+    void write_bson_document(const BasicJsonType& document, std::false_type /*single_pass*/)
+    {
+        write_bson_document_two_pass(document);
+    }
+
+    /*!
+    @brief Serializes the JSON object @a document as a BSON document: the
+           sizes of all objects and arrays first, then the entries
+
+    @param[in] document  the JSON object to serialize
+    @pre       document.type() == value_t::object
+    */
+    void write_bson_document_two_pass(const BasicJsonType& document)
     {
         std::vector<std::size_t> nested_sizes;
         const std::size_t document_size = calc_bson_sizes(document, nested_sizes);
@@ -3080,6 +3307,25 @@ class binary_writer
         }
     }
 
+    /// write the head and the characters of @a value, at most
+    /// max_text_in_reservation of them, in one reservation
+    JSON_HEDLEY_ALWAYS_INLINE void write_cbor_string_reserved(const string_t& value)
+    {
+        CharType* const out = oa.reserve(value.size() + 2);
+        CharType* p = out;
+        if (value.size() <= 0x17)
+        {
+            *p++ = to_char_type(static_cast<std::uint8_t>(0x60 + value.size()));
+        }
+        else
+        {
+            *p++ = to_char_type(0x78);
+            *p++ = to_char_type(static_cast<std::uint8_t>(value.size()));
+        }
+        copy_characters(p, reinterpret_cast<const CharType*>(value.data()), value.size());
+        oa.commit(static_cast<std::size_t>(p - out) + value.size());
+    }
+
     /*!
     @brief write a CBOR text string
 
@@ -3101,19 +3347,7 @@ class binary_writer
             // the string is written as it is: head and characters in one reservation
             if (JSON_HEDLEY_LIKELY(value.size() <= max_text_in_reservation))
             {
-                CharType* const out = oa.reserve(value.size() + 2);
+                write_cbor_string_reserved(value);
-                CharType* p = out;
-                if (value.size() <= 0x17)
-                {
-                    *p++ = to_char_type(static_cast<std::uint8_t>(0x60 + value.size()));
-                }
-                else
-                {
-                    *p++ = to_char_type(0x78);
-                    *p++ = to_char_type(static_cast<std::uint8_t>(value.size()));
-                }
-                copy_characters(p, reinterpret_cast<const CharType*>(value.data()), value.size());
-                oa.commit(static_cast<std::size_t>(p - out) + value.size());
                 return;
             }
         }
@@ -3127,6 +3361,13 @@ class binary_writer
     JSON_HEDLEY_NEVER_INLINE
     void write_cbor_checked_string(const string_t& value, const BasicJsonType& context)
     {
+        if (error_handler == error_handler_t::strict && value.size() <= max_text_in_reservation)
+        {
+            check_utf8(value, context);
+            write_cbor_string_reserved(value);
+            return;
+        }
+
         string_t storage;
         const string_t& sanitized = sanitize_utf8_for_write(value, context, storage);
 
@@ -3141,6 +3382,27 @@ class binary_writer
     // MsgPack //
     /////////////
 
+    /// write the head and the characters of @a value, at most
+    /// max_text_in_reservation of them, in one reservation
+    JSON_HEDLEY_ALWAYS_INLINE void write_msgpack_string_reserved(const string_t& value)
+    {
+        CharType* const out = oa.reserve(value.size() + 2);
+        CharType* p = out;
+        if (value.size() <= 31)
+        {
+            // fixstr
+            *p++ = to_char_type(static_cast<std::uint8_t>(0xA0 | value.size()));
+        }
+        else
+        {
+            // str 8
+            *p++ = to_char_type(0xD9);
+            *p++ = to_char_type(static_cast<std::uint8_t>(value.size()));
+        }
+        copy_characters(p, reinterpret_cast<const CharType*>(value.data()), value.size());
+        oa.commit(static_cast<std::size_t>(p - out) + value.size());
+    }
+
     /*!
     @brief write a MessagePack str
 
@@ -3158,21 +3420,7 @@ class binary_writer
             // the string is written as it is: head and characters in one reservation
             if (JSON_HEDLEY_LIKELY(value.size() <= max_text_in_reservation))
             {
-                CharType* const out = oa.reserve(value.size() + 2);
+                write_msgpack_string_reserved(value);
-                CharType* p = out;
-                if (value.size() <= 31)
-                {
-                    // fixstr
-                    *p++ = to_char_type(static_cast<std::uint8_t>(0xA0 | value.size()));
-                }
-                else
-                {
-                    // str 8
-                    *p++ = to_char_type(0xD9);
-                    *p++ = to_char_type(static_cast<std::uint8_t>(value.size()));
-                }
-                copy_characters(p, reinterpret_cast<const CharType*>(value.data()), value.size());
-                oa.commit(static_cast<std::size_t>(p - out) + value.size());
                 return;
             }
         }
@@ -3185,6 +3433,13 @@ class binary_writer
     JSON_HEDLEY_NEVER_INLINE
     void write_msgpack_checked_string(const string_t& value, const BasicJsonType& context)
     {
+        if (error_handler == error_handler_t::strict && value.size() <= max_text_in_reservation)
+        {
+            check_utf8(value, context);
+            write_msgpack_string_reserved(value);
+            return;
+        }
+
         string_t storage;
         const string_t& sanitized = sanitize_utf8_for_write(value, context, storage);
 
@@ -4794,10 +5049,9 @@ class binary_writer
     static void check_utf8(const string_t& s, const BasicJsonType& context)
     {
         const auto* data = reinterpret_cast<const unsigned char*>(s.data());
-        const std::size_t valid = valid_utf8_prefix(data, s.size());
+        if (JSON_HEDLEY_UNLIKELY(!valid_utf8_string(data, s.size())))
-        if (JSON_HEDLEY_UNLIKELY(valid != s.size()))
         {
-            throw_invalid_utf8(data, valid, context);
+            throw_invalid_utf8(data, valid_utf8_prefix(data, s.size()), context);
         }
     }
 

include/nlohmann/detail/output/output_adapters.hpp

@@ -11,6 +11,7 @@
 #include <algorithm> // max
 #include <array> // array
 #include <cstddef> // size_t
+#include <cstdint> // uint32_t
 #include <cstring> // memcpy
 #include <memory> // shared_ptr, make_shared
 #include <string> // basic_string
@@ -227,6 +228,31 @@ class output_vector_sink
     void flush() noexcept
     {}
 
+    /// the number of characters written so far, the position of the next one
+    std::size_t written() const noexcept
+    {
+        return position();
+    }
+
+    /// @brief replace the 4 bytes written before at position @a pos with
+    ///        those of @a n, in the byte order of the machine
+    ///
+    /// Lets a writer put a length in front of data it has written since, such
+    /// as the size of a BSON document, without computing it in advance.
+    void overwrite(const std::size_t pos, const std::uint32_t n) noexcept
+    {
+        static_assert(sizeof(CharType) == 1, "overwrite() writes bytes");
+        JSON_ASSERT(pos + sizeof(n) <= position());
+        std::memcpy(v->data() + pos, &n, sizeof(n));
+    }
+
+    /// drop the characters written from position @a pos on
+    void rewind(const std::size_t pos) noexcept
+    {
+        JSON_ASSERT(pos <= position());
+        cursor = v->data() + pos;
+    }
+
   private:
     /// number of characters of *v written so far
     std::size_t position() const noexcept

include/nlohmann/detail/output/serializer.hpp

@@ -1977,6 +1977,23 @@ class serializer
             return;
         }
 
+        // 17 to 20 digits: the first one to four, then twice eight
+        if (magnitude_fits_64_bits(abs_value, std::integral_constant<bool, magnitude_has_at_most_64_bits>()))
+        {
+            const auto value = static_cast<std::uint64_t>(abs_value);
+            const std::uint64_t top = value / 10000000000000000u; // 1 to 1844
+            const std::uint64_t low16 = value - (top * 10000000000000000u);
+            const std::uint64_t middle = low16 / 100000000u;
+            const std::uint64_t first = dtoa_impl::eight_digit_bytes(top);
+            const auto leading = static_cast<unsigned>(count_leading_zeros(first) / 8); // (first is not 0)
+            char* const p = buffer_ptr + n_chars;
+            dtoa_impl::store_msb_first(p, (first << (8 * leading)) + 0x3030303030303030u);
+            dtoa_impl::store_msb_first(p + 8 - leading, dtoa_impl::eight_digit_bytes(middle) + 0x3030303030303030u);
+            dtoa_impl::store_msb_first(p + 16 - leading, dtoa_impl::eight_digit_bytes(low16 - (middle * 100000000u)) + 0x3030303030303030u);
+            write_buffer_pos += n_chars + 24 - leading;
+            return;
+        }
+
         n_chars += count_digits(abs_value);
 
         // spare 1 byte for '\0'
@@ -2122,6 +2139,23 @@ class serializer
         }
     }
 
+    /// whether every number_magnitude_t fits 64 bits
+    static constexpr bool magnitude_has_at_most_64_bits = std::numeric_limits<number_magnitude_t>::is_specialized
+        && std::numeric_limits<number_magnitude_t>::digits <= 64;
+
+    /// whether @a x fits 64 bits: always if number_magnitude_t has at most
+    /// 64 bits (selected by type, as the comparison would be reported as
+    /// always true), else compared
+    static bool magnitude_fits_64_bits(const number_magnitude_t& /*unused*/, std::true_type /*unused*/) noexcept
+    {
+        return true;
+    }
+
+    static bool magnitude_fits_64_bits(const number_magnitude_t& x, std::false_type /*unused*/) noexcept
+    {
+        return x <= 18446744073709551615u;
+    }
+
     /*
      * Helper function for dump_integer
      *

include/nlohmann/detail/string_utils.hpp

@@ -13,6 +13,7 @@
 #include <cstddef> // size_t
 #include <cstdint> // uint8_t, uint32_t
 #include <string> // string, to_string
+#include <utility> // move
 
 #include <nlohmann/detail/abi_macros.hpp>
 #include <nlohmann/detail/macro_scope.hpp>

include/nlohmann/detail/view/builder.hpp

@@ -69,7 +69,8 @@ struct parse_failure
 };
 
 /// FloatType: the number_float_t of the document, whose overflow parse() rejects
-template<typename FloatType, bool Comments, bool TrailingCommas, bool NulIsEnd, bool Sentinel>
+/// Store: false for accept(): the grammar, number, string and UTF-8 checks run as for parse(), but no node is written
+template<typename FloatType, bool Comments, bool TrailingCommas, bool NulIsEnd, bool Sentinel, bool Store>
 class builder
 {
   public:
@@ -245,11 +246,87 @@ class builder
         return p;
     }
 
+    /// the string from p, the first backslash, control character, ill-formed UTF-8 byte, or the end, checked and not
+    /// decoded (accept() stores nothing, so the arena is not written): the checks, errors and error positions of
+    /// slow_string(), for everything up to the closing quote
+    NLOHMANN_VIEW_NOINLINE decoded check_string(const unsigned char* p)
+    {
+        const unsigned char* const end = m_end;
+        for (;;)
+        {
+            if (p == end)
+            {
+                return failed(error_code::string_missing_quote, p);
+            }
+            const unsigned char c = *p;
+            if (c == '"')
+            {
+                ++p;
+                return decoded{p, 0, 0};
+            }
+            if (c != '\\')
+            {
+                return failed(c < 0x20 ? error_code::string_control_character : error_code::string_utf8, p);
+            }
+            ++p;
+            if (p == end)
+            {
+                return failed(error_code::string_missing_quote, p);
+            }
+            const unsigned char d = *p++;
+            if (d == 'u')
+            {
+                const int cp = hex4(p); // (moves p past the four digits)
+                if (NLOHMANN_VIEW_UNLIKELY(cp < 0))
+                {
+                    return failed(error_code::string_unicode_hex, p);
+                }
+                if (NLOHMANN_VIEW_UNLIKELY((cp & 0xF800) == 0xD800)) // a surrogate
+                {
+                    if (cp >= 0xDC00)
+                    {
+                        return failed(error_code::string_surrogate_low, p);
+                    }
+                    if (end - p < 2 || p[0] != '\\' || p[1] != 'u')
+                    {
+                        return failed(error_code::string_surrogate_high, p);
+                    }
+                    p += 2;
+                    const int lo = hex4(p); // (moves p past the four digits)
+                    if (lo < 0)
+                    {
+                        return failed(error_code::string_unicode_hex, p);
+                    }
+                    if (lo < 0xDC00 || lo > 0xDFFF)
+                    {
+                        return failed(error_code::string_surrogate_high, p);
+                    }
+                }
+            }
+            else if (!(d == '"' || d == '\\' || d == '/' || d == 'b' || d == 'f' || d == 'n' || d == 'r' || d == 't'))
+            {
+                --p;
+                return failed(error_code::string_escape, p);
+            }
+            if (p != end && *p == '\\')
+            {
+                continue; // (escapes in a row: no run of text between them)
+            }
+            p = scan_string_run(p, end); // the text up to the next backslash, quote, control character, or ill-formed UTF-8
+        }
+    }
+
     /// s points to the start of the string, p to a backslash, a control
     /// character, an ill-formed UTF-8 byte, or the end: decode the string into
     /// the arena
     NLOHMANN_VIEW_NOINLINE decoded slow_string(const unsigned char* s, const unsigned char* p)
     {
+        if (!enabled(Store))
+        {
+            // (accept(): the string is checked, not decoded)
+            static_cast<void>(s);
+            return check_string(p);
+        }
         // single-character escapes; 0: invalid (and 'u', handled separately)
         static const std::array<char, 128> simple_escape =
         {
@@ -343,6 +420,21 @@ class builder
                 }
                 arena_ensure(w, w_end, 4);
                 w = put_utf8(w, cp);
+                // A chain of unicode escapes ("\u3042\u3044\u3046", as Python's json.dumps writes non-ASCII text) is
+                // decoded here, without going round the loop above for each of them. Anything else (a surrogate, a
+                // bad digit, fewer than six bytes) leaves p at the backslash: the loop above handles it, and reports
+                // the same error at the same position as it would without this loop.
+                while (end - p >= 6 && p[0] == '\\' && p[1] == 'u')
+                {
+                    const int next = hex_codepoint(p + 2);
+                    if (next < 0 || (next & 0xF800) == 0xD800)
+                    {
+                        break;
+                    }
+                    p += 6;
+                    arena_ensure(w, w_end, 4);
+                    w = put_utf8(w, next);
+                }
             }
             else if (NLOHMANN_VIEW_LIKELY(d < 128 && simple_escape[d] != 0))
             {
@@ -610,10 +702,13 @@ run_found:
 
         NLOHMANN_VIEW_ALWAYS_INLINE bool run()
         {
-            cold.m_doc.reserve(estimate_nodes(reinterpret_cast<const char*>(b), static_cast<std::size_t>(e - b)));
+            if (enabled(Store))
-            base = cold.m_doc.nodes;
+            {
-            out = base;
+                cold.m_doc.reserve(estimate_nodes(reinterpret_cast<const char*>(b), static_cast<std::size_t>(e - b)));
-            cap = base + cold.m_doc.node_capacity;
+                base = cold.m_doc.nodes;
+                out = base;
+                cap = base + cold.m_doc.node_capacity;
+            }
 
             if (e - p >= 3 && p[0] == 0xEF && p[1] == 0xBB && p[2] == 0xBF)
             {
@@ -797,7 +892,10 @@ close_object:
             // an array pays nothing for this)
             if (NLOHMANN_VIEW_UNLIKELY(cur_count >= document_data::index_min_members))
             {
-                cold.note_large_object(cur_idx);
+                if (enabled(Store))
+                {
+                    cold.note_large_object(cur_idx);
+                }
             }
 
 close_container:
@@ -957,6 +1055,12 @@ indent_done:
         /// an integer's value); two stores on little-endian targets
         NLOHMANN_VIEW_ALWAYS_INLINE node* push_node(node_kind k, std::uint8_t flags, std::uint16_t extra, std::size_t off, std::uint64_t second)
         {
+            if (!enabled(Store))
+            {
+                // accept(): nothing is stored, so what computes the arguments is not run either. base and out stay
+                // null: their difference, which open() takes, is 0.
+                return out;
+            }
             if (NLOHMANN_VIEW_UNLIKELY(out == cap))
             {
                 const auto n = static_cast<std::size_t>(out - base);
@@ -1015,9 +1119,12 @@ indent_done:
 
         NLOHMANN_VIEW_ALWAYS_INLINE void close()
         {
-            node& n = base[cur_idx];
+            if (enabled(Store))
-            n.len = cur_count;
+            {
-            n.next = static_cast<std::uint32_t>(out - base) - cur_idx;
+                node& n = base[cur_idx];
+                n.len = cur_count;
+                n.next = static_cast<std::uint32_t>(out - base) - cur_idx;
+            }
             if (--depth != 0)
             {
                 if (NLOHMANN_VIEW_LIKELY(depth <= 64))
@@ -1082,7 +1189,7 @@ indent_done:
             {
                 ++p;
                 const unsigned char* const f0 = p;
-                p = skip_digits(p, e);
+                p = skip_fraction_digits(p, e);
                 if (NLOHMANN_VIEW_UNLIKELY(p == f0))
                 {
                     return fail(error_code::number_after_dot);
@@ -1192,34 +1299,44 @@ indent_done:
 };
 
 /// run the builder with compile-time options
-template<typename FloatType, bool Comments, bool TrailingCommas, bool NulIsEnd>
+template<typename FloatType, bool Comments, bool TrailingCommas, bool NulIsEnd, bool Store>
 inline bool build_with(document_data& d, const char* src, std::size_t size, bool sentinel, parse_failure& failure)
 {
     if (sentinel)
     {
-        builder<FloatType, Comments, TrailingCommas, NulIsEnd, true> bld(d, src, size);
+        builder<FloatType, Comments, TrailingCommas, NulIsEnd, true, Store> bld(d, src, size);
         const bool ok = bld.run();
         failure = bld.failure();
         return ok;
     }
-    builder<FloatType, Comments, TrailingCommas, NulIsEnd, false> bld(d, src, size);
+    builder<FloatType, Comments, TrailingCommas, NulIsEnd, false, Store> bld(d, src, size);
     const bool ok = bld.run();
     failure = bld.failure();
     return ok;
 }
 
 /// sentinel: src[size] is readable and 0 (e.g. std::string); FloatType: the
-/// number_float_t of the document
+/// number_float_t of the document; validate_only: accept(), only the result is
+/// used
 template<typename FloatType, bool NulIsEnd>
-inline bool build(document_data& d, const char* src, std::size_t size, bool comments, bool trailing_commas, bool sentinel, parse_failure& failure)
+inline bool build(document_data& d, const char* src, std::size_t size, bool comments, bool trailing_commas, bool sentinel, bool validate_only,
+                  parse_failure& failure)
 {
+    if (validate_only && !comments && !trailing_commas)
+    {
+        // accept() of plain JSON: nothing is stored. The other option sets
+        // could skip the stores just as well; they keep the builder that
+        // stores only so that there are two more instantiations of the parse
+        // loop, not eight.
+        return build_with<FloatType, false, false, NulIsEnd, false>(d, src, size, sentinel, failure);
+    }
     if (comments)
     {
-        return trailing_commas ? build_with<FloatType, true, true, NulIsEnd>(d, src, size, sentinel, failure)
+        return trailing_commas ? build_with<FloatType, true, true, NulIsEnd, true>(d, src, size, sentinel, failure)
-               : build_with<FloatType, true, false, NulIsEnd>(d, src, size, sentinel, failure);
+               : build_with<FloatType, true, false, NulIsEnd, true>(d, src, size, sentinel, failure);
     }
-    return trailing_commas ? build_with<FloatType, false, true, NulIsEnd>(d, src, size, sentinel, failure)
+    return trailing_commas ? build_with<FloatType, false, true, NulIsEnd, true>(d, src, size, sentinel, failure)
-           : build_with<FloatType, false, false, NulIsEnd>(d, src, size, sentinel, failure);
+           : build_with<FloatType, false, false, NulIsEnd, true>(d, src, size, sentinel, failure);
 }
 
 }  // namespace view

include/nlohmann/detail/view/edit.hpp

@@ -157,7 +157,7 @@ class editor
         node* const o = own(object);
         if (o->kind != node_kind::object && o->kind != node_kind::null)
         {
-            throw_type_error(exception_id::subscript_wrong_type, "cannot use operator[] with a string argument with ", object.type_name());
+            throw_type_error(site_of(object), exception_id::subscript_wrong_type, "cannot use operator[] with a string argument with ", object.type_name());
         }
         check_utf8(key.data(), key.size());
         const encoded e = encode(std::forward<V>(value));
@@ -208,9 +208,9 @@ class editor
         node* const a = own(array);
         if (a->kind != node_kind::array)
         {
-            throw_type_error(exception_id::subscript_wrong_type, "cannot use operator[] with a numeric argument with ", array.type_name());
+            throw_type_error(site_of(array), exception_id::subscript_wrong_type, "cannot use operator[] with a numeric argument with ", array.type_name());
         }
-        check_index(idx, a->len);
+        check_index(a, idx, a->len);
         const encoded e = encode(std::forward<V>(value));
         node* const slot = const_cast<node*>(nav::value(element_at<true>(m_doc, a, idx))); // NOLINT(cppcoreguidelines-pro-type-const-cast)
         assign(slot, e, a, true);
@@ -225,7 +225,7 @@ class editor
         node* const a = own(array);
         if (a->kind != node_kind::array && a->kind != node_kind::null)
         {
-            throw_type_error(exception_id::push_back_wrong_type, "cannot use push_back() with ", array.type_name());
+            throw_type_error(site_of(array), exception_id::push_back_wrong_type, "cannot use push_back() with ", array.type_name());
         }
         const encoded e = encode(std::forward<V>(value));
         if (a->kind == node_kind::null)
@@ -249,9 +249,9 @@ class editor
         node* const a = own(array);
         if (a->kind != node_kind::array)
         {
-            throw_type_error(exception_id::insert_wrong_type, "cannot use insert() with ", array.type_name());
+            throw_type_error(site_of(array), exception_id::insert_wrong_type, "cannot use insert() with ", array.type_name());
         }
-        check_index(idx, a->len + 1);
+        check_index(a, idx, a->len + 1);
         const encoded e = encode(std::forward<V>(value));
         node* const slot = new_slot(e);
         node* const h = block_of(m_doc, a, 1);
@@ -269,7 +269,7 @@ class editor
         node* const o = own(object);
         if (o->kind != node_kind::object)
         {
-            throw_type_error(exception_id::erase_wrong_type, "cannot use erase() with ", object.type_name());
+            throw_type_error(site_of(object), exception_id::erase_wrong_type, "cannot use erase() with ", object.type_name());
         }
         for (const node* k = nav::first(m_doc, o), *end = nav::end(m_doc, o); k != end; k = document_data::after(k + 1))
         {
@@ -287,9 +287,9 @@ class editor
         node* const a = own(array);
         if (a->kind != node_kind::array)
         {
-            throw_type_error(exception_id::erase_wrong_type, "cannot use erase() with ", array.type_name());
+            throw_type_error(site_of(array), exception_id::erase_wrong_type, "cannot use erase() with ", array.type_name());
         }
-        check_index(idx, a->len);
+        check_index(a, idx, a->len);
         node* const h = block_of(m_doc, a, 0);
         std::memmove(h + 1 + idx, h + 2 + idx, (h->next - 2 - idx) * sizeof(node));
         --h->next;
@@ -316,11 +316,11 @@ class editor
         return const_cast<node*>(v.m_node); // NOLINT(cppcoreguidelines-pro-type-const-cast): the nodes belong to this document
     }
 
-    static void check_index(std::size_t idx, std::size_t limit)
+    void check_index(const node* array, std::size_t idx, std::size_t limit) const
     {
         if (idx >= limit)
         {
-            throw_array_index_out_of_range(idx);
+            throw_array_index_out_of_range(error_site(&m_doc, array), idx);
         }
     }
 

include/nlohmann/detail/view/errors.hpp

@@ -10,11 +10,15 @@
 
 #include <algorithm> // min
 #include <cstddef> // size_t
-#include <string> // string
+#include <string> // string, to_string
+#include <vector> // vector
 
 #include <nlohmann/json.hpp>
 #include <nlohmann/detail/view/builder.hpp>
+#include <nlohmann/detail/view/document_data.hpp>
 #include <nlohmann/detail/view/macro_scope.hpp>
+#include <nlohmann/detail/view/materialize.hpp>
+#include <nlohmann/detail/view/node.hpp>
 
 NLOHMANN_JSON_NAMESPACE_BEGIN
 namespace detail
@@ -22,8 +26,293 @@ namespace detail
 namespace view
 {
 
+/*!
+@brief the value an exception is about
+
+basic_json passes the value to the exception, which then names its path
+(JSON_DIAGNOSTICS) and its bytes in the source (JSON_DIAGNOSTIC_POSITIONS). A
+view's value has no parent pointer and no stored positions: both are found
+from the document and the node, when the exception is thrown (so only the
+error path pays for them). Without the two macros, a site is empty.
+*/
+class error_site
+{
+  public:
+    error_site() noexcept = default;
+
+#if JSON_DIAGNOSTICS || JSON_DIAGNOSTIC_POSITIONS
+    error_site(const document_data* d, const node* n) noexcept
+        : m_doc(d), m_node(n)
+    {}
+
+    const document_data* doc() const noexcept
+    {
+        return m_doc;
+    }
+    const node* value() const noexcept
+    {
+        return m_node;
+    }
+
+  private:
+    const document_data* m_doc = nullptr;
+    const node* m_node = nullptr;
+#else
+    error_site(const document_data* /*d*/, const node* /*n*/) noexcept {}
+#endif
+};
+
+/// the site of a view's value (the views make this a friend)
+struct view_access
+{
+    template<typename View>
+    static error_site site(const View& v) noexcept
+    {
+        return {v.m_doc, v.m_node};
+    }
+};
+
+template<typename View>
+error_site site_of(const View& v) noexcept
+{
+    static_assert(is_json_view<View>::value, "the site of a value is only defined for views");
+    return view_access::site(v);
+}
+
+#if JSON_DIAGNOSTICS
+/*!
+@brief the JSON pointer tokens from the root of document @a d to node @a target
+
+Nodes have no parent index, so the path is found by descending from the root:
+before the first edit, only into the subtree that contains the target (a
+container spans the next `next` nodes); after edits, elements can live in
+other sequences, so every container is searched. Iterative (the nesting depth
+is limited by memory only). False if the target is not found (then tokens is
+unspecified).
+*/
+template<bool Editable>
+bool value_path(const document_data& d, const node* target, std::vector<std::string>& tokens)
+{
+    using nav = navigation<Editable>;
+    struct frame
+    {
+        const node* pos;
+        const node* end;
+        std::size_t index;
+        bool object;
+    };
+    const node* const root = d.nodes;
+    if (target == root)
+    {
+        return true;
+    }
+    if (!is_container(*root))
+    {
+        return false;
+    }
+    std::vector<frame> stack;
+    stack.push_back(frame{nav::first(d, root), nav::end(d, root), 0, root->kind == node_kind::object});
+    // (tokens[i] leads from stack[i] to stack[i + 1], or to the target)
+    for (;;)
+    {
+        // a copy: pop_back() and push_back() below change the stack
+        const frame f = stack.back();
+        if (f.pos == f.end)
+        {
+            stack.pop_back();
+            if (stack.empty())
+            {
+                return false;
+            }
+            tokens.pop_back();
+            continue;
+        }
+        const node* entry = f.pos;
+        std::string token;
+        if (f.object)
+        {
+            // (as basic_json: the key up to a null byte)
+            token.assign(d.str(*entry), entry->len);
+            const std::size_t nul = token.find('\0');
+            if (nul != std::string::npos)
+            {
+                token.resize(nul);
+            }
+            ++entry;
+        }
+        else
+        {
+            token = std::to_string(f.index);
+        }
+        const node* const child = nav::value(entry);
+        if (child == target)
+        {
+            tokens.push_back(token);
+            return true;
+        }
+        stack.back().pos = document_data::after(entry);
+        ++stack.back().index;
+        if (is_container(*child) && (nav::editable() || (target > child && target < child + child->next)))
+        {
+            tokens.push_back(token);
+            stack.push_back(frame{nav::first(d, child), nav::end(d, child), 0, child->kind == node_kind::object});
+        }
+    }
+}
+#endif
+
+#if JSON_DIAGNOSTIC_POSITIONS
+/*!
+@brief the bytes [start, end) of node @a n in the source, as materialize() sets them
+
+Scalars as position_source::set_scalar; an array or object from its opening
+bracket to its closing bracket, which is found by following the chain of last
+elements to a scalar (or an empty container) and scanning from there to the
+closing brackets. None (npos) for values written by edits, and for containers
+whose last element has none, as with materialize().
+*/
+template<bool Editable>
+void value_range(const document_data& d, const node* n, std::size_t& start, std::size_t& end)
+{
+    using nav = navigation<Editable>;
+    position_source p;
+    start = std::string::npos;
+    end = std::string::npos;
+    if (!is_container(*n))
+    {
+        p.set_scalar(d, *n);
+        start = p.start;
+        end = p.end;
+        return;
+    }
+    if ((n->flags & node_flags::is_new) != 0 || n->off >= d.size)
+    {
+        return;
+    }
+    // the open brackets: n, then the last element as long as it is a non-empty container
+    std::size_t closing = 0;
+    std::size_t pos = std::string::npos;
+    for (const node* c = n;;)
+    {
+        if (!is_container(*c))
+        {
+            p.set_scalar(d, *c);
+            pos = p.end;
+            break;
+        }
+        if ((c->flags & node_flags::is_new) != 0 || c->off >= d.size)
+        {
+            return;
+        }
+        ++closing;
+        const node* const first = nav::first(d, c);
+        const node* const last_end = nav::end(d, c);
+        if (first == last_end)
+        {
+            pos = std::size_t{c->off} + 1;
+            break;
+        }
+        const std::size_t value_offset = c->kind == node_kind::object ? 1 : 0;
+        const node* last = first;
+        for (const node* e = document_data::after(first + value_offset); e != last_end; e = document_data::after(e + value_offset))
+        {
+            last = e;
+        }
+        c = nav::value(last + value_offset);
+    }
+    for (; closing != 0 && pos != std::string::npos; --closing)
+    {
+        pos = position_source::container_end(d.base[0], d.size, pos);
+    }
+    if (pos != std::string::npos)
+    {
+        start = n->off;
+        end = pos;
+    }
+}
+#endif
+
+/// the context basic_json puts in front of the message of an exception about
+/// a value: "(/path) " with JSON_DIAGNOSTICS (none for the root), then
+/// "(bytes a-b) " with JSON_DIAGNOSTIC_POSITIONS (with JSON_DIAGNOSTICS, only
+/// if there is a path)
+///
+/// This and the throw helpers that take an error_site are templates, so that
+/// the path and range code is instantiated only where an exception can be
+/// thrown: instantiated in every translation unit (as a non-template inline
+/// function makes it), it changed GCC's early inlining of unrelated view
+/// functions, so that GCC 13 no longer inlined a recursive walk over a view
+/// into itself (2.7 times the instructions).
+#if JSON_DIAGNOSTICS || JSON_DIAGNOSTIC_POSITIONS
+template<typename Unused = void>
+std::string diagnostics(const error_site site)
+{
+    if (site.doc() == nullptr || site.value() == nullptr)
+    {
+        return "";
+    }
+    const document_data& d = *site.doc();
+    const bool edited = d.edits != nullptr;
+    std::string context;
+#if JSON_DIAGNOSTICS
+    std::vector<std::string> tokens;
+    const bool found = edited ? value_path<true>(d, site.value(), tokens) : value_path<false>(d, site.value(), tokens);
+    if (!found || tokens.empty())
+    {
+        // context is still empty; returning it keeps the copy elided (-Wnrvo)
+        return context;
+    }
+    std::string path;
+    for (const std::string& token : tokens)
+    {
+        path.push_back('/');
+        path += detail::escape(token);
+    }
+    context = concat('(', path, ") ");
+#endif
+#if JSON_DIAGNOSTIC_POSITIONS
+    std::size_t start = 0;
+    std::size_t end = 0;
+    if (edited)
+    {
+        value_range<true>(d, site.value(), start, end);
+    }
+    else
+    {
+        value_range<false>(d, site.value(), start, end);
+    }
+    if (start != std::string::npos && end != std::string::npos)
+    {
+        context += concat("(bytes ", std::to_string(start), "-", std::to_string(end), ") ");
+    }
+#endif
+    return context;
+}
+#else
+template<typename Unused = void>
+std::string diagnostics(const error_site /*site*/)
+{
+    return "";
+}
+#endif
+
+/// the context of an exception about view @a v (see exception_prefix())
+template<typename View>
+std::string diagnostics_of(const View& v)
+{
+    return diagnostics<>(site_of(v));
+}
+
 // Exceptions are thrown out of line, so that the accessors that may throw stay
-// small enough to be inlined.
+// small enough to be inlined. The overloads with an error_site put its context
+// in front of the message.
+//
+// The overloads without one are for exceptions that are not about a value of a
+// document: limits and invalid input of the editing and image functions, a
+// discarded view (it has no node), a view of another document (its path would
+// be one in that document), and those that basic_json creates without a
+// context too (the errors of json_pointer, invalid UTF-8). The one exception
+// about a value that has no site is key(), see view_iterator::key().
 
 [[noreturn]] NLOHMANN_VIEW_NOINLINE inline void throw_type_error(exception_id id, const char* prefix, const char* type)
 {
@@ -45,6 +334,24 @@ namespace view
     NLOHMANN_VIEW_THROW(invalid_iterator::create(id, msg, nullptr));
 }
 
+template<typename Unused = void>
+[[noreturn]] NLOHMANN_VIEW_NOINLINE void throw_type_error(const error_site site, exception_id id, const char* prefix, const char* type)
+{
+    NLOHMANN_VIEW_THROW(type_error::create(id, concat(diagnostics<>(site), prefix, type), nullptr));
+}
+
+template<typename Unused = void>
+[[noreturn]] NLOHMANN_VIEW_NOINLINE void throw_out_of_range(const error_site site, exception_id id, const std::string& msg)
+{
+    NLOHMANN_VIEW_THROW(out_of_range::create(id, concat(diagnostics<>(site), msg), nullptr));
+}
+
+template<typename Unused = void>
+[[noreturn]] NLOHMANN_VIEW_NOINLINE void throw_invalid_iterator(const error_site site, exception_id id, const char* msg)
+{
+    NLOHMANN_VIEW_THROW(invalid_iterator::create(id, concat(diagnostics<>(site), msg), nullptr));
+}
+
 /// a parse error without a position (as those of json_pointer)
 [[noreturn]] NLOHMANN_VIEW_NOINLINE inline void throw_parse_error(exception_id id, const std::string& msg)
 {
@@ -53,22 +360,23 @@ namespace view
 
 /// out_of_range.401: the array index @a idx is not below the size
 template<typename Index>
-[[noreturn]] NLOHMANN_VIEW_NOINLINE void throw_array_index_out_of_range(const Index idx)
+[[noreturn]] NLOHMANN_VIEW_NOINLINE void throw_array_index_out_of_range(const error_site site, const Index idx)
 {
-    throw_out_of_range(exception_id::array_index_out_of_range, concat("array index ", std::to_string(idx), " is out of range"));
+    throw_out_of_range(site, exception_id::array_index_out_of_range, concat("array index ", std::to_string(idx), " is out of range"));
 }
 
 /// out_of_range.402: the JSON Pointer token '-' of an array of @a size elements
-[[noreturn]] NLOHMANN_VIEW_NOINLINE inline void throw_array_index_past_end(std::size_t size)
+template<typename Unused = void>
+[[noreturn]] NLOHMANN_VIEW_NOINLINE void throw_array_index_past_end(const error_site site, std::size_t size)
 {
-    throw_out_of_range(exception_id::pointer_past_the_end_index, concat("array index '-' (", std::to_string(size), ") is out of range"));
+    throw_out_of_range(site, exception_id::pointer_past_the_end_index, concat("array index '-' (", std::to_string(size), ") is out of range"));
 }
 
 /// out_of_range.403: the object has no member @a key
 template<typename Key>
-[[noreturn]] NLOHMANN_VIEW_NOINLINE void throw_key_not_found(const Key& key)
+[[noreturn]] NLOHMANN_VIEW_NOINLINE void throw_key_not_found(const error_site site, const Key& key)
 {
-    throw_out_of_range(exception_id::key_not_found, concat("key '", key, "' not found"));
+    throw_out_of_range(site, exception_id::key_not_found, concat("key '", key, "' not found"));
 }
 
 /// SAX handler of the error replay: accepts every event without storing it,
@@ -144,11 +452,21 @@ struct throwing_sax
 
     /// (a template, as in json_sax_dom_parser: the parser passes the concrete
     /// exception type, which is thrown as is)
+    ///
+    /// It is not [[noreturn]]: sax_parse() recovers from an error if
+    /// parse_error() returns true, and MSVC reports that code as unreachable
+    /// (C4702) for a handler that never returns. The flag hides that from it.
     template<class Exception>
-    [[noreturn]] bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/, const Exception& ex)
+    bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/, const Exception& ex)
     {
-        NLOHMANN_VIEW_THROW(ex);
+        if (throwing)
+        {
+            NLOHMANN_VIEW_THROW(ex);
+        }
+        return false; // LCOV_EXCL_LINE
     }
+
+    bool throwing = true;
 };
 
 /*!

include/nlohmann/detail/view/iterator.hpp

@@ -105,7 +105,10 @@ class view_iterator
     }
 
     /// the key of the current object member; throws invalid_iterator.207 for
-    /// other iterators, like basic_json's iterators
+    /// other iterators, like basic_json's iterators (but without the context
+    /// JSON_DIAGNOSTICS and JSON_DIAGNOSTIC_POSITIONS add: the iterator does not
+    /// know the value it iterates over, and keeping it would make iterators
+    /// larger, which keeps GCC from inlining recursive walks)
     string_view_t key() const
     {
         if (NLOHMANN_VIEW_UNLIKELY(m_value_offset == 0))

include/nlohmann/detail/view/macro_scope.hpp

@@ -40,6 +40,15 @@
     #define NLOHMANN_VIEW_NOINLINE
 #endif
 
+// GCC: a function compiled as written, without the clones in which interprocedural
+// constant propagation drops an argument that every caller passes as the same constant
+// (calling such a clone changes the code GCC generates for the parse loop that calls it)
+#if defined(__GNUC__) && !defined(__clang__)
+    #define NLOHMANN_VIEW_NOCLONE __attribute__((noclone))
+#else
+    #define NLOHMANN_VIEW_NOCLONE
+#endif
+
 #if defined(__GNUC__) || defined(__clang__)
     #define NLOHMANN_VIEW_NODISCARD __attribute__((warn_unused_result))
 #elif defined(_MSC_VER)

include/nlohmann/detail/view/macro_unscope.hpp

@@ -15,6 +15,7 @@
 #undef NLOHMANN_VIEW_UNLIKELY
 #undef NLOHMANN_VIEW_ALWAYS_INLINE
 #undef NLOHMANN_VIEW_NOINLINE
+#undef NLOHMANN_VIEW_NOCLONE
 #undef NLOHMANN_VIEW_NODISCARD
 #undef NLOHMANN_VIEW_THROW
 #undef NLOHMANN_VIEW_LITTLE_ENDIAN

include/nlohmann/detail/view/number.hpp

@@ -19,6 +19,7 @@
 #include <nlohmann/detail/view/macro_scope.hpp>
 #include <nlohmann/detail/view/node.hpp>
 #include <nlohmann/detail/view/scan.hpp>
+#include <nlohmann/detail/view/simd.hpp>
 
 NLOHMANN_JSON_NAMESPACE_BEGIN
 namespace detail
@@ -103,12 +104,56 @@ NLOHMANN_VIEW_NOINLINE FloatType float_value(const char* first, const node& n)
     return convert_float<FloatType>(first, last, dot, mantissa_end);
 }
 
+#if NLOHMANN_VIEW_VECTOR
+/*!
+@brief the value of the first 16 digits of a run of digits with a point after
+int_digits of them (or none, if int_digits >= 16)
+
+The 17 bytes at p are read at once: the bytes after the point move down by one,
+then pairs, quadruples, and octets of digits are combined by multiplications
+(after simdjson). The bytes [p, p + 32) must be readable. Bytes that are not
+digits (only in a damaged image loaded with image_check::bounds) give a wrong
+value, but no undefined behavior.
+*/
+NLOHMANN_VIEW_ALWAYS_INLINE std::uint64_t sixteen_digits(const unsigned char* p, unsigned int_digits) noexcept
+{
+#if NLOHMANN_VIEW_NEON
+    const uint8x16_t low = vld1q_u8(p);
+    const uint8x16_t index = vcombine_u8(vcreate_u8(0x0706050403020100u), vcreate_u8(0x0F0E0D0C0B0A0908u));
+    const uint8x16_t before_point = vcltq_u8(index, vdupq_n_u8(static_cast<std::uint8_t>(int_digits)));
+    const uint8x16_t digits = vsubq_u8(vbslq_u8(before_point, low, vextq_u8(low, vld1q_u8(p + 16), 1)), vdupq_n_u8('0'));
+    const uint16x8_t twos = vpaddlq_u8(vmulq_u8(digits, vreinterpretq_u8_u16(vdupq_n_u16(0x010A))));
+    const uint32x4_t fours = vpaddlq_u16(vmulq_u16(twos, vreinterpretq_u16_u32(vdupq_n_u32(0x00010064))));
+    const uint64x2_t eights = vpaddlq_u32(vmulq_u32(fours, vreinterpretq_u32_u64(vdupq_n_u64(0x0000000100002710u))));
+    return (vgetq_lane_u64(eights, 0) * 100000000u) + vgetq_lane_u64(eights, 1);
+#else
+    // NOLINTBEGIN(portability-simd-intrinsics)
+    const __m128i low = _mm_loadu_si128(static_cast<const __m128i*>(static_cast<const void*>(p))); // NOLINT(bugprone-casting-through-void)
+    const __m128i high = _mm_loadu_si128(static_cast<const __m128i*>(static_cast<const void*>(p + 16))); // NOLINT(bugprone-casting-through-void)
+    const __m128i before_point = _mm_cmpgt_epi8(_mm_set1_epi8(static_cast<char>(int_digits)), _mm_setr_epi8(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15));
+    const __m128i moved = _mm_or_si128(_mm_srli_si128(low, 1), _mm_slli_si128(high, 15));
+    const __m128i digits = _mm_sub_epi8(_mm_or_si128(_mm_and_si128(before_point, low), _mm_andnot_si128(before_point, moved)), _mm_set1_epi8('0'));
+    // (16-bit lanes: SSE2 has no multiply-add of bytes)
+    const __m128i tens = _mm_setr_epi16(10, 1, 10, 1, 10, 1, 10, 1);
+    const __m128i twos = _mm_packs_epi32(_mm_madd_epi16(_mm_unpacklo_epi8(digits, _mm_setzero_si128()), tens),
+                                         _mm_madd_epi16(_mm_unpackhi_epi8(digits, _mm_setzero_si128()), tens));
+    const __m128i fours = _mm_madd_epi16(twos, _mm_setr_epi16(100, 1, 100, 1, 100, 1, 100, 1));
+    const __m128i eights = _mm_madd_epi16(_mm_packs_epi32(fours, fours), _mm_setr_epi16(10000, 1, 10000, 1, 10000, 1, 10000, 1));
+    const auto upper = static_cast<std::uint32_t>(_mm_cvtsi128_si32(eights));
+    const auto lower = static_cast<std::uint32_t>(_mm_cvtsi128_si32(_mm_srli_si128(eights, 4)));
+    // NOLINTEND(portability-simd-intrinsics)
+    return (std::uint64_t{upper} * 100000000u) + lower;
+#endif
+}
+#endif
+
 /*!
 @brief the digits of a float token with at most 19 digits, from its layout
 
 The digit layout recorded while parsing says where the integer digits, the
 fraction digits, and the exponent are, so the digits are read eight at a
-time without scanning.
+time without scanning (with SIMD, a token of 16 to 19 digits sixteen at a
+time).
 
 @param[in] p  first character of the token
 @param[in] e  end of the token
@@ -118,15 +163,34 @@ NLOHMANN_VIEW_ALWAYS_INLINE float_significand layout_decimal(const unsigned char
 {
     const bool negative = *p == '-';
     p += negative ? 1 : 0;
-    std::uint64_t w = parse_upto19(p, int_digits, limit);
+    std::uint64_t w = 0;
-    p += int_digits;
     std::int64_t q = 0;
-    if (frac_digits != 0)
+#if NLOHMANN_VIEW_VECTOR
+    const unsigned count = int_digits + frac_digits;
+    if (count >= 16 && limit - p >= 32)
     {
-        w = (w * int_pow10(frac_digits)) + parse_upto19(p + 1, frac_digits, limit);
+        // (the point, if any, follows the integer digits; the digits after
-        p += 1 + frac_digits;
+        // the first 16 are at most three)
+        w = sixteen_digits(p, int_digits);
+        for (unsigned i = 16; i < count; ++i)
+        {
+            w = (w * 10) + static_cast<std::uint64_t>(p[i + (i >= int_digits ? 1 : 0)] - '0');
+        }
+        p += count + (frac_digits != 0 ? 1 : 0);
         q = -static_cast<std::int64_t>(frac_digits);
     }
+    else /* NOLINT(readability/braces) */
+#endif
+    {
+        w = parse_upto19(p, int_digits, limit);
+        p += int_digits;
+        if (frac_digits != 0)
+        {
+            w = (w * int_pow10(frac_digits)) + parse_upto19(p + 1, frac_digits, limit);
+            p += 1 + frac_digits;
+            q = -static_cast<std::int64_t>(frac_digits);
+        }
+    }
     if (p != e)
     {
         // [eE][+-]digits; huge exponents saturate (the parser rejected
@@ -161,12 +225,22 @@ NLOHMANN_VIEW_ALWAYS_INLINE float_significand layout_decimal(const unsigned char
 The result is correctly rounded by the lexer's conversion
 (detail::decimal_to_float(): Clinger's fast path where both operands are
 exact, else the Eisel-Lemire algorithm, which needs no fallback for up to 19
-digits), so it is the value parse() produces.
+digits), so it is the value parse() produces. Tokens of 16 or more digits go
+to Eisel-Lemire directly: the fast path takes few of them (none of 17 digits),
+and its check is a branch that data like GeoJSON coordinates mispredicts.
 */
 template<typename FloatType>
 NLOHMANN_VIEW_ALWAYS_INLINE FloatType layout_float(const unsigned char* p, const unsigned char* e, unsigned int_digits, unsigned frac_digits, const unsigned char* limit) noexcept
 {
-    return decimal_to_float<FloatType>(layout_decimal(p, e, int_digits, frac_digits, limit));
+    const float_significand d = layout_decimal(p, e, int_digits, frac_digits, limit);
+    if (int_digits + frac_digits >= 16)
+    {
+        using format = ieee_binary_format<std::numeric_limits<FloatType>::digits>;
+        native_float_t<FloatType> value{};
+        float_from_bits(eisel_lemire<format>(d.exponent, d.w) | (d.negative ? (std::uint64_t{1} << format::sign_bit()) : 0u), value);
+        return value;
+    }
+    return decimal_to_float<FloatType>(d);
 }
 
 /// the value of a float set by an edit: its token (the shortest round-trip
@@ -185,6 +259,80 @@ NLOHMANN_VIEW_NOINLINE FloatType edited_float(const char* token, const node& n)
     return float_value<FloatType>(token, n);
 }
 
+#if NLOHMANN_VIEW_VECTOR
+/// eisel_lemire() for the token w * 10^q of coordinate_float()
+template<typename Format>
+NLOHMANN_VIEW_ALWAYS_INLINE std::uint64_t coordinate_bits(std::int64_t q, std::uint64_t w, Format /*format*/) noexcept
+{
+    return eisel_lemire<Format>(q, w);
+}
+
+/*!
+@brief (binary64) eisel_lemire() without the cases that tokens of 16 or 17
+digits with 5 or more after the point cannot reach: a value of 0, a
+subnormal, an infinity, or a tie (q < -4)
+*/
+NLOHMANN_VIEW_ALWAYS_INLINE std::uint64_t coordinate_bits(std::int64_t q, std::uint64_t w, ieee_binary_format<53> format) noexcept
+{
+    if (NLOHMANN_VIEW_UNLIKELY(w == 0 || q > -5))
+    {
+        return eisel_lemire<ieee_binary_format<53>>(q, w);
+    }
+    static_cast<void>(format);
+    const int lz = count_leading_zeros(w);
+    const std::uint64_t normalized = w << static_cast<unsigned>(lz);
+    const auto index = static_cast<std::size_t>(2 * (q - pow5_128_smallest_power));
+    const uint128_parts product = full_multiplication(normalized, pow5_128()[index]);
+    std::uint64_t high = product.high;
+    if (NLOHMANN_VIEW_UNLIKELY((high & 0x1FFu) == 0x1FFu))
+    {
+        // (the lower bits may carry into the result)
+        const std::uint64_t second = full_multiplication(normalized, pow5_128()[index + 1]).high;
+        high += product.low + second < product.low ? 1u : 0u;
+    }
+    const auto upperbit = static_cast<unsigned>(high >> 63u);
+    std::uint64_t mantissa = high >> (upperbit + 9);
+    auto power2 = static_cast<std::uint64_t>(((217706 * q) >> 16) + 63 + static_cast<std::int64_t>(upperbit) - lz + 1023);
+    mantissa = (mantissa + (mantissa & 1u)) >> 1u;
+    const bool carry = mantissa >= (std::uint64_t{2} << 52u);
+    mantissa = carry ? (std::uint64_t{1} << 52u) : mantissa;
+    power2 += carry ? 1u : 0u;
+    return (mantissa & ~(std::uint64_t{1} << 52u)) | (power2 << 52u);
+}
+
+/*!
+@brief (binary32 and binary64) the value of a float token of the source with
+16 or 17 digits, a point, and no exponent, the shape of most coordinates
+written with full precision; whether the token of n has this shape
+
+These tokens skip the general layout of layout_decimal(): no digits beyond
+the 17th, no exponent (and Eisel-Lemire directly, as in layout_float()).
+*/
+template<typename FloatType>
+NLOHMANN_VIEW_ALWAYS_INLINE bool coordinate_float(const document_data& d, const node& n, FloatType& value) noexcept
+{
+    const unsigned int_digits = n.extra & 0xFFu;
+    const unsigned frac_digits = n.extra >> 8u;
+    const unsigned count = int_digits + frac_digits;
+    const auto* const first = reinterpret_cast<const unsigned char*>(d.src + n.off); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
+    const unsigned negative = first[0] == '-' ? 1u : 0u;
+    if (count - 16u > 1u || frac_digits == 0 || n.len != negative + count + 1 || d.size - n.off < 33)
+    {
+        return false;
+    }
+    const unsigned char* const digits = first + negative;
+    const std::uint64_t w = sixteen_digits(digits, int_digits);
+    // (the 17th digit follows the point)
+    const std::uint64_t w17 = (w * 10) + static_cast<std::uint64_t>(digits[17] - '0');
+    using format = ieee_binary_format<std::numeric_limits<FloatType>::digits>;
+    native_float_t<FloatType> result{};
+    float_from_bits(coordinate_bits(-static_cast<std::int64_t>(frac_digits), count == 17 ? w17 : w, format{}) | (static_cast<std::uint64_t>(negative) << format::sign_bit()), result);
+    value = result;
+    return true;
+}
+
+#endif
+
 /// the value of the float token of a node, as parse() converts it; floats and
 /// doubles with at most 19 digits are converted from the digit layout
 template<typename FloatType>
@@ -200,6 +348,13 @@ FloatType float_value(const document_data& d, const node& n)
 template<typename FloatType>
 FloatType float_value(const document_data& d, const node& n, std::true_type /*binary32 or binary64*/)
 {
+#if NLOHMANN_VIEW_VECTOR
+    FloatType value {};
+    if (coordinate_float(d, n, value))
+    {
+        return value;
+    }
+#endif
     const unsigned int_digits = n.extra & 0xFFu;
     const unsigned frac_digits = n.extra >> 8u;
     if (NLOHMANN_VIEW_LIKELY(int_digits + frac_digits <= 19)) // (255 marks "many")

include/nlohmann/detail/view/pointer.hpp

@@ -44,6 +44,7 @@ index_status array_index(const StringType& s, std::size_t& idx) noexcept
 }
 
 /// throws the exception json_pointer::array_index throws for this status
+/// (without a context, as json_pointer::array_index)
 template<typename StringType>
 [[noreturn]] NLOHMANN_VIEW_NOINLINE void throw_array_index_error(index_status status, const StringType& s)
 {
@@ -83,7 +84,7 @@ View resolve_pointer(View cur, const Tokens& tokens, pointer_mode mode)
             {
                 if (mode == pointer_mode::checked)
                 {
-                    throw_key_not_found(token);
+                    throw_key_not_found(site_of(cur), token);
                 }
                 return View();
             }
@@ -95,7 +96,7 @@ View resolve_pointer(View cur, const Tokens& tokens, pointer_mode mode)
             {
                 if (throwing)
                 {
-                    throw_array_index_past_end(cur.size());
+                    throw_array_index_past_end(site_of(cur), cur.size());
                 }
                 return View();
             }
@@ -114,7 +115,7 @@ View resolve_pointer(View cur, const Tokens& tokens, pointer_mode mode)
             {
                 if (mode == pointer_mode::checked)
                 {
-                    throw_array_index_out_of_range(idx);
+                    throw_array_index_out_of_range(site_of(cur), idx);
                 }
                 return View();
             }
@@ -124,7 +125,7 @@ View resolve_pointer(View cur, const Tokens& tokens, pointer_mode mode)
         {
             if (throwing)
             {
-                throw_out_of_range(exception_id::pointer_unresolved, concat("unresolved reference token '", token, "'"));
+                throw_out_of_range(site_of(cur), exception_id::pointer_unresolved, concat("unresolved reference token '", token, "'"));
             }
             return View();
         }

include/nlohmann/detail/view/scan.hpp

@@ -198,6 +198,57 @@ inline const unsigned char* find_block_comment_end(const unsigned char* p, const
     return p;
 }
 
+/// The digits of a fraction, whose number varies from one number to the next: skip_digits() with 16 bytes compared at
+/// once and the end of the run taken from a bit mask (SSE2), instead of a compare and a branch for each digit. That
+/// is about a third of the instructions for a fraction of 12 digits, and no branch that is mispredicted when the
+/// lengths vary. Where there is no SSE2, skip_digits() is used.
+#if NLOHMANN_VIEW_SSE2
+// The last bytes of the input, one at a time. GCC unrolls this loop in every inlined copy of the function below, which
+// makes the parse loop 4 KiB larger and slows down documents that have hardly any fractions; out of line, it does
+// not. Clang does better with the loop inline (a call in the parse loop costs it more than the unrolling).
+#if defined(__GNUC__) && !defined(__clang__)
+inline NLOHMANN_VIEW_NOINLINE const unsigned char* skip_digits_tail(const unsigned char* p, const unsigned char* e) noexcept
+{
+    while (p != e && is_digit(*p))
+    {
+        ++p;
+    }
+    return p;
+}
+#else
+NLOHMANN_VIEW_ALWAYS_INLINE const unsigned char* skip_digits_tail(const unsigned char* p, const unsigned char* e) noexcept
+{
+    while (p != e && is_digit(*p))
+    {
+        ++p;
+    }
+    return p;
+}
+#endif
+
+NLOHMANN_VIEW_ALWAYS_INLINE const unsigned char* skip_fraction_digits(const unsigned char* p, const unsigned char* e) noexcept
+{
+    while (e - p >= 16)
+    {
+        const __m128i in = _mm_loadu_si128(static_cast<const __m128i*>(static_cast<const void*>(p))); // NOLINT(bugprone-casting-through-void)
+        // a digit is above '/' and below ':' (signed compares: a byte with the top bit set is negative, and neither)
+        const __m128i digits = _mm_and_si128(_mm_cmpgt_epi8(in, _mm_set1_epi8('/')), _mm_cmplt_epi8(in, _mm_set1_epi8(':')));
+        const auto not_digit = static_cast<std::uint64_t>(~static_cast<unsigned>(_mm_movemask_epi8(digits)) & 0xFFFFu);
+        if (not_digit != 0)
+        {
+            return p + count_trailing_zeros(not_digit);
+        }
+        p += 16;
+    }
+    return skip_digits_tail(p, e);
+}
+#else
+NLOHMANN_VIEW_ALWAYS_INLINE const unsigned char* skip_fraction_digits(const unsigned char* p, const unsigned char* e) noexcept
+{
+    return skip_digits(p, e);
+}
+#endif
+
 /// powers of ten up to 10^19 as integers
 inline std::uint64_t int_pow10(unsigned k) noexcept
 {

include/nlohmann/detail/view/serializer.hpp

@@ -14,7 +14,9 @@
 #include <cstddef> // size_t
 #include <cstdint> // uint8_t, uint32_t
 #include <cstring> // memcpy, memset
+#include <functional> // less
 #include <limits> // numeric_limits
+#include <memory> // unique_ptr
 #include <type_traits> // integral_constant
 #include <vector> // vector
 
@@ -384,6 +386,143 @@ struct depth_counter<false>
     }
 };
 
+/*!
+@brief where the double x nearest to a token w * 10^q lies, for 10^15 <= w <
+10^17 and -17 <= q <= -1 (Eisel-Lemire, then a fixed-point position)
+
+dump() writes the shortest decimal in the rounding interval of x (the closest
+one if there are several, see zmij::to_shortest()). Its digits are those of a
+multiple of 10^g of the token's last digit, with g = dec_exp + 1 - q for x's
+decimal exponent dec_exp: the multiple of 10^g nearest to x if it lies in the
+interval (at most one does), else the multiple of 10^(g - 1) nearest to x
+(see token_digits_1() and token_digits_2()). For g <= 0, the token itself is
+in the interval.
+
+The position is exact in a 64-bit fixed-point number with 56 bits after the
+point (of the token's last digit): x / 10^q = 2c * 5^-q / 2^t for x = c * 2^e2
+and t = q + 1 - e2, with at most 56 bits after the point here, and |x / 10^q -
+w| is at most half the spacing of the doubles, below 12 units. So only the low
+64 bits of 2c * 5^-q are needed. c and e2 come from Eisel-Lemire's product
+directly (as eisel_lemire() rounds it: no tie is possible for q < -4, and its
+second product is added where it can carry into the result).
+
+@param[out] offset  x / 10^q - w + 32, with 56 bits after the point
+@param[out] radius  half the spacing of the doubles at x, likewise (+ 1 if
+                    the interval is closed)
+@return g; 3 if the digits are not decided here (a power of two, whose
+        interval is not symmetric, or out of range)
+*/
+NLOHMANN_VIEW_ALWAYS_INLINE int locate_token(std::uint64_t w, int q, std::uint64_t& offset, std::uint64_t& radius) noexcept
+{
+    const int lz = count_leading_zeros(w);
+    const std::uint64_t normalized = w << static_cast<unsigned>(lz);
+    const auto index = static_cast<std::size_t>(2 * (q - pow5_128_smallest_power));
+    const uint128_parts product = full_multiplication(normalized, pow5_128()[index]);
+    std::uint64_t high = product.high;
+    if (NLOHMANN_VIEW_UNLIKELY((high & 0x1FFu) == 0x1FFu))
+    {
+        // (as eisel_lemire(): the lower bits may carry into the result)
+        const std::uint64_t second = full_multiplication(normalized, pow5_128()[index + 1]).high;
+        high += product.low + second < product.low ? 1u : 0u;
+    }
+    const auto upperbit = static_cast<unsigned>(high >> 63u);
+    const std::uint64_t mantissa = high >> (upperbit + 9);
+    const std::uint64_t c = (mantissa + (mantissa & 1u)) >> 1u;
+    // e2 as eisel_lemire() computes the exponent (log2(10) ~ 217706 / 2^16)
+    const int e2 = static_cast<int>((217706 * static_cast<std::int64_t>(q)) >> 16) + 63 + static_cast<int>(upperbit) - lz - 52;
+    const int t = q + 1 - e2;
+    const int g = ((e2 * 315653) >> 20) + 1 - q; // (dec_exp as in zmij::to_shortest(), for a regular x)
+    static const std::array<std::uint64_t, 17> fives = {{1u, 5u, 25u, 125u, 625u, 3125u, 15625u, 78125u, 390625u, 1953125u, 9765625u, 48828125u, 244140625u, 1220703125u, 6103515625u, 30517578125u, 152587890625u}};
+    const std::uint64_t five = fives[static_cast<std::size_t>(-q)]; // 5^-q
+    const unsigned s = static_cast<unsigned>(56 - t) & 63u;
+    offset = ((2 * c * five) << s) - (w << 56u) + (std::uint64_t{32} << 56u);
+    radius = (five << s) + ((c & 1u) == 0 ? 1u : 0u);
+    // (c = 2^52: a power of two; c = 2^53: rounded up to one. With 8 to 16
+    // fraction digits, as the batch's tokens have, t is in [15, 40] and g at
+    // most 2. x lies halfway between two candidates only if its distance from
+    // the token is a whole unit (g == 2) or a half (g == 1): such tokens are
+    // left to the full conversion)
+    const bool decided = c - ((std::uint64_t{1} << 52u) + 1) < (std::uint64_t{1} << 52u) - 1
+                         && (offset & ((std::uint64_t{1} << 55u) - 1)) != 0;
+    return decided ? g : 3;
+}
+
+/// all bits set if b, else none
+NLOHMANN_VIEW_ALWAYS_INLINE std::uint64_t mask_if(bool b) noexcept
+{
+    return std::uint64_t{0} - static_cast<std::uint64_t>(b);
+}
+
+/// a if b, else c, without a branch (the conditions of token_digits_1/2() are
+/// as good as random, and GCC compiled their ?: as branches: one misprediction
+/// per float of canada.json on x86-64)
+NLOHMANN_VIEW_ALWAYS_INLINE std::uint64_t select_if(bool b, std::uint64_t a, std::uint64_t c) noexcept
+{
+    return c ^ ((a ^ c) & mask_if(b));
+}
+
+/// the distance of a position (see locate_token()) from the biased point 32
+NLOHMANN_VIEW_ALWAYS_INLINE std::uint64_t biased_distance(std::uint64_t position) noexcept
+{
+    constexpr std::uint64_t center = std::uint64_t{32} << 56u;
+    // |position - center|, without a branch: negated (complemented, plus 1) if below
+    const std::uint64_t below = mask_if(position < center);
+    return ((position - center) ^ below) - below;
+}
+
+/*!
+@brief (g == 1) the digits dump() writes for a token, from the position of its
+double (see locate_token()): the multiple of 10 nearest to x if it is in the
+interval, else w rounded to x (an exact tie between two candidates, like
+754570942.33984373, whose double is 754570942.33984375, is excluded by
+locate_token())
+
+With |x - w| at most the radius, which is below half the spacing of the
+candidates, the multiple of 10 nearest to x is w - r10 or the one above it.
+*/
+NLOHMANN_VIEW_ALWAYS_INLINE std::uint64_t token_digits_1(std::uint64_t w, std::uint64_t offset, std::uint64_t radius, std::uint64_t r10) noexcept
+{
+    constexpr std::uint64_t one = std::uint64_t{1} << 56u;
+    const std::uint64_t from = offset + (r10 * one); // x - (w - r10) (+ 32)
+    const bool up = from > 37 * one;                  // (nearer to the multiple above)
+    const bool shorter = biased_distance(from - (mask_if(up) & (10 * one))) < radius;
+    const std::uint64_t a = w - r10 + (mask_if(up) & 10u);
+    const std::uint64_t b = w + ((offset + (one / 2)) >> 56u) - 32;                // w rounded to x
+    return select_if(shorter, a, b);
+}
+
+/*!
+@brief (g == 2) the digits dump() writes for a token, as token_digits_1(): the
+multiple of 100 nearest to x if it is in the interval, else the multiple of 10
+nearest to x
+*/
+NLOHMANN_VIEW_ALWAYS_INLINE std::uint64_t token_digits_2(std::uint64_t w, std::uint64_t offset, std::uint64_t radius, std::uint64_t r10, std::uint64_t r100) noexcept
+{
+    constexpr std::uint64_t one = std::uint64_t{1} << 56u;
+    const std::uint64_t from = offset + (r100 * one);
+    const bool up = from > 82 * one;
+    const bool shorter = biased_distance(from - (mask_if(up) & (100 * one))) < radius;
+    const std::uint64_t a = w - r100 + (mask_if(up) & 100u);
+    // x - (w - r10), in [-12, 21]: the multiple of 10 nearest to x is that one,
+    // or one or two above, or one below
+    const std::uint64_t y = offset + (r10 * one);
+    const std::uint64_t steps = static_cast<std::uint64_t>(y > 37 * one) + static_cast<std::uint64_t>(y > 47 * one);
+    const bool down = y < 27 * one;
+    const std::uint64_t b = w - r10 + (10 * steps) - (mask_if(down) & 10u);
+    return select_if(shorter, a, b);
+}
+
+/// the four ASCII digits of x < 10^4, the first in the lowest byte (as
+/// little-endian memory holds them)
+NLOHMANN_VIEW_ALWAYS_INLINE std::uint32_t four_digit_text(std::uint32_t x) noexcept
+{
+    const std::uint32_t ab = (x * 5243u) >> 19u; // x / 100
+    std::uint32_t v = ab | ((x - (ab * 100u)) << 16u);
+    const std::uint32_t a = ((v * 103u) >> 10u) & 0x000F000Fu; // / 10 per 16-bit lane
+    v = a | ((v - (a * 10u)) << 8u);
+    return v + 0x30303030u;
+}
+
 /// how the view's dump() writes a value
 struct dump_style
 {
@@ -711,6 +850,7 @@ class view_serializer
         std::array<frame, 32> buffer; // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
         std::vector<frame> heap;
         small_stack<frame> stack(buffer.data(), buffer.size(), heap);
+        float_batch batch;
         const char* const src = m_doc.src;
         const char* const src_end = src + m_doc.size;
         char* w = m_out.cursor();
@@ -850,7 +990,7 @@ class view_serializer
                     else if (enabled(fast_double::value))
                     {
                         room(64);
-                        w = write_double_at(w, *n);
+                        w = write_double_ahead(w, src, n, batch);
                     }
                     else
                     {
@@ -875,6 +1015,20 @@ class view_serializer
                 case value_t::array:
                 {
                     const bool object = n->kind == node_kind::object;
+                    // (read-only documents, compact output) a pair of floats,
+                    // like the coordinates of GeoJSON, without a frame
+                    if (enabled(!Editable && !Pretty && !SourceNumbers && fast_double::value) && n->next == 3 && n->len == 2
+                            && n[1].kind == node_kind::number_float && n[2].kind == node_kind::number_float)
+                    {
+                        room(2 * 64);
+                        *w++ = '[';
+                        w = write_double_ahead(w, src, n + 1, batch);
+                        *w++ = ',';
+                        w = write_double_ahead(w, src, n + 2, batch);
+                        *w++ = ']';
+                        n += 2; // (the element before the next node)
+                        break;
+                    }
                     room(8);
                     if (n->len == 0)
                     {
@@ -984,6 +1138,254 @@ class view_serializer
         }
     }
 
+    /// (float_batch) the mark of digits whose rounding carried beyond the
+    /// token's last four (see mark())
+    static constexpr std::uint64_t carried() noexcept
+    {
+        return std::uint64_t{1} << 63u;
+    }
+
+    /// floats of 16 or 17 digits ahead of the walk, converted together (see
+    /// convert_ahead()), with their digits as mark() keeps them
+    struct float_batch
+    {
+        static constexpr std::size_t capacity() noexcept
+        {
+            return 256;
+        }
+        /// (on the heap, allocated by the first conversion: a frame of more
+        /// than a page made every dump of a small document slower)
+        struct entries
+        {
+            std::array<const node*, 256> at; // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
+            std::array<std::uint64_t, 256> digits; // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
+        };
+        std::unique_ptr<entries> storage{};
+        const node** at = nullptr;
+        std::uint64_t* digits = nullptr;
+        std::size_t count = 0;
+        std::size_t next = 0;
+        const node* scanned = nullptr; ///< where the last scan of the node array ended
+    };
+
+    /// (doubles) the float at n as dump() writes it, at w (64 bytes of room):
+    /// from the batch if it holds n
+    NLOHMANN_VIEW_ALWAYS_INLINE char* write_double_ahead(char* w, const char* src, const node* n, float_batch& batch)
+    {
+#if NLOHMANN_VIEW_VECTOR && NLOHMANN_VIEW_LITTLE_ENDIAN
+        // (digits 0: not written by the batch; carried(): rounding carried
+        // beyond the last four digits)
+        if (batch.next != batch.count && batch.at[batch.next] == n && batch.digits[batch.next] - 1 < carried() - 1)
+        {
+            return write_token_digits(w, src, *n, batch.digits[batch.next++]);
+        }
+        return write_double_beside(w, n, batch);
+#else
+        static_cast<void>(src);
+        static_cast<void>(batch);
+        return write_double_at(w, *n);
+#endif
+    }
+
+#if NLOHMANN_VIEW_VECTOR && NLOHMANN_VIEW_LITTLE_ENDIAN
+    /// write_double_ahead() for a float that is not the batch's next one (out
+    /// of line: the walk is faster with less code in its loop)
+    NLOHMANN_VIEW_NOINLINE char* write_double_beside(char* w, const node* n, float_batch& batch)
+    {
+        // (an editable document's walk can skip nodes of the array)
+        const std::less<const node*> before{}; // NOLINT(modernize-use-transparent-functors) C++11 has no std::less<>
+        while (batch.next != batch.count && before(batch.at[batch.next], n))
+        {
+            ++batch.next;
+        }
+        if ((batch.next != batch.count && batch.at[batch.next] == n) || convert_ahead(n, batch))
+        {
+            const std::uint64_t digits = batch.digits[batch.next++];
+            if ((digits & carried()) != 0)
+            {
+                return write_carried_digits(w, *n, digits & ~carried());
+            }
+            if (digits != 0)
+            {
+                return write_token_digits(w, m_doc.src, *n, digits);
+            }
+        }
+        return write_double_at(w, *n);
+    }
+
+    /*!
+    @brief convert the floats of 16 or 17 digits among the next nodes of the
+    array at once; whether n is the first of them
+
+    Such tokens (like the coordinates of GeoJSON, written with 17 digits) are
+    written as the token with its last four digits replaced
+    (write_token_digits()). Each step (the digits, locate_token(),
+    token_digits_1/2()) is a long chain of dependent operations; one loop per
+    step over many floats lets the processor overlap the floats, which it does
+    not across the steps of one float (on Apple M1, one loop for the digits and
+    locate_token() made the dump of canada.json 12 % slower). Each node of the array
+    is scanned once; a node that the walk does not reach in order (in an
+    editable document) is written without the batch.
+    */
+    NLOHMANN_VIEW_NOINLINE bool convert_ahead(const node* n, float_batch& batch)
+    {
+        const std::less<const node*> before{}; // NOLINT(modernize-use-transparent-functors) C++11 has no std::less<>
+        const node* const last = m_doc.nodes + m_doc.node_count;
+        if (before(n, batch.scanned) || before(n, m_doc.nodes) || !before(n, last))
+        {
+            return false;
+        }
+        const char* const src = m_doc.src;
+        const char* const src_end = src + m_doc.size;
+        const node* const stop = last - n > 1024 ? n + 1024 : last;
+        batch.count = 0;
+        batch.next = 0;
+        // the digits of each token (and its last four from the text)
+        std::array<std::uint64_t, 256> digits;     // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
+        std::array<int, 256> exponent;             // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
+        std::array<std::uint32_t, 256> last_four;  // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
+        const node* p = n;
+        if (!batch.storage)
+        {
+            batch.storage.reset(new typename float_batch::entries); // NOLINT(cppcoreguidelines-owning-memory,modernize-make-unique): C++11
+            batch.at = batch.storage->at.data();
+            batch.digits = batch.storage->digits.data();
+        }
+        for (; p != stop && batch.count != float_batch::capacity(); ++p)
+        {
+            // a token of the source: 16 or 17 digits, at least 8 after the
+            // point, no exponent, 32 bytes to read
+            const unsigned frac_digits = p->extra >> 8u;
+            const unsigned int_digits = p->extra & 0xFFu;
+            const unsigned count = int_digits + frac_digits;
+            if (p->kind == node_kind::number_float && (p->flags & node_flags::storage) == 0
+                    && count - 16u <= 1u && frac_digits >= 8 && src_end - (src + p->off) >= 33)
+            {
+                const auto* const token = reinterpret_cast<const unsigned char*>(src + p->off); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
+                const unsigned sign = token[0] == '-' ? 1 : 0;
+                const unsigned char* const first = token + sign;
+                const std::uint64_t w = sixteen_digits(first, int_digits);
+                // (the 17th digit follows the point; a token with an exponent
+                // gets digits outside the range of locate_token(), and is not
+                // written by the batch. A token 0.xxx with 16 significant
+                // digits is: its double is written as "0." and its digits too)
+                const std::uint64_t w17 = (w * 10) + static_cast<std::uint64_t>(first[17] - '0');
+                const std::size_t i = batch.count++;
+                batch.at[i] = p;
+                if (p->len != sign + count + 1)
+                {
+                    digits[i] = 0;
+                }
+                else
+                {
+                    digits[i] = count == 17 ? w17 : w;
+                }
+                exponent[i] = -static_cast<int>(frac_digits);
+                // (byte 0: the last digit, byte 1: the last two, bits 16-29: the last four)
+                std::uint32_t four = 0;
+                std::memcpy(&four, token + p->len - 4, 4);
+                four -= 0x30303030u;
+                const std::uint32_t pairs = ((four & 0x0F0F0F0Fu) * 2561u) >> 8u; // (byte 0, 2: the pairs)
+                last_four[i] = (four >> 24u) | ((pairs >> 16u & 0xFFu) << 8u) | ((((pairs & 0x00FF00FFu) * 6553601u) >> 16u) << 16u);
+            }
+        }
+        batch.scanned = p;
+        // the position of each double, and the floats of either kind
+        std::array<std::uint64_t, 256> offset; // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
+        std::array<std::uint64_t, 256> radius; // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
+        std::array<std::uint8_t, 256> kind_1;  // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
+        std::array<std::uint8_t, 256> kind_2;  // NOLINT(cppcoreguidelines-pro-type-member-init,hicpp-member-init): written before read
+        std::size_t count_1 = 0;
+        std::size_t count_2 = 0;
+        for (std::size_t i = 0; i != batch.count; ++i)
+        {
+            const std::uint64_t w = digits[i];
+            // (the range excludes the values of bytes that are not digits, only
+            // in a damaged image loaded with image_check::bounds)
+            const int g = w - 1000000000000000u < 99000000000000000u ? locate_token(w, exponent[i], offset[i], radius[i]) : 3;
+            batch.digits[i] = mask_if(g <= 0) & ((last_four[i] >> 16u) + 1);
+            kind_1[count_1] = static_cast<std::uint8_t>(i);
+            count_1 += static_cast<std::size_t>(g == 1);
+            kind_2[count_2] = static_cast<std::uint8_t>(i);
+            count_2 += static_cast<std::size_t>(g == 2);
+        }
+        // the digits to write, in one loop per kind (see mark())
+        for (std::size_t k = 0; k != count_1; ++k)
+        {
+            const std::size_t i = kind_1[k];
+            const std::uint64_t s = token_digits_1(digits[i], offset[i], radius[i], last_four[i] & 0xFFu);
+            batch.digits[i] = mark(digits[i], s, last_four[i] >> 16u);
+        }
+        for (std::size_t k = 0; k != count_2; ++k)
+        {
+            const std::size_t i = kind_2[k];
+            const std::uint64_t s = token_digits_2(digits[i], offset[i], radius[i], last_four[i] & 0xFFu, (last_four[i] >> 8u) & 0xFFu);
+            batch.digits[i] = mark(digits[i], s, last_four[i] >> 16u);
+        }
+        return batch.count != 0 && batch.at[0] == n;
+    }
+
+    /*!
+    @brief a float token of 16 or 17 digits as dump() writes it, at w (64 bytes
+    of room), from the digits of its double (see locate_token()) whose
+    rounding carried beyond the token's last four digits
+
+    (like 43.42027299999999 -> 43.420273): the digits go to the writer of
+    Zmij's results as its 15 or 16 leading digits and one more, which omits
+    the trailing zeros.
+    */
+    NLOHMANN_VIEW_NOINLINE char* write_carried_digits(char* w, const node& n, std::uint64_t digits) const noexcept
+    {
+        const int exponent = -(n.extra >> 8u);
+        const std::uint64_t leading = digits / 10;
+        const auto last = static_cast<unsigned char>(digits - (leading * 10));
+        const bool seventeen = digits >= 10000000000000000u;
+        *w = '-';
+        w += m_doc.src[n.off] == '-' ? 1 : 0;
+        return ::nlohmann::detail::dtoa_impl::write_shortest(w, seventeen ? ::nlohmann::detail::zmij::shortest_decimal{leading, exponent, last, last != 0}
+                : ::nlohmann::detail::zmij::shortest_decimal{digits, exponent - 1, 0, false});
+    }
+
+    /// the digits s of a token with the digits w and the last four digits
+    /// r4, as the batch keeps them: the new last four digits + 1 (only those
+    /// are written over the token's), or s marked if it differs from w beyond
+    /// them (0 stays for a float the batch does not write)
+    static std::uint64_t mark(std::uint64_t w, std::uint64_t s, std::uint64_t r4) noexcept
+    {
+        const std::uint64_t last = s - (w - r4);
+        return last < 10000 ? last + 1 : s | carried();
+    }
+
+    /*!
+    @brief a float token of 16 or 17 digits as dump() writes it, at w (64 bytes
+    of room): the token with its last four digits replaced by digits - 1 (see
+    mark(); the others are the token's), and without trailing zeros after the
+    point, but one
+    */
+    char* write_token_digits(char* w, const char* src, const node& n, std::uint64_t digits)
+    {
+        const char* const token = src + n.off;
+        std::memcpy(w, token, 32);
+        char* const end = w + n.len;
+        const std::uint32_t last = four_digit_text(static_cast<std::uint32_t>(digits - 1));
+        std::memcpy(end - 4, &last, 4);
+        // (the token's text before the new digits, from the source: reading
+        // the output back would wait for the stores)
+        std::uint64_t before = 0;
+        std::memcpy(&before, token + n.len - 12, 8);
+        const std::uint32_t last_zeros = last ^ 0x30303030u;
+        const std::uint64_t before_zeros = before ^ 0x3030303030303030u;
+        if (NLOHMANN_VIEW_UNLIKELY(last_zeros == 0 && before_zeros == 0))
+        {
+            return write_double_at(w, n); // (twelve or more trailing zeros)
+        }
+        const unsigned zeros = last_zeros != 0 ? (static_cast<unsigned>(count_leading_zeros(last_zeros)) / 8) - 4
+                               : 4 + (static_cast<unsigned>(count_leading_zeros(before_zeros)) / 8);
+        const unsigned frac_digits = n.extra >> 8u;
+        return end - (zeros < frac_digits ? zeros : frac_digits - 1);
+    }
+#endif
+
     /// a float node as dump() writes it
     void write_float_node(const node& n)
     {

include/nlohmann/detail/view/simd.hpp

@@ -289,9 +289,12 @@ compares, and the UTF-8 check covers the bytes up to it. Returns where the
 string scan stops, like scan_string_run: before ill-formed UTF-8 and for the
 last bytes of the input, the bytes are checked one sequence at a time. Out of
 line, so that no constants of the check occupy registers in the parse loop.
-On x86-64, it is compiled for SSSE3 (see cpu_has_ssse3()).
+On x86-64, it is compiled for SSSE3 (see cpu_has_ssse3()). Not cloned: with
+the calls of accept()'s builder, GCC 16 would clone it without plain (always
+string_plain()), and calling the clone made json_document::parse() run 0.7-1.2 %
+more instructions.
 */
-NLOHMANN_VIEW_SSSE3_TARGET NLOHMANN_VIEW_NOINLINE inline const unsigned char* scan_string_vector(const unsigned char* p, const unsigned char* e, const std::uint8_t* plain) noexcept
+NLOHMANN_VIEW_SSSE3_TARGET NLOHMANN_VIEW_NOINLINE NLOHMANN_VIEW_NOCLONE inline const unsigned char* scan_string_vector(const unsigned char* p, const unsigned char* e, const std::uint8_t* plain) noexcept
 {
     using lookup = utf8_lookup4<>;
     const unsigned char* block = p;

include/nlohmann/detail/view/value.hpp

@@ -60,7 +60,7 @@ NLOHMANN_VIEW_ALWAYS_INLINE T arithmetic_value(const document_data& d, const nod
         case value_t::binary:
         case value_t::discarded:
         default:
-            throw_type_error(exception_id::type_mismatch, "type must be number, but is ", value_type_name(static_cast<value_t>(n.kind)));
+            throw_type_error(error_site(&d, &n), exception_id::type_mismatch, "type must be number, but is ", value_type_name(static_cast<value_t>(n.kind)));
     }
 }
 
@@ -70,7 +70,7 @@ std::vector<U, A> vector_value(const View& v)
 {
     if (NLOHMANN_VIEW_UNLIKELY(!v.is_array()))
     {
-        throw_type_error(exception_id::type_mismatch, "type must be array, but is ", v.type_name());
+        throw_type_error(site_of(v), exception_id::type_mismatch, "type must be array, but is ", v.type_name());
     }
     std::vector<U, A> r;
     r.reserve(v.size());
@@ -88,7 +88,7 @@ Map map_value(const View& v)
 {
     if (NLOHMANN_VIEW_UNLIKELY(!v.is_object()))
     {
-        throw_type_error(exception_id::type_mismatch, "type must be object, but is ", v.type_name());
+        throw_type_error(site_of(v), exception_id::type_mismatch, "type must be object, but is ", v.type_name());
     }
     Map r;
     for (auto it = v.begin(); it != v.end(); ++it)

include/nlohmann/json.hpp

@@ -150,14 +150,14 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
     // can be restored when json_pointer backwards compatibility is removed
     // friend ::nlohmann::json_pointer<StringType>;
 
-    template<typename BasicJsonType, typename InputType>
+    template<typename BasicJsonType, typename InputType, bool DiscardValues>
     friend class ::nlohmann::detail::parser;
     friend ::nlohmann::detail::serializer<basic_json>;
     template<typename BasicJsonType>
     friend class ::nlohmann::detail::iter_impl;
     template<typename BasicJsonType, typename CharType, typename OutputSinkType>
     friend class ::nlohmann::detail::binary_writer;
-    template<typename BasicJsonType, typename InputType, typename SAX>
+    template<typename BasicJsonType, typename InputType, typename SAX, bool AllowRecovery>
     friend class ::nlohmann::detail::binary_reader;
     template<typename BasicJsonType, typename InputAdapterType, bool Collect, typename LexerType>
     friend class ::nlohmann::detail::json_sax_dom_parser;
@@ -183,12 +183,11 @@ class basic_json // NOLINT(cppcoreguidelines-special-member-functions,hicpp-spec
         detail::parser_callback_t<basic_json>cb = nullptr,
         const bool allow_exceptions = true,
         const bool ignore_comments = false,
-        const bool ignore_trailing_commas = false,
+        const bool ignore_trailing_commas = false
-        const bool discard_number_values = false
     )
     {
         return ::nlohmann::detail::parser<basic_json, InputAdapterType>(std::move(adapter),
-               std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas, discard_number_values);
+               std::move(cb), allow_exceptions, ignore_comments, ignore_trailing_commas);
     }
 
   private:
@@ -3117,7 +3116,9 @@ public:
                             const bool ignore_comments,
                             const bool ignore_trailing_commas)
     {
-        return parser(std::forward<InputAdapterType>(adapter), nullptr, false, ignore_comments, ignore_trailing_commas, true).accept(true);
+        // the parser and lexer that discard the values of tokens (see lexer)
+        return ::nlohmann::detail::parser<basic_json, typename std::decay<InputAdapterType>::type, true>(std::forward<InputAdapterType>(adapter),
+                nullptr, false, ignore_comments, ignore_trailing_commas).accept(true);
     }
 
     /// generate SAX events from the input behind an already created adapter
@@ -3133,7 +3134,8 @@ public:
         {
             return parser(std::move(ia), nullptr, true, ignore_comments, ignore_trailing_commas).sax_parse(sax, strict);
         }
-        return detail::binary_reader<basic_json, InputAdapterType, SAX>(std::move(ia), format).sax_parse(sax, strict, tag_handler);
+        // the SAX parser may recover from errors (see parse_error())
+        return detail::binary_reader<basic_json, InputAdapterType, SAX, true>(std::move(ia), format).sax_parse(sax, strict, tag_handler);
     }
 
     /// serialize @a j into a new vector; @a write receives the vector-backed writer

include/nlohmann/json_view.hpp

@@ -285,7 +285,7 @@ class basic_json_view
             {
                 return {};
             }
-            detail::view::throw_type_error(detail::exception_id::subscript_wrong_type, "cannot use operator[] with a string argument with ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::subscript_wrong_type, "cannot use operator[] with a string argument with ", type_name());
         }
         return lookup(key);
     }
@@ -311,7 +311,7 @@ class basic_json_view
             {
                 return {};
             }
-            detail::view::throw_type_error(detail::exception_id::subscript_wrong_type, "cannot use operator[] with a numeric argument with ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::subscript_wrong_type, "cannot use operator[] with a numeric argument with ", type_name());
         }
         return idx < m_node->len ? element(idx) : basic_json_view();
     }
@@ -344,12 +344,12 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_object()))
         {
-            detail::view::throw_type_error(detail::exception_id::at_wrong_type, "cannot use at() with ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::at_wrong_type, "cannot use at() with ", type_name());
         }
         const basic_json_view r = lookup(key);
         if (NLOHMANN_VIEW_UNLIKELY(r.is_discarded()))
         {
-            detail::view::throw_key_not_found(std::string(key.data(), key.size()));
+            detail::view::throw_key_not_found(site(), std::string(key.data(), key.size()));
         }
         return r;
     }
@@ -370,11 +370,11 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_array()))
         {
-            detail::view::throw_type_error(detail::exception_id::at_wrong_type, "cannot use at() with ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::at_wrong_type, "cannot use at() with ", type_name());
         }
         if (NLOHMANN_VIEW_UNLIKELY(idx >= m_node->len))
         {
-            detail::view::throw_array_index_out_of_range(idx);
+            detail::view::throw_array_index_out_of_range(site(), idx);
         }
         return element(idx);
     }
@@ -402,7 +402,7 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_object()))
         {
-            detail::view::throw_type_error(detail::exception_id::value_wrong_type, "cannot use value() with ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::value_wrong_type, "cannot use value() with ", type_name());
         }
         const basic_json_view r = lookup(key);
         return r.is_discarded() ? default_value : r.template get<T>();
@@ -412,7 +412,7 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_object()))
         {
-            detail::view::throw_type_error(detail::exception_id::value_wrong_type, "cannot use value() with ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::value_wrong_type, "cannot use value() with ", type_name());
         }
         // (the default is made only if there is no such member)
         const basic_json_view r = lookup(key);
@@ -427,7 +427,7 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_structured()))
         {
-            detail::view::throw_type_error(detail::exception_id::value_wrong_type, "cannot use value() with ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::value_wrong_type, "cannot use value() with ", type_name());
         }
         const basic_json_view r = detail::view::resolve_pointer(*this, detail::json_pointer_access::reference_tokens(ptr), detail::view::pointer_mode::value);
         return r.is_discarded() ? default_value : r.template get<T>();
@@ -437,7 +437,7 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_structured()))
         {
-            detail::view::throw_type_error(detail::exception_id::value_wrong_type, "cannot use value() with ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::value_wrong_type, "cannot use value() with ", type_name());
         }
         const basic_json_view r = detail::view::resolve_pointer(*this, detail::json_pointer_access::reference_tokens(ptr), detail::view::pointer_mode::value);
         return r.is_discarded() ? string_t(default_value) : r.template get<string_t>();
@@ -450,7 +450,7 @@ class basic_json_view
         const iterator it = begin();
         if (NLOHMANN_VIEW_UNLIKELY(it == end()))
         {
-            detail::view::throw_invalid_iterator(detail::exception_id::iterator_value_unavailable, "cannot get value");
+            detail::view::throw_invalid_iterator(site(), detail::exception_id::iterator_value_unavailable, "cannot get value");
         }
         return *it;
     }
@@ -606,7 +606,7 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_string()))
         {
-            detail::view::throw_type_error(detail::exception_id::type_mismatch, "type must be string, but is ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::type_mismatch, "type must be string, but is ", type_name());
         }
         return {m_doc->str(*m_node), m_node->len};
     }
@@ -618,7 +618,7 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_number()))
         {
-            detail::view::throw_type_error(detail::exception_id::type_mismatch, "type must be number, but is ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::type_mismatch, "type must be number, but is ", type_name());
         }
         return {m_doc->str(*m_node), detail::view::number_length(*m_node)};
     }
@@ -764,15 +764,19 @@ class basic_json_view
     template<typename, view_mode> friend class basic_json_view;
     template<typename, typename> friend class detail::view::editor;
     template<typename, typename> friend class detail::view::view_side;
+    friend struct detail::view::view_access;
     friend iterator;
 
-    /// whether the view is one of an editable document
-    static constexpr bool is_editable = Editable;
-
     basic_json_view(const document_data* d, const node* n) noexcept
         : m_doc(d), m_node(n)
     {}
 
+    /// this value as the context of an exception
+    detail::view::error_site site() const noexcept
+    {
+        return {m_doc, m_node};
+    }
+
     using json_side_t = detail::view::json_side<BasicJsonType, string_view_t>;
 
     static detail::view::view_side<BasicJsonType, basic_json_view> side(const basic_json_view& v) noexcept
@@ -879,7 +883,7 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_boolean()))
         {
-            detail::view::throw_type_error(detail::exception_id::type_mismatch, "type must be boolean, but is ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::type_mismatch, "type must be boolean, but is ", type_name());
         }
         return (m_node->flags & detail::view::node_flags::is_true) != 0;
     }
@@ -889,7 +893,7 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(m_node == nullptr))
         {
-            detail::view::throw_type_error(detail::exception_id::type_mismatch, "type must be number, but is ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::type_mismatch, "type must be number, but is ", type_name());
         }
         return detail::view::arithmetic_value<T, BasicJsonType>(*m_doc, *m_node);
     }
@@ -898,7 +902,7 @@ class basic_json_view
     {
         if (NLOHMANN_VIEW_UNLIKELY(!is_null()))
         {
-            detail::view::throw_type_error(detail::exception_id::type_mismatch, "type must be null, but is ", type_name());
+            detail::view::throw_type_error(site(), detail::exception_id::type_mismatch, "type must be null, but is ", type_name());
         }
         return nullptr;
     }
@@ -1063,7 +1067,7 @@ class basic_json_document
                                      const bool ignore_trailing_commas = false)
     {
         basic_json_document d;
-        d.read_range(first, last, allow_exceptions, ignore_comments, ignore_trailing_commas);
+        d.read_range(first, last, error_mode(allow_exceptions), ignore_comments, ignore_trailing_commas);
         return d;
     }
 
@@ -1078,7 +1082,7 @@ class basic_json_document
         basic_json_document d;
         std::string buf;
         append_input(buf, std::forward<InputType>(input));
-        d.build_owned(std::move(buf), allow_exceptions, ignore_comments, ignore_trailing_commas);
+        d.build_owned(std::move(buf), error_mode(allow_exceptions), ignore_comments, ignore_trailing_commas);
         return d;
     }
 
@@ -1091,7 +1095,8 @@ class basic_json_document
     static bool accept(InputType&& input, const bool ignore_comments = false, const bool ignore_trailing_commas = false)
     {
         basic_json_document d;
-        d.read(std::forward<InputType>(input), false, ignore_comments, ignore_trailing_commas);
+        d.read_kind(std::forward<InputType>(input), on_error::validate, ignore_comments, ignore_trailing_commas,
+                    std::integral_constant<detail::view::input_kind, detail::view::classify_input<InputType>::value> {});
         return !d.is_discarded();
     }
 
@@ -1107,7 +1112,7 @@ class basic_json_document
               const bool ignore_comments = false,
               const bool ignore_trailing_commas = false)
     {
-        read_kind(std::forward<InputType>(input), allow_exceptions, ignore_comments, ignore_trailing_commas,
+        read_kind(std::forward<InputType>(input), error_mode(allow_exceptions), ignore_comments, ignore_trailing_commas,
                   std::integral_constant<detail::view::input_kind, detail::view::classify_input<InputType>::value> {});
     }
 
@@ -1312,7 +1317,7 @@ class basic_json_document
     template<typename I, typename V, typename std::enable_if<detail::view::is_non_bool_integral<I>::value, int>::type = 0>
     view_type set(view_type array, I idx, V && value)
     {
-        return editor().set(array, index(idx), std::forward<V>(value));
+        return editor().set(array, index(array, idx), std::forward<V>(value));
     }
 
     /// set the value at a JSON pointer: its parent must exist; an object
@@ -1349,7 +1354,7 @@ class basic_json_document
                 const std::size_t idx = pointer_index(token);
                 if (idx >= 0xFFFFFFFFu)
                 {
-                    detail::view::throw_array_index_out_of_range(idx);
+                    detail::view::throw_array_index_out_of_range(detail::view::site_of(parent), idx);
                 }
                 for (std::size_t i = 0; i < idx; ++i)
                 {
@@ -1383,7 +1388,7 @@ class basic_json_document
     template<typename I, typename V, typename std::enable_if<detail::view::is_non_bool_integral<I>::value, int>::type = 0>
     view_type insert(view_type array, I idx, V && value)
     {
-        return editor().insert(array, index(idx), std::forward<V>(value));
+        return editor().insert(array, index(array, idx), std::forward<V>(value));
     }
 
     /// remove all members with this key; returns their number
@@ -1396,7 +1401,7 @@ class basic_json_document
     template<typename I, typename std::enable_if<detail::view::is_non_bool_integral<I>::value, int>::type = 0>
     void erase(view_type array, I idx)
     {
-        editor().erase(array, index(idx));
+        editor().erase(array, index(array, idx));
     }
 
     /// remove the value at a JSON pointer; returns the number of removed
@@ -1430,25 +1435,25 @@ class basic_json_document
         return detail::view::editor<BasicJsonType, view_type>(*m_data);
     }
 
-    /// an index (out_of_range.401 if negative)
+    /// an index into @a array (out_of_range.401 if negative)
     template<typename I>
-    static std::size_t index(I idx)
+    static std::size_t index(const view_type& array, I idx)
     {
-        return index(idx, std::is_signed<I> {});
+        return index(array, idx, std::is_signed<I> {});
     }
 
     template<typename I>
-    static std::size_t index(I idx, std::true_type /*signed*/)
+    static std::size_t index(const view_type& array, I idx, std::true_type /*signed*/)
     {
         if (idx < 0)
         {
-            detail::view::throw_array_index_out_of_range(idx);
+            detail::view::throw_array_index_out_of_range(detail::view::site_of(array), idx);
         }
         return static_cast<std::size_t>(idx);
     }
 
     template<typename I>
-    static std::size_t index(I idx, std::false_type /*unsigned*/)
+    static std::size_t index(const view_type& /*array*/, I idx, std::false_type /*unsigned*/)
     {
         return static_cast<std::size_t>(idx);
     }
@@ -1466,6 +1471,22 @@ class basic_json_document
         return idx;
     }
 
+    /// what a failed parse does
+    enum class on_error : std::uint8_t
+    {
+        discard, ///< the document is discarded
+        raise,   ///< parse_error is thrown
+        validate ///< accept(): the document is discarded, and no node is stored while parsing
+    };
+
+    static on_error error_mode(bool allow_exceptions) noexcept
+    {
+        return allow_exceptions ? on_error::raise : on_error::discard;
+    }
+
+    /// NulIsEnd of the builder: a null byte ends the input unless strict_nul_handling
+    static constexpr bool nul_is_end = !detail::abi_config::strict_nul_handling;
+
     /// create the storage (sized for the input) on first use
     void ensure_data(const char* src, std::size_t size)
     {
@@ -1477,7 +1498,7 @@ class basic_json_document
 
     /// parse a buffer the document takes ownership of (it replaces the
     /// previous owned copy)
-    void build_owned(std::string&& buf, bool allow_exceptions, bool comments, bool trailing_commas)
+    void build_owned(std::string&& buf, on_error allow_exceptions, bool comments, bool trailing_commas)
     {
         ensure_data(buf.data(), buf.size());
         m_data->owned = std::move(buf);
@@ -1487,7 +1508,7 @@ class basic_json_document
     /// parse a copy of the input; the copy is appended to the cleared owned
     /// buffer, so that its capacity is reused
     template<typename T>
-    void build_copy(T&& input, bool allow_exceptions, bool comments, bool trailing_commas)
+    void build_copy(T&& input, on_error allow_exceptions, bool comments, bool trailing_commas)
     {
         if (!m_data)
         {
@@ -1508,7 +1529,7 @@ class basic_json_document
     }
 
     /// sentinel: src[size] is readable and 0 (std::string, C strings)
-    void build(const char* src, std::size_t size, bool allow_exceptions, bool comments, bool trailing_commas, bool owned, bool sentinel)
+    void build(const char* src, std::size_t size, on_error allow_exceptions, bool comments, bool trailing_commas, bool owned, bool sentinel)
     {
         ensure_data(src, size);
         document_data& d = *m_data;
@@ -1535,7 +1556,8 @@ class basic_json_document
         }
         else
         {
-            ok = detail::view::build < typename BasicJsonType::number_float_t, !detail::abi_config::strict_nul_handling > (d, src, size, comments, trailing_commas, sentinel, failure);
+            ok = detail::view::build<typename BasicJsonType::number_float_t, nul_is_end>(d, src, size, comments, trailing_commas, sentinel,
+                 allow_exceptions == on_error::validate, failure);
         }
         if (NLOHMANN_VIEW_LIKELY(ok))
         {
@@ -1548,7 +1570,7 @@ class basic_json_document
             return;
         }
         std::vector<std::uint32_t>().swap(d.large_objects);
-        if (allow_exceptions)
+        if (allow_exceptions == on_error::raise)
         {
             detail::view::throw_parse_failure<BasicJsonType>(failure, src, size, comments, trailing_commas);
         }
@@ -1556,13 +1578,13 @@ class basic_json_document
 
     // --- input dispatch (see detail::view::input_kind) ---
 
-    void read_kind(std::string&& s, bool ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::move_string> /*unused*/)
+    void read_kind(std::string&& s, on_error ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::move_string> /*unused*/)
     {
         build_owned(std::move(s), ae, c, tc);
     }
 
     template<typename CharT>
-    void read_kind(CharT* s, bool ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::c_string> /*unused*/)
+    void read_kind(CharT* s, on_error ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::c_string> /*unused*/)
     {
         if (s == nullptr)
         {
@@ -1576,7 +1598,7 @@ class basic_json_document
     }
 
     template<typename Array>
-    void read_kind(Array& a, bool ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::char_array> /*unused*/)
+    void read_kind(Array& a, on_error ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::char_array> /*unused*/)
     {
         using CharT = typename std::remove_cv<typename std::remove_extent<Array>::type>::type;
         const std::size_t n = std::extent<Array>::value;
@@ -1587,7 +1609,7 @@ class basic_json_document
     }
 
     template<typename T>
-    void read_kind(const T& s, bool ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::borrow_range> /*unused*/)
+    void read_kind(const T& s, on_error ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::borrow_range> /*unused*/)
     {
         // std::basic_string guarantees data()[size()] == 0: use it as sentinel
         const auto size = static_cast<std::size_t>(s.size());
@@ -1596,13 +1618,13 @@ class basic_json_document
     }
 
     template<typename T>
-    void read_kind(const T& s, bool ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::copy_range> /*unused*/)
+    void read_kind(const T& s, on_error ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::copy_range> /*unused*/)
     {
         build_copy(s, ae, c, tc);
     }
 
     template<typename CharT>
-    void read_kind(CharT* s, bool ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::wide_c_string> /*unused*/)
+    void read_kind(CharT* s, on_error ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::wide_c_string> /*unused*/)
     {
         if (s == nullptr)
         {
@@ -1614,13 +1636,13 @@ class basic_json_document
     }
 
     template<typename T>
-    void read_kind(T&& input, bool ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::adapter> /*unused*/)
+    void read_kind(T&& input, on_error ae, bool c, bool tc, std::integral_constant<input_kind, input_kind::adapter> /*unused*/)
     {
         build_copy(std::forward<T>(input), ae, c, tc);
     }
 
     template<typename IteratorType>
-    void read_range(IteratorType first, IteratorType last, bool ae, bool c, bool tc)
+    void read_range(IteratorType first, IteratorType last, on_error ae, bool c, bool tc)
     {
         using value_type = typename std::remove_cv<typename std::iterator_traits<IteratorType>::value_type>::type;
         // borrow the range where the library's input adapter scans it as one
@@ -1631,14 +1653,14 @@ class basic_json_document
     }
 
     template<typename IteratorType>
-    void read_range_impl(IteratorType first, IteratorType last, bool ae, bool c, bool tc, std::true_type /*contiguous bytes*/)
+    void read_range_impl(IteratorType first, IteratorType last, on_error ae, bool c, bool tc, std::true_type /*contiguous bytes*/)
     {
         const auto size = static_cast<std::size_t>(std::distance(first, last));
         build(size == 0 ? "" : reinterpret_cast<const char*>(&*first), size, ae, c, tc, false, size == 0); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
     }
 
     template<typename IteratorType>
-    void read_range_impl(IteratorType first, IteratorType last, bool ae, bool c, bool tc, std::false_type /*other*/)
+    void read_range_impl(IteratorType first, IteratorType last, on_error ae, bool c, bool tc, std::false_type /*other*/)
     {
         build_copy(detail::view::make_adapter_input(detail::input_adapter(first, last)), ae, c, tc);
     }

tests/src/fuzzer-parse_bjdata.cpp

@@ -47,6 +47,10 @@ dumps is stable under exactly the same values that break operator==.
 The unit tests run the same checks on a fixed corpus (see the "BJData round-trip
 invariants" test case), so keep both in sync.
 
+Furthermore, it reads data with a SAX parser that recovers from every error
+and checks that the events are balanced, that reading ends, and that it
+reports an error exactly when from_bjdata() fails (see #3989).
+
 Finally, it reads the data from a stream, a non-contiguous iterator range, and
 a FILE*, which read byte by byte, and checks that each gives the same value or
 error (id, message, byte position) as the pointer range, which takes the bulk
@@ -57,6 +61,7 @@ drivers.
 */
 
 #include "fuzzer_common.hpp"
+#include "fuzzer-recovering_checker.hpp"
 
 namespace
 {
@@ -81,6 +86,9 @@ static bool is_value_stable(const json& lhs, const json& rhs)
 // see http://llvm.org/docs/LibFuzzer.html
 extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
 {
+    // recover from all errors, reading from memory and from a stream
+    const bool recovered_without_errors = check_recovering_parse(data, size, json::input_format_t::bjdata).errors == 0;
+
     std::vector<uint8_t> const vec1(data, data + size);
 
     // step 0: parse input without exceptions
@@ -99,6 +107,9 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
         // without exceptions, the same input must give the same value
         assert(!noexcept_threw && !j_noexcept.is_discarded() && same_value(j_noexcept, j1));
 
+        // the recovering parser must not have reported an error either
+        assert(recovered_without_errors);
+
         try
         {
             // step 2.1: round trip without adding size annotations to container types
@@ -132,6 +143,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // parse errors are ok, because input may be random bytes
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
     catch (const json::type_error&)
     {
@@ -142,6 +154,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // out of range errors may happen if provided sizes are excessive
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
 
     // step 3: the other input adapters must read the data alike

tests/src/fuzzer-parse_bon8.cpp

@@ -17,6 +17,10 @@ array data, it performs the following steps:
 - j2 = from_bon8(vec)
 - assert(to_bon8(j2) == vec)
 
+Furthermore, it reads data with a SAX parser that recovers from every error
+and checks that the events are balanced, that reading ends, and that it
+reports an error exactly when from_bon8() fails (see #3989).
+
 Finally, it reads the data from a stream, a non-contiguous iterator range, and
 a FILE*, which read byte by byte, and checks that each gives the same value or
 error (id, message, byte position) as the pointer range, which takes the bulk
@@ -27,6 +31,7 @@ drivers.
 */
 
 #include "fuzzer_common.hpp"
+#include "fuzzer-recovering_checker.hpp"
 
 namespace
 {
@@ -44,6 +49,9 @@ struct read_bon8
 // see http://llvm.org/docs/LibFuzzer.html
 extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
 {
+    // recover from all errors, reading from memory and from a stream
+    const bool recovered_without_errors = check_recovering_parse(data, size, json::input_format_t::bon8).errors == 0;
+
     std::vector<uint8_t> const vec1(data, data + size);
 
     // step 0: parse input without exceptions
@@ -62,6 +70,9 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
         // without exceptions, the same input must give the same value
         assert(!noexcept_threw && !j_noexcept.is_discarded() && same_value(j_noexcept, j1));
 
+        // the recovering parser must not have reported an error either
+        assert(recovered_without_errors);
+
         try
         {
             // step 2: round trip
@@ -83,6 +94,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // parse errors are ok, because input may be random bytes
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
     catch (const json::type_error&)
     {
@@ -93,6 +105,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // out of range errors may happen if provided sizes are excessive
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
 
     // step 3: the other input adapters must read the data alike

tests/src/fuzzer-parse_bson.cpp

@@ -17,6 +17,10 @@ array data, it performs the following steps:
 - j2 = from_bson(vec)
 - assert(to_bson(j2) == vec)
 
+Furthermore, it reads data with a SAX parser that recovers from every error
+and checks that the events are balanced, that reading ends, and that it
+reports an error exactly when from_bson() fails (see #3989).
+
 Finally, it reads the data from a stream, a non-contiguous iterator range, and
 a FILE*, which read byte by byte, and checks that each gives the same value or
 error (id, message, byte position) as the pointer range, which takes the bulk
@@ -27,6 +31,7 @@ drivers.
 */
 
 #include "fuzzer_common.hpp"
+#include "fuzzer-recovering_checker.hpp"
 
 namespace
 {
@@ -44,6 +49,9 @@ struct read_bson
 // see http://llvm.org/docs/LibFuzzer.html
 extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
 {
+    // recover from all errors, reading from memory and from a stream
+    const bool recovered_without_errors = check_recovering_parse(data, size, json::input_format_t::bson).errors == 0;
+
     std::vector<uint8_t> const vec1(data, data + size);
 
     // step 0: parse input without exceptions
@@ -62,6 +70,9 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
         // without exceptions, the same input must give the same value
         assert(!noexcept_threw && !j_noexcept.is_discarded() && same_value(j_noexcept, j1));
 
+        // the recovering parser must not have reported an error either
+        assert(recovered_without_errors);
+
         try
         {
             // step 2: round trip
@@ -83,6 +94,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // parse errors are ok, because input may be random bytes
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
     catch (const json::type_error&)
     {
@@ -93,6 +105,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // out of range errors can occur during parsing, too
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
 
     // step 3: the other input adapters must read the data alike

tests/src/fuzzer-parse_cbor.cpp

@@ -17,6 +17,10 @@ array data, it performs the following steps:
 - j2 = from_cbor(vec)
 - assert(to_cbor(j2) == vec)
 
+Furthermore, it reads data with a SAX parser that recovers from every error
+and checks that the events are balanced, that reading ends, and that it
+reports an error exactly when from_cbor() fails (see #3989).
+
 Finally, it reads the data from a stream, a non-contiguous iterator range, and
 a FILE*, which read byte by byte, and checks that each gives the same value or
 error (id, message, byte position) as the pointer range, which takes the bulk
@@ -27,6 +31,7 @@ drivers.
 */
 
 #include "fuzzer_common.hpp"
+#include "fuzzer-recovering_checker.hpp"
 
 namespace
 {
@@ -44,6 +49,9 @@ struct read_cbor
 // see http://llvm.org/docs/LibFuzzer.html
 extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
 {
+    // recover from all errors, reading from memory and from a stream
+    const bool recovered_without_errors = check_recovering_parse(data, size, json::input_format_t::cbor).errors == 0;
+
     std::vector<uint8_t> const vec1(data, data + size);
 
     // step 0: parse input without exceptions
@@ -62,6 +70,9 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
         // without exceptions, the same input must give the same value
         assert(!noexcept_threw && !j_noexcept.is_discarded() && same_value(j_noexcept, j1));
 
+        // the recovering parser must not have reported an error either
+        assert(recovered_without_errors);
+
         try
         {
             // step 2: round trip
@@ -83,6 +94,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // parse errors are ok, because input may be random bytes
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
     catch (const json::type_error&)
     {
@@ -93,6 +105,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // out of range errors can occur during parsing, too
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
 
     // step 3: the other input adapters must read the data alike

tests/src/fuzzer-parse_json.cpp

@@ -24,11 +24,16 @@ array data, it performs the following steps:
 The pointer range takes the bulk fast paths, the other input adapters read byte
 by byte; see check_input_adapters().
 
+Furthermore, it parses data with a SAX parser that recovers from every error
+and checks that the events are balanced, that parsing ends, and that valid
+input is parsed without errors (see #3989).
+
 The provided function `LLVMFuzzerTestOneInput` can be used in different fuzzer
 drivers.
 */
 
 #include "fuzzer_common.hpp"
+#include "fuzzer-recovering_checker.hpp"
 
 namespace
 {
@@ -59,6 +64,13 @@ struct accept_json
 // see http://llvm.org/docs/LibFuzzer.html
 extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
 {
+    // recover from all errors, reading from memory and from a stream
+    {
+        const auto checker = check_recovering_parse(data, size, json::input_format_t::json);
+        assert(checker.events <= (4 * size) + 4);
+        assert((checker.errors == 0) == json::accept(data, data + size));
+    }
+
     // step 0: parse input without exceptions
     bool noexcept_threw = false;
     json const j_noexcept = parse_without_exceptions([&] { return json::parse(data, data + size, nullptr, false); }, noexcept_threw);

tests/src/fuzzer-parse_msgpack.cpp

@@ -17,6 +17,10 @@ array data, it performs the following steps:
 - j2 = from_msgpack(vec)
 - assert(to_msgpack(j2) == vec)
 
+Furthermore, it reads data with a SAX parser that recovers from every error
+and checks that the events are balanced, that reading ends, and that it
+reports an error exactly when from_msgpack() fails (see #3989).
+
 Finally, it reads the data from a stream, a non-contiguous iterator range, and
 a FILE*, which read byte by byte, and checks that each gives the same value or
 error (id, message, byte position) as the pointer range, which takes the bulk
@@ -27,6 +31,7 @@ drivers.
 */
 
 #include "fuzzer_common.hpp"
+#include "fuzzer-recovering_checker.hpp"
 
 namespace
 {
@@ -44,6 +49,9 @@ struct read_msgpack
 // see http://llvm.org/docs/LibFuzzer.html
 extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
 {
+    // recover from all errors, reading from memory and from a stream
+    const bool recovered_without_errors = check_recovering_parse(data, size, json::input_format_t::msgpack).errors == 0;
+
     std::vector<uint8_t> const vec1(data, data + size);
 
     // step 0: parse input without exceptions
@@ -62,6 +70,9 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
         // without exceptions, the same input must give the same value
         assert(!noexcept_threw && !j_noexcept.is_discarded() && same_value(j_noexcept, j1));
 
+        // the recovering parser must not have reported an error either
+        assert(recovered_without_errors);
+
         try
         {
             // step 2: round trip
@@ -83,6 +94,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // parse errors are ok, because input may be random bytes
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
     catch (const json::type_error&)
     {
@@ -93,6 +105,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // out of range errors may happen if provided sizes are excessive
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
 
     // step 3: the other input adapters must read the data alike

tests/src/fuzzer-parse_ubjson.cpp

@@ -26,6 +26,10 @@ array data, it performs the following steps:
 The unit tests run the same checks on a fixed corpus (see the "UBJSON round-trip
 invariants" test case), so keep both in sync.
 
+Furthermore, it reads data with a SAX parser that recovers from every error
+and checks that the events are balanced, that reading ends, and that it
+reports an error exactly when from_ubjson() fails (see #3989).
+
 Finally, it reads the data from a stream, a non-contiguous iterator range, and
 a FILE*, which read byte by byte, and checks that each gives the same value or
 error (id, message, byte position) as the pointer range, which takes the bulk
@@ -36,6 +40,7 @@ drivers.
 */
 
 #include "fuzzer_common.hpp"
+#include "fuzzer-recovering_checker.hpp"
 
 namespace
 {
@@ -53,6 +58,9 @@ struct read_ubjson
 // see http://llvm.org/docs/LibFuzzer.html
 extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
 {
+    // recover from all errors, reading from memory and from a stream
+    const bool recovered_without_errors = check_recovering_parse(data, size, json::input_format_t::ubjson).errors == 0;
+
     std::vector<uint8_t> const vec1(data, data + size);
 
     // step 0: parse input without exceptions
@@ -71,6 +79,9 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
         // without exceptions, the same input must give the same value
         assert(!noexcept_threw && !j_noexcept.is_discarded() && same_value(j_noexcept, j1));
 
+        // the recovering parser must not have reported an error either
+        assert(recovered_without_errors);
+
         try
         {
             // step 2.1: round trip without adding size annotations to container types
@@ -102,6 +113,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // parse errors are ok, because input may be random bytes
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
     catch (const json::type_error&)
     {
@@ -112,6 +124,7 @@ extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size)
     {
         // out of range errors may happen if provided sizes are excessive
         assert(parsed || noexcept_threw || j_noexcept.is_discarded());
+        assert(parsed || !recovered_without_errors);
     }
 
     // step 3: the other input adapters must read the data alike

tests/src/fuzzer-recovering_checker.hpp

@@ -0,0 +1,154 @@
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++ (supporting code)
+// |  |  |__   |  |  | | | |  version 3.12.0
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013-2026 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+#pragma once
+
+#include <cassert>
+#include <cstddef>
+#include <cstdint>
+#include <sstream>
+#include <string>
+#include <vector>
+#include <nlohmann/json.hpp>
+
+namespace
+{
+// a SAX parser that recovers from every error and checks that the events are
+// balanced and that every key is followed by exactly one value
+class recovering_checker : public nlohmann::json_sax<nlohmann::json>
+{
+  public:
+    bool null() override
+    {
+        return value();
+    }
+
+    bool boolean(bool /*val*/) override
+    {
+        return value();
+    }
+
+    bool number_integer(number_integer_t /*val*/) override
+    {
+        return value();
+    }
+
+    bool number_unsigned(number_unsigned_t /*val*/) override
+    {
+        return value();
+    }
+
+    bool number_float(number_float_t /*val*/, const string_t& /*s*/) override
+    {
+        return value();
+    }
+
+    bool string(string_t& /*val*/) override
+    {
+        return value();
+    }
+
+    bool binary(binary_t& /*val*/) override
+    {
+        return value();
+    }
+
+    bool start_object(std::size_t /*elements*/) override
+    {
+        value();
+        stack.push_back('o');
+        return true;
+    }
+
+    bool key(string_t& /*val*/) override
+    {
+        ++events;
+        assert(!stack.empty() && stack.back() == 'o');
+        stack.back() = 'v';
+        return true;
+    }
+
+    bool end_object() override
+    {
+        ++events;
+        assert(!stack.empty() && stack.back() == 'o');
+        stack.pop_back();
+        return true;
+    }
+
+    bool start_array(std::size_t /*elements*/) override
+    {
+        value();
+        stack.push_back('a');
+        return true;
+    }
+
+    bool end_array() override
+    {
+        ++events;
+        assert(!stack.empty() && stack.back() == 'a');
+        stack.pop_back();
+        return true;
+    }
+
+    bool parse_error(std::size_t /*position*/, const std::string& /*last_token*/, const nlohmann::detail::exception& /*ex*/) override
+    {
+        ++errors;
+        return true;
+    }
+
+    bool complete() const
+    {
+        return stack.empty();
+    }
+
+    std::size_t events = 0;
+    std::size_t errors = 0;
+
+  private:
+    bool value()
+    {
+        ++events;
+        if (!stack.empty())
+        {
+            // an array element, or the value of a key
+            assert(stack.back() != 'o');
+            if (stack.back() == 'v')
+            {
+                stack.back() = 'o';
+            }
+        }
+        return true;
+    }
+
+    // 'a' for an array, 'o' for an object that expects a key, 'v' for an
+    // object that expects the value of a key
+    std::vector<char> stack {}; // NOLINT(readability-redundant-member-init)
+};
+
+/// parses @a data with a recovering_checker from memory and from a stream,
+/// checks that both see the same, that the events are balanced, and that the
+/// number of errors is bounded, and returns the checker (see #3989)
+inline recovering_checker check_recovering_parse(const std::uint8_t* data, const std::size_t size, const nlohmann::json::input_format_t format)
+{
+    recovering_checker checker;
+    const bool ok = nlohmann::json::sax_parse(data, data + size, &checker, format);
+    assert(checker.complete());
+    assert(checker.errors <= size + 1);
+    assert(ok == (checker.errors == 0));
+
+    std::istringstream stream(std::string(reinterpret_cast<const char*>(data), size));
+    recovering_checker stream_checker;
+    assert(nlohmann::json::sax_parse(stream, &stream_checker, format) == ok);
+    assert(stream_checker.complete());
+    assert(stream_checker.events == checker.events);
+    assert(stream_checker.errors == checker.errors);
+
+    return checker;
+}
+} // namespace

tests/src/json_view_test_helpers.hpp

@@ -15,6 +15,7 @@
 #include <nlohmann/json_view.hpp>
 
 #include <algorithm> // any_of
+#include <cstddef> // size_t
 #include <functional> // function
 #include <random> // mt19937
 #include <string> // string
@@ -88,6 +89,27 @@ inline bool has_duplicate_keys(const nlohmann::ordered_json_view& v)
     });
 }
 
+// the message of an exception about the value at path (a JSON pointer, "" for
+// the root) whose source bytes are [start, end): the context basic_json adds
+// with JSON_DIAGNOSTICS and JSON_DIAGNOSTIC_POSITIONS, if enabled, then what
+inline std::string message(const std::string& id, const std::string& path, std::size_t start, std::size_t end, const std::string& what)
+{
+#if JSON_DIAGNOSTIC_POSITIONS
+    const std::string positions = "(bytes " + std::to_string(start) + "-" + std::to_string(end) + ") ";
+#else
+    const std::string positions;
+    static_cast<void>(start);
+    static_cast<void>(end);
+#endif
+#if JSON_DIAGNOSTICS
+    const std::string context = path.empty() ? "" : "(" + path + ") " + positions;
+#else
+    const std::string& context = positions;
+    static_cast<void>(path);
+#endif
+    return "[json.exception." + id + "] " + context + what;
+}
+
 #if !defined(JSON_NOEXCEPTION)
 // the exception a call throws, or "" if it throws none
 inline std::string exception_of_call(const std::function<void()>& f)

tests/src/sax_event_loggers.hpp

@@ -131,11 +131,13 @@ struct SaxEventLogger : public nlohmann::json_sax<nlohmann::json>
     {
         errored = true;
         events.push_back("parse_error(" + std::to_string(position) + ")");
-        return false;
+        return recover;
     }
 
     std::vector<std::string> events {}; // NOLINT(readability-redundant-member-init)
     bool errored = false;
+    /// whether parse_error() asks the parser to recover from the error (see #3989)
+    bool recover = false;
 };
 
 struct SaxEventLoggerExitAfterStartObject : public SaxEventLogger

tests/src/unit-alt-string.cpp

@@ -432,6 +432,46 @@ TEST_CASE("alternative string type")
         CHECK(j2.dump() == R"({"/foo/0":"bar","/foo/1":"baz"})");
     }
 
+    SECTION("error recovery")
+    {
+        // a SAX parser that recovers from every error (see #3989)
+        struct recovering_parser : nlohmann::detail::json_sax_dom_parser<alt_json>
+        {
+            explicit recovering_parser(alt_json& j)
+                : nlohmann::detail::json_sax_dom_parser<alt_json>(j, false)
+            {}
+
+            // sax_parse() calls the SAX parser's own parse_error(), so hiding
+            // the one of the base class is what recovering takes
+            // NOLINTNEXTLINE(bugprone-derived-method-shadowing-base-method)
+            bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/, const nlohmann::detail::exception& /*unused*/)
+            {
+                ++errors;
+                return true;
+            }
+
+            std::size_t errors = 0;
+        };
+
+        alt_json j;
+        recovering_parser sax(j);
+        // not inside CHECK(): MSVC reads the escape in a stringized raw string
+        const std::string input = R"([1., "a\qb", tru, {"k" 2}])";
+        CHECK(!alt_json::sax_parse(input, &sax));
+        CHECK(sax.errors == 4);
+        CHECK(j.dump() == R"([1,"aqb",null,{"k":2}])");
+
+        // a UBJSON high-precision number, a CBOR key that is not a string
+        alt_json u;
+        recovering_parser ubjson_sax(u);
+        CHECK(!alt_json::sax_parse(std::vector<std::uint8_t> {'[', 'H', 'i', 2, '1', '.', ']'}, &ubjson_sax, alt_json::input_format_t::ubjson));
+        CHECK(u.dump() == "[1]");
+        alt_json c;
+        recovering_parser cbor_sax(c);
+        CHECK(!alt_json::sax_parse(std::vector<std::uint8_t> {0xA2, 0x01, 0x02, 0x61, 'a', 0x03}, &cbor_sax, alt_json::input_format_t::cbor));
+        CHECK(c.dump() == R"({"a":3})");
+    }
+
     SECTION("conversion between basic_json specializations (#2649)")
     {
         // explicit conversions are always possible

tests/src/unit-bjdata.cpp

@@ -26,11 +26,14 @@ using nlohmann::ordered_json;
 #include <limits>
 #include <iostream>
 #include <fstream>
+#include <list>
 #include <set>
 #include "make_test_data_available.hpp"
 #include "test_utils.hpp"
 #include "sax_countdown.hpp"
+#include "sax_event_loggers.hpp"
 using utils::SaxCountdown;
+using utils::SaxEventLogger;
 
 
 // trait_test_arg and the "value_in_range_of trait" TEST_CASE_TEMPLATE_DEFINE
@@ -4746,3 +4749,77 @@ TEST_CASE("BJData optimized container type marker is validated for count 0")
         CHECK(json::from_bjdata(vb) == json::binary({}));
     }
 }
+
+TEST_CASE("BJData optimized arrays from contiguous and non-contiguous input")
+{
+    // contiguous input reads the numbers of an array with a type marker in
+    // one loop, other input number by number; both must give the same events,
+    // value, or error for the complete document and for every truncation of
+    // it, and stop at every event a SAX handler rejects
+    json document =
+    {
+        {"u8", {0, 200, 255}},
+        {"i8", {-5, 100, -128}},
+        {"i16", {-30000, 1000, 30000}},
+        {"i32", {-2000000000, 5, 70000}},
+        {"i64", {-9000000000000000000LL, 1, 5000000000LL}},
+        {"f64", {1.5, -0.1, 2.5}},
+        {"pairs", {{1.25, -3.75}, {1.25, -3.75}, json::array()}},
+        {"mixed", {1, "a", 2.5, nullptr}},
+        {"u16", {40000, 1, 65535}},
+        {"u32", {4000000000U, 1, 70000}},
+        {"u64", {18446744073709551615ULL, 1, 5000000000ULL}},
+        {"nd", {{"_ArrayType_", "double"}, {"_ArraySize_", {2, 2}}, {"_ArrayData_", {1.5, -2.5, 3.5, -4.5}}}},
+        {"nd8", {{"_ArrayType_", "uint8"}, {"_ArraySize_", {3}}, {"_ArrayData_", {1, 2, 3}}}},
+    };
+    const std::vector<std::uint8_t> encoded = json::to_bjdata(document, true, true);
+    CHECK(json::from_bjdata(encoded) == document);
+
+    for (std::size_t length = 0; length <= encoded.size(); ++length)
+    {
+        CAPTURE(length)
+        const std::vector<std::uint8_t> contiguous(encoded.begin(), encoded.begin() + static_cast<std::ptrdiff_t>(length));
+        const std::list<std::uint8_t> bytewise(contiguous.begin(), contiguous.end());
+
+        CHECK(json::from_bjdata(contiguous, true, false).dump() == json::from_bjdata(bytewise.begin(), bytewise.end(), true, false).dump());
+#if !defined(JSON_NOEXCEPTION)
+        std::string contiguous_error;
+        std::string bytewise_error;
+        try
+        {
+            json _;
+            _ = json::from_bjdata(contiguous);
+        }
+        catch (const json::exception& e)
+        {
+            contiguous_error = e.what();
+        }
+        try
+        {
+            json _;
+            _ = json::from_bjdata(bytewise.begin(), bytewise.end());
+        }
+        catch (const json::exception& e)
+        {
+            bytewise_error = e.what();
+        }
+        CHECK(contiguous_error == bytewise_error);
+#endif
+
+        SaxEventLogger from_contiguous;
+        SaxEventLogger from_bytewise;
+        CHECK(json::sax_parse(contiguous, &from_contiguous, json::input_format_t::bjdata)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::bjdata));
+        CHECK(from_contiguous.events == from_bytewise.events);
+    }
+
+    for (int events = 0; events <= 80; ++events)
+    {
+        CAPTURE(events)
+        const std::list<std::uint8_t> bytewise(encoded.begin(), encoded.end());
+        SaxCountdown from_contiguous(events);
+        SaxCountdown from_bytewise(events);
+        CHECK(json::sax_parse(encoded, &from_contiguous, json::input_format_t::bjdata)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::bjdata));
+    }
+}

tests/src/unit-bson.cpp

@@ -19,9 +19,11 @@
 #include <nlohmann/json.hpp>
 using nlohmann::json;
 
+#include <array>
 #include <cstdint>
 #include <fstream>
 #include <limits>
+#include <list>
 #include <sstream>
 #include <vector>
 #include "make_test_data_available.hpp"
@@ -1164,7 +1166,13 @@ TEST_CASE("BSON keys from contiguous and stream input")
     // contiguous input reads a key up to its \x00-byte in one step, a stream
     // reads it byte by byte; both must give the same value or error for the
     // complete document and for every truncation of it
-    const json j = {{"", true}, {"k", {1, 2, 3}}, {std::string(40, 'x'), {{"nested key", "value"}}}};
+    json j = {{"", true}, {"k", {1, 2, 3}}, {std::string(40, 'x'), {{"nested key", "value"}}}};
+    // keys and array element names ("0" to "11") that end at every position of
+    // an 8-byte word, which contiguous input scans at a time
+    for (std::size_t length = 1; length <= 17; ++length)
+    {
+        j[std::string(length, static_cast<char>('a' + length))] = json::array({0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, {11}});
+    }
     const std::vector<std::uint8_t> bson = json::to_bson(j);
     CHECK(json::from_bson(bson) == j);
 
@@ -1184,7 +1192,7 @@ TEST_CASE("BSON keys from contiguous and stream input")
         }
         try
         {
-            std::istringstream stream(std::string(input.begin(), input.end()));
+            std::istringstream stream(std::string(reinterpret_cast<const char*>(input.data()), input.size()));
             from_stream = json::from_bson(stream).dump();
         }
         catch (const json::parse_error& e)
@@ -1192,8 +1200,72 @@ TEST_CASE("BSON keys from contiguous and stream input")
             from_stream = e.what();
         }
         CHECK(from_vector == from_stream);
+
+        // error_handler_t::strict checks the array element names in place on
+        // contiguous input and byte by byte otherwise; the names are
+        // well-formed, so both give the same value or error as above
+        std::string strict_from_vector;
+        std::string strict_from_list;
+        try
+        {
+            strict_from_vector = json::from_bson(input, true, true, json::error_handler_t::strict).dump();
+        }
+        catch (const json::parse_error& e)
+        {
+            strict_from_vector = e.what();
+        }
+        try
+        {
+            const std::list<std::uint8_t> list(input.begin(), input.end());
+            strict_from_list = json::from_bson(list.begin(), list.end(), true, true, json::error_handler_t::strict).dump();
+        }
+        catch (const json::parse_error& e)
+        {
+            strict_from_list = e.what();
+        }
+        CHECK(strict_from_vector == from_vector);
+        CHECK(strict_from_list == from_vector);
     }
 }
+
+TEST_CASE("BSON array element names")
+{
+    // an array's element names are not part of the value, so they are not
+    // passed on; error_handler_t::strict still rejects ill-formed UTF-8 in them
+    const std::vector<std::uint8_t> bson =
+    {
+        0x1D, 0x00, 0x00, 0x00, // size (little endian)
+        0x04, 'a', 0x00,        // entry: array "a"
+        0x15, 0x00, 0x00, 0x00, // size (little endian)
+        0x08, 0xFF, 0x00, 0x01, // entry: boolean named "\xFF", true
+        0x08, '1', '2', '3', '4', '5', '6', '7', '8', '9', 0x00, 0x00, // entry: boolean named "123456789", false
+        0x00,                   // end of the array
+        0x00                    // end of the document
+    };
+    const json expected = {{"a", {true, false}}};
+
+    CHECK(json::from_bson(bson) == expected);
+    CHECK(json::from_bson(bson, true, true, json::error_handler_t::replace) == expected);
+    CHECK(json::from_bson(bson, true, true, json::error_handler_t::ignore) == expected);
+
+    json _;
+    CHECK_THROWS_WITH_AS(_ = json::from_bson(bson, true, true, json::error_handler_t::strict), "[json.exception.parse_error.113] parse error at byte 14: syntax error while parsing BSON key: invalid string: ill-formed UTF-8 byte", json::parse_error&);
+
+    std::istringstream stream(std::string(reinterpret_cast<const char*>(bson.data()), bson.size()));
+    CHECK(json::from_bson(stream) == expected);
+
+    // input that is not contiguous reads the names byte by byte and must reject
+    // the same name
+    const std::list<std::uint8_t> list(bson.begin(), bson.end());
+    CHECK_THROWS_WITH_AS(_ = json::from_bson(list.begin(), list.end(), true, true, json::error_handler_t::strict), "[json.exception.parse_error.113] parse error at byte 14: syntax error while parsing BSON key: invalid string: ill-formed UTF-8 byte", json::parse_error&);
+
+    // without exceptions, the rejected name ends the parse
+    CHECK(json::from_bson(bson, true, false, json::error_handler_t::strict).is_discarded());
+
+    // well-formed names pass the check without being stored
+    const json valid = {{"a", {true, false, {{"k", json::array({1, 2})}}}}};
+    CHECK(json::from_bson(json::to_bson(valid), true, true, json::error_handler_t::strict) == valid);
+}
 #endif
 
 TEST_CASE("Negative size of binary value")
@@ -2267,3 +2339,153 @@ TEST_CASE("BSON large strings and binaries (chunked reader)")
         }
     }
 }
+
+TEST_CASE("BSON written in one pass and in two passes")
+{
+    // to_bson() into a returned vector writes each size once the object or
+    // array is complete; through an output adapter, all sizes are computed
+    // first. Both must produce the same bytes and throw the same exceptions.
+    const auto two_pass = [](const json & j, const json::error_handler_t handler)
+    {
+        std::vector<std::uint8_t> result;
+        json::to_bson(j, result, handler);
+        return result;
+    };
+
+    const auto check_same = [&](const json & j, const json::error_handler_t handler)
+    {
+        CAPTURE(j.dump(-1, ' ', true, json::error_handler_t::replace))
+        CAPTURE(static_cast<int>(handler))
+#if !defined(JSON_NOEXCEPTION)
+        std::string one_pass_error;
+        std::string two_pass_error;
+        std::vector<std::uint8_t> one;
+        std::vector<std::uint8_t> two;
+        try
+        {
+            one = json::to_bson(j, handler);
+        }
+        catch (const json::exception& e)
+        {
+            one_pass_error = e.what();
+        }
+        try
+        {
+            two = two_pass(j, handler);
+        }
+        catch (const json::exception& e)
+        {
+            two_pass_error = e.what();
+        }
+        CHECK(one_pass_error == two_pass_error);
+        CHECK(one == two);
+#else
+        CHECK(json::to_bson(j, handler) == two_pass(j, handler));
+#endif
+    };
+
+    const std::array<json::error_handler_t, 4> handlers = {{
+            json::error_handler_t::keep, json::error_handler_t::strict,
+            json::error_handler_t::replace, json::error_handler_t::ignore
+        }
+    };
+
+    SECTION("corpus")
+    {
+        for (const auto& value : utils::round_trip_corpus::values())
+        {
+            const json document = value.is_object() ? value : json{{"value", value}};
+            for (const auto handler : handlers)
+            {
+                check_same(document, handler);
+            }
+        }
+    }
+
+    SECTION("nesting and arrays of every length class")
+    {
+        json document = json::object();
+        document["empty object"] = json::object();
+        document["empty array"] = json::array();
+        for (const std::size_t n :
+                {
+                    1u, 9u, 10u, 11u, 99u, 100u, 101u, 1000u
+                })
+        {
+            json numbers = json::array();
+            json mixed = json::array();
+            for (std::size_t i = 0; i < n; ++i)
+            {
+                numbers.push_back(i % 3 == 0 ? json(static_cast<double>(i) / 3) : json(static_cast<std::int64_t>(i) * 1000003));
+                switch (i % 4)
+                {
+                    case 0:
+                        mixed.push_back(json("s" + std::to_string(i)));
+                        break;
+                    case 1:
+                        mixed.push_back(json::array({i, json::object({{"k", i}})}));
+                        break;
+                    case 2:
+                        mixed.push_back(json(nullptr));
+                        break;
+                    default:
+                        mixed.push_back(json(i % 2 == 0));
+                        break;
+                }
+            }
+            document["numbers " + std::to_string(n)] = numbers;
+            document["mixed " + std::to_string(n)] = mixed;
+            document["nested " + std::to_string(n)] = json::object({{"a", numbers}, {"b", json::array({mixed, json::object()})}});
+        }
+        document["binary"] = json::binary({1, 2, 3}, 0x80);
+        document["binary old"] = json::binary({1, 2, 3}, 0x02);
+        for (const auto handler : handlers)
+        {
+            check_same(document, handler);
+        }
+    }
+
+#if !defined(JSON_NOEXCEPTION)
+    // every entry here throws, which aborts without exceptions
+    SECTION("errors at every position")
+    {
+        // an entry that cannot be written, at each place among valid ones
+        const std::string bad_utf8 = "\xC3\x28";
+        const std::array<json, 7> bad_entries =
+        {
+            {
+                json::object({{std::string("a\0b", 3), 1}}),
+                json::object({{"s", bad_utf8}}),
+                json::object({{bad_utf8, 1}}),
+                json::array({1, 2, bad_utf8}),
+                json::object({{"b", json::binary({1}, 0x100)}}),
+                json::object({{"d", json(json::value_t::discarded)}}),
+                json::array({json(json::value_t::discarded)}),
+            }
+        };
+        for (const auto& bad : bad_entries)
+        {
+            for (std::size_t pos = 0; pos < 4; ++pos)
+            {
+                json document = json::object();
+                json list = json::array();
+                for (std::size_t i = 0; i < 4; ++i)
+                {
+                    if (i == pos)
+                    {
+                        document["x" + std::to_string(i)] = bad;
+                        list.push_back(bad);
+                    }
+                    document["n" + std::to_string(i)] = json::array({i, "str", json::object({{"k", i}})});
+                    list.push_back(i);
+                }
+                document["list"] = list;
+                for (const auto handler : handlers)
+                {
+                    check_same(document, handler);
+                }
+            }
+        }
+    }
+#endif
+}

tests/src/unit-cbor.cpp

@@ -29,7 +29,9 @@ using nlohmann::json;
 #include "round_trip_corpus.hpp"
 #include "test_utils.hpp"
 #include "sax_countdown.hpp"
+#include "sax_event_loggers.hpp"
 using utils::SaxCountdown;
+using utils::SaxEventLogger;
 
 
 TEST_CASE("CBOR")
@@ -2446,6 +2448,78 @@ TEST_CASE("CBOR SAX parsing stops at every event")
     CHECK(count_events(std::vector<std::uint8_t>({0xBF, 0x61, 'a', 0x9F, 0x01, 0xFF, 0xFF})) == 6);
 }
 
+TEST_CASE("CBOR arrays of scalars from contiguous and non-contiguous input")
+{
+    // contiguous input reads the integers, floats, booleans, and nulls an
+    // array begins with in one loop, other input value by value; both must
+    // give the same events, value, or error for the complete document and for
+    // every truncation of it, and stop at every event a SAX handler rejects
+    const std::vector<std::uint8_t> cbor =
+    {
+        0xA3,                                                       // map(3)
+        0x61, 'a',                                                  // "a"
+        0x98, 0x16,                                                 // array(22)
+        0x00, 0x17, 0x18, 0xFF, 0x19, 0x01, 0x00,                   // 0, 23, 255, 256
+        0x1A, 0x00, 0x01, 0x00, 0x00,                               // 65536
+        0x1B, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,       // 18446744073709551615
+        0x20, 0x37, 0x38, 0xFF, 0x39, 0x01, 0x00,                   // -1, -24, -256, -257
+        0x3A, 0x00, 0x01, 0x00, 0x00,                               // -65537
+        0x3B, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,       // -18446744073709551616 (a float)
+        0x3B, 0x7F, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,       // -9223372036854775808
+        0xF4, 0xF5, 0xF6,                                           // false, true, null
+        0xFA, 0x3F, 0xC0, 0x00, 0x00,                               // 1.5 (float32)
+        0xFB, 0xBF, 0xB9, 0x99, 0x99, 0x99, 0x99, 0x99, 0x9A,       // -0.1
+        0xF9, 0x3C, 0x00,                                           // 1.0 (float16, ends the loop)
+        0x01, 0x61, 'a',                                            // 1, "a"
+        0xFB, 0x40, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       // 2.5
+        0x61, 'b',                                                  // "b"
+        0x83,                                                       // array(3)
+        0x82, 0xFB, 0x3F, 0xF4, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // [1.25,
+        0xFB, 0xC0, 0x0E, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       //  -3.75]
+        0x82, 0xFB, 0x3F, 0xF4, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // [1.25,
+        0xFB, 0xC0, 0x0E, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       //  -3.75]
+        0x80,                                                       // []
+        0x61, 'c',                                                  // "c"
+        0x84, 0xA0, 0x01, 0xC1, 0x02, 0x81, 0x03                    // [{}, 1, tag(1) 2, [3]]
+    };
+    const auto tags = json::cbor_tag_handler_t::ignore;
+
+    const json j = json::from_cbor(cbor, true, true, tags);
+    CHECK(j["a"].size() == 22);
+    CHECK(j["a"][5] == 18446744073709551615u);
+    CHECK(j["a"][11].is_number_float());
+    CHECK(j["a"][11] == -18446744073709551616.0);
+    CHECK(j["a"][12] == (std::numeric_limits<std::int64_t>::min)());
+    CHECK(j["a"][17] == -0.1);
+    CHECK(j["b"] == json::parse("[[1.25, -3.75], [1.25, -3.75], []]"));
+    CHECK(j["c"] == json::parse("[{}, 1, 2, [3]]"));
+
+    for (std::size_t length = 0; length <= cbor.size(); ++length)
+    {
+        CAPTURE(length)
+        const std::vector<std::uint8_t> contiguous(cbor.begin(), cbor.begin() + static_cast<std::ptrdiff_t>(length));
+        const std::list<std::uint8_t> bytewise(contiguous.begin(), contiguous.end());
+
+        CHECK(json::from_cbor(contiguous, true, false, tags).dump() == json::from_cbor(bytewise.begin(), bytewise.end(), true, false, tags).dump());
+
+        SaxEventLogger from_contiguous;
+        SaxEventLogger from_bytewise;
+        CHECK(json::sax_parse(contiguous, &from_contiguous, json::input_format_t::cbor, true, false, false, tags)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::cbor, true, false, false, tags));
+        CHECK(from_contiguous.events == from_bytewise.events);
+    }
+
+    for (int events = 0; events <= 60; ++events)
+    {
+        CAPTURE(events)
+        const std::list<std::uint8_t> bytewise(cbor.begin(), cbor.end());
+        SaxCountdown from_contiguous(events);
+        SaxCountdown from_bytewise(events);
+        CHECK(json::sax_parse(cbor, &from_contiguous, json::input_format_t::cbor, true, false, false, tags)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::cbor, true, false, false, tags));
+    }
+}
+
 TEST_CASE("CBOR indefinite-length strings do not recurse per chunk")
 {
     // Reading an indefinite-length string or byte array used to call itself

tests/src/unit-class_parser.cpp

@@ -2953,6 +2953,651 @@ TEST_CASE("diagnostic positions: value lifetime, input adapters, and SAX")
 }
 #endif
 
+namespace
+{
+/// builds a value like json::parse(), but asks the parser to recover from
+/// errors (see #3989), and checks that the events it receives are balanced
+class RecoveringDomParser
+{
+  public:
+    explicit RecoveringDomParser(json& j, std::size_t max_errors_ = static_cast<std::size_t>(-1))
+        : dom(j, false)
+        , max_errors(max_errors_)
+    {}
+
+    bool null()
+    {
+        value();
+        return dom.null();
+    }
+
+    bool boolean(bool val)
+    {
+        value();
+        return dom.boolean(val);
+    }
+
+    bool number_integer(json::number_integer_t val)
+    {
+        value();
+        return dom.number_integer(val);
+    }
+
+    bool number_unsigned(json::number_unsigned_t val)
+    {
+        value();
+        return dom.number_unsigned(val);
+    }
+
+    bool number_float(json::number_float_t val, const std::string& s)
+    {
+        value();
+        return dom.number_float(val, s);
+    }
+
+    bool string(std::string& val)
+    {
+        value();
+        return dom.string(val);
+    }
+
+    bool binary(json::binary_t& val)
+    {
+        value();
+        return dom.binary(val);
+    }
+
+    bool start_object(std::size_t elements)
+    {
+        value();
+        stack.push_back('o');
+        return dom.start_object(elements);
+    }
+
+    bool key(std::string& val)
+    {
+        ++events;
+        if (stack.empty() || stack.back() != 'o')
+        {
+            well_formed = false;
+            return false;
+        }
+        stack.back() = 'v';
+        return dom.key(val);
+    }
+
+    bool end_object()
+    {
+        ++events;
+        if (stack.empty() || stack.back() != 'o')
+        {
+            well_formed = false;
+            return false;
+        }
+        stack.pop_back();
+        return dom.end_object();
+    }
+
+    bool start_array(std::size_t elements)
+    {
+        value();
+        stack.push_back('a');
+        return dom.start_array(elements);
+    }
+
+    bool end_array()
+    {
+        ++events;
+        if (stack.empty() || stack.back() != 'a')
+        {
+            well_formed = false;
+            return false;
+        }
+        stack.pop_back();
+        return dom.end_array();
+    }
+
+    bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/, const json::exception& ex)
+    {
+        errors.emplace_back(ex.what());
+        return errors.size() < max_errors;
+    }
+
+    /// whether the events were balanced and every key was followed by a value
+    bool balanced() const
+    {
+        return well_formed && stack.empty();
+    }
+
+    /// builds the value
+    nlohmann::detail::json_sax_dom_parser<json> dom;
+    std::vector<std::string> errors {}; // NOLINT(readability-redundant-member-init)
+    std::size_t events = 0;
+    /// the open containers: 'a' for an array, 'o' for an object that expects
+    /// a key, 'v' for an object that expects the value of a key
+    std::vector<char> stack {}; // NOLINT(readability-redundant-member-init)
+    bool well_formed = true;
+    std::size_t max_errors;
+
+  private:
+    /// a value is passed: it is an array element, or the value of a key
+    void value()
+    {
+        ++events;
+        if (!stack.empty())
+        {
+            if (stack.back() == 'v')
+            {
+                stack.back() = 'o';
+            }
+            else if (stack.back() == 'o')
+            {
+                // a value without a key
+                well_formed = false;
+            }
+        }
+    }
+
+};
+
+struct RecoveryResult
+{
+    json value;
+    std::vector<std::string> errors;
+    std::size_t events;
+    bool ok;
+    bool balanced;
+};
+
+template<typename InputType>
+RecoveryResult parse_recovering(InputType&& input, const bool strict = true,
+                                const bool ignore_comments = false, const bool ignore_trailing_commas = false)
+{
+    json j;
+    RecoveringDomParser sax(j);
+    const bool ok = json::sax_parse(std::forward<InputType>(input), &sax, json::input_format_t::json,
+                                    strict, ignore_comments, ignore_trailing_commas);
+    return {j, sax.errors, sax.events, ok, sax.balanced()};
+}
+
+/// stops after a number of events, but recovers from errors
+class RecoveringCountdown
+{
+  public:
+    explicit RecoveringCountdown(const int count)
+        : countdown(count)
+    {}
+
+    bool null()
+    {
+        return countdown.null();
+    }
+
+    bool boolean(bool val)
+    {
+        return countdown.boolean(val);
+    }
+
+    bool number_integer(json::number_integer_t val)
+    {
+        return countdown.number_integer(val);
+    }
+
+    bool number_unsigned(json::number_unsigned_t val)
+    {
+        return countdown.number_unsigned(val);
+    }
+
+    bool number_float(json::number_float_t val, const std::string& s)
+    {
+        return countdown.number_float(val, s);
+    }
+
+    bool string(std::string& val)
+    {
+        return countdown.string(val);
+    }
+
+    bool binary(json::binary_t& val)
+    {
+        return countdown.binary(val);
+    }
+
+    bool start_object(std::size_t elements)
+    {
+        return countdown.start_object(elements);
+    }
+
+    bool key(std::string& val)
+    {
+        return countdown.key(val);
+    }
+
+    bool end_object()
+    {
+        return countdown.end_object();
+    }
+
+    bool start_array(std::size_t elements)
+    {
+        return countdown.start_array(elements);
+    }
+
+    bool end_array()
+    {
+        return countdown.end_array();
+    }
+
+    bool parse_error(std::size_t /*position*/, const std::string& /*last_token*/, const json::exception& /*ex*/) // NOLINT(readability-convert-member-functions-to-static)
+    {
+        return true;
+    }
+
+  private:
+    SaxCountdown countdown;
+};
+
+/// a repaired input: the value it is repaired to, and the number of errors
+struct Repair
+{
+    const char* input;
+    const char* expected;
+    std::size_t errors;
+};
+} // namespace
+
+TEST_CASE("parser error recovery (#3989)")
+{
+    SECTION("repairs")
+    {
+        const std::vector<Repair> repairs =
+        {
+            // a missing separator is inserted
+            {"[1 2]", "[1,2]", 1},
+            {R"({"a":1 "b":2})", R"({"a":1,"b":2})", 1},
+            {R"({"a" 1})", R"({"a":1})", 1},
+            {"[1 tru 2]", "[1,null,2]", 2},
+            {R"({"a" "b": 1})", R"({"a":"b"})", 2},
+
+            // a missing value is null in an object; in an array, a ',' stands
+            // for null, while an array that ends there just ends
+            {R"({"a":})", R"({"a":null})", 1},
+            {R"({"a"})", R"({"a":null})", 1},
+            {R"({"a","b":1})", R"({"a":null,"b":1})", 1},
+            {"[1,,2]", "[1,null,2]", 1},
+            {"[,1]", "[null,1]", 1},
+            {"[1,]", "[1]", 1},
+            {"[1,2,3,]", "[1,2,3]", 1},
+            {R"({"a":1,})", R"({"a":1})", 1},
+
+            // a broken string keeps what can be read
+            {R"(["a\qb"])", R"(["aqb"])", 1},
+            {R"({"na\me":1})", R"({"name":1})", 1},
+            {"[\"\xFF\"]", R"(["\uFFFD"])", 1},
+            {"[\"a\xC3(\"]", R"(["a\uFFFD("])", 1},
+            {"[\"\xE2\x82\"]", R"(["\uFFFD"])", 1},
+            {"[\"\xC3\\\\\", 1]", R"(["\uFFFD\\",1])", 1},
+            {R"(["\u12"])", R"(["\uFFFD"])", 1},
+            {R"(["\u12G4"])", R"(["\uFFFDG4"])", 1},
+            {R"(["\uDC00x"])", R"(["\uFFFDx"])", 1},
+            {R"(["\uD800x"])", R"(["\uFFFDx"])", 1},
+            {R"(["\uD800\u0041"])", R"(["\uFFFDA"])", 1},
+            {R"(["\uD800\uD800\uDC00"])", R"(["\uFFFD\uD800\uDC00"])", 1},
+            {R"(["\uD800\uD800\uD800x"])", R"(["\uFFFD\uFFFD\uFFFDx"])", 1},
+            {
+                R"(["\uD800\"x", 1])", R"(["\uFFFD\"x",1])", 1
+            },
+            {R"(["\uD800\q"])", R"(["\uFFFDq"])", 1},
+            {"[\"a\tb\"]", R"(["a\tb"])", 1},
+            {R"(["a\qb\u0041\x"])", R"(["aqbAx"])", 1},
+
+            // a broken number keeps its longest valid prefix
+            {"[1.]", "[1]", 1},
+            {"[-2.]", "[-2]", 1},
+            {"[1.5e]", "[1.5]", 1},
+            {"[1e+]", "[1]", 1},
+            {"[1.x2, 3]", "[1,3]", 1},
+
+            // what cannot be read at all is null
+            {"[1,NaN,3]", "[1,null,3]", 1},
+            {"[tru]", "[null]", 1},
+            {"[-]", "[null]", 1},
+            {R"({"a":Infinity})", R"({"a":null})", 1},
+
+            // a stray token is dropped
+            {"[:1]", "[1]", 1},
+            {R"(["a":1])", R"(["a",1])", 1},
+            {R"({"a"::1})", R"({"a":1})", 1},
+
+            // a member that cannot be read is skipped
+            {R"({1:2,"b":3})", R"({"b":3})", 1},
+            {R"({"a":1 2})", R"({"a":1})", 1},
+            {R"({,"a":1})", R"({"a":1})", 1},
+            {R"({"a":1,,"b":2})", R"({"a":1,"b":2})", 1},
+            {"{a:1}", "{}", 1},
+            {R"({"a":1 [1,{"b":2}], "c":3})", R"({"a":1,"c":3})", 1},
+            {R"([{1}, "a"])", R"([{},"a"])", 1},
+
+            // a wrong closing bracket closes the innermost container
+            {R"({"a":[1,2}, "b":3})", R"({"a":[1,2],"b":3})", 1},
+            {R"([{"a":1], 2])", R"([{"a":1},2])", 1},
+            {"{]", "{}", 1},
+            {"[}", "[]", 1},
+
+            // the end of the input closes all containers
+            {R"({"a":[1,2)", R"({"a":[1,2]})", 1},
+            {"[", "[]", 1},
+            {"{", "{}", 1},
+            {R"({"a")", R"({"a":null})", 1},
+            {R"({"a":)", R"({"a":null})", 1},
+            {"[1,", "[1]", 1},
+            {"[[[1", "[[[1]]]", 1},
+            {
+                R"(["abc)", R"(["abc"])", 2
+            },
+            {"[1,tr", "[1,null]", 2},
+            {"\"abc", "\"abc\"", 1},
+            {"[\"ab\ncd\"]", R"(["ab",null,"]"])", 4},
+
+            // what comes before the top-level value is skipped
+            {")]}'\n{\"a\":1}", R"({"a":1})", 1},
+            {R"(data: {"a":1})", R"({"a":1})", 1},
+            {"\xEF\xBB[1]", "[1]", 1},
+
+            // what comes after it is an error that ends parsing
+            {R"({"a":1}})", R"({"a":1})", 1},
+            {"[1}]", "[1]", 2},
+            {"[1] [2]", "[1]", 1},
+        };
+
+        for (const auto& repair : repairs)
+        {
+            CAPTURE(repair.input)
+            const auto result = parse_recovering(std::string(repair.input));
+            CHECK(!result.ok);
+            CHECK(result.balanced);
+            CHECK(result.value == json::parse(repair.expected));
+            CHECK(result.errors.size() == repair.errors);
+        }
+    }
+
+    SECTION("number overflow")
+    {
+        const auto result = parse_recovering(std::string("[1e999,-1e999]"));
+        CHECK(!result.ok);
+        CHECK(result.balanced);
+        CHECK(result.errors.size() == 2);
+        CHECK(result.errors[0] == "[json.exception.out_of_range.406] number overflow parsing '1e999'");
+        REQUIRE(result.value.size() == 2);
+        CHECK(result.value[0].is_number_float());
+        CHECK(result.value[0].get<double>() == std::numeric_limits<double>::infinity());
+        CHECK(result.value[1].get<double>() == -std::numeric_limits<double>::infinity());
+
+        // the SAX parser gets the number's text
+        SaxEventLogger logger;
+        logger.recover = true;
+        CHECK(!json::sax_parse("1e999", &logger));
+        CHECK(logger.events == std::vector<std::string>({"parse_error(5)", "number_float(1e999)"}));
+    }
+
+    SECTION("nothing to recover")
+    {
+        for (const std::string s :
+                {
+                    "", "   ", "]", "tru", "NaN", ",:", "/* comment"
+                })
+        {
+            CAPTURE(s)
+            const auto result = parse_recovering(s, true, true);
+            CHECK(!result.ok);
+            CHECK(result.balanced);
+            CHECK(result.events == 0);
+            CHECK(result.value == nullptr);
+            CHECK(result.errors.size() == 1);
+        }
+    }
+
+    SECTION("error messages")
+    {
+        // the first error is reported as without recovery
+        for (const std::string s :
+                {
+                    "[1 2]", R"({"a":1 "b":2})", R"({"a" 1})", R"({"a":})", "[1,]", "[1.]",
+                    R"(["a\qb"])", "[1e999]", "{1:2}", R"({"a":[1,2}})", "[1,", "[1] [2]", "{a:1}"
+                })
+        {
+            CAPTURE(s)
+            const auto result = parse_recovering(s);
+            REQUIRE(!result.errors.empty());
+            json _;
+            CHECK_THROWS_WITH_STD_STR(_ = json::parse(s), result.errors.front());
+        }
+
+        // the token of an error begins where the previous error was
+        const auto result = parse_recovering(std::string("[tru, fals, nul]"));
+        CHECK(result.errors == std::vector<std::string>(
+        {
+            "[json.exception.parse_error.101] parse error at line 1, column 5: syntax error while parsing value - invalid literal; last read: '[tru,'",
+            "[json.exception.parse_error.101] parse error at line 1, column 11: syntax error while parsing value - invalid literal; last read: ', fals,'",
+            "[json.exception.parse_error.101] parse error at line 1, column 16: syntax error while parsing value - invalid literal; last read: ', nul]'"
+        }));
+        CHECK(result.value == json::parse("[null,null,null]"));
+    }
+
+    SECTION("events")
+    {
+        // see #4522
+        SaxEventLogger logger;
+        logger.recover = true;
+        CHECK(!json::sax_parse(R"([{1}, "a"])", &logger));
+        CHECK(logger.events == std::vector<std::string>(
+        {
+            "start_array()", "start_object()", "parse_error(3)", "end_object()", "string(a)", "end_array()"
+        }));
+    }
+
+    SECTION("options")
+    {
+        SECTION("strict")
+        {
+            const auto result = parse_recovering(std::string("[1 2] [3]"), false);
+            CHECK(!result.ok);
+            CHECK(result.value == json::parse("[1,2]"));
+            CHECK(result.errors.size() == 1);
+        }
+
+        SECTION("ignore_trailing_commas")
+        {
+            for (const std::string s :
+                    {
+                        "[1,]", R"({"a":1,})", "[[1,],]"
+                    })
+            {
+                CAPTURE(s)
+                const auto result = parse_recovering(s, true, false, true);
+                CHECK(result.ok);
+                CHECK(result.errors.empty());
+            }
+
+            auto result = parse_recovering(std::string("[1,,]"), true, false, true);
+            CHECK(result.value == json::parse("[1,null]"));
+            CHECK(result.errors.size() == 1);
+
+            result = parse_recovering(std::string(R"({"a":1,,})"), true, false, true);
+            CHECK(result.value == json::parse(R"({"a":1})"));
+            CHECK(result.errors.size() == 1);
+        }
+
+        SECTION("ignore_comments")
+        {
+            auto result = parse_recovering(std::string("[1 /* one */ 2]"), true, true);
+            CHECK(result.value == json::parse("[1,2]"));
+            CHECK(result.errors.size() == 1);
+
+            // a comment that is not closed runs to the end of the input, which
+            // is not reported again
+            result = parse_recovering(std::string("[1, 2 /* unterminated"), true, true);
+            CHECK(result.balanced);
+            CHECK(result.value == json::parse("[1,2]"));
+            CHECK(result.errors.size() == 1);
+
+            // a '/' that does not begin a comment is garbage
+            result = parse_recovering(std::string("[1, /x, 2]"), true, true);
+            CHECK(result.balanced);
+            CHECK(result.value == json::parse("[1,null,2]"));
+            CHECK(result.errors.size() == 1);
+        }
+    }
+
+    SECTION("null bytes")
+    {
+        // a null byte ends the input, unless JSON_STRICT_NUL_HANDLING is set
+        const auto result = parse_recovering(std::string("[1,\0x", 5));
+        CHECK(result.balanced);
+        CHECK(!result.ok);
+#ifdef JSON_TEST_STRICT_NUL_HANDLING_ENABLED
+        CHECK(result.value == json::parse("[1,null]"));
+#else
+        CHECK(result.value == json::parse("[1]"));
+        CHECK(result.errors.size() == 1);
+#endif
+
+        const auto in_string = parse_recovering(std::string("[\"a\0b\"]", 7));
+        CHECK(in_string.balanced);
+#ifdef JSON_TEST_STRICT_NUL_HANDLING_ENABLED
+        CHECK(in_string.value == json::array({std::string("a\0b", 3)}));
+#else
+        CHECK(in_string.value == json::parse(R"(["a"])"));
+#endif
+    }
+
+    SECTION("the SAX parser stops recovering")
+    {
+        json j;
+        RecoveringDomParser sax(j, 2);
+        CHECK(!json::sax_parse("[1 2 3 4 5]", &sax));
+        CHECK(sax.errors.size() == 2);
+
+        // an error at a delimiter that an invalid token consumed is reported
+        // to the SAX parser, too
+        json j2;
+        RecoveringDomParser sax2(j2, 2);
+        CHECK(!json::sax_parse("[tru}, 1]", &sax2));
+        CHECK(sax2.errors.size() == 2);
+    }
+
+    SECTION("an event stops parsing during a repair")
+    {
+        // start_object() and key() are passed, then null() for the missing
+        // value returns false
+        RecoveringCountdown countdown(2);
+        CHECK(!json::sax_parse(R"({"a":})", &countdown));
+
+        // the end of the input: end_array() for the second array returns false
+        RecoveringCountdown countdown2(4);
+        CHECK(!json::sax_parse("[[1", &countdown2));
+    }
+
+    SECTION("input adapters")
+    {
+        // the lexer reads contiguous and streaming input differently, and it
+        // puts back a character that ended an invalid token
+        for (const std::string s :
+                {
+                    "[1 2]", "[tru}, 1]", R"({"a" "b\q", "c":[1.x, 2}})", "[\"\xFF\xC3(\", -, 1e+]", "{a:1,\"b\":2", ")]}' [1]"
+                })
+        {
+            CAPTURE(s)
+            const auto reference = parse_recovering(s);
+            CHECK(reference.balanced);
+
+            const auto from_c_string = parse_recovering(s.c_str());
+            CHECK(from_c_string.value == reference.value);
+            CHECK(from_c_string.errors == reference.errors);
+
+            const std::list<char> l(s.begin(), s.end());
+            json j;
+            RecoveringDomParser sax(j);
+            CHECK(!json::sax_parse(l.begin(), l.end(), &sax));
+            CHECK(j == reference.value);
+            CHECK(sax.errors == reference.errors);
+
+            std::istringstream ss(s);
+            const auto from_stream = parse_recovering(ss);
+            CHECK(from_stream.value == reference.value);
+            CHECK(from_stream.errors == reference.errors);
+        }
+    }
+
+    SECTION("long runs of errors")
+    {
+        // no error may copy all the input read before it
+        const auto closing = parse_recovering("[" + std::string(100000, '}'));
+        CHECK(closing.balanced);
+        CHECK(closing.value == json::array());
+
+        const auto garbage = parse_recovering("[" + std::string(100000, 'x') + "]");
+        CHECK(garbage.balanced);
+        CHECK(garbage.errors.size() == 1);
+
+        const auto commas = parse_recovering("{" + std::string(100000, ',') + "}");
+        CHECK(commas.balanced);
+        CHECK(commas.value == json::object());
+    }
+
+    SECTION("mutations of valid input")
+    {
+        // whatever the input, the events are balanced, every error is reported
+        // at most once, and valid input is parsed as usual
+        const std::vector<std::string> documents =
+        {
+            R"({"name": "value", "list": [1, -2.5, true, null, {"x": [[]]}], "e": "\u00e9"})",
+            R"([{"a": [1, 2, {"b": "c"}]}, [], {}, "\ud83d\ude00", 1e10])",
+            "{\"\xC3\xA9\": \"\xF0\x9F\x98\x80\"}",
+            R"(  {"k" : [ "v" , 0 ] }  )",
+        };
+        // each character that can be inserted, including a null byte
+        const std::string insertions("[]{},:\"x\\\0\xFF", 11);
+
+        std::vector<std::string> inputs;
+        for (const auto& doc : documents)
+        {
+            for (std::size_t i = 0; i <= doc.size(); ++i)
+            {
+                inputs.push_back(doc.substr(0, i));
+                if (i < doc.size())
+                {
+                    inputs.push_back(doc.substr(0, i) + doc.substr(i + 1));
+                }
+                for (const char c : insertions)
+                {
+                    inputs.push_back(doc.substr(0, i) + c + doc.substr(i));
+                }
+            }
+        }
+
+        for (const auto& s : inputs)
+        {
+            CAPTURE(s)
+            const auto result = parse_recovering(s);
+            CHECK(result.balanced);
+            CHECK(result.errors.size() <= s.size() + 1);
+            CHECK(result.events <= (4 * s.size()) + 4);
+            if (json::accept(s))
+            {
+                CHECK(result.ok);
+                CHECK(result.errors.empty());
+                CHECK(result.value == json::parse(s));
+            }
+            else
+            {
+                CHECK(!result.ok);
+                CHECK(!result.errors.empty());
+            }
+        }
+    }
+}
+
 TEST_CASE("accept() agrees with parse() on floats around the overflow limit")
 {
     // accept() does not convert a float that cannot overflow; the tokens
@@ -2982,6 +3627,67 @@ TEST_CASE("accept() agrees with parse() on floats around the overflow limit")
     }
 }
 
+TEST_CASE("accept() agrees with parse() on strings, which it checks without copying them")
+{
+    // accept() scans strings and object keys with a separate instantiation of
+    // the string scanner that does not copy them into the token buffer; it
+    // must accept and reject exactly what parse() does
+    const std::vector<std::string> pieces =
+    {
+        "", "abc", R"(\"\\\/\b\f\n\r\t)", R"(\u0041\u00e4\u20AC)", "\\ud83d\\ude00", "\\uD83D\\uDE00",
+        "\\ud800", "\\udc00", "\\ud800\\u0041", "\\ud800x", "\\udbff\\udfff", "\\u12", "\\u12g4", "\\x", "\\",
+        "\xc3\xa4", "\xe2\x82\xac", "\xf0\x9f\x98\x80", "\xf4\x8f\xbf\xbf", "\xc0\xaf", "\xed\xa0\x80", "\xf4\x90\x80\x80",
+        "\xff", "\xe2\x82", "\xc3", "\x01", "\x1f", "\x7f", "\n", std::string(1, '\0')
+    };
+
+    for (std::size_t run = 0; run <= 70; ++run)
+    {
+        for (const std::string& piece : pieces)
+        {
+            // a run of ordinary characters puts the piece at every offset of the bulk scanners
+            const std::string str = std::string(run, 'a') + piece;
+            std::string opened = "\"";
+            opened += str;
+            std::string closed = opened;
+            closed += '"';
+            std::vector<std::string> texts;
+            texts.push_back(closed);
+            texts.emplace_back("[");
+            texts.back() += closed;
+            texts.back() += ']';
+            texts.emplace_back("{");
+            texts.back() += closed;
+            texts.back() += ":1}";
+            texts.emplace_back("{\"k\":");
+            texts.back() += closed;
+            texts.back() += ',';
+            texts.back() += closed;
+            texts.back() += ":[";
+            texts.back() += closed;
+            texts.back() += "]}";
+            texts.emplace_back("[");
+            texts.back() += opened;
+            for (const std::string& text : texts)
+            {
+                CAPTURE(text)
+                for (int options = 0; options < 4; ++options)
+                {
+                    const bool ignore_comments = (options & 1) != 0;
+                    const bool ignore_trailing_commas = (options & 2) != 0;
+                    const bool parses = !json::parse(text, nullptr, false, ignore_comments, ignore_trailing_commas).is_discarded();
+                    CHECK(json::accept(text, ignore_comments, ignore_trailing_commas) == parses);
+
+                    std::istringstream stream(text);
+                    CHECK(json::accept(stream, ignore_comments, ignore_trailing_commas) == parses);
+
+                    const std::list<char> chars(text.begin(), text.end());
+                    CHECK(json::accept(chars.begin(), chars.end(), ignore_comments, ignore_trailing_commas) == parses);
+                }
+            }
+        }
+    }
+}
+
 namespace
 {
 // a DOM parser that reports parse errors without throwing, as in the sax_no_exception example

tests/src/unit-guard-page.cpp

@@ -403,6 +403,7 @@ TEST_CASE("guard page")
                     CHECK(nlohmann::detail::find_string_special(p, n) <= n);
                     CHECK(nlohmann::detail::find_ascii_copyable_run(p, n) <= n);
                     CHECK(nlohmann::detail::valid_utf8_prefix(p, n) <= n);
+                    CHECK(nlohmann::detail::valid_utf8_string(p, n) == (nlohmann::detail::valid_utf8_prefix(p, n) == n));
 #if defined(JSON_USE_SIMDUTF) && defined(JSON_HAS_CPP_17)
                     CHECK(nlohmann::detail::find_string_delimiter(p, n) <= n);
 #endif

tests/src/unit-json_view.cpp

@@ -225,7 +225,7 @@ TEST_CASE("json_view")
 #if JSON_DIAGNOSTICS
         // the parents are set, so errors name the path
         const json m = materialized<json_document>(R"({"a":{"b":[1]}})");
-        CHECK_THROWS_WITH_AS(m.at("a").at("b").at(0).at("x"), "[json.exception.type_error.304] (/a/b/0) cannot use at() with number", json::type_error&);
+        CHECK_THROWS_WITH_AS(m.at("a").at("b").at(0).at("x"), json_view_test::message("type_error.304", "/a/b/0", 11, 12, "cannot use at() with number").c_str(), json::type_error&);
 #endif
     }
 
@@ -750,19 +750,6 @@ std::string exception_of(F f)
     return "";
 }
 
-// the exception a call on a basic_json value throws, without the byte range of
-// the value (JSON_DIAGNOSTIC_POSITIONS), which the exceptions of views lack
-template<typename F>
-std::string json_exception_of(F f)
-{
-    std::string what = exception_of(f);
-    const std::size_t at = what.find("] (bytes ");
-    if (at != std::string::npos)
-    {
-        what.erase(at + 2, what.find(") ", at) - at);
-    }
-    return what;
-}
 #endif
 
 // compares a view with the ordered_json value materialize() gives for it:
@@ -971,24 +958,25 @@ TEST_CASE("json_view element access and iteration")
 #if !defined(JSON_NOEXCEPTION)
             if (!j.is_object())
             {
-                CHECK(exception_of([&] { static_cast<void>(v["a"]); }) == json_exception_of([&] { static_cast<void>(j["a"]); }));
+                CHECK(exception_of([&] { static_cast<void>(v["a"]); }) == exception_of([&] { static_cast<void>(j["a"]); }));
             }
             if (!j.is_array())
             {
-                CHECK(exception_of([&] { static_cast<void>(v[0]); }) == json_exception_of([&] { static_cast<void>(j[0]); }));
+                CHECK(exception_of([&] { static_cast<void>(v[0]); }) == exception_of([&] { static_cast<void>(j[0]); }));
             }
-            CHECK(exception_of([&] { static_cast<void>(v.at("a")); }) == json_exception_of([&] { static_cast<void>(j.at("a")); }));
+            CHECK(exception_of([&] { static_cast<void>(v.at("a")); }) == exception_of([&] { static_cast<void>(j.at("a")); }));
-            CHECK(exception_of([&] { static_cast<void>(v.at("missing")); }) == json_exception_of([&] { static_cast<void>(j.at("missing")); }));
+            CHECK(exception_of([&] { static_cast<void>(v.at("missing")); }) == exception_of([&] { static_cast<void>(j.at("missing")); }));
-            CHECK(exception_of([&] { static_cast<void>(v.at(0)); }) == json_exception_of([&] { static_cast<void>(j.at(0)); }));
+            CHECK(exception_of([&] { static_cast<void>(v.at(0)); }) == exception_of([&] { static_cast<void>(j.at(0)); }));
-            CHECK(exception_of([&] { static_cast<void>(v.at(5)); }) == json_exception_of([&] { static_cast<void>(j.at(5)); }));
+            CHECK(exception_of([&] { static_cast<void>(v.at(5)); }) == exception_of([&] { static_cast<void>(j.at(5)); }));
             if (!(j.is_object() && j.empty())) // (key() of an end iterator)
             {
-                CHECK(exception_of([&] { static_cast<void>(v.begin().key()); }) == json_exception_of([&] { static_cast<void>(j.begin().key()); }));
+                // (without the context of basic_json: view iterators do not know the value they iterate over)
+                CHECK(exception_of([&] { static_cast<void>(v.begin().key()); }) == (j.is_object() ? "" : "[json.exception.invalid_iterator.207] cannot use key() for non-object iterators"));
             }
             if (!j.empty() || j.is_null())
             {
-                CHECK(exception_of([&] { static_cast<void>(v.front()); }) == json_exception_of([&] { static_cast<void>(j.front()); }));
+                CHECK(exception_of([&] { static_cast<void>(v.front()); }) == exception_of([&] { static_cast<void>(j.front()); }));
-                CHECK(exception_of([&] { static_cast<void>(v.back()); }) == json_exception_of([&] { static_cast<void>(j.back()); }));
+                CHECK(exception_of([&] { static_cast<void>(v.back()); }) == exception_of([&] { static_cast<void>(j.back()); }));
             }
 #endif
             CHECK(v.contains("a") == j.contains("a"));
@@ -1000,8 +988,8 @@ TEST_CASE("json_view element access and iteration")
         const json_document d = json_document::parse(R"({"a":[]})");
         CHECK(d.root()["b"].is_discarded());
         CHECK(d.root()["a"][0].is_discarded());
-        CHECK_THROWS_WITH_AS(d.root()["a"].front(), "[json.exception.invalid_iterator.214] cannot get value", json::invalid_iterator&);
+        CHECK_THROWS_WITH_AS(d.root()["a"].front(), json_view_test::message("invalid_iterator.214", "/a", 5, 7, "cannot get value").c_str(), json::invalid_iterator&);
-        CHECK_THROWS_WITH_AS(d.root()["a"].back(), "[json.exception.invalid_iterator.214] cannot get value", json::invalid_iterator&);
+        CHECK_THROWS_WITH_AS(d.root()["a"].back(), json_view_test::message("invalid_iterator.214", "/a", 5, 7, "cannot get value").c_str(), json::invalid_iterator&);
         const json_view invalid{};
         CHECK(invalid.begin() == invalid.end());
         CHECK(std::string(invalid.type_name()) == "discarded");
@@ -1032,10 +1020,10 @@ TEST_CASE("json_view element access and iteration")
         CHECK(v[json_view::json_pointer("/a/b/5")][0].is_discarded());
 #if !defined(JSON_NOEXCEPTION)
         // type errors on values that are not discarded stay
-        CHECK_THROWS_WITH_AS(v[0], "[json.exception.type_error.305] cannot use operator[] with a numeric argument with object", json::type_error&);
+        CHECK_THROWS_WITH_AS(v[0], json_view_test::message("type_error.305", "", 0, 29, "cannot use operator[] with a numeric argument with object").c_str(), json::type_error&);
-        CHECK_THROWS_WITH_AS(v["a"]["b"]["c"], "[json.exception.type_error.305] cannot use operator[] with a string argument with array", json::type_error&);
+        CHECK_THROWS_WITH_AS(v["a"]["b"]["c"], json_view_test::message("type_error.305", "/a/b", 10, 17, "cannot use operator[] with a string argument with array").c_str(), json::type_error&);
-        CHECK_THROWS_WITH_AS(v["s"]["c"], "[json.exception.type_error.305] cannot use operator[] with a string argument with string", json::type_error&);
+        CHECK_THROWS_WITH_AS(v["s"]["c"], json_view_test::message("type_error.305", "/s", 23, 28, "cannot use operator[] with a string argument with string").c_str(), json::type_error&);
-        CHECK_THROWS_WITH_AS(v["s"][0], "[json.exception.type_error.305] cannot use operator[] with a numeric argument with string", json::type_error&);
+        CHECK_THROWS_WITH_AS(v["s"][0], json_view_test::message("type_error.305", "/s", 23, 28, "cannot use operator[] with a numeric argument with string").c_str(), json::type_error&);
         CHECK_THROWS_AS(v["a"]["b"][0]["c"], json::type_error&);
         CHECK_THROWS_AS(v["s"][json_view::json_pointer("/x")], json::out_of_range&);
         // at() keeps throwing on a discarded view
@@ -1110,9 +1098,9 @@ TEST_CASE("json_view element access and iteration")
         CHECK(v[(std::numeric_limits<std::size_t>::max)()].is_discarded());
         CHECK(v[std::numeric_limits<int>::max()].is_discarded());
 #if !defined(JSON_NOEXCEPTION)
-        CHECK_THROWS_WITH_AS(v.at(3), "[json.exception.out_of_range.401] array index 3 is out of range", json::out_of_range&);
+        CHECK_THROWS_WITH_AS(v.at(3), json_view_test::message("out_of_range.401", "", 0, 10, "array index 3 is out of range").c_str(), json::out_of_range&);
-        CHECK_THROWS_WITH_AS(v.at(3u), "[json.exception.out_of_range.401] array index 3 is out of range", json::out_of_range&);
+        CHECK_THROWS_WITH_AS(v.at(3u), json_view_test::message("out_of_range.401", "", 0, 10, "array index 3 is out of range").c_str(), json::out_of_range&);
-        CHECK_THROWS_WITH_AS(v.at(std::int64_t(3)), "[json.exception.out_of_range.401] array index 3 is out of range", json::out_of_range&);
+        CHECK_THROWS_WITH_AS(v.at(std::int64_t(3)), json_view_test::message("out_of_range.401", "", 0, 10, "array index 3 is out of range").c_str(), json::out_of_range&);
         CHECK_THROWS_AS(v.at(-1), json::out_of_range&);
         CHECK_THROWS_AS(v.at(-1L), json::out_of_range&);
         CHECK_THROWS_AS(v.at(std::int64_t(-1)), json::out_of_range&);
@@ -1246,26 +1234,6 @@ TEST_CASE("json_view element access and iteration")
 
 namespace
 {
-#if !defined(JSON_NOEXCEPTION)
-// an exception message without the context that basic_json adds with
-// JSON_DIAGNOSTICS ("(/path) ") and JSON_DIAGNOSTIC_POSITIONS ("(bytes 1-2) ");
-// the view's exceptions have no such context
-std::string without_path(std::string msg)
-{
-    for (const char* prefix :
-            {"] (/", "] (bytes "
-            })
-    {
-        const std::size_t open = msg.find(prefix);
-        if (open != std::string::npos)
-        {
-            msg.erase(open + 2, msg.find(") ", open) + 2 - (open + 2));
-        }
-    }
-    return msg;
-}
-#endif
-
 // the bits of a float, to compare values bit for bit
 std::uint64_t bits(double x)
 {
@@ -1336,18 +1304,18 @@ void check_values(const ordered_json_view& v, const ordered_json& j, const std::
     // conversions to the wrong type throw what basic_json throws
     if (!j.is_number())
     {
-        CHECK(exception_of([&] { static_cast<void>(v.get<int>()); }) == without_path(exception_of([&] { static_cast<void>(j.get<int>()); })));
+        CHECK(exception_of([&] { static_cast<void>(v.get<int>()); }) == exception_of([&] { static_cast<void>(j.get<int>()); }));
     }
-    CHECK(exception_of([&] { static_cast<void>(v.get<bool>()); }) == without_path(exception_of([&] { static_cast<void>(j.get<bool>()); })));
+    CHECK(exception_of([&] { static_cast<void>(v.get<bool>()); }) == exception_of([&] { static_cast<void>(j.get<bool>()); }));
-    CHECK(exception_of([&] { static_cast<void>(v.get<std::string>()); }) == without_path(exception_of([&] { static_cast<void>(j.get<std::string>()); })));
+    CHECK(exception_of([&] { static_cast<void>(v.get<std::string>()); }) == exception_of([&] { static_cast<void>(j.get<std::string>()); }));
-    CHECK(exception_of([&] { static_cast<void>(v.get<std::nullptr_t>()); }) == without_path(exception_of([&] { static_cast<void>(j.get<std::nullptr_t>()); })));
+    CHECK(exception_of([&] { static_cast<void>(v.get<std::nullptr_t>()); }) == exception_of([&] { static_cast<void>(j.get<std::nullptr_t>()); }));
     if (!j.is_array())
     {
-        CHECK(exception_of([&] { static_cast<void>(v.get<std::vector<int>>()); }) == without_path(exception_of([&] { static_cast<void>(j.get<std::vector<int>>()); })));
+        CHECK(exception_of([&] { static_cast<void>(v.get<std::vector<int>>()); }) == exception_of([&] { static_cast<void>(j.get<std::vector<int>>()); }));
     }
     if (!j.is_object())
     {
-        CHECK(exception_of([&] { static_cast<void>(v.get<std::map<std::string, int>>()); }) == without_path(exception_of([&] { static_cast<void>(j.get<std::map<std::string, int>>()); })));
+        CHECK(exception_of([&] { static_cast<void>(v.get<std::map<std::string, int>>()); }) == exception_of([&] { static_cast<void>(j.get<std::map<std::string, int>>()); }));
     }
 #endif
 
@@ -1490,8 +1458,8 @@ TEST_CASE("json_view values")
         CHECK(v[3].number_token() == "123456789012345678901234567890");
         CHECK(v[4].number_token() == "-12");
         CHECK(v[5].number_token() == "7");
-        CHECK_THROWS_WITH_AS(v[6].number_token(), "[json.exception.type_error.302] type must be number, but is string", json::type_error&);
+        CHECK_THROWS_WITH_AS(v[6].number_token(), json_view_test::message("type_error.302", "/6", 56, 59, "type must be number, but is string").c_str(), json::type_error&);
-        CHECK_THROWS_WITH_AS(v.get_string(), "[json.exception.type_error.302] type must be string, but is array", json::type_error&);
+        CHECK_THROWS_WITH_AS(v.get_string(), json_view_test::message("type_error.302", "", 0, 60, "type must be string, but is array").c_str(), json::type_error&);
     }
 
     SECTION("conversions")
@@ -1565,15 +1533,15 @@ TEST_CASE("json_view values")
         CHECK(v["o"]["x"].value(json::json_pointer("/1"), 0) == j["o"]["x"].value(json::json_pointer("/1"), 0));
         CHECK(v["o"]["x"].value(json::json_pointer("/7"), 3) == j["o"]["x"].value(json::json_pointer("/7"), 3));
 #if !defined(JSON_NOEXCEPTION)
-        CHECK(exception_of([&] { static_cast<void>(v["o"]["x"].value("k", 0)); }) == without_path(exception_of([&] { static_cast<void>(j["o"]["x"].value("k", 0)); })));
+        CHECK(exception_of([&] { static_cast<void>(v["o"]["x"].value("k", 0)); }) == exception_of([&] { static_cast<void>(j["o"]["x"].value("k", 0)); }));
-        CHECK(exception_of([&] { static_cast<void>(v.value("s", 0)); }) == without_path(exception_of([&] { static_cast<void>(j.value("s", 0)); })));
+        CHECK(exception_of([&] { static_cast<void>(v.value("s", 0)); }) == exception_of([&] { static_cast<void>(j.value("s", 0)); }));
-        CHECK(exception_of([&] { static_cast<void>(v["n"].value("x", 0)); }) == without_path(exception_of([&] { static_cast<void>(j["n"].value("x", 0)); })));
+        CHECK(exception_of([&] { static_cast<void>(v["n"].value("x", 0)); }) == exception_of([&] { static_cast<void>(j["n"].value("x", 0)); }));
-        CHECK(exception_of([&] { static_cast<void>(v["n"].value(json::json_pointer("/x"), 0)); }) == without_path(exception_of([&] { static_cast<void>(j["n"].value(json::json_pointer("/x"), 0)); })));
+        CHECK(exception_of([&] { static_cast<void>(v["n"].value(json::json_pointer("/x"), 0)); }) == exception_of([&] { static_cast<void>(j["n"].value(json::json_pointer("/x"), 0)); }));
         // string defaults: the member is not a string, or the value is neither an object nor an array
-        CHECK(exception_of([&] { static_cast<void>(v.value("n", "default")); }) == without_path(exception_of([&] { static_cast<void>(j.value("n", "default")); })));
+        CHECK(exception_of([&] { static_cast<void>(v.value("n", "default")); }) == exception_of([&] { static_cast<void>(j.value("n", "default")); }));
-        CHECK(exception_of([&] { static_cast<void>(v["n"].value("x", "default")); }) == without_path(exception_of([&] { static_cast<void>(j["n"].value("x", "default")); })));
+        CHECK(exception_of([&] { static_cast<void>(v["n"].value("x", "default")); }) == exception_of([&] { static_cast<void>(j["n"].value("x", "default")); }));
-        CHECK(exception_of([&] { static_cast<void>(v.value(json::json_pointer("/n"), "default")); }) == without_path(exception_of([&] { static_cast<void>(j.value(json::json_pointer("/n"), "default")); })));
+        CHECK(exception_of([&] { static_cast<void>(v.value(json::json_pointer("/n"), "default")); }) == exception_of([&] { static_cast<void>(j.value(json::json_pointer("/n"), "default")); }));
-        CHECK(exception_of([&] { static_cast<void>(v["s"].value(json::json_pointer("/x"), "default")); }) == without_path(exception_of([&] { static_cast<void>(j["s"].value(json::json_pointer("/x"), "default")); })));
+        CHECK(exception_of([&] { static_cast<void>(v["s"].value(json::json_pointer("/x"), "default")); }) == exception_of([&] { static_cast<void>(j["s"].value(json::json_pointer("/x"), "default")); }));
 #endif
         // (defaults and members longer than the short string optimization)
         CHECK(v.value("s", "a default longer than fifteen characters") == "text");
@@ -1632,7 +1600,7 @@ TEST_CASE("json_view JSON pointers")
         {
             CAPTURE(pointer)
             const json::json_pointer p(pointer);
-            const std::string at_error = without_path(exception_of([&] { static_cast<void>(j.at(p)); }));
+            const std::string at_error = exception_of([&] { static_cast<void>(j.at(p)); });
             CHECK(exception_of([&] { static_cast<void>(v.at(p)); }) == at_error);
             if (at_error.empty())
             {
@@ -1646,7 +1614,7 @@ TEST_CASE("json_view JSON pointers")
             }
             else
             {
-                CHECK(exception_of([&] { static_cast<void>(v[p]); }) == without_path(exception_of([&] { static_cast<void>(j[p]); })));
+                CHECK(exception_of([&] { static_cast<void>(v[p]); }) == exception_of([&] { static_cast<void>(j[p]); }));
             }
             // (basic_json::contains() throws out_of_range.404 for an empty
             // array index token, although it is not meant to throw; the view
@@ -1657,7 +1625,7 @@ TEST_CASE("json_view JSON pointers")
                 static_cast<void>(found);
             });
             CHECK(v.contains(p) == (contains_error.empty() && j.contains(p)));
-            CHECK(exception_of([&] { static_cast<void>(v.value(p, 5)); }) == without_path(exception_of([&] { static_cast<void>(j.value(p, 5)); })));
+            CHECK(exception_of([&] { static_cast<void>(v.value(p, 5)); }) == exception_of([&] { static_cast<void>(j.value(p, 5)); }));
         }
     }
 #endif

tests/src/unit-json_view_builder.cpp

@@ -13,6 +13,7 @@
     #include <nlohmann/detail/view/builder.hpp>
     #include <nlohmann/detail/view/materialize.hpp>
     #include <nlohmann/detail/view/string_ref.hpp>
+    #include <nlohmann/json_view.hpp> // json_document::accept
 #else
     #include <nlohmann/json_view.hpp> // the single header contains the internal headers
 #endif
@@ -59,7 +60,8 @@ built build(const std::string& text, bool comments, bool trailing_commas, bool s
         r.copy.assign(text.begin(), text.end());
         src = r.copy.data();
     }
-    r.ok = nlohmann::detail::view::build < FloatType, !nlohmann::detail::abi_config::strict_nul_handling > (*r.data, src, text.size(), comments, trailing_commas, sentinel, r.failure);
+    constexpr bool nul_is_end = !nlohmann::detail::abi_config::strict_nul_handling;
+    r.ok = nlohmann::detail::view::build<FloatType, nul_is_end>(*r.data, src, text.size(), comments, trailing_commas, sentinel, false, r.failure);
     r.data->src = src;
     r.data->base[0] = src;
     r.data->base[1] = r.data->arena.data();
@@ -172,6 +174,22 @@ void check_string(const std::string& content)
     }
 }
 
+// accept() stores no nodes: it must accept and reject what json::accept does, for all options and for inputs with
+// and without a terminating NUL
+void check_accept(const std::string& text)
+{
+    CAPTURE(text)
+    const std::vector<char> bytes(text.begin(), text.end());
+    for (int options = 0; options < 4; ++options)
+    {
+        const bool comments = (options & 1) != 0;
+        const bool trailing_commas = (options & 2) != 0;
+        const bool accepted = json::accept(text, comments, trailing_commas);
+        CHECK(nlohmann::json_document::accept(text, comments, trailing_commas) == accepted);
+        CHECK(nlohmann::json_document::accept(bytes, comments, trailing_commas) == accepted);
+    }
+}
+
 // a small deterministic generator of documents
 struct generator
 {
@@ -423,6 +441,133 @@ TEST_CASE("json_view builder")
         }
     }
 
+    SECTION("accept() without an index")
+    {
+        for (const char* text :
+                {
+                    "", " ", "\n", "null", "nul", "true", "[", "]", "[]", "{}", "[1,2,3]", "[1,2,3", "[1,2,3,]", "{\"a\":1,}", "{\"a\"}", "{\"a\":}",
+                    "[1] x", "[1][2]", "\xEF\xBB\xBF[1]", "[\"a\\", R"(["\u12"])", R"(["\ud800"])", "[\"\x01\"]", "[\"\xff\"]", "[01]", "[1.]",
+                    "[.5]", "[-]", "[1e]", "// c\n[1]", "/* c */ {\"a\" /* x */ : 1}", "/* c", "[1] // c", "1e400", "-1e400", "123456789012345678901234567890",
+                    R"({"a":1,"a":2})", "[[[]]]", R"("\u00e4\n\ud83d\ude00")"
+                })
+        {
+            check_accept(text);
+        }
+        // a null byte inside the text, and after it
+        check_accept(std::string("[1,\0]", 5));
+        check_accept(std::string("[1]\0", 4));
+        check_accept(std::string("\0", 1));
+
+        // generated documents and damaged copies, as for the builder above
+        generator g;
+        for (int i = 0; i < 2000; ++i)
+        {
+            std::string text;
+            g.value(text, 0);
+            check_accept(text);
+            std::string damaged = text;
+            const auto at = static_cast<std::size_t>(g.r(static_cast<int>(damaged.size())));
+            static const char replacements[] = {'x', '"', '\\', ',', ':', ']', '}', '[', '{', '1', '-', '.', 'e', '\0', '\n', '/'}; // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+            damaged[at] = replacements[g.r(16)];
+            check_accept(damaged);
+            check_accept(text.substr(0, at));
+        }
+
+        // nesting beyond the 64 levels that are kept on the builder's own stack, closed and not closed
+        for (const std::size_t depth :
+                {
+                    63u, 64u, 65u, 66u, 200u, 1000u
+                })
+        {
+            check_accept(std::string(depth, '[') + std::string(depth, ']'));
+            check_accept(std::string(depth, '[') + std::string(depth - 1, ']'));
+            std::string objects;
+            for (std::size_t i = 0; i < depth; ++i)
+            {
+                objects += "{\"k\":";
+            }
+            objects += "1" + std::string(depth, '}');
+            check_accept(objects);
+            check_accept(objects.substr(0, objects.size() - 1));
+        }
+
+        // objects with enough members to get a hash index when they are indexed (and duplicate keys)
+        std::string large = "{";
+        for (int i = 0; i < 300; ++i)
+        {
+            large += (i == 0 ? "\"k" : ",\"k") + std::to_string(i % 200) + "\":" + std::to_string(i);
+        }
+        check_accept(large + "}");
+        check_accept(large);
+        check_accept("[" + large + "},{" + large.substr(1) + "}]");
+
+        // the overflow check uses the floating-point type of the document
+        using float_json = nlohmann::basic_json<std::map, std::vector, std::string, bool, std::int64_t, std::uint64_t, float>;
+        for (const char* text :
+                {
+                    "1e39", "-1e39", "3.4028235e38", "3.4028236e38", "[1.5e38, 3.5e38]", R"({"a": 1e-50, "b": 1e39})", "340282356779733661637539395458142568448"
+                })
+        {
+            CAPTURE(text)
+            CHECK(nlohmann::basic_json_document<float_json>::accept(text) == float_json::accept(text));
+        }
+
+        // strings are checked and not decoded: runs of text of every length from 0 to 40 between valid and invalid
+        // escapes, UTF-8 sequences and control characters, in values and in keys, cut near the end of the text
+        const std::array<const char*, 9> valid = {{"\\n", "\\\"", "\\\\", "\\/", "\\t", "\\u00e4", "\\u20ac", "\\ud83d\\ude00", "\\u0000"}};
+        const std::array<const char*, 12> invalid = {{"\\x", "\\ud800", "\\udc00", "\\u12G4", "\\u12", "\\ud83dx", "\\ud83d\\u0041", "\\ud83d\\ud83d", "\\", "\x01", "\x1f", "\x7f"}};
+        const std::array<const char*, 6> utf8 = {{"\xc3\xa4", "\xe2\x82\xac", "\xf0\x9f\x98\x80", "\xc3", "\xff", "\x80"}};
+        std::mt19937 rng(2029); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed): reproducible
+        for (std::size_t run = 0; run <= 40; ++run)
+        {
+            for (int i = 0; i < 30; ++i)
+            {
+                std::string content = std::string(rng() % 20, 'a');
+                const auto pieces = 1 + (rng() % 6);
+                for (unsigned k = 0; k < pieces; ++k)
+                {
+                    content += valid[rng() % valid.size()];
+                    content += std::string(k == 0 ? run : rng() % 41, static_cast<char>('b' + (rng() % 20)));
+                    switch (rng() % 8)
+                    {
+                        case 0:
+                            content += utf8[rng() % 3];
+                            break;
+                        case 1:
+                            content += utf8[rng() % utf8.size()];
+                            break;
+                        case 2:
+                            content += invalid[rng() % invalid.size()];
+                            break;
+                        default:
+                            break;
+                    }
+                }
+                for (const std::string& text :
+                        {
+                            "[\"" + content + "\"]", "{\"" + content + "\":1}", "[\"" + content, "[\"" + content.substr(0, content.size() - (rng() % 4)) + "\"]"
+                        })
+                {
+                    check_accept(text);
+                }
+            }
+        }
+    }
+
+    SECTION("escapes in a row")
+    {
+        for (const char* content :
+                {
+                    "\\n", R"(\n\t\\)", R"(\u00e9\u00e9\u4e2d)", "\\u00e9\\u00e", "\\u00e9\\", "\\n\\q", R"(\n\ud83d\ude00\n)",
+                    R"(\ud83d\ude00\ud83d)", R"(\n\n\n abc \n\n)", "\\n\x01", "\\n\xC3\x28", "\\n\\"
+                })
+        {
+            check_accept(std::string("[\"") + content + "\"]");
+            check_accept(std::string("[\"") + content);
+            check_accept(std::string("{\"") + content + "\":1}");
+        }
+    }
+
     SECTION("test files")
     {
         for (const char* name :
@@ -523,6 +668,54 @@ TEST_CASE("json_view builder: strings across vector blocks")
     }
 }
 
+TEST_CASE("json_view builder: fraction digits")
+{
+    // The digits of a fraction are skipped 16 bytes at a time where there are
+    // that many left (SSE2), and one at a time after that. Fractions of every
+    // length from 0 to 40 digits, followed by every kind of byte that ends a
+    // number or lies next to the digits in ASCII ('/' and ':'), as values, as
+    // keys' neighbours, and at the very end of the text (a vector without a
+    // terminating NUL). The length of the number is part of the value that
+    // the test compares with json::parse().
+    const std::string followers = ",]}/:.eE+-x \n\t\"09";
+    std::mt19937 rng(2028); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed): reproducible
+    for (std::size_t digits = 0; digits <= 40; ++digits)
+    {
+        for (int i = 0; i < 3; ++i)
+        {
+            std::string fraction;
+            for (std::size_t k = 0; k < digits; ++k)
+            {
+                fraction += static_cast<char>('0' + (rng() % 10));
+            }
+            const std::string number = std::string(rng() % 2 == 0 ? "-" : "") + std::to_string(rng() % 100000) + "." + fraction;
+            check_same(number);
+            check_same("[" + number + "]");
+            std::string pair;
+            pair.reserve((number.size() * 2) + 3);
+            pair += '[';
+            pair += number;
+            pair += ',';
+            pair += number;
+            pair += ']';
+            check_same(pair);
+            check_same("{\"a\":" + number + "}");
+            for (const char follower : followers)
+            {
+                check_same(number + follower);
+                check_same("[" + number + follower + "]");
+                check_same("[" + number + follower + std::string(rng() % 20, '1'));
+            }
+            // everything near the end of the text: the number cut in every place
+            for (std::size_t cut = 1; cut <= 18 && cut <= number.size(); ++cut)
+            {
+                check_same(number.substr(0, number.size() - cut));
+                check_same("[" + number.substr(0, number.size() - cut));
+            }
+        }
+    }
+}
+
 TEST_CASE("json_view builder: escapes between runs of text")
 {
     // The text between escapes is copied 16 bytes at a time where there are
@@ -575,6 +768,59 @@ TEST_CASE("json_view builder: escapes between runs of text")
     }
 }
 
+TEST_CASE("json_view builder: chains of unicode escapes")
+{
+    // Consecutive \uXXXX escapes are decoded in a loop of their own. Chains of
+    // every length from 1 to 40 of escapes that give one, two, and three
+    // bytes of UTF-8, in upper and lower case, followed by what ends a chain:
+    // the end of the string, text, another escape, UTF-8, and everything that
+    // is not a valid escape (a bad digit at every position of the last
+    // escape, a lone surrogate, a surrogate pair, a cut in every place).
+    const std::array<const char*, 8> chain = {{"\\u0041", "\\u00e4", "\\u00E4", "\\u20ac", "\\u3042", "\\uFFFF", "\\u0000", "\\u07ff"}};
+    const std::array<const char*, 12> ends = {{"", "abc", "\\n", "\\u00e9", "\\ud83d\\ude00", "\\ud83d", "\\ude00", "\\ud83d\\u0041", "\\u12G4", "\\u12", "\xc3\xa4", "\\"}};
+    std::mt19937 rng(2027); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed): reproducible
+    for (std::size_t length = 1; length <= 40; ++length)
+    {
+        for (int i = 0; i < 20; ++i)
+        {
+            std::string content = std::string(rng() % 20, 'a');
+            for (std::size_t k = 0; k < length; ++k)
+            {
+                content += chain[rng() % chain.size()];
+            }
+            const std::string whole = content;
+            for (const char* end : ends)
+            {
+                check_string(whole + end);
+                check_string(whole + end + std::string(rng() % 19, 'z'));
+            }
+            // a bad digit at every position of every escape near the end of the chain
+            for (std::size_t back = 1; back <= 6 && back <= length; ++back)
+            {
+                for (std::size_t digit = 2; digit < 6; ++digit)
+                {
+                    std::string broken = whole;
+                    broken[broken.size() - (6 * back) + digit] = "gG\"\x01 z"[rng() % 6];
+                    check_string(broken);
+                }
+            }
+            // the text cut in every place of the last three escapes
+            for (std::size_t cut = 0; cut <= 18 && cut <= whole.size(); ++cut)
+            {
+                check_string(whole.substr(0, whole.size() - cut));
+            }
+        }
+    }
+    // a chain longer than the first growth of the arena, in a key and in a value
+    std::string long_chain;
+    for (int i = 0; i < 3000; ++i)
+    {
+        long_chain += chain[static_cast<std::size_t>(i) % chain.size()];
+    }
+    check_string(long_chain);
+    check_string(long_chain + "\\ud83d\\ude00" + long_chain);
+}
+
 TEST_CASE("json_view node integer bits")
 {
     using nlohmann::detail::view::integer_bits;

tests/src/unit-json_view_dump.cpp

@@ -35,9 +35,9 @@ using json_view_test::has_duplicate_keys;
 #include <string>
 #include <vector>
 
-// These tests were split off unit-json_view.cpp, whose object file got too large
+// These tests were split off unit-json_view.cpp, whose object file got too large for the MinGW
-// for the MinGW linker (see json_view_test_helpers.hpp). Like that file, this one
+// linker (see json_view_test_helpers.hpp). Like that file, this one is also built
-// is also built with C++17, as it mentions JSON_HAS_CPP_17.
+// with C++17, as it mentions JSON_HAS_CPP_17.
 
 TEST_CASE("json_view dump")
 {
@@ -150,6 +150,88 @@ TEST_CASE("json_view dump")
             CHECK(many_doc.root().dump() == json::parse(many_tokens).dump());
         }
 
+        // tokens of 16 and 17 digits without an exponent (like GeoJSON
+        // coordinates), which dump() converts in batches: doubles exactly
+        // halfway between two candidates for the shortest digits, random
+        // digits, and digits near a shorter decimal (...99999997,
+        // ...00000003), whose rounding carries; in pairs and among other
+        // values, so that the batches are interrupted; the last token near the
+        // end of the text
+        {
+            std::mt19937_64 tokens(5627); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed)
+            const auto draw = [&tokens](std::size_t n)
+            {
+                const std::uint64_t r = tokens() % n;
+                return static_cast<std::size_t>(r);
+            };
+            std::string coordinates = "[754570942.33984373,257769445.56835938,858106901.94921875,97944964.876953125,"
+                                      "[0.5,1e-7],[1.5,-0.0,2.5],[1E2,12.5,0.1,3.25,4.5,5.75,6.125,7.0,8.5],[],[1.5],[1,2.5],[2.5,\"x\"],[[1.5],2.5],";
+            for (int i = 0; i < 20000; ++i)
+            {
+                const std::size_t length = 16 + draw(2);
+                std::string digits(1, static_cast<char>('1' + draw(9)));
+                const std::size_t kept = draw(3) == 0 ? length : 2 + draw(length - 3);
+                for (std::size_t k = 1; k < kept; ++k)
+                {
+                    digits += static_cast<char>('0' + draw(10));
+                }
+                digits += std::string(length - kept, draw(2) == 0 ? '9' : '0');
+                digits.back() = static_cast<char>('0' + draw(10));
+                const std::size_t point = 1 + draw(9);
+                const std::string token = (draw(2) == 0 ? "-" : "") + digits.substr(0, point) + "." + digits.substr(point);
+                if (i != 0)
+                {
+                    coordinates += ',';
+                }
+                if (draw(8) == 0)
+                {
+                    coordinates += R"({"x":)";
+                    coordinates += token;
+                    coordinates += R"(,"s":"a"})";
+                }
+                else
+                {
+                    coordinates += '[';
+                    coordinates += token;
+                    coordinates += ',';
+                    coordinates += token;
+                    coordinates += ']';
+                }
+            }
+            coordinates += ",-65.613616999999977]";
+            const json_document parsed = json_document::parse(coordinates);
+            const ordered_json j = ordered_json::parse(coordinates);
+            CHECK(parsed.root().dump() == j.dump());
+            CHECK(parsed.root().dump(1) == j.dump(1));
+            CHECK(parsed.root().dump(-1, ' ', true) == j.dump(-1, ' ', true));
+            CHECK(parsed.root().back().get<double>() == j.back().get<double>());
+            CHECK(parsed.root().back().dump() == "-65.61361699999998");
+            std::size_t i = 0;
+            for (const json_view v : parsed.root())
+            {
+                const ordered_json& e = j[i++];
+                CHECK(v.dump() == e.dump());
+                if (v.is_array() && !v.empty() && v[0].is_number())
+                {
+                    CHECK(v[0].get<double>() == e[0].get<double>());
+                }
+            }
+
+            // an editable document: edits move some arrays, so that the walk
+            // leaves the order of the parsed nodes
+            ordered_json_editable_document editable = ordered_json_editable_document::parse(coordinates);
+            ordered_json expected = j;
+            for (std::size_t k = 0; k < 20000; k += 97)
+            {
+                if (expected[k].is_array())
+                {
+                    editable.push_back(editable.root()[k], ordered_json(1.5));
+                    expected[k].push_back(1.5);
+                }
+            }
+            CHECK(editable.root().dump() == expected.dump());
+        }
+
         // random doubles, written as parse() and dump() would
         std::mt19937_64 rng(1170); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed)
         std::string many = "[";
@@ -171,673 +253,4 @@ TEST_CASE("json_view dump")
         const nlohmann::basic_json_document<json_float> float_document = nlohmann::basic_json_document<json_float>::parse("[0.1, 1.5e10, 3.4028235e38]");
         CHECK(float_document.root().dump() == json_float::parse("[0.1, 1.5e10, 3.4028235e38]").dump());
     }
-
-    SECTION("members in document order, all of them")
-    {
-        const json_document d = json_document::parse(R"({"b": 1, "a": 2, "b": 3})");
-        CHECK(d.root().dump() == R"({"b":1,"a":2,"b":3})");
-        CHECK(d.root().dump(1) == "{\n \"b\": 1,\n \"a\": 2,\n \"b\": 3\n}");
-        // dump() writes all members, though the lookup finds the first
-        CHECK(d.root()["b"].dump() == "1");
-    }
-
-    SECTION("deep nesting")
-    {
-        const std::string deep = std::string(100000, '[') + std::string(100000, ']');
-        const json_document deep_document = json_document::parse(deep);
-        CHECK(deep_document.root().dump() == deep);
-    }
-
-    SECTION("ensure_ascii: a character that needs an escape at each position of a string")
-    {
-        // string lengths around the width of the vector steps, in a string value, a key, and a value near the end of the text
-        for (const char* const special :
-                {
-                    "\xC3\xA9", "\xE4\xB8\xAD", "\xF0\x9F\x98\x80", "\x7F", "\\n", "\\u0001", "\\\"", "\\\\"
-                })
-        {
-            for (std::size_t length = 0; length <= 40; ++length)
-            {
-                for (std::size_t position = 0; position <= length; ++position)
-                {
-                    std::string content(length, 'a');
-                    content.insert(position, special);
-                    const std::array<std::string, 3> texts =
-                    {
-                        {
-                            "\"" + content + "\"",
-                            "[\"" + content + "\"]",
-                            "{\"" + content + "\":1}",
-                        }
-                    };
-                    for (const std::string& text : texts)
-                    {
-                        CAPTURE(text)
-                        const ordered_json_document d = ordered_json_document::parse(text);
-                        const ordered_json j = ordered_json::parse(text);
-                        CHECK(d.root().dump(-1, ' ', true) == j.dump(-1, ' ', true));
-                        CHECK(d.root().dump(2, ' ', true) == j.dump(2, ' ', true));
-                        CHECK(d.root().dump() == j.dump());
-                    }
-                }
-            }
-        }
-    }
-
-    SECTION("ensure_ascii: every character of the Basic Multilingual Plane and some beyond")
-    {
-        // (the two- and three-byte sequences, including the ranges that need a check: E0, ED, EF BF BF)
-        std::string text = "[";
-        std::size_t in_chunk = 0;
-        const auto append = [&](std::uint32_t cp)
-        {
-            if (cp < 0x80)
-            {
-                if (cp < 0x20 || cp == '"' || cp == '\\')
-                {
-                    return;
-                }
-                text += static_cast<char>(cp);
-            }
-            else if (cp < 0x800)
-            {
-                text += static_cast<char>(0xC0 | (cp >> 6));
-                text += static_cast<char>(0x80 | (cp & 0x3F));
-            }
-            else if (cp < 0x10000)
-            {
-                text += static_cast<char>(0xE0 | (cp >> 12));
-                text += static_cast<char>(0x80 | ((cp >> 6) & 0x3F));
-                text += static_cast<char>(0x80 | (cp & 0x3F));
-            }
-            else
-            {
-                text += static_cast<char>(0xF0 | (cp >> 18));
-                text += static_cast<char>(0x80 | ((cp >> 12) & 0x3F));
-                text += static_cast<char>(0x80 | ((cp >> 6) & 0x3F));
-                text += static_cast<char>(0x80 | (cp & 0x3F));
-            }
-        };
-        for (std::uint32_t cp = 0x20; cp < 0x10400; ++cp)
-        {
-            if (cp >= 0xD800 && cp < 0xE000)
-            {
-                continue;
-            }
-            if (in_chunk == 0)
-            {
-                text += text.size() > 1 ? ",\"" : "\"";
-            }
-            append(cp);
-            if (++in_chunk == 37)
-            {
-                text += "\"";
-                in_chunk = 0;
-            }
-        }
-        text += in_chunk != 0 ? "\"]" : "]";
-        const ordered_json_document d = ordered_json_document::parse(text);
-        const ordered_json j = ordered_json::parse(text);
-        CHECK(d.root().dump(-1, ' ', true) == j.dump(-1, ' ', true));
-        CHECK(d.root().dump(2, ' ', true) == j.dump(2, ' ', true));
-        CHECK(d.root().dump() == j.dump());
-        CHECK(d.root().dump(2) == j.dump(2));
-    }
-
-    SECTION("deep nesting, indented")
-    {
-        // indentations of up to and beyond the width of the writer's fixed-size copy
-        for (const int depth :
-                {
-                    1, 5, 20, 70, 300
-                })
-        {
-            std::string text;
-            for (int i = 0; i < depth; ++i)
-            {
-                text += i % 2 == 0 ? "[1," : R"({"k":2,"\u00e9":)";
-            }
-            text += "[]";
-            for (int i = depth - 1; i >= 0; --i)
-            {
-                text += i % 2 == 0 ? "]" : "}";
-            }
-            const ordered_json_document d = ordered_json_document::parse(text);
-            const ordered_json j = ordered_json::parse(text);
-            for (const int indent :
-                    {
-                        0, 1, 2, 3, 8
-                    })
-            {
-                CHECK(d.root().dump(indent) == j.dump(indent));
-                CHECK(d.root().dump(indent, '\t') == j.dump(indent, '\t'));
-            }
-        }
-    }
-
-    SECTION("the output buffer of a small value is small")
-    {
-        // an escaped key after the value: its node lies in the arena, so the
-        // source extent of the value cannot be read from the next node
-        const std::string big(100000, 'a');
-        const std::string text = R"({"small":1,"list":[1,2,3],"k\n":")" + big + R"("})";
-        const json_document d = json_document::parse(text);
-        const auto small = d.root()["small"].dump();
-        CHECK(small == "1");
-        CHECK(small.capacity() < 4096);
-        const auto list = d.root()["list"].dump();
-        CHECK(list == "[1,2,3]");
-        CHECK(list.capacity() < 4096);
-        CHECK(d.root()["list"].dump(2).capacity() < 4096);
-
-        // the whole document and the large value are unaffected
-        CHECK(d.root().dump() == ordered_json::parse(text).dump());
-        CHECK(d.root()["k\n"].dump() == "\"" + big + "\"");
-    }
-
-    SECTION("output that outgrows the estimate")
-    {
-        // ensure_ascii writes six bytes for each two-byte character
-        std::string chars;
-        for (int i = 0; i < 5000; ++i)
-        {
-            chars += "\xC3\xA9";
-        }
-        const json_document d = json_document::parse(R"({"a":")" + chars + R"(","k\n":1})");
-        const json expected = json::parse("\"" + chars + "\"");
-        CHECK(d.root()["a"].dump(-1, ' ', true) == expected.dump(-1, ' ', true));
-        CHECK(d.root()["a"].dump(-1, ' ', true).size() == 2 + (5000 * 6));
-    }
-
-    SECTION("ensure_ascii with strings of the source, decoded strings, and edited strings")
-    {
-        // plain ASCII, non-ASCII, escaped (decoded), and mixed strings, as values and as keys;
-        // the second text has a DEL, which ensure_ascii writes as an escape
-        for (const char* const del :
-                {"", "\x7F"
-                })
-        {
-            const std::string text = std::string("{\"plain\":\"abc\",\"caf\xC3\xA9\":\"\xC3\xA9\xE2\x82\xAC\xF0\x9F\x98\x80\",")
-                                     + "\"esc\\n\\\"\":\"a\\tb\\u00e4\\ud83d\\ude00\",\"del\":\"a" + del + "b\",\"\":\"\"," // NOLINT(modernize-raw-string-literal)
-                                     + "\"list\":[\"x\",\"\xC3\xBC\",\"\\\\\",\"long ASCII string with more than thirty-two bytes in it\","
-                                     + "\"long string with more than thirty-two bytes and one non-ASCII byte \xC3\xB6\"],"
-                                     + "\"n\":[1,2.5,-0,true,null]}";
-            const ordered_json j = ordered_json::parse(text);
-            const ordered_json_document d = ordered_json_document::parse(text);
-            CHECK(d.root().dump(-1, ' ', true) == j.dump(-1, ' ', true));
-            CHECK(d.root().dump(4, ' ', true) == j.dump(4, ' ', true));
-            CHECK(d.root().dump() == j.dump());
-            CHECK(d.root()["list"].dump(-1, ' ', true) == j["list"].dump(-1, ' ', true));
-            CHECK(d.root()["del"].dump(-1, ' ', true) == j["del"].dump(-1, ' ', true));
-
-            // strings that an edit wrote (the source strings that they replace were plain)
-            ordered_json_editable_document e = ordered_json_editable_document::parse(text);
-            ordered_json k = j;
-            e.set(e.root()["plain"], std::string("\xC3\xA4gypten"));
-            k["plain"] = "\xC3\xA4gypten";
-            e.set(e.root()["list"][0], std::string("ascii after edit"));
-            k["list"][0] = "ascii after edit";
-            e.set(e.root()["list"][1], std::string("\xC3\xA4"));
-            k["list"][1] = "\xC3\xA4";
-            e.set(e.root()["del"], std::string("tab\there"));
-            k["del"] = "tab\there";
-            CHECK(e.root().dump(-1, ' ', true) == k.dump(-1, ' ', true));
-            CHECK(e.root().dump(2, ' ', true) == k.dump(2, ' ', true));
-            CHECK(e.root().dump() == k.dump());
-
-            // strings that were edited from or to ASCII text
-            e.set(e.root()["list"][2], std::string("\xC3\xA4"));
-            k["list"][2] = "\xC3\xA4";
-            e.set(e.root()["list"][4], std::string("plain again"));
-            k["list"][4] = "plain again";
-            CHECK(e.root().dump(-1, ' ', true) == k.dump(-1, ' ', true));
-        }
-    }
-
-    SECTION("ensure_ascii with DEL or a non-ASCII character at every position")
-    {
-        // the check whether a string of the source has no byte >= 0x7F reads
-        // eight bytes at a time and then the rest one by one
-        for (std::size_t length = 1; length <= 20; ++length)
-        {
-            for (std::size_t pos = 0; pos < length; ++pos)
-            {
-                for (const char* const c :
-                        {
-                            "\x7F", "\x7E", "\xC3\xA4", "\xC3\xBF" // NOLINT(modernize-raw-string-literal)
-                        })
-                {
-                    const std::string s = std::string(pos, 'a') + c + std::string(length - pos - 1, 'b');
-                    std::string text = "[\"";
-                    text.append(s).append("\",{\"").append(s).append("\":1}]");
-                    const json_document d = json_document::parse(text);
-                    const json j = json::parse(text);
-                    CHECK(d.root().dump(-1, ' ', true) == j.dump(-1, ' ', true));
-                    CHECK(d.root().dump(1, ' ', true) == j.dump(1, ' ', true));
-                }
-            }
-        }
-    }
-
-    SECTION("streams and discarded views")
-    {
-        const json_document d = json_document::parse(R"({"a": [1, 2]})");
-        std::ostringstream compact;
-        compact << d.root();
-        CHECK(compact.str() == R"({"a":[1,2]})");
-        std::ostringstream pretty;
-        pretty << std::setw(2) << std::setfill('.') << d.root() << d.root()["a"];
-        CHECK(pretty.str() == "{\n..\"a\": [\n....1,\n....2\n..]\n}[1,2]");
-        CHECK(json_view().dump() == json(json::value_t::discarded).dump());
-    }
-}
-
-TEST_CASE("json_view large objects")
-{
-    // objects with 128 members or more are looked up with a hash index
-    for (const std::size_t members :
-            {
-                127u, 128u, 129u, 10000u
-            })
-    {
-        CAPTURE(members)
-        std::string text = "{";
-        for (std::size_t i = 0; i < members; ++i)
-        {
-            text += (i != 0 ? ",\"" : "\"") + std::string(i % 23, 'k') + std::to_string(i) + (i % 7 == 0 ? "\\n" : "") + "\":" + std::to_string(i);
-        }
-        text += R"(,"":"empty key","k1":"a duplicate of an earlier key"})";
-        const json_document d = json_document::parse(text);
-        const json_view v = d.root();
-        const json j = json::parse(text);
-        for (std::size_t i = 0; i < members; ++i)
-        {
-            const std::string key = std::string(i % 23, 'k') + std::to_string(i) + (i % 7 == 0 ? "\n" : "");
-            CHECK(v[key].get<std::size_t>() == i);
-            CHECK(v.contains(key));
-            CHECK(v.find(key).key() == key);
-            CHECK(v.at(key).get<std::size_t>() == i);
-            CHECK(!v.contains(key + "x"));
-        }
-        CHECK(v[""].get_string() == "empty key");
-        CHECK(v["k1"].get<int>() == 1); // the first of duplicate keys, as for small objects
-        CHECK(!v.contains("missing"));
-        CHECK_THROWS_WITH_AS(v.at("missing"), "[json.exception.out_of_range.403] key 'missing' not found", json::out_of_range&);
-        CHECK(v == j);
-        CHECK(v.materialize() == j);
-    }
-
-    SECTION("duplicate keys: the first member wins, with and without a table")
-    {
-        // an object of `total` members: the keys "k0".."k<n-1>" in order, then
-        // three keys repeated twice more (one copy in the middle, one at the
-        // end), and two keys repeated once; the value of a member is its
-        // position, so that the first member of a key can be told apart
-        struct member
-        {
-            std::string key;
-            std::size_t position;
-        };
-        const auto make_members = [](std::size_t total)
-        {
-            std::vector<std::string> keys;
-            for (std::size_t i = 0; i + 8 < total; ++i)
-            {
-                keys.push_back("k" + std::to_string(i));
-            }
-            const std::size_t n = keys.size();
-            const std::array<std::size_t, 3> triple = {{3, 17, n - 1}};
-            const std::array<std::size_t, 2> twice = {{5, n / 2}};
-            std::vector<std::string> ordered = keys;
-            for (const std::size_t i : triple)
-            {
-                ordered.insert(ordered.begin() + static_cast<std::ptrdiff_t>(ordered.size() / 2), keys[i]);
-            }
-            for (const std::size_t i : twice)
-            {
-                ordered.insert(ordered.begin() + static_cast<std::ptrdiff_t>(ordered.size() / 3), keys[i]);
-            }
-            for (const std::size_t i : triple)
-            {
-                ordered.push_back(keys[i]);
-            }
-            std::vector<member> result;
-            result.reserve(ordered.size());
-            for (std::size_t i = 0; i < ordered.size(); ++i)
-            {
-                result.push_back({ordered[i], i});
-            }
-            return result;
-        };
-
-        // 100 and 127 members: no table; 128 members and more: a table
-        for (const std::size_t total :
-                {
-                    100u, 127u, 128u, 200u, 5000u
-                })
-        {
-            CAPTURE(total)
-            const std::vector<member> members = make_members(total);
-            REQUIRE(members.size() >= total);
-            REQUIRE(members.size() >= 100);
-            std::string text = "{";
-            std::map<std::string, std::size_t> first;
-            std::map<std::string, std::size_t> last;
-            std::map<std::string, std::size_t> occurrences;
-            for (const member& m : members)
-            {
-                text += (text.size() > 1 ? ",\"" : "\"") + m.key + "\":" + std::to_string(m.position);
-                first.insert({m.key, m.position});
-                last[m.key] = m.position;
-                ++occurrences[m.key];
-            }
-            text += '}';
-            REQUIRE(first.size() < members.size());
-
-            const json_document d = json_document::parse(text);
-            const json_view v = d.root();
-            const json j = json::parse(text);
-            CHECK(v.size() == members.size()); // every occurrence is visited
-            const json m = v.materialize();
-            for (const auto& entry : first)
-            {
-                CAPTURE(entry.first)
-                const std::size_t expected = entry.second;
-                CHECK(v[entry.first].get<std::size_t>() == expected);
-                CHECK(v.at(entry.first).get<std::size_t>() == expected);
-                CHECK(v.find(entry.first).value().get<std::size_t>() == expected);
-                CHECK(v.value(entry.first, std::size_t{0}) == expected);
-                CHECK(v.contains(entry.first));
-                CHECK(v.count(entry.first) == occurrences[entry.first]);
-                const std::string pointer = "/" + entry.first;
-                CHECK(v[json::json_pointer(pointer)].get<std::size_t>() == expected);
-                CHECK(v.at(json::json_pointer(pointer)).get<std::size_t>() == expected);
-                CHECK(v.value(json::json_pointer(pointer), std::size_t{0}) == expected);
-                CHECK(v.contains(json::json_pointer(pointer)));
-                // materialize() and parse() keep the last value instead
-                CHECK(j[entry.first].get<std::size_t>() == last.at(entry.first));
-                CHECK(m[entry.first].get<std::size_t>() == last.at(entry.first));
-            }
-            CHECK(!v.contains("k"));
-            CHECK(v["missing"].is_discarded());
-            CHECK(v.materialize() == j);
-            CHECK(v == j);
-        }
-    }
-
-    SECTION("colliding keys")
-    {
-        // the hash is not seeded, so keys that all land in one place must not
-        // make the table build quadratic: such an object gets no table and is
-        // searched linearly
-        constexpr std::size_t members = 300;
-        constexpr std::size_t table_size = 1024; // the table size for 300 members: the next power of two >= 600
-        std::vector<std::string> colliding;
-        std::vector<std::string> spread;
-        for (std::uint64_t counter = 0; colliding.size() < members || spread.size() < members; ++counter)
-        {
-            std::string key(8, 'a');
-            std::uint64_t x = counter;
-            for (std::size_t i = 0; i < 8; ++i, x /= 26)
-            {
-                key[i] = static_cast<char>('a' + (x % 26));
-            }
-            const bool lands_in_slot_zero = (nlohmann::detail::view::key_hash(key.data(), key.size()) & (table_size - 1)) == 0;
-            if (lands_in_slot_zero && colliding.size() < members)
-            {
-                colliding.push_back(key);
-            }
-            else if (!lands_in_slot_zero && spread.size() < members)
-            {
-                spread.push_back(key);
-            }
-        }
-
-        const auto make_text = [](const std::vector<std::string>& keys)
-        {
-            std::string text = "{";
-            for (std::size_t i = 0; i < keys.size(); ++i)
-            {
-                text += (i != 0 ? ",\"" : "\"") + keys[i] + "\":" + std::to_string(i % 10);
-            }
-            return text + "}";
-        };
-        const std::string colliding_text = make_text(colliding);
-        const std::string spread_text = make_text(spread);
-
-        json_document with_collisions = json_document::parse(colliding_text);
-        json_document without_collisions = json_document::parse(spread_text);
-        for (const auto* pair :
-                {
-                    &colliding, &spread
-                })
-        {
-            const json_view v = (pair == &colliding ? with_collisions : without_collisions).root();
-            for (std::size_t i = 0; i < members; ++i)
-            {
-                CAPTURE(i)
-                CHECK(v[(*pair)[i]].get<std::size_t>() == i % 10);
-                CHECK(v.at((*pair)[i]).get<std::size_t>() == i % 10);
-                CHECK(v.find((*pair)[i]).key() == (*pair)[i]);
-                CHECK(!v.contains((*pair)[i] + "x"));
-            }
-            CHECK(!v.contains("missing"));
-        }
-        CHECK(with_collisions.root() == json::parse(colliding_text));
-
-        // only the object with the spread keys got a table (both texts have the
-        // same length, so the tables are the only difference)
-        with_collisions.shrink_to_fit();
-        without_collisions.shrink_to_fit();
-        CHECK(without_collisions.memory_usage() >= with_collisions.memory_usage() + (table_size * sizeof(std::uint32_t)));
-    }
-
-    SECTION("keys that differ in their last bytes are spread over the table")
-    {
-        // The hash must spread the usual families of keys (numbered names, hex
-        // strings, dates, short names): a key farther than index_max_displacement
-        // from its home slot would leave the object without a table. The keys
-        // are placed with linear probing, as the table does.
-        const auto max_displacement = [](const std::vector<std::string>& keys)
-        {
-            std::size_t slots = 16;
-            while (slots < 2 * keys.size())
-            {
-                slots *= 2;
-            }
-            std::vector<bool> used(slots, false);
-            std::size_t farthest = 0;
-            for (const std::string& key : keys)
-            {
-                std::size_t i = nlohmann::detail::conditional_static_cast<std::size_t>(nlohmann::detail::view::key_hash(key.data(), key.size())) & (slots - 1);
-                std::size_t distance = 0;
-                while (used[i])
-                {
-                    i = (i + 1) & (slots - 1);
-                    ++distance;
-                }
-                used[i] = true;
-                farthest = (std::max)(farthest, distance);
-            }
-            return farthest;
-        };
-        // the keys of a printf-like pattern with one %d, %0Nd or %0Nx, formatted with a stream
-        // (a format string that is not a literal is a Flawfinder error)
-        const auto numbered = [](const char* format, int count, int step)
-        {
-            const std::string pattern(format);
-            const std::size_t percent = pattern.find('%');
-            std::size_t conversion = percent + 1;
-            std::size_t width = 0;
-            while (pattern[conversion] >= '0' && pattern[conversion] <= '9')
-            {
-                width = (width * 10) + static_cast<std::size_t>(pattern[conversion] - '0');
-                ++conversion;
-            }
-            const bool hex = pattern[conversion] == 'x';
-            std::vector<std::string> keys;
-            for (int i = 0; i < count; ++i)
-            {
-                std::ostringstream key;
-                key << pattern.substr(0, percent) << std::setfill('0') << std::setw(static_cast<int>(width))
-                    << (hex ? std::hex : std::dec) << (i * step) << pattern.substr(conversion + 1);
-                keys.push_back(key.str());
-            }
-            return keys;
-        };
-        for (const char* format :
-                {"key%d", "%d", "item_%04d", "a%05d", "%03dxyz", "id%d", "user_%d", "%d_suffix", "0x%04x", "A%07d", "12345678%d", "1234567%d", "123456789%d",
-                 "configuration_parameter_value_%04d_extra", "2026-01-%02dT00:00:00Z"
-                })
-        {
-            CAPTURE(format)
-            CHECK(max_displacement(numbered(format, std::string(format).find("%02d") != std::string::npos ? 60 : 1000, 1)) < nlohmann::detail::view::index_max_displacement); // NOLINT(abseil-string-find-str-contains)
-            CHECK(max_displacement(numbered(format, 600, 7)) < nlohmann::detail::view::index_max_displacement);
-        }
-        std::mt19937_64 rng(7); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed)
-        for (const std::size_t length :
-                {
-                    std::size_t{1}, std::size_t{2}, std::size_t{3}, std::size_t{5}, std::size_t{7}, std::size_t{8}, std::size_t{9}, std::size_t{12}, std::size_t{16}, std::size_t{17}, std::size_t{33}
-                })
-        {
-            CAPTURE(length)
-            std::set<std::string> distinct;
-            // (there are 26 and 676 keys of one and two letters)
-            std::size_t wanted = 700;
-            if (length == 1)
-            {
-                wanted = 26;
-            }
-            else if (length == 2)
-            {
-                wanted = 500;
-            }
-            while (distinct.size() < wanted)
-            {
-                std::string key(length, 'a');
-                for (char& c : key)
-                {
-                    c = static_cast<char>('a' + (rng() % 26));
-                }
-                distinct.insert(key);
-            }
-            CHECK(max_displacement(std::vector<std::string>(distinct.begin(), distinct.end())) < nlohmann::detail::view::index_max_displacement);
-        }
-    }
-
-    SECTION("keys of every length that differ in one byte")
-    {
-        // In a large object, a key is found by comparing its bytes with those of the member the
-        // table points to: up to 16 bytes as two overlapping words, longer ones with memcmp.
-        // Keys of every length from 0 to 40, and keys that differ from them in one byte.
-        std::mt19937_64 rng(2026); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed)
-        std::vector<std::string> keys;
-        std::set<std::string> distinct;
-        for (std::size_t length = 0; length <= 40; ++length)
-        {
-            for (int copy = 0; copy < 5; ++copy)
-            {
-                std::string key(length, 'a');
-                for (char& c : key)
-                {
-                    c = static_cast<char>('a' + (rng() % 26));
-                }
-                if (distinct.insert(key).second)
-                {
-                    keys.push_back(key);
-                }
-            }
-        }
-        std::string text = "{";
-        for (std::size_t i = 0; i < keys.size(); ++i)
-        {
-            text += (i != 0 ? ",\"" : "\"") + keys[i] + "\":" + std::to_string(i);
-        }
-        text += "}";
-        REQUIRE(keys.size() >= 128); // (a table is built for 128 members)
-        const json_document d = json_document::parse(text);
-        const json_view v = d.root();
-        for (std::size_t i = 0; i < keys.size(); ++i)
-        {
-            CAPTURE(keys[i])
-            CHECK(v[keys[i]].get<std::size_t>() == i);
-            CHECK(v.at(keys[i]).get<std::size_t>() == i);
-            CHECK(v.contains(keys[i]));
-            for (std::size_t at = 0; at < keys[i].size(); ++at)
-            {
-                std::string other = keys[i];
-                other[at] = other[at] == 'z' ? 'y' : 'z';
-                CAPTURE(other)
-                CHECK(v.contains(other) == (distinct.count(other) != 0));
-            }
-            CHECK(v.contains(keys[i] + "a") == (distinct.count(keys[i] + "a") != 0));
-            if (!keys[i].empty())
-            {
-                CHECK(v.contains(keys[i].substr(1)) == (distinct.count(keys[i].substr(1)) != 0));
-            }
-        }
-    }
-
-    SECTION("shrink_to_fit releases the tables' spare capacity")
-    {
-        const auto make_text = [](int objects, int members)
-        {
-            std::string text = "[";
-            for (int object = 0; object < objects; ++object)
-            {
-                text += object != 0 ? ",{" : "{";
-                for (int i = 0, n = members + object; i < n; ++i)
-                {
-                    text += (i != 0 ? ",\"" : "\"") + std::to_string(i) + "\":" + std::to_string(i);
-                }
-                text += '}';
-            }
-            return text + "]";
-        };
-        const std::string small_text = make_text(5, 150);
-        const std::string big_text = make_text(40, 400);
-
-        // reading a big text, and then a small one, leaves the spare capacity
-        // of the big one: shrink_to_fit() brings the document to the size of
-        // one parsed from the small text alone
-        json_document d = json_document::parse(big_text);
-        const std::size_t big = d.memory_usage();
-        d.read(small_text);
-        CHECK(d.memory_usage() >= big);
-        d.shrink_to_fit();
-        json_document fresh = json_document::parse(small_text);
-        fresh.shrink_to_fit();
-        CHECK(d.memory_usage() == fresh.memory_usage());
-        CHECK(d.memory_usage() < big / 2);
-        CHECK(d.root() == json::parse(small_text));
-        for (int object = 0; object < 5; ++object)
-        {
-            const json_view v = d.root()[static_cast<std::size_t>(object)];
-            for (int i = 0, n = 150 + object; i < n; ++i)
-            {
-                CHECK(v[std::to_string(i)].get<int>() == i);
-            }
-        }
-    }
-
-    SECTION("nested, reused, and in arrays")
-    {
-        std::string inner = "{";
-        for (int i = 0; i < 300; ++i)
-        {
-            inner += (i != 0 ? ",\"m" : "\"m") + std::to_string(i) + "\":" + std::to_string(i);
-        }
-        inner += '}';
-        const std::string text = "[" + inner + ",{\"x\":" + inner + "}," + inner + "]";
-        json_document d = json_document::parse(text);
-        CHECK(d.root()[0]["m299"].get<int>() == 299);
-        CHECK(d.root()[1]["x"]["m150"].get<int>() == 150);
-        CHECK(d.root()[2]["m0"].get<int>() == 0);
-        const std::size_t with_index = d.memory_usage();
-        d.read(std::string("{\"small\": 1}"));
-        CHECK(d.root()["small"].get<int>() == 1);
-        d.read(text);
-        CHECK(d.root()[2]["m7"].get<int>() == 7);
-        CHECK(d.memory_usage() >= with_index / 2);
-    }
 }

tests/src/unit-json_view_dump_escape.cpp

@@ -0,0 +1,298 @@
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++ (supporting code)
+// |  |  |__   |  |  | | | |  version 3.12.0
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013-2026 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+#include "doctest_compatibility.h"
+
+#include <nlohmann/json_view.hpp>
+using nlohmann::json;
+using nlohmann::ordered_json;
+using nlohmann::json_document;
+using nlohmann::ordered_json_editable_document;
+using nlohmann::json_view;
+using nlohmann::ordered_json_document;
+using nlohmann::ordered_json_view;
+
+#include <algorithm>
+#include <array>
+#include <cmath>
+#include <cstddef>
+#include <cstdint>
+#include <cstring>
+#include <iomanip>
+#include <map>
+#include <random>
+#include <set>
+#include <sstream>
+#include <string>
+#include <vector>
+
+// These tests were split off unit-json_view_dump.cpp, whose object file got too large for the MinGW
+// linker (see json_view_test_helpers.hpp). Like that file, this one is also built
+// with C++17, as it mentions JSON_HAS_CPP_17.
+
+TEST_CASE("json_view dump: escaping, nesting, and output buffers")
+{
+    SECTION("members in document order, all of them")
+    {
+        const json_document d = json_document::parse(R"({"b": 1, "a": 2, "b": 3})");
+        CHECK(d.root().dump() == R"({"b":1,"a":2,"b":3})");
+        CHECK(d.root().dump(1) == "{\n \"b\": 1,\n \"a\": 2,\n \"b\": 3\n}");
+        // dump() writes all members, though the lookup finds the first
+        CHECK(d.root()["b"].dump() == "1");
+    }
+
+    SECTION("deep nesting")
+    {
+        const std::string deep = std::string(100000, '[') + std::string(100000, ']');
+        const json_document deep_document = json_document::parse(deep);
+        CHECK(deep_document.root().dump() == deep);
+    }
+
+    SECTION("ensure_ascii: a character that needs an escape at each position of a string")
+    {
+        // string lengths around the width of the vector steps, in a string value, a key, and a value near the end of the text
+        for (const char* const special :
+                {
+                    "\xC3\xA9", "\xE4\xB8\xAD", "\xF0\x9F\x98\x80", "\x7F", "\\n", "\\u0001", "\\\"", "\\\\"
+                })
+        {
+            for (std::size_t length = 0; length <= 40; ++length)
+            {
+                for (std::size_t position = 0; position <= length; ++position)
+                {
+                    std::string content(length, 'a');
+                    content.insert(position, special);
+                    const std::array<std::string, 3> texts =
+                    {
+                        {
+                            "\"" + content + "\"",
+                            "[\"" + content + "\"]",
+                            "{\"" + content + "\":1}",
+                        }
+                    };
+                    for (const std::string& text : texts)
+                    {
+                        CAPTURE(text)
+                        const ordered_json_document d = ordered_json_document::parse(text);
+                        const ordered_json j = ordered_json::parse(text);
+                        CHECK(d.root().dump(-1, ' ', true) == j.dump(-1, ' ', true));
+                        CHECK(d.root().dump(2, ' ', true) == j.dump(2, ' ', true));
+                        CHECK(d.root().dump() == j.dump());
+                    }
+                }
+            }
+        }
+    }
+
+    SECTION("ensure_ascii: every character of the Basic Multilingual Plane and some beyond")
+    {
+        // (the two- and three-byte sequences, including the ranges that need a check: E0, ED, EF BF BF)
+        std::string text = "[";
+        std::size_t in_chunk = 0;
+        const auto append = [&](std::uint32_t cp)
+        {
+            if (cp < 0x80)
+            {
+                if (cp < 0x20 || cp == '"' || cp == '\\')
+                {
+                    return;
+                }
+                text += static_cast<char>(cp);
+            }
+            else if (cp < 0x800)
+            {
+                text += static_cast<char>(0xC0 | (cp >> 6));
+                text += static_cast<char>(0x80 | (cp & 0x3F));
+            }
+            else if (cp < 0x10000)
+            {
+                text += static_cast<char>(0xE0 | (cp >> 12));
+                text += static_cast<char>(0x80 | ((cp >> 6) & 0x3F));
+                text += static_cast<char>(0x80 | (cp & 0x3F));
+            }
+            else
+            {
+                text += static_cast<char>(0xF0 | (cp >> 18));
+                text += static_cast<char>(0x80 | ((cp >> 12) & 0x3F));
+                text += static_cast<char>(0x80 | ((cp >> 6) & 0x3F));
+                text += static_cast<char>(0x80 | (cp & 0x3F));
+            }
+        };
+        for (std::uint32_t cp = 0x20; cp < 0x10400; ++cp)
+        {
+            if (cp >= 0xD800 && cp < 0xE000)
+            {
+                continue;
+            }
+            if (in_chunk == 0)
+            {
+                text += text.size() > 1 ? ",\"" : "\"";
+            }
+            append(cp);
+            if (++in_chunk == 37)
+            {
+                text += "\"";
+                in_chunk = 0;
+            }
+        }
+        text += in_chunk != 0 ? "\"]" : "]";
+        const ordered_json_document d = ordered_json_document::parse(text);
+        const ordered_json j = ordered_json::parse(text);
+        CHECK(d.root().dump(-1, ' ', true) == j.dump(-1, ' ', true));
+        CHECK(d.root().dump(2, ' ', true) == j.dump(2, ' ', true));
+        CHECK(d.root().dump() == j.dump());
+        CHECK(d.root().dump(2) == j.dump(2));
+    }
+
+    SECTION("deep nesting, indented")
+    {
+        // indentations of up to and beyond the width of the writer's fixed-size copy
+        for (const int depth :
+                {
+                    1, 5, 20, 70, 300
+                })
+        {
+            std::string text;
+            for (int i = 0; i < depth; ++i)
+            {
+                text += i % 2 == 0 ? "[1," : R"({"k":2,"\u00e9":)";
+            }
+            text += "[]";
+            for (int i = depth - 1; i >= 0; --i)
+            {
+                text += i % 2 == 0 ? "]" : "}";
+            }
+            const ordered_json_document d = ordered_json_document::parse(text);
+            const ordered_json j = ordered_json::parse(text);
+            for (const int indent :
+                    {
+                        0, 1, 2, 3, 8
+                    })
+            {
+                CHECK(d.root().dump(indent) == j.dump(indent));
+                CHECK(d.root().dump(indent, '\t') == j.dump(indent, '\t'));
+            }
+        }
+    }
+
+    SECTION("the output buffer of a small value is small")
+    {
+        // an escaped key after the value: its node lies in the arena, so the
+        // source extent of the value cannot be read from the next node
+        const std::string big(100000, 'a');
+        const std::string text = R"({"small":1,"list":[1,2,3],"k\n":")" + big + R"("})";
+        const json_document d = json_document::parse(text);
+        const auto small = d.root()["small"].dump();
+        CHECK(small == "1");
+        CHECK(small.capacity() < 4096);
+        const auto list = d.root()["list"].dump();
+        CHECK(list == "[1,2,3]");
+        CHECK(list.capacity() < 4096);
+        CHECK(d.root()["list"].dump(2).capacity() < 4096);
+
+        // the whole document and the large value are unaffected
+        CHECK(d.root().dump() == ordered_json::parse(text).dump());
+        CHECK(d.root()["k\n"].dump() == "\"" + big + "\"");
+    }
+
+    SECTION("output that outgrows the estimate")
+    {
+        // ensure_ascii writes six bytes for each two-byte character
+        std::string chars;
+        for (int i = 0; i < 5000; ++i)
+        {
+            chars += "\xC3\xA9";
+        }
+        const json_document d = json_document::parse(R"({"a":")" + chars + R"(","k\n":1})");
+        const json expected = json::parse("\"" + chars + "\"");
+        CHECK(d.root()["a"].dump(-1, ' ', true) == expected.dump(-1, ' ', true));
+        CHECK(d.root()["a"].dump(-1, ' ', true).size() == 2 + (5000 * 6));
+    }
+
+    SECTION("ensure_ascii with strings of the source, decoded strings, and edited strings")
+    {
+        // plain ASCII, non-ASCII, escaped (decoded), and mixed strings, as values and as keys;
+        // the second text has a DEL, which ensure_ascii writes as an escape
+        for (const char* const del :
+                {"", "\x7F"
+                })
+        {
+            const std::string text = std::string("{\"plain\":\"abc\",\"caf\xC3\xA9\":\"\xC3\xA9\xE2\x82\xAC\xF0\x9F\x98\x80\",")
+                                     + "\"esc\\n\\\"\":\"a\\tb\\u00e4\\ud83d\\ude00\",\"del\":\"a" + del + "b\",\"\":\"\"," // NOLINT(modernize-raw-string-literal)
+                                     + "\"list\":[\"x\",\"\xC3\xBC\",\"\\\\\",\"long ASCII string with more than thirty-two bytes in it\","
+                                     + "\"long string with more than thirty-two bytes and one non-ASCII byte \xC3\xB6\"],"
+                                     + "\"n\":[1,2.5,-0,true,null]}";
+            const ordered_json j = ordered_json::parse(text);
+            const ordered_json_document d = ordered_json_document::parse(text);
+            CHECK(d.root().dump(-1, ' ', true) == j.dump(-1, ' ', true));
+            CHECK(d.root().dump(4, ' ', true) == j.dump(4, ' ', true));
+            CHECK(d.root().dump() == j.dump());
+            CHECK(d.root()["list"].dump(-1, ' ', true) == j["list"].dump(-1, ' ', true));
+            CHECK(d.root()["del"].dump(-1, ' ', true) == j["del"].dump(-1, ' ', true));
+
+            // strings that an edit wrote (the source strings that they replace were plain)
+            ordered_json_editable_document e = ordered_json_editable_document::parse(text);
+            ordered_json k = j;
+            e.set(e.root()["plain"], std::string("\xC3\xA4gypten"));
+            k["plain"] = "\xC3\xA4gypten";
+            e.set(e.root()["list"][0], std::string("ascii after edit"));
+            k["list"][0] = "ascii after edit";
+            e.set(e.root()["list"][1], std::string("\xC3\xA4"));
+            k["list"][1] = "\xC3\xA4";
+            e.set(e.root()["del"], std::string("tab\there"));
+            k["del"] = "tab\there";
+            CHECK(e.root().dump(-1, ' ', true) == k.dump(-1, ' ', true));
+            CHECK(e.root().dump(2, ' ', true) == k.dump(2, ' ', true));
+            CHECK(e.root().dump() == k.dump());
+
+            // strings that were edited from or to ASCII text
+            e.set(e.root()["list"][2], std::string("\xC3\xA4"));
+            k["list"][2] = "\xC3\xA4";
+            e.set(e.root()["list"][4], std::string("plain again"));
+            k["list"][4] = "plain again";
+            CHECK(e.root().dump(-1, ' ', true) == k.dump(-1, ' ', true));
+        }
+    }
+
+    SECTION("ensure_ascii with DEL or a non-ASCII character at every position")
+    {
+        // the check whether a string of the source has no byte >= 0x7F reads
+        // eight bytes at a time and then the rest one by one
+        for (std::size_t length = 1; length <= 20; ++length)
+        {
+            for (std::size_t pos = 0; pos < length; ++pos)
+            {
+                for (const char* const c :
+                        {
+                            "\x7F", "\x7E", "\xC3\xA4", "\xC3\xBF" // NOLINT(modernize-raw-string-literal)
+                        })
+                {
+                    const std::string s = std::string(pos, 'a') + c + std::string(length - pos - 1, 'b');
+                    std::string text = "[\"";
+                    text.append(s).append("\",{\"").append(s).append("\":1}]");
+                    const json_document d = json_document::parse(text);
+                    const json j = json::parse(text);
+                    CHECK(d.root().dump(-1, ' ', true) == j.dump(-1, ' ', true));
+                    CHECK(d.root().dump(1, ' ', true) == j.dump(1, ' ', true));
+                }
+            }
+        }
+    }
+
+    SECTION("streams and discarded views")
+    {
+        const json_document d = json_document::parse(R"({"a": [1, 2]})");
+        std::ostringstream compact;
+        compact << d.root();
+        CHECK(compact.str() == R"({"a":[1,2]})");
+        std::ostringstream pretty;
+        pretty << std::setw(2) << std::setfill('.') << d.root() << d.root()["a"];
+        CHECK(pretty.str() == "{\n..\"a\": [\n....1,\n....2\n..]\n}[1,2]");
+        CHECK(json_view().dump() == json(json::value_t::discarded).dump());
+    }
+}

tests/src/unit-json_view_edit.cpp

@@ -532,18 +532,18 @@ TEST_CASE("json_view edits: errors")
 
     CHECK_THROWS_WITH_AS(d.set(other_editable.root(), 1), "[json.exception.invalid_iterator.202] view does not belong to this document", json::invalid_iterator&);
     CHECK_THROWS_WITH_AS(d.set(json_editable_view(), 1), "[json.exception.invalid_iterator.202] view does not belong to this document", json::invalid_iterator&);
-    CHECK_THROWS_WITH_AS(d.set(root["a"], "k", 1), "[json.exception.type_error.305] cannot use operator[] with a string argument with array", json::type_error&);
+    CHECK_THROWS_WITH_AS(d.set(root["a"], "k", 1), json_view_test::message("type_error.305", "/a", 21, 27, "cannot use operator[] with a string argument with array").c_str(), json::type_error&);
-    CHECK_THROWS_WITH_AS(d.set(root["o"], 0, 1), "[json.exception.type_error.305] cannot use operator[] with a numeric argument with object", json::type_error&);
+    CHECK_THROWS_WITH_AS(d.set(root["o"], 0, 1), json_view_test::message("type_error.305", "/o", 6, 14, "cannot use operator[] with a numeric argument with object").c_str(), json::type_error&);
-    CHECK_THROWS_WITH_AS(d.set(root["a"], 2, 1), "[json.exception.out_of_range.401] array index 2 is out of range", json::out_of_range&);
+    CHECK_THROWS_WITH_AS(d.set(root["a"], 2, 1), json_view_test::message("out_of_range.401", "/a", 21, 27, "array index 2 is out of range").c_str(), json::out_of_range&);
-    CHECK_THROWS_WITH_AS(d.set(root["a"], -1, 1), "[json.exception.out_of_range.401] array index -1 is out of range", json::out_of_range&);
+    CHECK_THROWS_WITH_AS(d.set(root["a"], -1, 1), json_view_test::message("out_of_range.401", "/a", 21, 27, "array index -1 is out of range").c_str(), json::out_of_range&);
-    CHECK_THROWS_WITH_AS(d.push_back(root["o"], 1), "[json.exception.type_error.308] cannot use push_back() with object", json::type_error&);
+    CHECK_THROWS_WITH_AS(d.push_back(root["o"], 1), json_view_test::message("type_error.308", "/o", 6, 14, "cannot use push_back() with object").c_str(), json::type_error&);
-    CHECK_THROWS_WITH_AS(d.insert(root["n"], 0, 1), "[json.exception.type_error.309] cannot use insert() with number", json::type_error&);
+    CHECK_THROWS_WITH_AS(d.insert(root["n"], 0, 1), json_view_test::message("type_error.309", "/n", 34, 35, "cannot use insert() with number").c_str(), json::type_error&);
-    CHECK_THROWS_WITH_AS(d.insert(root["a"], 3, 1), "[json.exception.out_of_range.401] array index 3 is out of range", json::out_of_range&);
+    CHECK_THROWS_WITH_AS(d.insert(root["a"], 3, 1), json_view_test::message("out_of_range.401", "/a", 21, 27, "array index 3 is out of range").c_str(), json::out_of_range&);
-    CHECK_THROWS_WITH_AS(d.erase(root["n"], "k"), "[json.exception.type_error.307] cannot use erase() with number", json::type_error&);
+    CHECK_THROWS_WITH_AS(d.erase(root["n"], "k"), json_view_test::message("type_error.307", "/n", 34, 35, "cannot use erase() with number").c_str(), json::type_error&);
-    CHECK_THROWS_WITH_AS(d.erase(root["o"], 0), "[json.exception.type_error.307] cannot use erase() with object", json::type_error&);
+    CHECK_THROWS_WITH_AS(d.erase(root["o"], 0), json_view_test::message("type_error.307", "/o", 6, 14, "cannot use erase() with object").c_str(), json::type_error&);
-    CHECK_THROWS_WITH_AS(d.erase(root["a"], 2), "[json.exception.out_of_range.401] array index 2 is out of range", json::out_of_range&);
+    CHECK_THROWS_WITH_AS(d.erase(root["a"], 2), json_view_test::message("out_of_range.401", "/a", 21, 27, "array index 2 is out of range").c_str(), json::out_of_range&);
     CHECK_THROWS_WITH_AS(d.erase(json::json_pointer("")), "[json.exception.out_of_range.405] JSON pointer has no parent", json::out_of_range&);
-    CHECK_THROWS_WITH_AS(d.erase(json::json_pointer("/missing/x")), "[json.exception.out_of_range.403] key 'missing' not found", json::out_of_range&);
+    CHECK_THROWS_WITH_AS(d.erase(json::json_pointer("/missing/x")), json_view_test::message("out_of_range.403", "", 0, 47, "key 'missing' not found").c_str(), json::out_of_range&);
     CHECK_THROWS_WITH_AS(d.set(json::json_pointer("/a/01"), 1), "[json.exception.parse_error.106] parse error: array index '01' must not begin with '0'", json::parse_error&);
     CHECK_THROWS_WITH_AS(d.set(root, json_editable_view()), "[json.exception.type_error.302] type must be a value, but is discarded", json::type_error&);
     CHECK_THROWS_WITH_AS(d.set(root, json::binary({1, 2})), "[json.exception.type_error.319] cannot store a binary value in a json_document", json::type_error&);

tests/src/unit-json_view_from_json.cpp

@@ -39,6 +39,7 @@ using nlohmann::ordered_json_document;
 
 #ifdef JSON_HAS_CPP_17
     #include <optional>
+    #include <string_view>
 #endif
 
 // NLOHMANN_JSON_SERIALIZE_ENUM uses a static std::pair; same() compares floating-point values on purpose
@@ -337,8 +338,7 @@ TEST_CASE("json_view from_json")
         CHECK(d.root().get<std::unordered_map<std::string, int>>().at("a") == 3);
     }
 
-    // (the exceptions of views have neither the path nor the byte range of the value)
+#if !defined(JSON_NOEXCEPTION)
-#if !defined(JSON_NOEXCEPTION) && !JSON_DIAGNOSTICS && !JSON_DIAGNOSTIC_POSITIONS
     SECTION("std types: the same errors as from the materialized value")
     {
         const json_document d = json_document::parse(R"([[1, "x"], {"a": 1}, "text", [1], 5, null])");
@@ -403,10 +403,10 @@ TEST_CASE("json_view from_json")
         CHECK(d.root()[4].get<strict_color>() == strict_color::red);
 #if !defined(JSON_NOEXCEPTION)
         const json j = d.root().materialize();
-        CHECK(exception_of_call([&] { static_cast<void>(d.root()[2].get<strict_color>()); }) == "[json.exception.out_of_range.410] enum value out of range for strict_color: \"purple\"");
 #if !JSON_DIAGNOSTICS && !JSON_DIAGNOSTIC_POSITIONS
-        CHECK(exception_of_call([&] { static_cast<void>(d.root()[2].get<strict_color>()); }) == exception_of_call([&] { static_cast<void>(j[2].get<strict_color>()); }));
+        CHECK(exception_of_call([&] { static_cast<void>(d.root()[2].get<strict_color>()); }) == "[json.exception.out_of_range.410] enum value out of range for strict_color: \"purple\"");
 #endif
+        CHECK(exception_of_call([&] { static_cast<void>(d.root()[2].get<strict_color>()); }) == exception_of_call([&] { static_cast<void>(j[2].get<strict_color>()); }));
 #endif
     }
 
@@ -443,6 +443,30 @@ TEST_CASE("json_view from_json")
         CHECK(d.root().materialize().get<point>().x == 3);
     }
 
+    SECTION("escaped keys: members are found by their decoded names")
+    {
+        // keys are decoded before they are compared, as by parse(); a lookup
+        // that compares the raw bytes of the source would miss these members
+        const json_document d = json_document::parse(R"([{"\u0078": 1, "y": 2}, {"V\u0061lue": 3, "Te\u0078t": "t"}, {"\u0062": "escaped"}, {"x": 1, "\u0078": 3, "y": 2}])");
+        const json j = d.root().materialize();
+        CHECK(d.root()[0].get<point>().x == 1);
+        CHECK(d.root()[0].get<point>().y == 2);
+        CHECK((d.root()[0].get<point>() == j[0].get<point>()));
+        CHECK(d.root()[1].get<with_names>().value == 3);
+        CHECK(d.root()[1].get<with_names>().text == "t");
+        CHECK(d.root()[2].get<with_defaults>().b == "escaped");
+        CHECK(d.root()[2].get<with_defaults>().a == 1);
+        // a duplicate key written once plain and once escaped: still the first member
+        CHECK(d.root()[3].get<point>().x == 1);
+        CHECK(j[3].get<point>().x == 3);
+#if !defined(JSON_NOEXCEPTION) && !JSON_DIAGNOSTICS && !JSON_DIAGNOSTIC_POSITIONS
+        // an escaped backslash is part of the decoded key: "x\\" is not "x"
+        const json_document e = json_document::parse(R"({"x\\": 1, "y": 2})");
+        CHECK(exception_of_call([&] { static_cast<void>(e.root().get<point>()); }) == exception_of_call([&] { static_cast<void>(e.root().materialize().get<point>()); }));
+        CHECK(exception_of_call([&] { static_cast<void>(e.root().get<point>()); }) == "[json.exception.out_of_range.403] key 'x' not found");
+#endif
+    }
+
     SECTION("editable documents and ordered_json")
     {
         json_editable_document e = json_editable_document::parse(R"({"x": 1, "y": 2})");
@@ -452,11 +476,9 @@ TEST_CASE("json_view from_json")
         CHECK(o.root().get<shape>().center.y == 2);
     }
 
-#if !defined(JSON_NOEXCEPTION) && !JSON_DIAGNOSTICS && !JSON_DIAGNOSTIC_POSITIONS
+#if !defined(JSON_NOEXCEPTION)
     SECTION("errors are those of the same conversion of a json value")
     {
-        // (with JSON_DIAGNOSTICS, those of json name the path of the value,
-        // and with JSON_DIAGNOSTIC_POSITIONS its bytes)
         const json_document d = json_document::parse(R"([{"x": 1}, {"x": "1", "y": 2}, [1, 2], {"id": "7", "c": "red"}])");
         const json j = d.root().materialize();
         for (std::size_t i = 0; i < 3; ++i)
@@ -469,5 +491,70 @@ TEST_CASE("json_view from_json")
 #endif
 }
 
+#ifdef JSON_HAS_CPP_17
+namespace view_from_json
+{
+// a string_view member is not copied: it points into the source text or, for a
+// string with escape sequences, into the buffer of decoded strings of the document
+struct named_view
+{
+    // initialized for GCC -Weffc++
+    // NOLINTNEXTLINE(readability-redundant-member-init)
+    std::string_view name{};
+    int n = 0;
+    NLOHMANN_DEFINE_TYPE_INTRUSIVE(named_view, name, n)
+};
+}  // namespace view_from_json
+
+TEST_CASE("json_view from_json: string_view members borrow from the document")
+{
+    using view_from_json::named_view;
+
+    // whether all of s is inside the bytes of text
+    const auto within = [](std::string_view s, std::string_view text)
+    {
+        const auto first = reinterpret_cast<std::uintptr_t>(text.data());
+        const auto begin = reinterpret_cast<std::uintptr_t>(s.data());
+        return begin >= first && begin + s.size() <= first + text.size();
+    };
+
+    SECTION("a string without escapes points into the source text")
+    {
+        const std::string text = R"({"name": "hello", "n": 1})";
+        const json_document d = json_document::parse(text);
+        REQUIRE_FALSE(d.owns_source());
+        const auto r = d.root().get<named_view>();
+        CHECK(r.name == "hello");
+        CHECK(r.n == 1);
+        CHECK(within(r.name, text));
+        CHECK(within(r.name, d.source()));
+    }
+
+    SECTION("a string with escapes points into the document, not into the source text")
+    {
+        const std::string text = R"({"name": "a\nb", "n": 2})";
+        json_document d = json_document::parse(text);
+        REQUIRE_FALSE(d.owns_source());
+        const auto r = d.root().get<named_view>();
+        CHECK(r.name == "a\nb");
+        CHECK(r.n == 2);
+        CHECK_FALSE(within(r.name, text));
+        // valid as long as the document is alive, even if the document is moved
+        const json_document moved = std::move(d);
+        CHECK(r.name == "a\nb");
+        CHECK(moved.root().get<named_view>().name == r.name);
+    }
+
+    SECTION("a document that owns its text keeps the members readable while it is alive")
+    {
+        const json_document d = json_document::parse(std::string(R"({"name": "owned", "n": 3})"));
+        REQUIRE(d.owns_source());
+        const auto r = d.root().get<named_view>();
+        CHECK(r.name == "owned");
+        CHECK(within(r.name, d.source()));
+    }
+}
+#endif
+
 DOCTEST_CLANG_SUPPRESS_WARNING_POP
 DOCTEST_GCC_SUPPRESS_WARNING_POP

tests/src/unit-json_view_large.cpp

@@ -0,0 +1,544 @@
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++ (supporting code)
+// |  |  |__   |  |  | | | |  version 3.12.0
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013-2026 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+#include "doctest_compatibility.h"
+
+#include <nlohmann/json_view.hpp>
+using nlohmann::json;
+using nlohmann::ordered_json;
+using nlohmann::json_document;
+using nlohmann::ordered_json_editable_document;
+using nlohmann::json_view;
+using nlohmann::ordered_json_document;
+using nlohmann::ordered_json_view;
+
+#include "json_view_test_helpers.hpp"
+using json_view_test::generator;
+
+#include <algorithm>
+#include <array>
+#include <cmath>
+#include <cstddef>
+#include <cstdint>
+#include <cstring>
+#include <iomanip>
+#include <map>
+#include <random>
+#include <set>
+#include <sstream>
+#include <string>
+#include <vector>
+
+// These tests were split off unit-json_view_dump.cpp, whose object file got too large for the MinGW
+// linker (see json_view_test_helpers.hpp). Like that file, this one is also built
+// with C++17, as it mentions JSON_HAS_CPP_17.
+
+TEST_CASE("json_view comparison")
+{
+    SECTION("equality of the values parse() produces")
+    {
+        generator g;
+        std::vector<std::string> texts;
+        for (int i = 0; i < 600; ++i)
+        {
+            std::string text;
+            g.value(text, 0);
+            texts.push_back(text);
+            // the same value written differently: sorted keys, canonical numbers
+            texts.push_back(json::parse(text).dump(1));
+        }
+        for (std::size_t i = 0; i + 2 < texts.size(); ++i)
+        {
+            for (std::size_t k = i; k < i + 3; ++k)
+            {
+                CAPTURE(texts[i])
+                CAPTURE(texts[k])
+                const json_document a = json_document::parse(texts[i]);
+                const json_document b = json_document::parse(texts[k]);
+                const json ja = json::parse(texts[i]);
+                const json jb = json::parse(texts[k]);
+                CHECK((a.root() == b.root()) == (ja == jb));
+                CHECK((a.root() != b.root()) == (ja != jb));
+                CHECK((a.root() == jb) == (ja == jb));
+                CHECK((jb == a.root()) == (ja == jb));
+                CHECK((a.root() != jb) == (ja != jb));
+                CHECK((jb != a.root()) == (ja != jb));
+
+                // ordered_json compares members in order
+                const ordered_json_document oa = ordered_json_document::parse(texts[i]);
+                const ordered_json_document ob = ordered_json_document::parse(texts[k]);
+                const ordered_json oja = ordered_json::parse(texts[i]);
+                const ordered_json ojb = ordered_json::parse(texts[k]);
+                CHECK((oa.root() == ob.root()) == (oja == ojb));
+                CHECK((oa.root() == ojb) == (oja == ojb));
+            }
+        }
+    }
+
+    SECTION("numbers, duplicate keys, member order")
+    {
+        const auto same = [](const char* x, const char* y)
+        {
+            const json_document dx = json_document::parse(x);
+            const json_document dy = json_document::parse(y);
+            return dx.root() == dy.root();
+        };
+        CHECK(same("1", "1.0"));
+        CHECK(same("[1, -1, 2.5]", "[1.0, -1.0, 25e-1]"));
+        CHECK(!same("1", "1.5"));
+        CHECK(same("18446744073709551615", "18446744073709551615"));
+        CHECK(same(R"({"a": 1, "a": 2})", R"({"a": 2})"));
+        CHECK(!same(R"({"a": 1, "a": 2})", R"({"a": 1})"));
+        CHECK(same(R"({"a": 1, "b": 2})", R"({"b": 2, "a": 1})"));
+        CHECK(!same(R"({"a": 1})", R"({"a": 1, "b": 2})"));
+        CHECK(!same("[1, 2]", "[2, 1]"));
+        CHECK(!same("\"a\"", "\"b\""));
+        CHECK(same("\"\\u00e9\"", "\"\xc3\xa9\""));
+        CHECK(!same("null", "false"));
+        CHECK(!same("[]", "{}"));
+        const ordered_json_document dup = ordered_json_document::parse(R"({"a": 1, "b": 2, "a": 3})");
+        const ordered_json_document last = ordered_json_document::parse(R"({"a": 3, "b": 2})");
+        CHECK(dup.root() == last.root());
+        // (== compares as parse() resolves duplicates, a lookup finds the first)
+        CHECK(dup.root()["a"].materialize() == 1);
+        CHECK(dup.root()["a"] != last.root()["a"]);
+        const ordered_json_document ab = ordered_json_document::parse(R"({"a": 1, "b": 2})");
+        const ordered_json_document ba = ordered_json_document::parse(R"({"b": 2, "a": 1})");
+        CHECK(ab.root() != ba.root());
+
+        // discarded values compare as basic_json's do
+        const json discarded(json::value_t::discarded);
+        CHECK((json_view() == json_view()) == (discarded == discarded)); // NOLINT(readability-container-size-empty): operator== is tested
+        CHECK((json_view() == discarded) == (discarded == discarded)); // NOLINT(readability-container-size-empty)
+        const json_document null_document = json_document::parse("null");
+        CHECK(!(json_view() == null_document.root())); // NOLINT(readability-container-size-empty)
+        CHECK(!(null_document.root() == discarded));
+    }
+
+    SECTION("deep nesting")
+    {
+        const std::string deep = std::string(100000, '[') + std::string(100000, ']');
+        const json_document a = json_document::parse(deep);
+        const json_document b = json_document::parse(deep);
+        CHECK(a.root() == b.root());
+        CHECK(a.root() == json::parse(deep));
+        const std::string other = std::string(100000, '[') + "1" + std::string(100000, ']');
+        const json_document c = json_document::parse(other);
+        CHECK(a.root() != c.root());
+    }
+}
+
+TEST_CASE("json_view large objects")
+{
+    // objects with 128 members or more are looked up with a hash index
+    for (const std::size_t members :
+            {
+                127u, 128u, 129u, 10000u
+            })
+    {
+        CAPTURE(members)
+        std::string text = "{";
+        for (std::size_t i = 0; i < members; ++i)
+        {
+            text += (i != 0 ? ",\"" : "\"") + std::string(i % 23, 'k') + std::to_string(i) + (i % 7 == 0 ? "\\n" : "") + "\":" + std::to_string(i);
+        }
+        text += R"(,"":"empty key","k1":"a duplicate of an earlier key"})";
+        const json_document d = json_document::parse(text);
+        const json_view v = d.root();
+        const json j = json::parse(text);
+        for (std::size_t i = 0; i < members; ++i)
+        {
+            const std::string key = std::string(i % 23, 'k') + std::to_string(i) + (i % 7 == 0 ? "\n" : "");
+            CHECK(v[key].get<std::size_t>() == i);
+            CHECK(v.contains(key));
+            CHECK(v.find(key).key() == key);
+            CHECK(v.at(key).get<std::size_t>() == i);
+            CHECK(!v.contains(key + "x"));
+        }
+        CHECK(v[""].get_string() == "empty key");
+        CHECK(v["k1"].get<int>() == 1); // the first of duplicate keys, as for small objects
+        CHECK(!v.contains("missing"));
+        CHECK_THROWS_WITH_AS(v.at("missing"), json_view_test::message("out_of_range.403", "", 0, text.size(), "key 'missing' not found").c_str(), json::out_of_range&);
+        CHECK(v == j);
+        CHECK(v.materialize() == j);
+    }
+
+    SECTION("duplicate keys: the first member wins, with and without a table")
+    {
+        // an object of `total` members: the keys "k0".."k<n-1>" in order, then
+        // three keys repeated twice more (one copy in the middle, one at the
+        // end), and two keys repeated once; the value of a member is its
+        // position, so that the first member of a key can be told apart
+        struct member
+        {
+            std::string key;
+            std::size_t position;
+        };
+        const auto make_members = [](std::size_t total)
+        {
+            std::vector<std::string> keys;
+            for (std::size_t i = 0; i + 8 < total; ++i)
+            {
+                keys.push_back("k" + std::to_string(i));
+            }
+            const std::size_t n = keys.size();
+            const std::array<std::size_t, 3> triple = {{3, 17, n - 1}};
+            const std::array<std::size_t, 2> twice = {{5, n / 2}};
+            std::vector<std::string> ordered = keys;
+            for (const std::size_t i : triple)
+            {
+                ordered.insert(ordered.begin() + static_cast<std::ptrdiff_t>(ordered.size() / 2), keys[i]);
+            }
+            for (const std::size_t i : twice)
+            {
+                ordered.insert(ordered.begin() + static_cast<std::ptrdiff_t>(ordered.size() / 3), keys[i]);
+            }
+            for (const std::size_t i : triple)
+            {
+                ordered.push_back(keys[i]);
+            }
+            std::vector<member> result;
+            result.reserve(ordered.size());
+            for (std::size_t i = 0; i < ordered.size(); ++i)
+            {
+                result.push_back({ordered[i], i});
+            }
+            return result;
+        };
+
+        // 100 and 127 members: no table; 128 members and more: a table
+        for (const std::size_t total :
+                {
+                    100u, 127u, 128u, 200u, 5000u
+                })
+        {
+            CAPTURE(total)
+            const std::vector<member> members = make_members(total);
+            REQUIRE(members.size() >= total);
+            REQUIRE(members.size() >= 100);
+            std::string text = "{";
+            std::map<std::string, std::size_t> first;
+            std::map<std::string, std::size_t> last;
+            std::map<std::string, std::size_t> occurrences;
+            for (const member& m : members)
+            {
+                text += (text.size() > 1 ? ",\"" : "\"") + m.key + "\":" + std::to_string(m.position);
+                first.insert({m.key, m.position});
+                last[m.key] = m.position;
+                ++occurrences[m.key];
+            }
+            text += '}';
+            REQUIRE(first.size() < members.size());
+
+            const json_document d = json_document::parse(text);
+            const json_view v = d.root();
+            const json j = json::parse(text);
+            CHECK(v.size() == members.size()); // every occurrence is visited
+            const json m = v.materialize();
+            for (const auto& entry : first)
+            {
+                CAPTURE(entry.first)
+                const std::size_t expected = entry.second;
+                CHECK(v[entry.first].get<std::size_t>() == expected);
+                CHECK(v.at(entry.first).get<std::size_t>() == expected);
+                CHECK(v.find(entry.first).value().get<std::size_t>() == expected);
+                CHECK(v.value(entry.first, std::size_t{0}) == expected);
+                CHECK(v.contains(entry.first));
+                CHECK(v.count(entry.first) == occurrences[entry.first]);
+                const std::string pointer = "/" + entry.first;
+                CHECK(v[json::json_pointer(pointer)].get<std::size_t>() == expected);
+                CHECK(v.at(json::json_pointer(pointer)).get<std::size_t>() == expected);
+                CHECK(v.value(json::json_pointer(pointer), std::size_t{0}) == expected);
+                CHECK(v.contains(json::json_pointer(pointer)));
+                // materialize() and parse() keep the last value instead
+                CHECK(j[entry.first].get<std::size_t>() == last.at(entry.first));
+                CHECK(m[entry.first].get<std::size_t>() == last.at(entry.first));
+            }
+            CHECK(!v.contains("k"));
+            CHECK(v["missing"].is_discarded());
+            CHECK(v.materialize() == j);
+            CHECK(v == j);
+        }
+    }
+
+    SECTION("colliding keys")
+    {
+        // the hash is not seeded, so keys that all land in one place must not
+        // make the table build quadratic: such an object gets no table and is
+        // searched linearly
+        constexpr std::size_t members = 300;
+        constexpr std::size_t table_size = 1024; // the table size for 300 members: the next power of two >= 600
+        std::vector<std::string> colliding;
+        std::vector<std::string> spread;
+        for (std::uint64_t counter = 0; colliding.size() < members || spread.size() < members; ++counter)
+        {
+            std::string key(8, 'a');
+            std::uint64_t x = counter;
+            for (std::size_t i = 0; i < 8; ++i, x /= 26)
+            {
+                key[i] = static_cast<char>('a' + (x % 26));
+            }
+            const bool lands_in_slot_zero = (nlohmann::detail::view::key_hash(key.data(), key.size()) & (table_size - 1)) == 0;
+            if (lands_in_slot_zero && colliding.size() < members)
+            {
+                colliding.push_back(key);
+            }
+            else if (!lands_in_slot_zero && spread.size() < members)
+            {
+                spread.push_back(key);
+            }
+        }
+
+        const auto make_text = [](const std::vector<std::string>& keys)
+        {
+            std::string text = "{";
+            for (std::size_t i = 0; i < keys.size(); ++i)
+            {
+                text += (i != 0 ? ",\"" : "\"") + keys[i] + "\":" + std::to_string(i % 10);
+            }
+            return text + "}";
+        };
+        const std::string colliding_text = make_text(colliding);
+        const std::string spread_text = make_text(spread);
+
+        json_document with_collisions = json_document::parse(colliding_text);
+        json_document without_collisions = json_document::parse(spread_text);
+        for (const auto* pair :
+                {
+                    &colliding, &spread
+                })
+        {
+            const json_view v = (pair == &colliding ? with_collisions : without_collisions).root();
+            for (std::size_t i = 0; i < members; ++i)
+            {
+                CAPTURE(i)
+                CHECK(v[(*pair)[i]].get<std::size_t>() == i % 10);
+                CHECK(v.at((*pair)[i]).get<std::size_t>() == i % 10);
+                CHECK(v.find((*pair)[i]).key() == (*pair)[i]);
+                CHECK(!v.contains((*pair)[i] + "x"));
+            }
+            CHECK(!v.contains("missing"));
+        }
+        CHECK(with_collisions.root() == json::parse(colliding_text));
+
+        // only the object with the spread keys got a table (both texts have the
+        // same length, so the tables are the only difference)
+        with_collisions.shrink_to_fit();
+        without_collisions.shrink_to_fit();
+        CHECK(without_collisions.memory_usage() >= with_collisions.memory_usage() + (table_size * sizeof(std::uint32_t)));
+    }
+
+    SECTION("keys that differ in their last bytes are spread over the table")
+    {
+        // The hash must spread the usual families of keys (numbered names, hex
+        // strings, dates, short names): a key farther than index_max_displacement
+        // from its home slot would leave the object without a table. The keys
+        // are placed with linear probing, as the table does.
+        const auto max_displacement = [](const std::vector<std::string>& keys)
+        {
+            std::size_t slots = 16;
+            while (slots < 2 * keys.size())
+            {
+                slots *= 2;
+            }
+            std::vector<bool> used(slots, false);
+            std::size_t farthest = 0;
+            for (const std::string& key : keys)
+            {
+                std::size_t i = nlohmann::detail::conditional_static_cast<std::size_t>(nlohmann::detail::view::key_hash(key.data(), key.size())) & (slots - 1);
+                std::size_t distance = 0;
+                while (used[i])
+                {
+                    i = (i + 1) & (slots - 1);
+                    ++distance;
+                }
+                used[i] = true;
+                farthest = (std::max)(farthest, distance);
+            }
+            return farthest;
+        };
+        // the keys of a printf-like pattern with one %d, %0Nd or %0Nx, formatted with a stream
+        // (a format string that is not a literal is a Flawfinder error)
+        const auto numbered = [](const char* format, int count, int step)
+        {
+            const std::string pattern(format);
+            const std::size_t percent = pattern.find('%');
+            std::size_t conversion = percent + 1;
+            std::size_t width = 0;
+            while (pattern[conversion] >= '0' && pattern[conversion] <= '9')
+            {
+                width = (width * 10) + static_cast<std::size_t>(pattern[conversion] - '0');
+                ++conversion;
+            }
+            const bool hex = pattern[conversion] == 'x';
+            std::vector<std::string> keys;
+            for (int i = 0; i < count; ++i)
+            {
+                std::ostringstream key;
+                key << pattern.substr(0, percent) << std::setfill('0') << std::setw(static_cast<int>(width))
+                    << (hex ? std::hex : std::dec) << (i * step) << pattern.substr(conversion + 1);
+                keys.push_back(key.str());
+            }
+            return keys;
+        };
+        for (const char* format :
+                {"key%d", "%d", "item_%04d", "a%05d", "%03dxyz", "id%d", "user_%d", "%d_suffix", "0x%04x", "A%07d", "12345678%d", "1234567%d", "123456789%d",
+                 "configuration_parameter_value_%04d_extra", "2026-01-%02dT00:00:00Z"
+                })
+        {
+            CAPTURE(format)
+            CHECK(max_displacement(numbered(format, std::string(format).find("%02d") != std::string::npos ? 60 : 1000, 1)) < nlohmann::detail::view::index_max_displacement); // NOLINT(abseil-string-find-str-contains)
+            CHECK(max_displacement(numbered(format, 600, 7)) < nlohmann::detail::view::index_max_displacement);
+        }
+        std::mt19937_64 rng(7); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed)
+        for (const std::size_t length :
+                {
+                    std::size_t{1}, std::size_t{2}, std::size_t{3}, std::size_t{5}, std::size_t{7}, std::size_t{8}, std::size_t{9}, std::size_t{12}, std::size_t{16}, std::size_t{17}, std::size_t{33}
+                })
+        {
+            CAPTURE(length)
+            std::set<std::string> distinct;
+            // (there are 26 and 676 keys of one and two letters)
+            std::size_t wanted = 700;
+            if (length == 1)
+            {
+                wanted = 26;
+            }
+            else if (length == 2)
+            {
+                wanted = 500;
+            }
+            while (distinct.size() < wanted)
+            {
+                std::string key(length, 'a');
+                for (char& c : key)
+                {
+                    c = static_cast<char>('a' + (rng() % 26));
+                }
+                distinct.insert(key);
+            }
+            CHECK(max_displacement(std::vector<std::string>(distinct.begin(), distinct.end())) < nlohmann::detail::view::index_max_displacement);
+        }
+    }
+
+    SECTION("keys of every length that differ in one byte")
+    {
+        // In a large object, a key is found by comparing its bytes with those of the member the
+        // table points to: up to 16 bytes as two overlapping words, longer ones with memcmp.
+        // Keys of every length from 0 to 40, and keys that differ from them in one byte.
+        std::mt19937_64 rng(2026); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed)
+        std::vector<std::string> keys;
+        std::set<std::string> distinct;
+        for (std::size_t length = 0; length <= 40; ++length)
+        {
+            for (int copy = 0; copy < 5; ++copy)
+            {
+                std::string key(length, 'a');
+                for (char& c : key)
+                {
+                    c = static_cast<char>('a' + (rng() % 26));
+                }
+                if (distinct.insert(key).second)
+                {
+                    keys.push_back(key);
+                }
+            }
+        }
+        std::string text = "{";
+        for (std::size_t i = 0; i < keys.size(); ++i)
+        {
+            text += (i != 0 ? ",\"" : "\"") + keys[i] + "\":" + std::to_string(i);
+        }
+        text += "}";
+        REQUIRE(keys.size() >= 128); // (a table is built for 128 members)
+        const json_document d = json_document::parse(text);
+        const json_view v = d.root();
+        for (std::size_t i = 0; i < keys.size(); ++i)
+        {
+            CAPTURE(keys[i])
+            CHECK(v[keys[i]].get<std::size_t>() == i);
+            CHECK(v.at(keys[i]).get<std::size_t>() == i);
+            CHECK(v.contains(keys[i]));
+            for (std::size_t at = 0; at < keys[i].size(); ++at)
+            {
+                std::string other = keys[i];
+                other[at] = other[at] == 'z' ? 'y' : 'z';
+                CAPTURE(other)
+                CHECK(v.contains(other) == (distinct.count(other) != 0));
+            }
+            CHECK(v.contains(keys[i] + "a") == (distinct.count(keys[i] + "a") != 0));
+            if (!keys[i].empty())
+            {
+                CHECK(v.contains(keys[i].substr(1)) == (distinct.count(keys[i].substr(1)) != 0));
+            }
+        }
+    }
+
+    SECTION("shrink_to_fit releases the tables' spare capacity")
+    {
+        const auto make_text = [](int objects, int members)
+        {
+            std::string text = "[";
+            for (int object = 0; object < objects; ++object)
+            {
+                text += object != 0 ? ",{" : "{";
+                for (int i = 0, n = members + object; i < n; ++i)
+                {
+                    text += (i != 0 ? ",\"" : "\"") + std::to_string(i) + "\":" + std::to_string(i);
+                }
+                text += '}';
+            }
+            return text + "]";
+        };
+        const std::string small_text = make_text(5, 150);
+        const std::string big_text = make_text(40, 400);
+
+        // reading a big text, and then a small one, leaves the spare capacity
+        // of the big one: shrink_to_fit() brings the document to the size of
+        // one parsed from the small text alone
+        json_document d = json_document::parse(big_text);
+        const std::size_t big = d.memory_usage();
+        d.read(small_text);
+        CHECK(d.memory_usage() >= big);
+        d.shrink_to_fit();
+        json_document fresh = json_document::parse(small_text);
+        fresh.shrink_to_fit();
+        CHECK(d.memory_usage() == fresh.memory_usage());
+        CHECK(d.memory_usage() < big / 2);
+        CHECK(d.root() == json::parse(small_text));
+        for (int object = 0; object < 5; ++object)
+        {
+            const json_view v = d.root()[static_cast<std::size_t>(object)];
+            for (int i = 0, n = 150 + object; i < n; ++i)
+            {
+                CHECK(v[std::to_string(i)].get<int>() == i);
+            }
+        }
+    }
+
+    SECTION("nested, reused, and in arrays")
+    {
+        std::string inner = "{";
+        for (int i = 0; i < 300; ++i)
+        {
+            inner += (i != 0 ? ",\"m" : "\"m") + std::to_string(i) + "\":" + std::to_string(i);
+        }
+        inner += '}';
+        const std::string text = "[" + inner + ",{\"x\":" + inner + "}," + inner + "]";
+        json_document d = json_document::parse(text);
+        CHECK(d.root()[0]["m299"].get<int>() == 299);
+        CHECK(d.root()[1]["x"]["m150"].get<int>() == 150);
+        CHECK(d.root()[2]["m0"].get<int>() == 0);
+        const std::size_t with_index = d.memory_usage();
+        d.read(std::string("{\"small\": 1}"));
+        CHECK(d.root()["small"].get<int>() == 1);
+        d.read(text);
+        CHECK(d.root()[2]["m7"].get<int>() == 7);
+        CHECK(d.memory_usage() >= with_index / 2);
+    }
+}

tests/src/unit-json_view_macros.cpp

@@ -25,7 +25,7 @@
 #endif
 
 // the view's own macros do not leak
-#if defined(NLOHMANN_VIEW_LIKELY) || defined(NLOHMANN_VIEW_UNLIKELY) || defined(NLOHMANN_VIEW_ALWAYS_INLINE) || defined(NLOHMANN_VIEW_NOINLINE) \
+#if defined(NLOHMANN_VIEW_LIKELY) || defined(NLOHMANN_VIEW_UNLIKELY) || defined(NLOHMANN_VIEW_ALWAYS_INLINE) || defined(NLOHMANN_VIEW_NOINLINE) || defined(NLOHMANN_VIEW_NOCLONE) \
     || defined(NLOHMANN_VIEW_NODISCARD) || defined(NLOHMANN_VIEW_THROW) || defined(NLOHMANN_VIEW_HAS_CPP_17) || defined(NLOHMANN_VIEW_LITTLE_ENDIAN) \
     || defined(NLOHMANN_VIEW_REPEAT16)
     #error "json_view.hpp leaks a macro"

tests/src/unit-json_view_positions.cpp

@@ -81,6 +81,112 @@ std::string range(const json& j)
 {
     return j.start_pos() == npos && j.end_pos() == npos ? "none" : std::to_string(j.start_pos()) + "-" + std::to_string(j.end_pos());
 }
+
+#if !defined(JSON_NOEXCEPTION)
+using json_view_test::exception_of_call;
+
+// the exceptions about every value of v are those about the value at the same
+// place in j, a value of the whole document from parse() or materialize() (so
+// that its parents and positions are set); members whose key repeats are
+// skipped (the view finds the first one, j holds the last)
+template<typename View>
+void check_exceptions(const View& v, const json& j)
+{
+    CAPTURE(j.dump())
+    if (!v.is_number())
+    {
+        CHECK(exception_of_call([&] { static_cast<void>(v.template get<int>()); }) == exception_of_call([&] { static_cast<void>(j.get<int>()); }));
+    }
+    if (!v.is_string())
+    {
+        CHECK(exception_of_call([&] { static_cast<void>(v.template get<std::string>()); }) == exception_of_call([&] { static_cast<void>(j.get<std::string>()); }));
+    }
+    if (!v.is_array())
+    {
+        CHECK(exception_of_call([&] { static_cast<void>(v.template get<std::vector<int>>()); }) == exception_of_call([&] { static_cast<void>(j.get<std::vector<int>>()); }));
+    }
+    CHECK(exception_of_call([&] { static_cast<void>(v.at("missing")); }) == exception_of_call([&] { static_cast<void>(j.at("missing")); }));
+    CHECK(exception_of_call([&] { static_cast<void>(v.at(1000)); }) == exception_of_call([&] { static_cast<void>(j.at(1000)); }));
+    CHECK(exception_of_call([&] { static_cast<void>(v.at(json::json_pointer("/missing"))); }) == exception_of_call([&] { static_cast<void>(j.at(json::json_pointer("/missing"))); }));
+    CHECK(exception_of_call([&] { static_cast<void>(v.at(json::json_pointer("/-"))); }) == exception_of_call([&] { static_cast<void>(j.at(json::json_pointer("/-"))); }));
+    CHECK(exception_of_call([&] { static_cast<void>(v.value("k", 0)); }) == exception_of_call([&] { static_cast<void>(j.value("k", 0)); }));
+    if (!v.is_object())
+    {
+        CHECK(exception_of_call([&] { static_cast<void>(v["k"]); }) == exception_of_call([&] { static_cast<void>(j["k"]); }));
+        // (without the context of basic_json: view iterators do not know the value they iterate over)
+        CHECK(exception_of_call([&] { static_cast<void>(v.begin().key()); }) == "[json.exception.invalid_iterator.207] cannot use key() for non-object iterators");
+    }
+    if (!v.is_array())
+    {
+        CHECK(exception_of_call([&] { static_cast<void>(v[0]); }) == exception_of_call([&] { static_cast<void>(j[0]); }));
+    }
+    if (v.is_null())
+    {
+        CHECK(exception_of_call([&] { static_cast<void>(v.front()); }) == exception_of_call([&] { static_cast<void>(j.front()); }));
+    }
+    if (v.is_object())
+    {
+        for (auto it = v.begin(); it != v.end(); ++it)
+        {
+            const std::string key(it.key().data(), it.key().size());
+            if (v.count(key) == 1)
+            {
+                check_exceptions(it.value(), j.at(key));
+            }
+        }
+    }
+    else if (v.is_array())
+    {
+        for (std::size_t i = 0; i < v.size(); ++i)
+        {
+            check_exceptions(v[i], j[i]);
+        }
+    }
+}
+
+// as check_exceptions, for edits that fail (they change neither d nor j)
+void check_edit_exceptions(json_editable_document& d, const json_editable_document::view_type& v, json& j)
+{
+    CAPTURE(j.dump())
+    if (!v.is_object() && !v.is_null())
+    {
+        CHECK(exception_of_call([&] { d.set(v, "k", 1); }) == exception_of_call([&] { j["k"] = 1; }));
+        CHECK(exception_of_call([&] { d.erase(v, "k"); }) == exception_of_call([&] { j.erase("k"); }));
+    }
+    if (!v.is_array() && !v.is_null())
+    {
+        CHECK(exception_of_call([&] { d.set(v, 0, 1); }) == exception_of_call([&] { j[0] = 1; }));
+        CHECK(exception_of_call([&] { d.push_back(v, 1); }) == exception_of_call([&] { j.push_back(1); }));
+    }
+    if (!v.is_array())
+    {
+        CHECK(exception_of_call([&] { d.insert(v, 0, 1); }) == exception_of_call([&] { j.insert(j.cbegin(), 1); }));
+        CHECK(exception_of_call([&] { d.erase(v, 0); }) == exception_of_call([&] { j.erase(0); }));
+    }
+    else
+    {
+        CHECK(exception_of_call([&] { d.erase(v, 1000); }) == exception_of_call([&] { j.erase(1000); }));
+    }
+    if (v.is_object())
+    {
+        for (auto it = v.begin(); it != v.end(); ++it)
+        {
+            const std::string key(it.key().data(), it.key().size());
+            if (v.count(key) == 1)
+            {
+                check_edit_exceptions(d, it.value(), j.at(key));
+            }
+        }
+    }
+    else if (v.is_array())
+    {
+        for (std::size_t i = 0; i < v.size(); ++i)
+        {
+            check_edit_exceptions(d, v[i], j[i]);
+        }
+    }
+}
+#endif
 } // namespace
 
 TEST_CASE("json_view materialize() source positions")
@@ -345,3 +451,124 @@ TEST_CASE("json_view materialize() source positions")
     }
 #endif
 }
+
+#if !defined(JSON_NOEXCEPTION)
+TEST_CASE("json_view exceptions name the path and the bytes of the value")
+{
+    SECTION("as those of basic_json")
+    {
+        const std::vector<std::string> texts =
+        {
+            "null", "true", "42", "\"s\"", "[]", "{}", "[1, [2, [3, []]]]", R"({"a": {"b": [1, {"c": null}]}, "d": "e"})",
+            R"({"a/b": {"~": [1]}, "\u0041\n": {"x": "\t"}, "": [{}]})",
+            R"([1 /* ] */, {"a": [] // }
+, "b": {} } , ])",
+            "\n\t[\r\n  1 ,\n  {\n    \"a\" : [ ]\n  }\n]\n",
+        };
+        for (const auto& text : texts)
+        {
+            CAPTURE(text)
+            const json j = json::parse(text, nullptr, true, true, true);
+            const json_document d = json_document::parse(text, true, true, true);
+            check_exceptions(d.root(), j);
+            json_editable_document e = json_editable_document::parse(text, true, true, true);
+            check_exceptions(e.root(), j);
+            json m = e.root().materialize();
+            check_edit_exceptions(e, e.root(), m);
+        }
+    }
+
+    SECTION("generated documents")
+    {
+        json_view_test::generator gen;
+        for (int i = 0; i < 100; ++i)
+        {
+            std::string text;
+            gen.value(text, 0);
+            CAPTURE(text)
+            const json_document d = json_document::parse(text);
+            check_exceptions(d.root(), json::parse(text));
+        }
+    }
+
+    SECTION("edited documents: as those of materialize()")
+    {
+        const std::string text = R"({"a": [1, 2], "b": {"c": "\n", "d": true}, "e": [3, [4]], "f": "x"})";
+        json_editable_document d = json_editable_document::parse(text);
+        d.set(d.root()["f"], 5);
+        d.push_back(d.root()["a"], json::parse(R"({"n": [null]})"));
+        d.erase(d.root()["e"], 0);
+        d.set(d.root()["b"], "c2", d.root()["e"]);
+        d.insert(d.root()["e"], 0, "new");
+        json m = d.root().materialize();
+        check_exceptions(d.root(), m);
+        check_edit_exceptions(d, d.root(), m);
+        // (a value inside a new container: a path, but no bytes)
+        std::string expected = "[json.exception.type_error.302] ";
+#if JSON_DIAGNOSTICS
+        expected += "(/a/2/n/0) ";
+#endif
+        expected += "type must be number, but is null";
+        CHECK(exception_of_call([&] { static_cast<void>(d.root()["a"][2]["n"][0].get<int>()); }) == expected);
+    }
+
+    SECTION("deep documents with siblings after every nested container")
+    {
+        // the search for the path pushes and pops one frame per level (the
+        // stack reallocates while it grows, and shrinks again at every
+        // sibling): run under a sanitizer, a reference to a popped or moved
+        // frame would show here
+        std::string text = "7";
+        for (int level = 0; level < 40; ++level)
+        {
+            text.insert(0, R"([[1, [2]], )");
+            text += R"(, {"k": [3, {"l": [4]}]}, [5, [6, [7]]]])";
+        }
+        CAPTURE(text)
+        const json j = json::parse(text);
+        const json_document d = json_document::parse(text);
+        check_exceptions(d.root(), j);
+        json_editable_document e = json_editable_document::parse(text);
+        e.set(e.root()[0][0], 9);
+        check_exceptions(e.root(), e.root().materialize());
+    }
+
+    SECTION("examples")
+    {
+        const std::string text = R"({"people": [{"name": "a", "age": 1}, {"name": "c", "age": "x"}], "o": {"k": 1}})";
+        const json_document d = json_document::parse(text);
+        std::string expected_age = "[json.exception.type_error.302] ";
+#if JSON_DIAGNOSTICS
+        expected_age += "(/people/1/age) ";
+#endif
+        expected_age += "(bytes 58-61) type must be number, but is string";
+        CHECK(exception_of_call([&] { static_cast<void>(d.root()["people"][1]["age"].get<int>()); }) == expected_age);
+
+        std::string expected_zz = "[json.exception.out_of_range.403] ";
+#if JSON_DIAGNOSTICS
+        expected_zz += "(/people/0) ";
+#endif
+        expected_zz += "(bytes 12-35) key 'zz' not found";
+        CHECK(exception_of_call([&] { static_cast<void>(d.root().at(json::json_pointer("/people/0/zz"))); }) == expected_zz);
+
+        // the root has no path, and with JSON_DIAGNOSTICS then no bytes either (as basic_json)
+        std::string expected_missing = "[json.exception.out_of_range.403] ";
+#if !JSON_DIAGNOSTICS
+        expected_missing += "(bytes 0-79) ";
+#endif
+        expected_missing += "key 'missing' not found";
+        CHECK(exception_of_call([&] { static_cast<void>(d.root().at("missing")); }) == expected_missing);
+    }
+
+    SECTION("duplicate keys: the value the view finds")
+    {
+        const json_document d = json_document::parse(R"({"x": "1", "x": "22"})");
+        std::string expected = "[json.exception.type_error.302] ";
+#if JSON_DIAGNOSTICS
+        expected += "(/x) ";
+#endif
+        expected += "(bytes 6-9) type must be number, but is string";
+        CHECK(exception_of_call([&] { static_cast<void>(d.root()["x"].get<int>()); }) == expected);
+    }
+}
+#endif

tests/src/unit-msgpack.cpp

@@ -27,12 +27,15 @@ using nlohmann::json;
 #include <sstream>
 #include <iomanip>
 #include <limits>
+#include <list>
 #include <set>
 #include "make_test_data_available.hpp"
 #include "round_trip_corpus.hpp"
 #include "test_utils.hpp"
 #include "sax_countdown.hpp"
+#include "sax_event_loggers.hpp"
 using utils::SaxCountdown;
+using utils::SaxEventLogger;
 
 
 TEST_CASE("MessagePack")
@@ -1645,30 +1648,6 @@ TEST_CASE("MessagePack")
             }
         }
     }
-
-    SECTION("SAX aborts")
-    {
-        SECTION("start_array(len)")
-        {
-            std::vector<uint8_t> const v = {0x93, 0x01, 0x02, 0x03};
-            SaxCountdown scp(0);
-            CHECK(!json::sax_parse(v, &scp, json::input_format_t::msgpack));
-        }
-
-        SECTION("start_object(len)")
-        {
-            std::vector<uint8_t> const v = {0x81, 0xa3, 0x66, 0x6F, 0x6F, 0xc2};
-            SaxCountdown scp(0);
-            CHECK(!json::sax_parse(v, &scp, json::input_format_t::msgpack));
-        }
-
-        SECTION("key()")
-        {
-            std::vector<uint8_t> const v = {0x81, 0xa3, 0x66, 0x6F, 0x6F, 0xc2};
-            SaxCountdown scp(1);
-            CHECK(!json::sax_parse(v, &scp, json::input_format_t::msgpack));
-        }
-    }
 }
 
 TEST_CASE("issue #5405 - array reserve for definite-length MessagePack arrays")
@@ -1739,21 +1718,6 @@ TEST_CASE("issue #5405 - array reserve for definite-length MessagePack arrays")
             CHECK(json::from_msgpack(packed) == j);
         }
     }
-
-    SECTION("a user-defined SAX consumer is unaffected by the internal DOM reserve optimization")
-    {
-        // the reserve() call is local to json_sax_dom_parser / json_sax_dom_callback_parser;
-        // a custom SAX consumer that does not touch a DOM array sees identical events
-        json j = json::array();
-        for (int i = 0; i < 100; ++i)
-        {
-            j.push_back(i);
-        }
-        const auto packed = json::to_msgpack(j);
-
-        SaxCountdown scp(1000000); // large enough to never trigger an abort
-        CHECK(json::sax_parse(packed, &scp, json::input_format_t::msgpack));
-    }
 }
 
 TEST_CASE("regression test - MessagePack ext type rejects a subtype that doesn't fit a single byte")
@@ -1792,15 +1756,6 @@ TEST_CASE("MessagePack nesting does not consume the call stack")
         CHECK(json::from_msgpack(input, true, false).is_discarded());
     }
 
-    SECTION("a well-formed deep value is read through the SAX interface")
-    {
-        std::vector<uint8_t> input(300000, 0x91);
-        input.push_back(0x01); // innermost value
-
-        SaxCountdown accept_all(600001);
-        CHECK(json::sax_parse(input, &accept_all, json::input_format_t::msgpack));
-    }
-
     SECTION("a well-formed deep value is read into a value")
     {
         const std::size_t depth = 10000;
@@ -1848,31 +1803,6 @@ TEST_CASE("MessagePack input that cannot be read is discarded by every overload"
 #endif
 }
 
-TEST_CASE("MessagePack SAX parsing stops at every event")
-{
-    // Containers are opened and closed by the loop that reads them; a SAX
-    // handler that rejects any event - including the end of a nested
-    // container - must stop the parse right there.
-    const auto count_events = [](const std::vector<std::uint8_t>& input)
-    {
-        int events = 0;
-        while (true)
-        {
-            SaxCountdown scp(events);
-            if (json::sax_parse(input, &scp, json::input_format_t::msgpack))
-            {
-                return events;
-            }
-            ++events;
-            REQUIRE(events < 1000);
-        }
-    };
-
-    // 20 events: every container kind closes inside another one
-    const json j = json::parse(R"({"a": [1, {"b": []}], "c": {"d": [[2]]}})");
-    CHECK(count_events(json::to_msgpack(j)) == 20);
-}
-
 TEST_CASE("single MessagePack roundtrip")
 {
     SECTION("sample.json")
@@ -2212,3 +2142,99 @@ TEST_CASE("MessagePack large strings and binaries (chunked reader)")
         }
     }
 }
+
+TEST_CASE("MessagePack arrays of scalars from contiguous and non-contiguous input")
+{
+    // contiguous input reads the integers, floats, booleans, and nils an
+    // array begins with in one loop, other input value by value; both must
+    // give the same events, value, or error for the complete document and for
+    // every truncation of it, and stop at every event a SAX handler rejects
+    const std::vector<std::uint8_t> msgpack =
+    {
+        0x83,                                                       // map(3)
+        0xA1, 'a',                                                  // "a"
+        0xDC, 0x00, 0x16,                                           // array 16 (22)
+        0x00, 0x7F, 0xCC, 0xFF, 0xCD, 0x01, 0x00,                   // 0, 127, 255, 256
+        0xCE, 0x00, 0x01, 0x00, 0x00,                               // 65536
+        0xCF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF,       // 18446744073709551615
+        0xFF, 0xE0, 0xD0, 0x80, 0xD1, 0xFF, 0x00,                   // -1, -32, -128, -256
+        0xD2, 0xFF, 0xFE, 0xFF, 0xFF,                               // -65537
+        0xD3, 0x80, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       // -9223372036854775808
+        0xC2, 0xC3, 0xC0,                                           // false, true, nil
+        0xCA, 0x3F, 0xC0, 0x00, 0x00,                               // 1.5 (float 32)
+        0xCB, 0xBF, 0xB9, 0x99, 0x99, 0x99, 0x99, 0x99, 0x9A,       // -0.1
+        0xA1, 'x',                                                  // "x" (ends the loop)
+        0x01,                                                       // 1
+        0xCB, 0x40, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       // 2.5
+        0xD9, 0x01, 'y',                                            // "y" (str 8)
+        0xC0,                                                       // nil
+        0xA1, 'b',                                                  // "b"
+        0x93,                                                       // fixarray(3)
+        0x92, 0xCB, 0x3F, 0xF4, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // [1.25,
+        0xCB, 0xC0, 0x0E, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       //  -3.75]
+        0x92, 0xCB, 0x3F, 0xF4, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // [1.25,
+        0xCB, 0xC0, 0x0E, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,       //  -3.75]
+        0x90,                                                       // []
+        0xA1, 'c',                                                  // "c"
+        0x94, 0x80, 0x01, 0xC4, 0x01, 0x05, 0x91, 0x03              // [{}, 1, bin 8, [3]]
+    };
+
+    const json j = json::from_msgpack(msgpack);
+    CHECK(j["a"].size() == 22);
+    CHECK(j["a"][5] == 18446744073709551615u);
+    CHECK(j["a"][7] == -32);
+    CHECK(j["a"][11] == (std::numeric_limits<std::int64_t>::min)());
+    CHECK(j["a"][15] == 1.5);
+    CHECK(j["a"][16] == -0.1);
+    CHECK(j["a"][17] == "x");
+    CHECK(j["b"] == json::parse("[[1.25, -3.75], [1.25, -3.75], []]"));
+    CHECK(j["c"][2].is_binary());
+
+    for (std::size_t length = 0; length <= msgpack.size(); ++length)
+    {
+        CAPTURE(length)
+        const std::vector<std::uint8_t> contiguous(msgpack.begin(), msgpack.begin() + static_cast<std::ptrdiff_t>(length));
+        const std::list<std::uint8_t> bytewise(contiguous.begin(), contiguous.end());
+
+        CHECK(json::from_msgpack(contiguous, true, false).dump() == json::from_msgpack(bytewise.begin(), bytewise.end(), true, false).dump());
+#if !defined(JSON_NOEXCEPTION)
+        std::string contiguous_error;
+        std::string bytewise_error;
+        try
+        {
+            json _;
+            _ = json::from_msgpack(contiguous);
+        }
+        catch (const json::parse_error& e)
+        {
+            contiguous_error = e.what();
+        }
+        try
+        {
+            json _;
+            _ = json::from_msgpack(bytewise.begin(), bytewise.end());
+        }
+        catch (const json::parse_error& e)
+        {
+            bytewise_error = e.what();
+        }
+        CHECK(contiguous_error == bytewise_error);
+#endif
+
+        SaxEventLogger from_contiguous;
+        SaxEventLogger from_bytewise;
+        CHECK(json::sax_parse(contiguous, &from_contiguous, json::input_format_t::msgpack)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::msgpack));
+        CHECK(from_contiguous.events == from_bytewise.events);
+    }
+
+    for (int events = 0; events <= 60; ++events)
+    {
+        CAPTURE(events)
+        const std::list<std::uint8_t> bytewise(msgpack.begin(), msgpack.end());
+        SaxCountdown from_contiguous(events);
+        SaxCountdown from_bytewise(events);
+        CHECK(json::sax_parse(msgpack, &from_contiguous, json::input_format_t::msgpack)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::msgpack));
+    }
+}

tests/src/unit-regression3.cpp

@@ -22,14 +22,6 @@
 // scoped enum, so get<std::byte>() (needed below to get<std::vector<std::byte>>()
 // from a plain JSON array, not just from an already-binary value) relies on
 // enum serialization being enabled
-// capture whether JSON_DELETE_DEPRECATED_FUNCTIONS was enabled on the command
-// line *before* including json.hpp, since the library #undefs it once the header
-// has been fully processed (see include/nlohmann/detail/macro_unscope.hpp); the
-// tests of deprecated functions are skipped if these functions are deleted
-#if defined(JSON_DELETE_DEPRECATED_FUNCTIONS) && (JSON_DELETE_DEPRECATED_FUNCTIONS == 1)
-    #define JSON_TEST_DEPRECATED_FUNCTIONS_DELETED
-#endif
-
 #if defined(JSON_DISABLE_ENUM_SERIALIZATION) && (JSON_DISABLE_ENUM_SERIALIZATION == 1)
     #define SKIP_TESTS_FOR_ENUM_SERIALIZATION
 #endif
@@ -810,50 +802,6 @@ TEST_CASE("issue #5338 - truncated CBOR tagged binary subtype is rejected")
     }
 }
 
-TEST_CASE("issue #5676 - SAX parsing of CBOR tags")
-{
-    const json expected = json::binary({1, 2, 3}, 42);
-    const auto cbor = json::to_cbor(expected);
-
-    nlohmann::detail::json_sax_acceptor<json> acceptor;
-    CHECK_FALSE(json::sax_parse(cbor, &acceptor, json::input_format_t::cbor));
-    CHECK_FALSE(json::sax_parse(cbor, &acceptor, json::input_format_t::cbor,
-                                true, false, false, json::cbor_tag_handler_t::error));
-
-    CHECK(json::sax_parse(cbor, &acceptor, json::input_format_t::cbor,
-                          true, false, false, json::cbor_tag_handler_t::ignore));
-
-    json parsed;
-    nlohmann::detail::json_sax_dom_parser<json, nlohmann::detail::string_input_adapter_type> sax(parsed);
-    CHECK(json::sax_parse(cbor, &sax, json::input_format_t::cbor,
-                          true, false, false, json::cbor_tag_handler_t::store));
-    CHECK(parsed == expected);
-
-    json iterator_parsed;
-    nlohmann::detail::json_sax_dom_parser<json, nlohmann::detail::string_input_adapter_type> iterator_sax(iterator_parsed);
-    CHECK(json::sax_parse(cbor.begin(), cbor.end(), &iterator_sax, json::input_format_t::cbor,
-                          true, false, false, json::cbor_tag_handler_t::store));
-    CHECK(iterator_parsed == expected);
-
-#ifndef JSON_TEST_DEPRECATED_FUNCTIONS_DELETED
-    json span_parsed;
-    nlohmann::detail::json_sax_dom_parser<json, nlohmann::detail::string_input_adapter_type> span_sax(span_parsed);
-    CHECK(json::sax_parse(nlohmann::detail::span_input_adapter(cbor.data(), cbor.size()), &span_sax,
-                          json::input_format_t::cbor, true, false, false, json::cbor_tag_handler_t::store));
-    CHECK(span_parsed == expected);
-#endif
-
-    const std::string text = "null";
-    CHECK(json::sax_parse(text, &acceptor, json::input_format_t::json,
-                          true, false, false, json::cbor_tag_handler_t::store));
-    CHECK(json::sax_parse(text.begin(), text.end(), &acceptor, json::input_format_t::json,
-                          true, false, false, json::cbor_tag_handler_t::store));
-#ifndef JSON_TEST_DEPRECATED_FUNCTIONS_DELETED
-    CHECK(json::sax_parse(nlohmann::detail::span_input_adapter(text.data(), text.size()), &acceptor,
-                          json::input_format_t::json, true, false, false, json::cbor_tag_handler_t::store));
-#endif
-}
-
 TEST_CASE("issue #5402 - update(merge_objects=true) overwrites a primitive with an object")
 {
     json t = {{"k", 1}};

tests/src/unit-sax_parse.cpp

@@ -0,0 +1,781 @@
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++ (supporting code)
+// |  |  |__   |  |  | | | |  version 3.12.0
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013-2026 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+/////////////////////////////////////////////////////////////////////
+// Tests that call basic_json::sax_parse have a file of their own: every
+// sax_parse call instantiates the parser and binary reader that recover from
+// errors (see #3989), and in unit-regression2.cpp, unit-regression3.cpp, and
+// unit-msgpack.cpp this made the objects too large for the MinGW linker to
+// relocate (see #5511).
+/////////////////////////////////////////////////////////////////////
+
+#include "doctest_compatibility.h"
+
+// capture whether JSON_DELETE_DEPRECATED_FUNCTIONS was enabled on the command
+// line *before* including json.hpp, since the library #undefs it once the header
+// has been fully processed (see include/nlohmann/detail/macro_unscope.hpp); the
+// tests of deprecated functions are skipped if these functions are deleted
+#if defined(JSON_DELETE_DEPRECATED_FUNCTIONS) && (JSON_DELETE_DEPRECATED_FUNCTIONS == 1)
+    #define JSON_TEST_DEPRECATED_FUNCTIONS_DELETED
+#endif
+
+#include <nlohmann/json.hpp>
+using json = nlohmann::json;
+
+#include <cmath>
+#include <cstddef>
+#include <cstdint>
+#include <initializer_list>
+#include <map>
+#include <string>
+#include <utility>
+#include <vector>
+
+#include "sax_countdown.hpp"
+using utils::SaxCountdown;
+
+// a narrow number_float_t, so that a double read from binary input can
+// overflow it
+using float_json = nlohmann::basic_json<std::map, std::vector, std::string, bool, std::int64_t, std::uint64_t, float>;
+
+DOCTEST_CLANG_SUPPRESS_WARNING_PUSH
+DOCTEST_CLANG_SUPPRESS_WARNING("-Wexit-time-destructors")
+
+namespace
+{
+/// builds a value from SAX events, asks the parser to recover from its first
+/// 100 errors, and checks that the events are balanced (see #3989)
+template<typename BasicJsonType>
+class BasicRecoveringParser
+{
+  public:
+    explicit BasicRecoveringParser(BasicJsonType& j)
+        : dom(j, false)
+    {}
+
+    bool null()
+    {
+        value();
+        return dom.null();
+    }
+
+    bool boolean(bool val)
+    {
+        value();
+        return dom.boolean(val);
+    }
+
+    bool number_integer(typename BasicJsonType::number_integer_t val)
+    {
+        value();
+        return dom.number_integer(val);
+    }
+
+    bool number_unsigned(typename BasicJsonType::number_unsigned_t val)
+    {
+        value();
+        return dom.number_unsigned(val);
+    }
+
+    bool number_float(typename BasicJsonType::number_float_t val, const std::string& s)
+    {
+        value();
+        return dom.number_float(val, s);
+    }
+
+    bool string(std::string& val)
+    {
+        value();
+        return dom.string(val);
+    }
+
+    bool binary(typename BasicJsonType::binary_t& val)
+    {
+        value();
+        return dom.binary(val);
+    }
+
+    bool start_object(std::size_t elements)
+    {
+        value();
+        stack.push_back('o');
+        return dom.start_object(elements);
+    }
+
+    bool key(std::string& val)
+    {
+        if (stack.empty() || stack.back() != 'o')
+        {
+            well_formed = false;
+            return false;
+        }
+        stack.back() = 'v';
+        return dom.key(val);
+    }
+
+    bool end_object()
+    {
+        if (stack.empty() || stack.back() != 'o')
+        {
+            well_formed = false;
+            return false;
+        }
+        stack.pop_back();
+        return dom.end_object();
+    }
+
+    bool start_array(std::size_t elements)
+    {
+        value();
+        stack.push_back('a');
+        return dom.start_array(elements);
+    }
+
+    bool end_array()
+    {
+        if (stack.empty() || stack.back() != 'a')
+        {
+            well_formed = false;
+            return false;
+        }
+        stack.pop_back();
+        return dom.end_array();
+    }
+
+    bool parse_error(std::size_t /*unused*/, const std::string& /*unused*/, const json::exception& ex)
+    {
+        messages.emplace_back(ex.what());
+        // a limit, so that a reader that does not stop fails the test
+        // instead of making it hang
+        return ++errors < 100;
+    }
+
+    /// whether the events were balanced and every key was followed by a value
+    bool balanced() const
+    {
+        return well_formed && stack.empty();
+    }
+
+    /// builds the value
+    nlohmann::detail::json_sax_dom_parser<BasicJsonType> dom;
+    std::size_t errors = 0;
+    std::vector<std::string> messages {}; // NOLINT(readability-redundant-member-init)
+    std::vector<char> stack {}; // NOLINT(readability-redundant-member-init)
+    bool well_formed = true;
+
+  private:
+    void value()
+    {
+        if (!stack.empty())
+        {
+            if (stack.back() == 'v')
+            {
+                stack.back() = 'o';
+            }
+            else if (stack.back() == 'o')
+            {
+                well_formed = false;
+            }
+        }
+    }
+
+};
+
+using RecoveringParser = BasicRecoveringParser<json>;
+
+struct BinaryParseResult
+{
+    json value;
+    std::size_t errors;
+    std::vector<std::string> messages;
+    bool ok;
+    bool balanced;
+};
+
+BinaryParseResult parse_binary_recovering(const std::vector<std::uint8_t>& input, const json::input_format_t format)
+{
+    json j;
+    RecoveringParser sax(j);
+    const bool ok = json::sax_parse(input, &sax, format);
+    return {j, sax.errors, sax.messages, ok, sax.balanced()};
+}
+
+#if !defined(JSON_NOEXCEPTION)
+/// the message of the exception that reading @a input into a JSON value
+/// throws, or an empty string if reading succeeds
+std::string binary_error_message(const std::vector<std::uint8_t>& input, const json::input_format_t format)
+{
+    try
+    {
+        json _;
+        switch (format)
+        {
+            case json::input_format_t::cbor:
+                _ = json::from_cbor(input);
+                break;
+            case json::input_format_t::msgpack:
+                _ = json::from_msgpack(input);
+                break;
+            case json::input_format_t::ubjson:
+                _ = json::from_ubjson(input);
+                break;
+            case json::input_format_t::bjdata:
+                _ = json::from_bjdata(input);
+                break;
+            case json::input_format_t::bson:
+                _ = json::from_bson(input);
+                break;
+            case json::input_format_t::bon8:
+                _ = json::from_bon8(input);
+                break;
+            case json::input_format_t::json:
+            default:
+                break;
+        }
+    }
+    catch (const json::exception& e)
+    {
+        return e.what();
+    }
+    return "";
+}
+#endif
+
+/// a BSON element: its type, its name, and its value
+std::vector<std::uint8_t> bson_element(const std::uint8_t type, const std::string& name, const std::vector<std::uint8_t>& value)
+{
+    std::vector<std::uint8_t> result = {type};
+    result.insert(result.end(), name.begin(), name.end());
+    result.push_back(0x00);
+    result.insert(result.end(), value.begin(), value.end());
+    return result;
+}
+
+/// a BSON document of the given elements; @a size_offset is added to the
+/// size it declares
+std::vector<std::uint8_t> bson_document(const std::vector<std::vector<std::uint8_t>>& elements, const int size_offset = 0)
+{
+    std::vector<std::uint8_t> body;
+    for (const auto& element : elements)
+    {
+        body.insert(body.end(), element.begin(), element.end());
+    }
+    const auto size = static_cast<std::uint32_t>(static_cast<int>(body.size()) + 5 + size_offset);
+    std::vector<std::uint8_t> result = {static_cast<std::uint8_t>(size & 0xFFu), static_cast<std::uint8_t>((size >> 8u) & 0xFFu),
+                                        static_cast<std::uint8_t>((size >> 16u) & 0xFFu), static_cast<std::uint8_t>((size >> 24u) & 0xFFu)
+                                       };
+    result.insert(result.end(), body.begin(), body.end());
+    result.push_back(0x00);
+    return result;
+}
+
+/// a BSON int32 value
+std::vector<std::uint8_t> bson_int32(const std::int32_t value)
+{
+    const auto u = static_cast<std::uint32_t>(value);
+    return {static_cast<std::uint8_t>(u & 0xFFu), static_cast<std::uint8_t>((u >> 8u) & 0xFFu),
+            static_cast<std::uint8_t>((u >> 16u) & 0xFFu), static_cast<std::uint8_t>((u >> 24u) & 0xFFu)};
+}
+
+/// a BSON string value, whose length is @a length_offset off
+std::vector<std::uint8_t> bson_string(const std::string& value, const std::int32_t length_offset = 0)
+{
+    auto result = bson_int32(static_cast<std::int32_t>(value.size() + 1) + length_offset);
+    result.insert(result.end(), value.begin(), value.end());
+    result.push_back(0x00);
+    return result;
+}
+
+/// @a count bytes of value 0xAB
+std::vector<std::uint8_t> bytes(const std::size_t count)
+{
+    return std::vector<std::uint8_t>(count, 0xAB);
+}
+
+/// the bytes of @a parts, one after the other
+std::vector<std::uint8_t> concatenated(std::initializer_list<std::vector<std::uint8_t>> parts)
+{
+    std::vector<std::uint8_t> result;
+    for (const auto& part : parts)
+    {
+        result.insert(result.end(), part.begin(), part.end());
+    }
+    return result;
+}
+
+/// U+FFFD REPLACEMENT CHARACTER
+std::string replacement_character()
+{
+    return "\xEF\xBF\xBD";
+}
+}  // namespace
+
+TEST_CASE("regression test - #3989 SAX parse_error() returning true")
+{
+    SECTION("binary formats complete what was read before the input ends")
+    {
+        const json j = {{"a", {1, -2, {{"b", "c"}}, json::array()}}, {"d", {{"e", nullptr}, {"f", true}}}, {"g", 1.5}, {"h", json::binary({1, 2, 3})}};
+
+        const std::vector<std::pair<json::input_format_t, std::vector<std::uint8_t>>> encodings =
+        {
+            {json::input_format_t::cbor, json::to_cbor(j)},
+            {json::input_format_t::msgpack, json::to_msgpack(j)},
+            {json::input_format_t::ubjson, json::to_ubjson(j)},
+            {json::input_format_t::ubjson, json::to_ubjson(j, true, true)},
+            {json::input_format_t::bjdata, json::to_bjdata(j)},
+            {json::input_format_t::bjdata, json::to_bjdata(j, true, true)},
+            {json::input_format_t::bson, json::to_bson(j)},
+            {json::input_format_t::bon8, json::to_bon8(j)},
+        };
+
+        for (const auto& encoding : encodings)
+        {
+            const auto format = encoding.first;
+            const auto& bytes = encoding.second;
+            CAPTURE(format)
+
+            // every prefix is truncated input
+            for (std::size_t length = 0; length < bytes.size(); ++length)
+            {
+                CAPTURE(length)
+                const auto result = parse_binary_recovering(std::vector<std::uint8_t>(bytes.begin(), bytes.begin() + static_cast<std::ptrdiff_t>(length)), format);
+                CHECK(!result.ok);
+                CHECK(result.errors == 1);
+                CHECK(result.balanced);
+            }
+
+            // the complete input is read as usual (binary values do not
+            // round-trip through every format, so compare with a plain parse)
+            json expected;
+            nlohmann::detail::json_sax_dom_parser<json> dom(expected);
+            CHECK(json::sax_parse(bytes, &dom, format));
+            const auto complete = parse_binary_recovering(bytes, format);
+            CHECK(complete.ok);
+            CHECK(complete.errors == 0);
+            CHECK(complete.value == expected);
+
+            // a byte after the value
+            auto trailing_bytes = bytes;
+            trailing_bytes.push_back(0x01);
+            const auto trailing = parse_binary_recovering(trailing_bytes, format);
+            CHECK(!trailing.ok);
+            CHECK(trailing.errors == 1);
+            CHECK(trailing.value == expected);
+        }
+    }
+
+    SECTION("containers without an end")
+    {
+        // these made the readers loop, or read on, after the error
+        const auto cbor_array = parse_binary_recovering({0x9F}, json::input_format_t::cbor);
+        CHECK(cbor_array.errors == 1);
+        CHECK(cbor_array.value == json::array());
+
+        const auto cbor_map = parse_binary_recovering({0xBF, 0x61, 'a'}, json::input_format_t::cbor);
+        CHECK(cbor_map.errors == 1);
+        CHECK(cbor_map.value == json({{"a", nullptr}}));
+
+        const auto msgpack_array = parse_binary_recovering({0xDD, 0xFF, 0xFF, 0xFF, 0xFF}, json::input_format_t::msgpack);
+        CHECK(msgpack_array.errors == 1);
+        CHECK(msgpack_array.value == json::array());
+
+        const auto msgpack_map = parse_binary_recovering({0x81, 0xA1, 'a', 0x92, 0x01}, json::input_format_t::msgpack);
+        CHECK(msgpack_map.errors == 1);
+        CHECK(msgpack_map.value == json({{"a", {1}}}));
+    }
+
+    SECTION("BJData ndarray")
+    {
+        // a 2x3 int8 array with two of its six elements; the annotated array
+        // format opens an object and two arrays of its own
+        const auto result = parse_binary_recovering({'[', '$', 'i', '#', '[', '$', 'i', '#', 'i', 2, 2, 3, 1, 2}, json::input_format_t::bjdata);
+        CHECK(result.errors == 1);
+        CHECK(result.balanced);
+        CHECK(result.value == json({{"_ArrayType_", "int8"}, {"_ArraySize_", {2, 3}}, {"_ArrayData_", {1, 2}}}));
+    }
+
+    SECTION("binary formats repair items whose end is known")
+    {
+        struct Repair
+        {
+            json::input_format_t format;
+            std::vector<std::uint8_t> input;
+            json expected;
+            std::size_t errors;
+        };
+
+        const std::vector<Repair> repairs =
+        {
+            // CBOR: tags are ignored (here tag 1 and the self-describe tag 55799)
+            {json::input_format_t::cbor, {0x82, 0xC1, 0x05, 0xD9, 0xD9, 0xF7, 0x06}, {5, 6}, 2},
+            // CBOR: undefined and other simple values become null
+            {json::input_format_t::cbor, {0x84, 0xF7, 0xE0, 0xF8, 0x20, 0x01}, {nullptr, nullptr, nullptr, 1}, 3},
+            // CBOR: members whose key is not a string are skipped, whatever their key and value
+            {json::input_format_t::cbor, {0xA4, 0x01, 0x02, 0x82, 0x01, 0x02, 0xA1, 0x61, 'x', 0x9F, 0xFF, 0xC1, 0x01, 0x5F, 0x41, 0x00, 0xFF, 0x61, 'a', 0x03}, {{"a", 3}}, 3},
+            {json::input_format_t::cbor, {0xBF, 0xF5, 0xBF, 0x61, 'x', 0x7F, 0x61, 'y', 0xFF, 0xFF, 0x61, 'a', 0x03, 0xFF}, {{"a", 3}}, 1},
+            // MessagePack: members whose key is not a string are skipped
+            {json::input_format_t::msgpack, {0x84, 0x01, 0x02, 0x81, 0xA1, 'x', 0x01, 0x92, 0x01, 0x02, 0xD4, 0x01, 0x02, 0xC0, 0xA1, 'a', 0x04}, {{"a", 4}}, 3},
+            // UBJSON: a char that is not ASCII becomes U+FFFD
+            {json::input_format_t::ubjson, {'[', 'C', 0x80, 'C', 'A', ']'}, {replacement_character(), "A"}, 1},
+            // UBJSON: the longest beginning of a high-precision number is kept
+            {json::input_format_t::ubjson, {'[', 'H', 'i', 5, '1', '2', 'a', 'b', 'c', 'H', 'i', 2, '1', '.', 'H', 'i', 3, 'a', 'b', 'c', 'H', 'i', 3, '4', '.', '5', ']'}, {12, 1, nullptr, 4.5}, 3},
+            // BJData, too
+            {json::input_format_t::bjdata, {'[', 'C', 0xFF, 'H', 'i', 2, '-', '1', 'H', 'i', 2, '-', 'x', ']'}, {replacement_character(), -1, nullptr}, 2},
+            // a NUL ends a high-precision number, as it ends JSON text
+            {json::input_format_t::ubjson, {'[', 'H', 'i', 4, '1', '2', 0, '9', 'H', 'i', 2, 0, '1', 'i', 3, ']'}, {12, nullptr, 3}, 2},
+            // BON8: members whose key is not a string are skipped
+            {json::input_format_t::bon8, {0x89, 0x91, 0x92, 0xC9, 0x40, 0x82, 0x91, 0x92, 0x61, 0x93}, {{"a", 3}}, 2},
+            {json::input_format_t::bon8, {0x8B, 0x91, 0x85, 0x91, 0xFE, 0xFA, 0x8B, 'x', 0x91, 0xFE, 0x61, 0x93, 0xFE}, {{"a", 3}}, 2},
+            // BSON: elements of types the library does not read become null
+            {
+                json::input_format_t::bson, bson_document(
+                {
+                    bson_element(0x07, "_id", bytes(12)),                         // ObjectId
+                    bson_element(0x09, "date", bytes(8)),                         // UTC datetime
+                    bson_element(0x13, "decimal", bytes(16)),                     // 128-bit decimal
+                    bson_element(0x0B, "regex", {'a', '+', 0, 'i', 0}),           // regular expression
+                    bson_element(0x0D, "code", bson_string("f()")),               // JavaScript code
+                    bson_element(0x0E, "symbol", bson_string("s")),               // symbol
+                    bson_element(0x0C, "pointer", concatenated({bson_string("c"), bytes(12)})), // DBPointer
+                    bson_element(0x0F, "scope", concatenated({bson_int32(15), bson_string("g"), bson_document({})})), // code with scope
+                    bson_element(0x06, "undefined", {}),                          // undefined
+                    bson_element(0xFF, "min", {}),                                // min key
+                    bson_element(0x7F, "max", {}),                                // max key
+                    bson_element(0x10, "z", bson_int32(7)),
+                }),
+                {{"_id", nullptr}, {"date", nullptr}, {"decimal", nullptr}, {"regex", nullptr}, {"code", nullptr}, {"symbol", nullptr}, {"pointer", nullptr}, {"scope", nullptr}, {"undefined", nullptr}, {"min", nullptr}, {"max", nullptr}, {"z", 7}},
+                11
+            },
+            // BSON: an element of an unknown type becomes null, and the rest of its document is skipped
+            {
+                json::input_format_t::bson, bson_document(
+                {
+                    bson_element(0x03, "inner", bson_document({bson_element(0x10, "a", bson_int32(1)), bson_element(0x42, "x", bytes(3)), bson_element(0x10, "b", bson_int32(2))})),
+                    bson_element(0x04, "array", bson_document({bson_element(0x10, "0", bson_int32(1)), bson_element(0x42, "1", bytes(3))})),
+                    bson_element(0x10, "after", bson_int32(3)),
+                }),
+                {{"inner", {{"a", 1}, {"x", nullptr}}}, {"array", {1, nullptr}}, {"after", 3}},
+                2
+            },
+            // BSON: so does a string or byte array whose length cannot be right
+            {
+                json::input_format_t::bson, bson_document(
+                {
+                    bson_element(0x03, "inner", bson_document({bson_element(0x02, "s", bson_string("abc", -10)), bson_element(0x10, "b", bson_int32(2))})),
+                    bson_element(0x03, "bin", bson_document({bson_element(0x05, "b", concatenated({bson_int32(-1), bytes(1)})), bson_element(0x10, "b", bson_int32(2))})),
+                    bson_element(0x10, "after", bson_int32(3)),
+                }),
+                {{"inner", {{"s", nullptr}}}, {"bin", {{"b", nullptr}}}, {"after", 3}},
+                2
+            },
+            // BSON: a string without its terminator, and a document whose size does not match, are kept
+            {
+                json::input_format_t::bson, bson_document(
+                {
+                    bson_element(0x02, "s", {2, 0, 0, 0, 'a', 'X'}),
+                    bson_element(0x03, "inner", bson_document({bson_element(0x10, "a", bson_int32(1))}, 1)),
+                }),
+                {{"s", "a"}, {"inner", {{"a", 1}}}},
+                2
+            },
+            // BSON: so is a document whose size is too small for its content,
+            // which is read up to its terminator
+            {
+                json::input_format_t::bson, bson_document(
+                {
+                    bson_element(0x03, "inner", bson_document({bson_element(0x10, "a", bson_int32(1)), bson_element(0x10, "b", bson_int32(2))}, -7)),
+                    bson_element(0x10, "after", bson_int32(3)),
+                }),
+                {{"inner", {{"a", 1}, {"b", 2}}}, {"after", 3}},
+                1
+            },
+        };
+
+        for (const auto& repair : repairs)
+        {
+            CAPTURE(repair.format)
+            CAPTURE(repair.input)
+            const auto result = parse_binary_recovering(repair.input, repair.format);
+            CHECK(!result.ok);
+            CHECK(result.balanced);
+            CHECK(result.errors == repair.errors);
+            CHECK(result.value == repair.expected);
+            REQUIRE(!result.messages.empty());
+#if !defined(JSON_NOEXCEPTION)
+            // the first error is the one reported without recovering; under
+            // JSON_NOEXCEPTION, reading without recovering aborts instead of
+            // throwing, so there is no message to compare with
+            CHECK(result.messages.front() == binary_error_message(repair.input, repair.format));
+#endif
+        }
+    }
+
+    SECTION("binary formats repair numbers that are out of range")
+    {
+        // CBOR: a double too large for a float number_float_t
+        float_json cbor;
+        BasicRecoveringParser<float_json> sax(cbor);
+        const std::vector<std::uint8_t> cbor_input = {0x82, 0xFB, 0x7E, 0x37, 0xE4, 0x3C, 0x88, 0x00, 0x75, 0x9C, 0x01}; // [1e300, 1]
+        CHECK(!float_json::sax_parse(cbor_input, &sax, float_json::input_format_t::cbor));
+        CHECK(sax.errors == 1);
+        CHECK(sax.messages.front() == "[json.exception.out_of_range.406] syntax error while parsing CBOR value: number overflow");
+        REQUIRE(cbor.size() == 2);
+        CHECK(std::isinf(cbor[0].get<float>()));
+        CHECK(cbor[1] == 1);
+
+        // UBJSON: a high-precision number too large for number_float_t
+        const auto ubjson = parse_binary_recovering({'H', 'i', 5, '1', 'e', '9', '9', '9'}, json::input_format_t::ubjson);
+        CHECK(ubjson.errors == 1);
+        CHECK(ubjson.value.is_number_float());
+        CHECK(std::isinf(ubjson.value.get<double>()));
+    }
+
+    SECTION("binary formats stop where the end of an item is not known")
+    {
+        // a byte that begins no item
+        const auto cbor = parse_binary_recovering({0x82, 0x01, 0x1C, 0x02}, json::input_format_t::cbor);
+        CHECK(cbor.errors == 1);
+        CHECK(cbor.value == json({1}));
+
+        // a key that is no item: the unused MessagePack byte, a CBOR break
+        // in a map of known size, and the end of a BON8 container
+        const auto msgpack = parse_binary_recovering({0x82, 0xA1, 'a', 0x01, 0xC1, 0x02}, json::input_format_t::msgpack);
+        CHECK(msgpack.errors == 1);
+        CHECK(msgpack.value == json({{"a", 1}}));
+        const auto cbor_break = parse_binary_recovering({0xA2, 0x61, 'a', 0x01, 0xFF, 0x02}, json::input_format_t::cbor);
+        CHECK(cbor_break.errors == 1);
+        CHECK(cbor_break.value == json({{"a", 1}}));
+        const auto bon8 = parse_binary_recovering({0x88, 0x61, 0x91, 0xFE}, json::input_format_t::bon8);
+        CHECK(bon8.errors == 1);
+        CHECK(bon8.value == json({{"a", 1}}));
+
+        // an indefinite-length string inside an indefinite-length string
+        const auto nested = parse_binary_recovering({0x82, 0x01, 0x7F, 0x7F, 0x61, 'a', 0xFF, 0xFF}, json::input_format_t::cbor);
+        CHECK(nested.errors == 1);
+        CHECK(nested.value == json({1}));
+
+        // a BJData ndarray whose element type is not valid: the type is
+        // checked before the object that would hold the ndarray is begun
+        const auto ndarray = parse_binary_recovering({'[', '$', 0x01, '#', '[', '$', 'i', '#', 'i', 2, 2, 3}, json::input_format_t::bjdata);
+        CHECK(ndarray.errors == 1);
+        CHECK(ndarray.balanced);
+        CHECK(ndarray.value == nullptr);
+
+        // a skipped member that the input ends in
+        const auto truncated = parse_binary_recovering({0xA2, 0x01, 0x82, 0x01}, json::input_format_t::cbor);
+        CHECK(truncated.errors == 2);
+        CHECK(truncated.balanced);
+        CHECK(truncated.value == json::object());
+
+        // a BSON element of an unknown type in a document whose size cannot be
+        // right: the size is reported when it is read, and the element ends
+        // the document, as its end is not known
+        const auto bson = parse_binary_recovering(bson_document({bson_element(0x10, "a", bson_int32(1)), bson_element(0x42, "x", bytes(3))}, -10), json::input_format_t::bson);
+        CHECK(bson.errors == 2);
+        CHECK(bson.value == json({{"a", 1}, {"x", nullptr}}));
+    }
+
+    SECTION("changed bytes in binary input")
+    {
+        const json j = {{"a", {1, -2, {{"b", "c"}}, json::array()}}, {"d", {{"e", nullptr}, {"f", true}}}, {"g", 1.5}, {"h", json::binary({1, 2, 3})}, {"i", "\xC3\xA4"}};
+
+        const std::vector<std::pair<json::input_format_t, std::vector<std::uint8_t>>> encodings =
+        {
+            {json::input_format_t::cbor, json::to_cbor(j)},
+            {json::input_format_t::msgpack, json::to_msgpack(j)},
+            {json::input_format_t::ubjson, json::to_ubjson(j)},
+            {json::input_format_t::ubjson, json::to_ubjson(j, true, true)},
+            {json::input_format_t::bjdata, json::to_bjdata(j)},
+            {json::input_format_t::bjdata, json::to_bjdata(j, true, true)},
+            {json::input_format_t::bson, json::to_bson(j)},
+            {json::input_format_t::bon8, json::to_bon8(j)},
+        };
+        const std::vector<std::uint8_t> replacements = {0x00, 0x01, 0x7F, 0x80, 0xC1, 0xD9, 0xE0, 0xF7, 0xFE, 0xFF};
+
+        for (const auto& encoding : encodings)
+        {
+            const auto format = encoding.first;
+            const auto& original = encoding.second;
+            CAPTURE(format)
+
+            std::vector<std::vector<std::uint8_t>> inputs;
+            for (std::size_t position = 0; position < original.size(); ++position)
+            {
+                for (const auto replacement : replacements)
+                {
+                    auto changed = original;
+                    changed[position] = replacement;
+                    inputs.push_back(changed);
+                }
+                auto removed = original;
+                removed.erase(removed.begin() + static_cast<std::ptrdiff_t>(position));
+                inputs.push_back(removed);
+            }
+
+            for (const auto& input : inputs)
+            {
+                CAPTURE(input)
+                const auto result = parse_binary_recovering(input, format);
+                CHECK(result.balanced);
+                CHECK(result.errors <= input.size() + 1);
+#if !defined(JSON_NOEXCEPTION)
+                // an error is reported exactly if reading into a JSON value
+                // fails, and the first one is the same (under JSON_NOEXCEPTION,
+                // that reading aborts instead of throwing)
+                const auto message = binary_error_message(input, format);
+                CHECK(result.ok == message.empty());
+                if (!result.ok && result.errors < 100)
+                {
+                    CHECK(result.messages.front() == message);
+                }
+#endif
+            }
+        }
+    }
+
+    SECTION("JSON text")
+    {
+        // the parser stopped, but reported success
+        json j;
+        RecoveringParser sax(j);
+        CHECK(!json::sax_parse("[1,2,3,]", &sax));
+        CHECK(sax.errors == 1);
+        CHECK(j == json({1, 2, 3}));
+    }
+
+    SECTION("the SAX parsers of the library stop")
+    {
+        json _;
+        CHECK(json::from_cbor(std::vector<std::uint8_t> {0x9F}, true, false).is_discarded());
+        CHECK_THROWS_WITH_AS(_ = json::from_cbor(std::vector<std::uint8_t> {0x9F}), "[json.exception.parse_error.110] parse error at byte 2: syntax error while parsing CBOR value: unexpected end of input", json::parse_error&);
+        CHECK(json::parse("[1,2,3,]", nullptr, false).is_discarded());
+        CHECK(!json::accept("[1,2,3,]"));
+    }
+}
+
+
+TEST_CASE("issue #5676 - SAX parsing of CBOR tags")
+{
+    const json expected = json::binary({1, 2, 3}, 42);
+    const auto cbor = json::to_cbor(expected);
+
+    nlohmann::detail::json_sax_acceptor<json> acceptor;
+    CHECK_FALSE(json::sax_parse(cbor, &acceptor, json::input_format_t::cbor));
+    CHECK_FALSE(json::sax_parse(cbor, &acceptor, json::input_format_t::cbor,
+                                true, false, false, json::cbor_tag_handler_t::error));
+
+    CHECK(json::sax_parse(cbor, &acceptor, json::input_format_t::cbor,
+                          true, false, false, json::cbor_tag_handler_t::ignore));
+
+    json parsed;
+    nlohmann::detail::json_sax_dom_parser<json, nlohmann::detail::string_input_adapter_type> sax(parsed);
+    CHECK(json::sax_parse(cbor, &sax, json::input_format_t::cbor,
+                          true, false, false, json::cbor_tag_handler_t::store));
+    CHECK(parsed == expected);
+
+    json iterator_parsed;
+    nlohmann::detail::json_sax_dom_parser<json, nlohmann::detail::string_input_adapter_type> iterator_sax(iterator_parsed);
+    CHECK(json::sax_parse(cbor.begin(), cbor.end(), &iterator_sax, json::input_format_t::cbor,
+                          true, false, false, json::cbor_tag_handler_t::store));
+    CHECK(iterator_parsed == expected);
+
+#ifndef JSON_TEST_DEPRECATED_FUNCTIONS_DELETED
+    json span_parsed;
+    nlohmann::detail::json_sax_dom_parser<json, nlohmann::detail::string_input_adapter_type> span_sax(span_parsed);
+    CHECK(json::sax_parse(nlohmann::detail::span_input_adapter(cbor.data(), cbor.size()), &span_sax,
+                          json::input_format_t::cbor, true, false, false, json::cbor_tag_handler_t::store));
+    CHECK(span_parsed == expected);
+#endif
+
+    const std::string text = "null";
+    CHECK(json::sax_parse(text, &acceptor, json::input_format_t::json,
+                          true, false, false, json::cbor_tag_handler_t::store));
+    CHECK(json::sax_parse(text.begin(), text.end(), &acceptor, json::input_format_t::json,
+                          true, false, false, json::cbor_tag_handler_t::store));
+#ifndef JSON_TEST_DEPRECATED_FUNCTIONS_DELETED
+    CHECK(json::sax_parse(nlohmann::detail::span_input_adapter(text.data(), text.size()), &acceptor,
+                          json::input_format_t::json, true, false, false, json::cbor_tag_handler_t::store));
+#endif
+}
+
+
+TEST_CASE("MessagePack SAX aborts")
+{
+    SECTION("start_array(len)")
+    {
+        std::vector<uint8_t> const v = {0x93, 0x01, 0x02, 0x03};
+        SaxCountdown scp(0);
+        CHECK(!json::sax_parse(v, &scp, json::input_format_t::msgpack));
+    }
+
+    SECTION("start_object(len)")
+    {
+        std::vector<uint8_t> const v = {0x81, 0xa3, 0x66, 0x6F, 0x6F, 0xc2};
+        SaxCountdown scp(0);
+        CHECK(!json::sax_parse(v, &scp, json::input_format_t::msgpack));
+    }
+
+    SECTION("key()")
+    {
+        std::vector<uint8_t> const v = {0x81, 0xa3, 0x66, 0x6F, 0x6F, 0xc2};
+        SaxCountdown scp(1);
+        CHECK(!json::sax_parse(v, &scp, json::input_format_t::msgpack));
+    }
+}
+
+TEST_CASE("issue #5405 - a user-defined SAX consumer is unaffected by the internal DOM reserve optimization")
+{
+    // the reserve() call is local to json_sax_dom_parser / json_sax_dom_callback_parser;
+    // a custom SAX consumer that does not touch a DOM array sees identical events
+    json j = json::array();
+    for (int i = 0; i < 100; ++i)
+    {
+        j.push_back(i);
+    }
+    const auto packed = json::to_msgpack(j);
+
+    SaxCountdown scp(1000000); // large enough to never trigger an abort
+    CHECK(json::sax_parse(packed, &scp, json::input_format_t::msgpack));
+}
+
+TEST_CASE("MessagePack nesting does not consume the call stack - SAX interface")
+{
+    // see the test case of the same name in unit-msgpack.cpp (#5104)
+    std::vector<uint8_t> input(300000, 0x91);
+    input.push_back(0x01); // innermost value
+
+    SaxCountdown accept_all(600001);
+    CHECK(json::sax_parse(input, &accept_all, json::input_format_t::msgpack));
+}
+
+TEST_CASE("MessagePack SAX parsing stops at every event")
+{
+    // Containers are opened and closed by the loop that reads them; a SAX
+    // handler that rejects any event - including the end of a nested
+    // container - must stop the parse right there.
+    const auto count_events = [](const std::vector<std::uint8_t>& input)
+    {
+        int events = 0;
+        while (true)
+        {
+            SaxCountdown scp(events);
+            if (json::sax_parse(input, &scp, json::input_format_t::msgpack))
+            {
+                return events;
+            }
+            ++events;
+            REQUIRE(events < 1000);
+        }
+    };
+
+    // 20 events: every container kind closes inside another one
+    const json j = json::parse(R"({"a": [1, {"b": []}], "c": {"d": [[2]]}})");
+    CHECK(count_events(json::to_msgpack(j)) == 20);
+}
+
+DOCTEST_CLANG_SUPPRESS_WARNING_POP

tests/src/unit-serialization.cpp

@@ -23,6 +23,7 @@ using nlohmann::json;
 #include <cstdint>
 #include <sstream>
 #include <iomanip>
+#include <limits>
 #include <string>
 #include <vector>
 
@@ -248,6 +249,51 @@ TEST_CASE_TEMPLATE("serialization for extreme integer values", T, int32_t, uint3
     }
 }
 
+TEST_CASE("integers of every digit count are written like std::to_string")
+{
+    // dump_integer() writes up to 16 digits in two groups of eight and 17 to
+    // 20 digits as one to four digits and two groups of eight
+    std::vector<std::uint64_t> values;
+    std::uint64_t power = 1;
+    for (int digits = 1; digits <= 20; ++digits)
+    {
+        values.push_back(power);                     // 1 followed by zeros
+        values.push_back(power + 1);
+        values.push_back((power * 2) - 1);
+        if (digits < 20)
+        {
+            values.push_back((power * 10) - 1);      // all 9s
+            values.push_back((power * 10) - power);  // 9 followed by zeros
+            power *= 10;
+        }
+    }
+    values.push_back((std::numeric_limits<std::uint64_t>::max)());
+    values.push_back(static_cast<std::uint64_t>((std::numeric_limits<std::int64_t>::max)()));
+    values.push_back(static_cast<std::uint64_t>((std::numeric_limits<std::int64_t>::max)()) + 1);
+    std::uint64_t state = 0x9E3779B97F4A7C15u;
+    for (int i = 0; i < 20000; ++i)
+    {
+        state = (state * 6364136223846793005u) + 1442695040888963407u;
+        values.push_back(state >> (state % 61));     // every bit length
+    }
+
+    for (const auto v : values)
+    {
+        CAPTURE(v)
+        CHECK(json(v).dump() == std::to_string(v));
+        if (v <= static_cast<std::uint64_t>((std::numeric_limits<std::int64_t>::max)()))
+        {
+            const auto s = static_cast<std::int64_t>(v);
+            CHECK(json(s).dump() == std::to_string(s));
+            CHECK(json(-s).dump() == std::to_string(-s));
+        }
+    }
+    CHECK(json((std::numeric_limits<std::int64_t>::min)()).dump() == "-9223372036854775808");
+
+    // in a container, so the digits are followed by other output
+    CHECK(json::array({std::uint64_t{18446744073709551615u}, std::int64_t{-1234567890123456789}, 1}).dump() == "[18446744073709551615,-1234567890123456789,1]");
+}
+
 TEST_CASE("dump with binary values")
 {
     auto binary = json::binary({1, 2, 3, 4});

tests/src/unit-ubjson.cpp

@@ -21,12 +21,15 @@ using nlohmann::json;
 
 #include <iostream>
 #include <fstream>
+#include <list>
 #include <set>
 #include "make_test_data_available.hpp"
 #include "round_trip_corpus.hpp"
 #include "test_utils.hpp"
 #include "sax_countdown.hpp"
+#include "sax_event_loggers.hpp"
 using utils::SaxCountdown;
+using utils::SaxEventLogger;
 
 
 TEST_CASE("UBJSON")
@@ -3400,3 +3403,72 @@ TEST_CASE("UBJSON optimized container type marker is validated for count 0")
         }
     }
 }
+
+TEST_CASE("UBJSON optimized arrays from contiguous and non-contiguous input")
+{
+    // contiguous input reads the numbers of an array with a type marker in
+    // one loop, other input number by number; both must give the same events,
+    // value, or error for the complete document and for every truncation of
+    // it, and stop at every event a SAX handler rejects
+    json document =
+    {
+        {"u8", {0, 200, 255}},
+        {"i8", {-5, 100, -128}},
+        {"i16", {-30000, 1000, 30000}},
+        {"i32", {-2000000000, 5, 70000}},
+        {"i64", {-9000000000000000000LL, 1, 5000000000LL}},
+        {"f64", {1.5, -0.1, 2.5}},
+        {"pairs", {{1.25, -3.75}, {1.25, -3.75}, json::array()}},
+        {"mixed", {1, "a", 2.5, nullptr}},
+    };
+    const std::vector<std::uint8_t> encoded = json::to_ubjson(document, true, true);
+    CHECK(json::from_ubjson(encoded) == document);
+
+    for (std::size_t length = 0; length <= encoded.size(); ++length)
+    {
+        CAPTURE(length)
+        const std::vector<std::uint8_t> contiguous(encoded.begin(), encoded.begin() + static_cast<std::ptrdiff_t>(length));
+        const std::list<std::uint8_t> bytewise(contiguous.begin(), contiguous.end());
+
+        CHECK(json::from_ubjson(contiguous, true, false).dump() == json::from_ubjson(bytewise.begin(), bytewise.end(), true, false).dump());
+#if !defined(JSON_NOEXCEPTION)
+        std::string contiguous_error;
+        std::string bytewise_error;
+        try
+        {
+            json _;
+            _ = json::from_ubjson(contiguous);
+        }
+        catch (const json::exception& e)
+        {
+            contiguous_error = e.what();
+        }
+        try
+        {
+            json _;
+            _ = json::from_ubjson(bytewise.begin(), bytewise.end());
+        }
+        catch (const json::exception& e)
+        {
+            bytewise_error = e.what();
+        }
+        CHECK(contiguous_error == bytewise_error);
+#endif
+
+        SaxEventLogger from_contiguous;
+        SaxEventLogger from_bytewise;
+        CHECK(json::sax_parse(contiguous, &from_contiguous, json::input_format_t::ubjson)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::ubjson));
+        CHECK(from_contiguous.events == from_bytewise.events);
+    }
+
+    for (int events = 0; events <= 80; ++events)
+    {
+        CAPTURE(events)
+        const std::list<std::uint8_t> bytewise(encoded.begin(), encoded.end());
+        SaxCountdown from_contiguous(events);
+        SaxCountdown from_bytewise(events);
+        CHECK(json::sax_parse(encoded, &from_contiguous, json::input_format_t::ubjson)
+              == json::sax_parse(bytewise.begin(), bytewise.end(), &from_bytewise, json::input_format_t::ubjson));
+    }
+}

tests/src/unit-unicode.cpp

@@ -13,6 +13,7 @@ using nlohmann::json;
 
 #include <algorithm>
 #include <fstream>
+#include <random>
 #include <string>
 #include "make_test_data_available.hpp"
 #include "test_utils.hpp"
@@ -1631,3 +1632,40 @@ TEST_CASE("Unicode RFC 3629")
 }
 
 DOCTEST_CLANG_SUPPRESS_WARNING_POP
+
+TEST_CASE("UTF-8 check of whole strings")
+{
+    // valid_utf8_string() decides short strings with a word test and long ones
+    // 16 bytes at a time where a vector check is available; it must agree with
+    // the byte-wise check for every mix of ASCII, sequences of each length,
+    // ill-formed and truncated sequences, at every length and offset
+    const char* const pieces[] = // NOLINT(cppcoreguidelines-avoid-c-arrays,hicpp-avoid-c-arrays,modernize-avoid-c-arrays)
+    {
+        "a", "\x7F", "\xC3\xA9", "\xE2\x82\xAC", "\xE0\xA0\x80", "\xEF\xBF\xBF", "\xF0\x9F\x98\x80", "\xF4\x8F\xBF\xBF",
+        "\x80", "\xC0\xAF", "\xC3", "\xE2\x82", "\xED\xA0\x80", "\xF0\x9F", "\xF4\x90\x80\x80", "\xFF"
+    };
+    // std::mt19937's sequence is fixed by the standard; no distribution is used
+    std::mt19937 rng(5); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed)
+    std::size_t valid = 0;
+    for (int i = 0; i < 100000; ++i)
+    {
+        std::string s;
+        const std::size_t length = rng() % 100;
+        const auto ascii = static_cast<std::uint32_t>(rng() % 101);
+        while (s.size() < length)
+        {
+            // mostly well-formed pieces (the first eight), so that long valid
+            // strings are common
+            s += (rng() % 100 < ascii) ? std::string("x") : std::string(pieces[rng() % (rng() % 8 == 0 ? 16 : 8)]);
+        }
+        const auto* data = reinterpret_cast<const unsigned char*>(s.data());
+        const bool expected = nlohmann::detail::is_valid_utf8(s);
+        valid += expected ? 1 : 0;
+        CAPTURE(s)
+        CHECK(nlohmann::detail::valid_utf8_string(data, s.size()) == expected);
+        CHECK((nlohmann::detail::valid_utf8_prefix(data, s.size()) == s.size()) == expected);
+    }
+    // both outcomes are common enough to be tested
+    CHECK(valid > 20000);
+    CHECK(valid < 80000);
+}

tools/generate_natvis/README.md

@@ -6,6 +6,11 @@ The checked-in `nlohmann_json.natvis` covers the default namespace (no ABI tags)
 of ABI tags, with both versioned and unversioned namespace variants. Configurations that enable three or more ABI tags
 are not covered by this reduced file; use `--abi-tags` or `--all` (see below) to generate a file for them.
 
+The visualizers of `basic_json_document` and `basic_json_view` live in the same inline namespaces as `basic_json`, so the
+ABI tags apply to them in the same way, and the `view_mode` template argument is covered by a wildcard: it does not
+multiply the namespaces. Each namespace gets three more types (`basic_json_view<*,*>`, `basic_json_document<*,*>`, and
+`detail::view::node`), which roughly doubles the size of the file.
+
 The ABI tag list and the library version are parsed from
 `include/nlohmann/detail/abi_macros.hpp`, so this script must be re-run (via
 `make natvis`) whenever an `NLOHMANN_JSON_ABI_TAG_*` macro is added to that
@@ -35,7 +40,7 @@ version other than the one in `abi_macros.hpp`.
 |-------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
 | `--list-tags`     | Print every ABI tag (in namespace order) with the macro that enables it and the `NLOHMANN_JSON_ABI_TAG_*` identifier, then exit.                                                                                                                    |
 | `--abi-tags TAGS` | Add the namespaces of one configuration, given as a comma-separated list of ABI tags (`_diag,_dp,_ns`; the leading underscore is optional, and the order does not matter). Can be repeated for several configurations. Unknown tags are an error. |
-| `--all`           | Generate the namespaces for every combination of ABI tags. The file size doubles with each tag (about 60 MB for 12 tags). Cannot be combined with `--abi-tags`.                                                                                    |
+| `--all`           | Generate the namespaces for every combination of ABI tags. The file size doubles with each tag (about 60 MB for 11 tags). Cannot be combined with `--abi-tags`.                                                                                    |
 | `--output FILE`   | Write the Natvis file to `FILE` instead of `nlohmann_json.natvis` in the repository root.                                                                                                                                                           |
 
 Each configuration gets both the versioned (`json_abi_diag_dp_ns_v3_12_0`) and the unversioned (`json_abi_diag_dp_ns`,

tools/generate_natvis/generate_natvis.py

@@ -95,7 +95,7 @@ if __name__ == '__main__':
                                 'optional) of one configuration to add to the default namespaces; can be repeated')
     selection.add_argument('--all', action='store_true',
                            help='Generate the namespaces for every combination of ABI tags; the file size doubles '
-                                'with each ABI tag (about 60 MB for 12 tags)')
+                                'with each ABI tag (about 60 MB for 11 tags)')
     parser.add_argument('--list-tags', action='store_true',
                         help='List the ABI tags with the macros that enable them, and exit')
     parser.add_argument('--output', metavar='FILE',

tools/generate_natvis/nlohmann_json.natvis.j2

@@ -5,6 +5,39 @@
 <!-- * * * * * * * * AUTO-GENERATED FILE  * * * * * * * * -->
 
 <AutoVisualizer xmlns="http://schemas.microsoft.com/vstudio/debugger/natvis/2010">
+{# The children of the container node `node` (a pointer expression) of the document `doc` (a document_data pointer
+   expression). A node is only an index into the source text: strings are read from `doc`, so the strings of the
+   direct children are readable here, but the children of a nested container are shown as nodes (see the node
+   visualizer below). Containers with edits (flag 8, "moved") are not expanded. The node kinds are the numbers of
+   node_kind: 1 object, 2 array, 3 string; the float token is shown only by the node visualizer. #}
+{% macro children(node, doc=none) %}
+            <CustomListItems MaxItemsPerView="1000">
+              <Variable Name="nd" InitialValue="{{ node }}" />
+              <Variable Name="obj" InitialValue="nd != 0 &amp;&amp; (int)nd-&gt;kind == 1" />
+              <Variable Name="n" InitialValue="(nd != 0 &amp;&amp; ((int)nd-&gt;kind == 1 || (int)nd-&gt;kind == 2) &amp;&amp; (nd-&gt;flags &amp; 8) == 0) ? (int)nd-&gt;len : 0" />
+              <Variable Name="p" InitialValue="nd + 1" />
+              <Variable Name="i" InitialValue="0" />
+              <Size>obj ? 2 * n : n</Size>
+              <Loop Condition="i &lt; n">
+{% if doc %}
+                <Item Name="key" Condition="obj">{{ doc }}-&gt;base._Elems[p-&gt;flags &amp; 3] + p-&gt;off,[p-&gt;len]s8</Item>
+{% else %}
+                <Item Name="key" Condition="obj">*p</Item>
+{% endif %}
+                <Exec Condition="obj">p = p + 1</Exec>
+{% if doc %}
+                <Item Name="value" Condition="obj &amp;&amp; (int)p-&gt;kind == 3">{{ doc }}-&gt;base._Elems[p-&gt;flags &amp; 3] + p-&gt;off,[p-&gt;len]s8</Item>
+{% endif %}
+                <Item Name="value" Condition="obj{% if doc %} &amp;&amp; (int)p-&gt;kind != 3{% endif %}">*p</Item>
+{% if doc %}
+                <Item Condition="!obj &amp;&amp; (int)p-&gt;kind == 3">{{ doc }}-&gt;base._Elems[p-&gt;flags &amp; 3] + p-&gt;off,[p-&gt;len]s8</Item>
+{% endif %}
+                <Item Condition="!obj{% if doc %} &amp;&amp; (int)p-&gt;kind != 3{% endif %}">*p</Item>
+                <Exec>p = p + (((int)p-&gt;kind == 1 || (int)p-&gt;kind == 2) ? p-&gt;next : 1)</Exec>
+                <Exec>i = i + 1</Exec>
+              </Loop>
+            </CustomListItems>
+{%- endmacro %}
 {% for ns in namespaces %}
     <!-- Namespace {{ ns }} -->
     <Type Name="{{ ns }}::basic_json&lt;*&gt;">
@@ -62,5 +95,44 @@
     </Type>
 
 {% endfor %}
+
+    <!-- json_view / json_document (include/nlohmann/json_view.hpp). They live in the same inline namespaces as
+         basic_json, so the ABI tags need no extra entries; the view_mode is a template argument that the wildcard
+         covers. A view is a pointer to the document's data and a pointer to a node of its flat index; the
+         strings of the source are found through document_data::base (std::array of the MSVC STL: _Elems). -->
+    <Type Name="{{ ns }}::detail::view::node">
+        <DisplayString Condition="(int)kind == 0">null</DisplayString>
+        <DisplayString Condition="(int)kind == 1">object, size={len}</DisplayString>
+        <DisplayString Condition="(int)kind == 2">array, size={len}</DisplayString>
+        <DisplayString Condition="(int)kind == 3">string, length={len}</DisplayString>
+        <DisplayString Condition="(int)kind == 4">boolean: {(flags &amp; 4) != 0}</DisplayString>
+        <DisplayString Condition="(int)kind == 5">int64: {(__int64)((unsigned __int64)len | ((unsigned __int64)next &lt;&lt; 32))}</DisplayString>
+        <DisplayString Condition="(int)kind == 6">uint64: {(unsigned __int64)len | ((unsigned __int64)next &lt;&lt; 32)}</DisplayString>
+        <DisplayString Condition="(int)kind == 7">float, length={len}</DisplayString>
+        <DisplayString>link</DisplayString>
+        <Expand>
+{{ children('this') }}
+        </Expand>
+    </Type>
+
+    <Type Name="{{ ns }}::basic_json_view&lt;*,*&gt;">
+        <DisplayString Condition="m_node == 0">discarded</DisplayString>
+        <DisplayString Condition="(int)m_node-&gt;kind == 3">{m_doc-&gt;base._Elems[m_node-&gt;flags &amp; 3] + m_node-&gt;off,[m_node-&gt;len]s8}</DisplayString>
+        <DisplayString Condition="(int)m_node-&gt;kind == 7">{m_doc-&gt;base._Elems[m_node-&gt;flags &amp; 3] + m_node-&gt;off,[m_node-&gt;len]s8b}</DisplayString>
+        <DisplayString>{*m_node}</DisplayString>
+        <Expand>
+{{ children('m_node', 'm_doc') }}
+        </Expand>
+    </Type>
+
+    <Type Name="{{ ns }}::basic_json_document&lt;*,*&gt;">
+        <DisplayString Condition="m_data._Mypair._Myval2 == 0 || m_data._Mypair._Myval2-&gt;discarded">discarded</DisplayString>
+        <DisplayString>{*m_data._Mypair._Myval2-&gt;nodes}</DisplayString>
+        <Expand>
+            <Item Name="[source]" Condition="m_data._Mypair._Myval2 != 0">m_data._Mypair._Myval2-&gt;src,[m_data._Mypair._Myval2-&gt;size]s8b</Item>
+            <Item Name="[nodes]" Condition="m_data._Mypair._Myval2 != 0">m_data._Mypair._Myval2-&gt;node_count</Item>
+{{ children('m_data._Mypair._Myval2->nodes', 'm_data._Mypair._Myval2') }}
+        </Expand>
+    </Type>
 {% endfor %}
 </AutoVisualizer>