json_view: skip the digits of a fraction 16 bytes at a time (SSE2)

The digits after the decimal point were skipped by skip_digits(), 16 compare-and-branch steps in a row: about 3.5 instructions per digit, and an exit branch that is mispredicted when the lengths of the fractions vary. On x86 with SSE2, the fraction is now skipped by skip_fraction_digits(): 16 bytes are compared at once and the end of the run is taken from the bit mask, which is about a third of the instructions for a fraction of 12 digits. The last bytes of the input are skipped one at a time as before. Where there is no SSE2 (AArch64, other targets, JSON_NO_SIMD) skip_fraction_digits() is skip_digits(): nothing changes there. Integer digits keep skip_digits(), whose exit is predictable. GCC unrolls the loop over the last bytes in every inlined copy of the new function, which made the parse loop 4 KiB larger and parse() of documents with hardly any fractions about 4 % slower. For GCC that loop is therefore out of line (skip_digits_tail()); Clang is faster with it inline. json_document::parse(), A/B in one binary, 6 code layouts, x86-64 Haswell, time after / before (GCC 16.2 / Clang 23): - numbers.json (10,001 floats): 0.715 / 0.680 - canada.json: 0.778 / 0.790 - mesh.json: 0.908 / 0.883 - marine_ik.json: 0.937 / 0.926 - mesh.pretty.json: 0.923 / 0.960 Files with few or no fractions and files with fractions of a fixed length (20,000 prices with two decimals, 60,000 coordinates with six): between 0.98 and 1.014 on both compilers, except citm_catalog.json with GCC (1.05: the instruction count is +0.3 %, and that row varies by +-7 % between builds of unrelated variants on this machine). New test: 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, in several positions, and cut in every place near the end of the text, against json::parse() (the length of the number is part of what is compared), for all option combinations and with and without a terminating NUL. Checked on x86-64 with GCC 16.2, Clang 23 and Apple clang under Rosetta, and with a deliberate off-by-one at the 16th byte. Signed-off-by: Niels Lohmann <mail@nlohmann.me>
Fix the red CI on develop: remove the duplicate is_editable in basic_json_view (#5966 )
2026-10-31 01:45:43 +00:00 · 2026-10-30 22:17:00 +01:00 · 2026-10-30 16:48:31 +01:00 · 2026-10-30 15:20:12 +01:00 · 2026-10-30 15:19:42 +01:00 · 2026-10-30 15:19:08 +01:00
40 changed files with 3448 additions and 835 deletions
@@ -28,7 +28,7 @@ jobs:

    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -22,7 +22,7 @@ jobs:
    runs-on: ubuntu-latest
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -46,7 +46,7 @@ jobs:

    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -20,7 +20,7 @@ jobs:
    timeout-minutes: 30
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -13,7 +13,7 @@ jobs:
    runs-on: ubuntu-22.04
    steps:
    - name: Harden Runner
-      uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+      uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
      with:
        egress-policy: audit

@@ -27,7 +27,7 @@ jobs:

    steps:
    - name: Harden Runner
-      uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+      uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
      with:
        egress-policy: audit

@@ -38,14 +38,14 @@ jobs:

    # Initializes the CodeQL tools for scanning.
    - name: Initialize CodeQL
-      uses: github/codeql-action/init@6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee # v4.39.0
+      uses: github/codeql-action/init@9b3254085494b55a0f7c4c2b3d2d52cc712ce370 # v4.39.1
      with:
        languages: c-cpp

    # Autobuild attempts to build any compiled languages  (C/C++, C#, or Java).
    # If this step fails, then you should remove it and run the build manually (see below)
    - name: Autobuild
-      uses: github/codeql-action/autobuild@6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee # v4.39.0
+      uses: github/codeql-action/autobuild@9b3254085494b55a0f7c4c2b3d2d52cc712ce370 # v4.39.1

    - name: Perform CodeQL Analysis
-      uses: github/codeql-action/analyze@6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee # v4.39.0
+      uses: github/codeql-action/analyze@9b3254085494b55a0f7c4c2b3d2d52cc712ce370 # v4.39.1
@@ -20,7 +20,7 @@ jobs:
      pull-requests: write
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -21,7 +21,7 @@ jobs:
    runs-on: ubuntu-latest
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -31,7 +31,7 @@ jobs:
      security-events: write
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -48,6 +48,6 @@ jobs:
          output: 'flawfinder_results.sarif'

      - name: Upload analysis results to GitHub Security tab
-        uses: github/codeql-action/upload-sarif@6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee # v4.39.0
+        uses: github/codeql-action/upload-sarif@9b3254085494b55a0f7c4c2b3d2d52cc712ce370 # v4.39.1
        with:
          sarif_file: ${{github.workspace}}/flawfinder_results.sarif
@@ -54,7 +54,7 @@ jobs:
    timeout-minutes: 60
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -23,7 +23,7 @@ jobs:

    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -37,7 +37,7 @@ jobs:
    runs-on: ubuntu-latest
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -40,7 +40,7 @@ jobs:

    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -80,6 +80,6 @@ jobs:

      # Upload the results to GitHub's code scanning dashboard.
      - name: "Upload to code-scanning"
-        uses: github/codeql-action/upload-sarif@6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee # v4.39.0
+        uses: github/codeql-action/upload-sarif@9b3254085494b55a0f7c4c2b3d2d52cc712ce370 # v4.39.1
        with:
          sarif_file: results.sarif
@@ -36,7 +36,7 @@ jobs:
    runs-on: ubuntu-26.04
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -65,7 +65,7 @@ jobs:

      # Upload SARIF file generated in previous step
      - name: Upload SARIF file
-        uses: github/codeql-action/upload-sarif@6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee # v4.39.0
+        uses: github/codeql-action/upload-sarif@9b3254085494b55a0f7c4c2b3d2d52cc712ce370 # v4.39.1
        with:
          sarif_file: semgrep.sarif
        if: always()
@@ -16,7 +16,7 @@ jobs:

    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -80,7 +80,7 @@ jobs:
    runs-on: ubuntu-26.04
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -112,7 +112,7 @@ jobs:
        target: [ci_test_single_header, ci_cppcheck, ci_cpplint, ci_reproducible_tests, ci_non_git_tests, ci_offline_testdata, ci_reuse_compliance, ci_test_valgrind]
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -178,7 +178,7 @@ jobs:
    runs-on: ubuntu-26.04
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -433,7 +433,7 @@ jobs:
    runs-on: ubuntu-26.04
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -456,7 +456,7 @@ jobs:
        target: [ci_test_examples, ci_test_build_documentation, ci_test_documentation_mermaid]
    steps:
      - name: Harden Runner
-        uses: step-security/harden-runner@ccd8616d44fd3846e67624a50d5aad6d37bf2d25 # v2.22.1
+        uses: step-security/harden-runner@dc2769abcc0fcfbadfc1df54081c1608579ef6b2 # v2.22.2
        with:
          egress-policy: audit

@@ -28,7 +28,7 @@ jobs:
        with:
          persist-credentials: false
      - name: Set up MinGW
-        uses: egor-tensin/setup-mingw@41b837e47d7f85214629d255b9c4bc3fcbe9fd63 # v3.0
+        uses: egor-tensin/setup-mingw@d54f5ef0ce585bd3fbb1f8a71b5b254f59dbf931 # v3.0
        with:
          platform: ${{ matrix.architecture }}
          version: 12.2.0 # https://github.com/egor-tensin/setup-mingw/issues/14
@@ -153,7 +153,7 @@ jobs:
      - name: Install Clang
        run: curl -fsSL -o LLVM${{ matrix.version }}.exe https://github.com/llvm/llvm-project/releases/download/llvmorg-${{ matrix.version }}/LLVM-${{ matrix.version }}-win64.exe ; 7z x LLVM${{ matrix.version }}.exe -y -o"C:/Program Files/LLVM"
      - name: Set up MinGW
-        uses: egor-tensin/setup-mingw@41b837e47d7f85214629d255b9c4bc3fcbe9fd63 # v3.0
+        uses: egor-tensin/setup-mingw@d54f5ef0ce585bd3fbb1f8a71b5b254f59dbf931 # v3.0
        with:
          platform: x64
          version: 12.2.0 # https://github.com/egor-tensin/setup-mingw/issues/14
@@ -1408,9 +1408,9 @@ THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR I
 - The class contains a port of the shortest double-to-decimal conversion of [Żmij](https://github.com/vitaut/zmij) by Victor Zverovich, including the conversion of the digits to text by Xiang JunBo and the SIMD instruction sequence of Dougall Johnson, which is licensed under the [MIT License](https://opensource.org/licenses/MIT) (see above). Copyright &copy; 2025 [Victor Zverovich](https://github.com/vitaut)
 - The class contains a copy of [Hedley](https://nemequ.github.io/hedley/) from Evan Nemerson which is licensed as [CC0-1.0](https://creativecommons.org/publicdomain/zero/1.0/).
 - The class contains parts of [Google Abseil](https://github.com/abseil/abseil-cpp) which is licensed under the [Apache 2.0 License](https://opensource.org/licenses/Apache-2.0).
- The class contains an adapted version of the Eisel-Lemire algorithm, its table of powers of five, and its digit comparison for long numbers from [fast_float](https://github.com/fastfloat/fast_float) by Daniel Lemire and contributors, which is available under the [MIT License](https://opensource.org/licenses/MIT) (used here), the Apache 2.0 License, and the Boost Software License. Copyright &copy; 2021 The fast_float authors
+- The class contains an adapted version of the Eisel-Lemire algorithm, its table of powers of five, its digit comparison for long numbers, and its parsing of eight digits at a time from [fast_float](https://github.com/fastfloat/fast_float) by Daniel Lemire and contributors, which is available under the [MIT License](https://opensource.org/licenses/MIT) (used here), the Apache 2.0 License, and the Boost Software License. Copyright &copy; 2021 The fast_float authors
 - The view's parser (`<nlohmann/json_view.hpp>`) contains techniques and code adapted from [yyjson](https://github.com/ibireme/yyjson) by YaoYuan, which is licensed under the [MIT License](https://opensource.org/licenses/MIT) (see above): table-driven decoding of `\u` escapes and fixed-offset unrolled checks.
- The view's parser (`<nlohmann/json_view.hpp>`) validates non-ASCII strings with the vector UTF-8 check of [simdjson](https://github.com/simdjson/simdjson) by Daniel Lemire, Geoff Langdale, John Keiser, and contributors (its "lookup4" algorithm and tables, after J. Keiser and D. Lemire, "Validating UTF-8 In Less Than One Instruction Per Byte", 2021), which is available under the [MIT License](https://opensource.org/licenses/MIT) (used here) and the Apache 2.0 License. Copyright &copy; 2018-2025 The simdjson authors
+- The class's lexer and the view's parser (`<nlohmann/json_view.hpp>`) validate non-ASCII strings with the vector UTF-8 check of [simdjson](https://github.com/simdjson/simdjson) by Daniel Lemire, Geoff Langdale, John Keiser, and contributors (its "lookup4" algorithm and tables, after J. Keiser and D. Lemire, "Validating UTF-8 In Less Than One Instruction Per Byte", 2021), which is available under the [MIT License](https://opensource.org/licenses/MIT) (used here) and the Apache 2.0 License. Copyright &copy; 2018-2025 The simdjson authors

 <img align="right" src="https://git.fsfe.org/reuse/reuse-ci/raw/branch/master/reuse-horizontal.png" alt="REUSE Software">

@@ -22,8 +22,8 @@ The class contains a port of the shortest double-to-decimal conversion of [Żmij

 The class contains a copy of [Hedley](https://nemequ.github.io/hedley/) from Evan Nemerson which is licensed as [CC0-1.0](https://creativecommons.org/publicdomain/zero/1.0/).

-The class contains an adapted version of the Eisel-Lemire algorithm, its table of powers of five, and its digit comparison for long numbers from [fast_float](https://github.com/fastfloat/fast_float) by Daniel Lemire and contributors, which is available under the [MIT License](https://opensource.org/licenses/MIT) (used here), the Apache 2.0 License, and the Boost Software License. Copyright &copy; 2021 The fast_float authors
+The class contains an adapted version of the Eisel-Lemire algorithm, its table of powers of five, its digit comparison for long numbers, and its parsing of eight digits at a time from [fast_float](https://github.com/fastfloat/fast_float) by Daniel Lemire and contributors, which is available under the [MIT License](https://opensource.org/licenses/MIT) (used here), the Apache 2.0 License, and the Boost Software License. Copyright &copy; 2021 The fast_float authors

 The view's parser (`<nlohmann/json_view.hpp>`) contains techniques and code adapted from [yyjson](https://github.com/ibireme/yyjson) by YaoYuan, which is licensed under the [MIT License](https://opensource.org/licenses/MIT) (see above): table-driven decoding of `\u` escapes and fixed-offset unrolled checks.

-The view's parser (`<nlohmann/json_view.hpp>`) validates non-ASCII strings with the vector UTF-8 check of [simdjson](https://github.com/simdjson/simdjson) by Daniel Lemire, Geoff Langdale, John Keiser, and contributors (its "lookup4" algorithm and tables, after J. Keiser and D. Lemire, "Validating UTF-8 In Less Than One Instruction Per Byte", 2021), which is available under the [MIT License](https://opensource.org/licenses/MIT) (used here) and the Apache 2.0 License. Copyright &copy; 2018-2025 The simdjson authors
+The class's lexer and the view's parser (`<nlohmann/json_view.hpp>`) validate non-ASCII strings with the vector UTF-8 check of [simdjson](https://github.com/simdjson/simdjson) by Daniel Lemire, Geoff Langdale, John Keiser, and contributors (its "lookup4" algorithm and tables, after J. Keiser and D. Lemire, "Validating UTF-8 In Less Than One Instruction Per Byte", 2021), which is available under the [MIT License](https://opensource.org/licenses/MIT) (used here) and the Apache 2.0 License. Copyright &copy; 2018-2025 The simdjson authors
@@ -19,6 +19,11 @@
 #include <type_traits> // conditional


+// The shortest-digit conversion of doubles (see zmij.hpp) and the writing of
+// its digits, including the conversion of the digits to text by Xiang JunBo
+// and the SIMD instruction sequence by Dougall Johnson, are ported from Zmij
+// by Victor Zverovich (https://github.com/vitaut/zmij, MIT license).
+//
 // SSE2 (every x86-64 CPU) and NEON (every 64-bit Arm CPU) convert the 16
 // digits of a double at once (unless JSON_NO_SIMD selects the portable code)
 #if !defined(JSON_NO_SIMD) && (defined(__x86_64__) || (defined(_M_X64) && !defined(_M_ARM64EC)))
@@ -44,6 +44,13 @@
 // use std::from_chars or std::strtold. They are free functions so the lexer
 // stays focused on scanning (see lexer::convert_number()) and so that other
 // parsers of JSON text can convert tokens exactly like it does.
+//
+// The Eisel-Lemire conversion (eisel_lemire()), the digit comparison for the
+// cases it cannot decide, and the eight-digits-at-a-time helpers
+// (is_eight_digits(), parse_eight_digits()) are adapted from fast_float
+// (https://github.com/fastfloat/fast_float, Daniel Lemire and contributors,
+// available under the MIT, Apache 2.0, and Boost Software licenses; used here
+// under the MIT license). The table of powers of five is in pow5_table.hpp.

 NLOHMANN_JSON_NAMESPACE_BEGIN
 namespace detail
@@ -30,6 +30,13 @@
 // NLOHMANN_JSON_NO_TARGET_ATTRIBUTE (for compilers that ignore the target
 // attribute and then reject the SSSE3 intrinsics) keeps the check portable
 // without disabling the SSE2 code.
+//
+// The vector UTF-8 check (utf8_lookup4) is adapted from simdjson
+// (https://github.com/simdjson/simdjson, Daniel Lemire, Geoff Langdale, John
+// Keiser, and contributors, available under the Apache 2.0 and MIT licenses;
+// used here under the MIT license): the "lookup4" algorithm and tables of
+// J. Keiser and D. Lemire, "Validating UTF-8 In Less Than One Instruction Per
+// Byte" (2021).
 #if !defined(JSON_NO_SIMD) && defined(__aarch64__) && (defined(__GNUC__) || defined(__clang__)) && defined(__BYTE_ORDER__) && defined(__ORDER_LITTLE_ENDIAN__) && __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
    #include <arm_neon.h>
    #define JSON_SCAN_NEON 1
@@ -1977,23 +1977,6 @@ 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'
@@ -2139,23 +2122,6 @@ 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
     *
@@ -1082,7 +1082,7 @@ 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);
@@ -26,6 +26,7 @@
 #include <nlohmann/detail/view/lookup.hpp>
 #include <nlohmann/detail/view/macro_scope.hpp>
 #include <nlohmann/detail/view/node.hpp>
+#include <nlohmann/detail/view/scan.hpp>
 #include <nlohmann/detail/view/view_mode.hpp>

 NLOHMANN_JSON_NAMESPACE_BEGIN
@@ -46,10 +47,9 @@ struct first_true<false, Conditions...> : std::integral_constant < int, 1 + firs
 /// (type_error.316 with the same message), so that an editable document
 /// only holds valid UTF-8: the error is at the first byte that no
 /// well-formed sequence can continue with (Unicode, Table 3-7).
-inline void check_utf8(const char* s, std::size_t n)
+NLOHMANN_VIEW_ALWAYS_INLINE void check_utf8_bytewise(const unsigned char* p, std::size_t n, std::size_t i)
 {
-    const auto* const p = reinterpret_cast<const unsigned char*>(s); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
-    for (std::size_t i = 0; i < n;)
+    while (i < n)
    {
        const unsigned char c = p[i];
        if (c < 0x80)
@@ -79,6 +79,43 @@ inline void check_utf8(const char* s, std::size_t n)
    }
 }

+/// The long-string path of check_utf8(): skips well-formed runs with the
+/// parser's string scan (which stops at a quote, a backslash, or a control
+/// character, all legal here, and at ill-formed UTF-8); the byte-wise check
+/// runs only from where the scan found ill-formed UTF-8, to report it.
+NLOHMANN_VIEW_NOINLINE inline void check_utf8_scan(const unsigned char* p, std::size_t n)
+{
+    const unsigned char* const e = p + n;
+    const unsigned char* q = p;
+    for (;;)
+    {
+        q = scan_string_run<true>(q, e);
+        if (q == e)
+        {
+            return;
+        }
+        if (*q >= 0x80)
+        {
+            check_utf8_bytewise(p, n, static_cast<std::size_t>(q - p));
+            return; // not reached: the scan stopped at ill-formed UTF-8
+        }
+        ++q; // a quote, backslash, or control character
+    }
+}
+
+inline void check_utf8(const char* s, std::size_t n)
+{
+    const auto* const p = reinterpret_cast<const unsigned char*>(s); // NOLINT(cppcoreguidelines-pro-type-reinterpret-cast)
+    if (NLOHMANN_VIEW_UNLIKELY(n >= 32))
+    {
+        check_utf8_scan(p, n);
+        return;
+    }
+    // short strings: the vector scan costs more than it saves; (the loop is
+    // inlined with a constant start, so this path is the plain byte-wise check)
+    check_utf8_bytewise(p, n, 0);
+}
+
 /*!
@brief the edits of an editable basic_json_document

@@ -258,6 +258,81 @@ struct number_token
    bool is_float; ///< a fraction or an exponent
 };

+/// advance over ASCII digits, eight bytes at a time
+inline const unsigned char* digit_run(const unsigned char* p, const unsigned char* e) noexcept
+{
+    constexpr std::uint64_t ones = 0x0101010101010101ull;
+    while (e - p >= 8)
+    {
+        // (the digits become 0..9; a byte above 9, or with bit 7 set, is no digit. A carry
+        // out of a byte with bit 7 set only affects the bytes above, after the first one
+        // that count_trailing_zeros finds.)
+        const std::uint64_t x = read_eight_bytes(p) ^ (ones * 0x30u);
+        const std::uint64_t not_digit = ((x + (ones * 0x76u)) | x) & (ones * 0x80u);
+        if (not_digit != 0)
+        {
+            return p + (static_cast<std::size_t>(count_trailing_zeros(not_digit)) / 8);
+        }
+        p += 8;
+    }
+    while (p != e && is_digit(*p))
+    {
+        ++p;
+    }
+    return p;
+}
+
+/// whether the n bytes at p are ASCII digits (n > 0)
+inline bool all_digits(const unsigned char* p, std::size_t n) noexcept
+{
+    const auto digits = [](const unsigned char* q)
+    {
+        constexpr std::uint64_t ones = 0x0101010101010101ull;
+        const std::uint64_t x = read_eight_bytes(q) ^ (ones * 0x30u);
+        return (((x + (ones * 0x76u)) | x) & (ones * 0x80u)) == 0;
+    };
+    if (n >= 8)
+    {
+        for (std::size_t i = 0; i + 8 < n; i += 8)
+        {
+            if (!digits(p + i))
+            {
+                return false;
+            }
+        }
+        return digits(p + n - 8); // (overlapping the bytes before: they are digits)
+    }
+    for (std::size_t i = 0; i < n; ++i)
+    {
+        if (!is_digit(p[i]))
+        {
+            return false;
+        }
+    }
+    return true;
+}
+
+/// Whether a float token has the shape that the parser records for most of
+/// them: [-]digits.digits, with the numbers of digits of the node's layout
+/// (extra). Then the token is a number of that layout, and the general scan
+/// that follows is not needed.
+inline bool plain_float_token(const unsigned char* s, std::size_t len, std::uint16_t extra) noexcept
+{
+    const std::size_t int_digits = extra & 0xFFu;
+    const std::size_t frac_digits = extra >> 8u;
+    if (int_digits == 0 || int_digits >= 255 || frac_digits == 0 || frac_digits >= 255 || len == 0)
+    {
+        return false;
+    }
+    const std::size_t negative = *s == '-' ? 1 : 0;
+    if (len != negative + int_digits + 1 + frac_digits)
+    {
+        return false;
+    }
+    const unsigned char* const p = s + negative;
+    return !(int_digits > 1 && *p == '0') && p[int_digits] == '.' && all_digits(p, int_digits) && all_digits(p + int_digits + 1, frac_digits);
+}
+
 /// whether [s, s + len) is a JSON number (the grammar the parser accepts)
 inline bool scan_number_token(const unsigned char* s, std::size_t len, number_token& t)
 {
@@ -276,11 +351,7 @@ inline bool scan_number_token(const unsigned char* s, std::size_t len, number_to
    }
    else if (*p >= '1' && *p <= '9')
    {
-        do
-        {
-            ++p;
-        }
-        while (p != e && is_digit(*p));
+        p = digit_run(p + 1, e);
    }
    else
    {
@@ -292,10 +363,7 @@ inline bool scan_number_token(const unsigned char* s, std::size_t len, number_to
    if (p != e && *p == '.')
    {
        const unsigned char* const f0 = ++p;
-        while (p != e && is_digit(*p))
-        {
-            ++p;
-        }
+        p = digit_run(p, e);
        if (p == f0)
        {
            return false;
@@ -406,10 +474,15 @@ struct float_range
 template<typename Check>
 bool check_distinct_ranges(std::vector<byte_range>& ranges, Check check)
 {
-    std::sort(ranges.begin(), ranges.end(), [](const byte_range & a, const byte_range & b)
+    // (save() writes the ranges in the order of the nodes, which is sorted already)
+    const auto before = [](const byte_range & a, const byte_range & b)
    {
        return a.off != b.off ? a.off < b.off : a.len < b.len;
-    });
+    };
+    if (!std::is_sorted(ranges.begin(), ranges.end(), before))
+    {
+        std::sort(ranges.begin(), ranges.end(), before);
+    }
    std::size_t end = 0;
    for (std::size_t i = 0; i < ranges.size(); ++i)
    {
@@ -431,7 +504,8 @@ bool check_distinct_ranges(std::vector<byte_range>& ranges, Check check)
 /// record layouts that match it), tokens must not overlap.
 inline bool check_float_ranges(std::vector<float_range>& ranges, const unsigned char* text)
 {
-    std::sort(ranges.begin(), ranges.end(), [](const float_range & a, const float_range & b)
+    // (save() writes the tokens in the order of the nodes, which is sorted already)
+    const auto before = [](const float_range & a, const float_range & b)
    {
        if (a.off != b.off)
        {
@@ -442,7 +516,11 @@ inline bool check_float_ranges(std::vector<float_range>& ranges, const unsigned
            return a.len < b.len;
        }
        return a.extra < b.extra;
-    });
+    };
+    if (!std::is_sorted(ranges.begin(), ranges.end(), before))
+    {
+        std::sort(ranges.begin(), ranges.end(), before);
+    }
    std::size_t end = 0;
    std::size_t i = 0;
    while (i < ranges.size())
@@ -452,14 +530,18 @@ inline bool check_float_ranges(std::vector<float_range>& ranges, const unsigned
        {
            return false;
        }
-        number_token t{};
        const unsigned char* const s = text + r.off;
-        if (!scan_number_token(s, r.len, t) || !float_token_finite(s, r.len, t))
+        std::uint16_t layout = r.extra;
+        if (!plain_float_token(s, r.len, r.extra))
        {
-            return false;
+            number_token t{};
+            if (!scan_number_token(s, r.len, t) || !float_token_finite(s, r.len, t))
+            {
+                return false;
+            }
+            // the digit layout the parser records (or "many", as compaction writes it)
+            layout = float_layout(t);
        }
-        // the digit layout the parser records (or "many", as compaction writes it)
-        const std::uint16_t layout = float_layout(t);
        for (; i < ranges.size() && ranges[i].off == r.off && ranges[i].len == r.len; ++i)
        {
            if (ranges[i].extra != layout && ranges[i].extra != 0xFFFFu)
@@ -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
 {
@@ -29,6 +29,13 @@
 // NLOHMANN_VIEW_NO_TARGET_ATTRIBUTE (set by the GCC module interface, where GCC
 // ignores the target attribute and then rejects the SSSE3 intrinsics) keeps the
 // check portable without disabling the SSE2 code.
+//
+// The vector UTF-8 check (utf8_lookup4) is adapted from simdjson
+// (https://github.com/simdjson/simdjson, Daniel Lemire, Geoff Langdale, John
+// Keiser, and contributors, available under the Apache 2.0 and MIT licenses;
+// used here under the MIT license): the "lookup4" algorithm and tables of
+// J. Keiser and D. Lemire, "Validating UTF-8 In Less Than One Instruction Per
+// Byte" (2021).
 #if !defined(JSON_NO_SIMD) && defined(__aarch64__) && (defined(__GNUC__) || defined(__clang__)) && NLOHMANN_VIEW_LITTLE_ENDIAN
    #include <arm_neon.h>
    #define NLOHMANN_VIEW_NEON 1
@@ -763,6 +763,7 @@ class basic_json_view
    template<typename, view_mode> friend class basic_json_document;
    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 iterator;

    basic_json_view(const document_data* d, const node* n) noexcept
@@ -10970,6 +10970,13 @@ NLOHMANN_JSON_NAMESPACE_END
 // use std::from_chars or std::strtold. They are free functions so the lexer
 // stays focused on scanning (see lexer::convert_number()) and so that other
 // parsers of JSON text can convert tokens exactly like it does.
+//
+// The Eisel-Lemire conversion (eisel_lemire()), the digit comparison for the
+// cases it cannot decide, and the eight-digits-at-a-time helpers
+// (is_eight_digits(), parse_eight_digits()) are adapted from fast_float
+// (https://github.com/fastfloat/fast_float, Daniel Lemire and contributors,
+// available under the MIT, Apache 2.0, and Boost Software licenses; used here
+// under the MIT license). The table of powers of five is in pow5_table.hpp.

 NLOHMANN_JSON_NAMESPACE_BEGIN
 namespace detail
@@ -12078,6 +12085,13 @@ NLOHMANN_JSON_NAMESPACE_END
 // NLOHMANN_JSON_NO_TARGET_ATTRIBUTE (for compilers that ignore the target
 // attribute and then reject the SSSE3 intrinsics) keeps the check portable
 // without disabling the SSE2 code.
+//
+// The vector UTF-8 check (utf8_lookup4) is adapted from simdjson
+// (https://github.com/simdjson/simdjson, Daniel Lemire, Geoff Langdale, John
+// Keiser, and contributors, available under the Apache 2.0 and MIT licenses;
+// used here under the MIT license): the "lookup4" algorithm and tables of
+// J. Keiser and D. Lemire, "Validating UTF-8 In Less Than One Instruction Per
+// Byte" (2021).
 #if !defined(JSON_NO_SIMD) && defined(__aarch64__) && (defined(__GNUC__) || defined(__clang__)) && defined(__BYTE_ORDER__) && defined(__ORDER_LITTLE_ENDIAN__) && __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
    #include <arm_neon.h>
    #define JSON_SCAN_NEON 1
@@ -32495,6 +32509,11 @@ NLOHMANN_JSON_NAMESPACE_END
 #include <type_traits> // conditional


+// The shortest-digit conversion of doubles (see zmij.hpp) and the writing of
+// its digits, including the conversion of the digits to text by Xiang JunBo
+// and the SIMD instruction sequence by Dougall Johnson, are ported from Zmij
+// by Victor Zverovich (https://github.com/vitaut/zmij, MIT license).
+//
 // SSE2 (every x86-64 CPU) and NEON (every 64-bit Arm CPU) convert the 16
 // digits of a double at once (unless JSON_NO_SIMD selects the portable code)
 #if !defined(JSON_NO_SIMD) && (defined(__x86_64__) || (defined(_M_X64) && !defined(_M_ARM64EC)))
@@ -36205,23 +36224,6 @@ 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'
@@ -36367,23 +36369,6 @@ 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
     *
@@ -21,6 +21,10 @@ python3 tests/benchmarks/json_view/compare.py --data <json_test_data directory>
 - The other libraries come from the system: pkg-config, or Homebrew (`brew install yyjson simdjson boost`). With
  `--download`, pinned releases are downloaded instead and checked against their SHA-256. Without Boost headers (or
  with `--no-boost`), the Boost.JSON columns are skipped, and the results say so.
+- `--aligned` additionally builds and runs everything with all functions aligned to 64 bytes (`-mllvm
+  -align-all-functions=6` for Clang, `-falign-functions=64 -falign-loops=64` for GCC), and reports both result
+  tables as "default build" and "aligned build". Code layout alone can move a result by a few percent; a difference
+  that appears in only one of the two tables is a layout artifact. Off by default, which doubles the run time.
 - `--corpus file...` adds files to the corpus benchmark, e.g. those of
  [simdjson-data](https://github.com/simdjson/simdjson-data) or the
  [yyjson benchmark](https://github.com/ibireme/yyjson_benchmark).
@@ -39,7 +43,7 @@ where `json_view` stands on another architecture; they are not meant for publica

 ## What is measured

-`bench_view.cpp` runs four workloads on twitter, citm_catalog, canada, jeopardy, a single tweet (`status`), and a
+`bench_view.cpp` runs these workloads on twitter, citm_catalog, canada, jeopardy, a single tweet (`status`), and a
 JSON-RPC request (`rpc`):

 | workload | what it does |
@@ -48,6 +52,7 @@ JSON-RPC request (`rpc`):
 | traverse | parse, then visit every value, convert every number, touch every string and key |
 | select | parse, then read a few fields per record (e.g. id, user name, and retweet count of each tweet) |
 | dump | serialize a parsed document (compact) |
+| parse+dump | build a fresh document, then serialize it (compact), in one timed call |

 `bench_corpus.cpp` runs parse, traverse, and dump on any list of files, so that no library is tuned to a handful of
 documents. Its dump also writes the numbers as they are in the input: `json_view` with `number_format::source`, and
@@ -72,11 +77,32 @@ call follows an untimed call of the same engine: otherwise the engine after `jso
 cleaning up the tens of thousands of nodes `json::parse` just freed (with glibc, this made `json_view` look 1.7 times
 slower on citm_catalog traverse).

+`json_view` is an index into the source text, not a tree of owning nodes. Its like-for-like peers are yyjson and
+simdjson DOM (immutable, indexed, no per-node allocation). Boost.JSON and `json::parse` build mutable, owning DOMs and
+are reference points for what that costs, not equals; simdjson On-Demand is lazy and forward-only, so it is the other
+way round: it has no random access to pay for. All engines validate UTF-8 and syntax by default.
+
+Number conversion differs: `json_view` converts integers while parsing but converts floating-point numbers lazily,
+when they are read or dumped; yyjson (default flags) and simdjson DOM convert every number while parsing. The cost of
+float conversion therefore appears in the traverse and dump rows for `json_view`, and in the parse row for the others.
+For float-heavy documents (canada), the fair reading is a sum of parse and use: the `traverse` workload already
+includes the parse, and `parse+dump` measures a fresh document and its dump in one timed call; the `dump` row alone
+leaves the conversion in `json_view` and out of the others. Eager conversion at parse time was prototyped as an opt-in
+and measured: canada traverse (with its parse) was 4-7 % faster, but parse+dump 11-30 % slower and parse 36-90 % slower. It was
+dropped, so lazy conversion is the only mode.
+
+`json_view` can also be edited: `json_editable_document` has the same read API, and the `(editable)` rows in
+`bench_view` measure it on the same workloads. They are the fair counterpart of Boost.JSON and `json::parse`, which are
+mutable too; the plain `json_view` rows are the counterpart of yyjson and simdjson DOM, which are not. `bench_edit` adds
+the read-modify-write workloads. Read the tables with that in mind: an immutable-only library is not compared with an
+editable `json_view` unless it says so.
+
 The engines do not all offer the same features, which the numbers should be read with:

 | engine | document | random access | editable | notes |
 |---|---|---|---|---|
 | `json_view` | immutable index into the text | yes | no | a fresh document per parse; "reused" parses into the same document |
+| `json_view` (editable) | index into the text plus edits | yes | yes, in place | `json_editable_document`; keeps member order and number spellings |
 | yyjson | immutable (`yyjson_read`) | yes | via a mutable copy | |
 | simdjson DOM | immutable, parser reused | yes | no | "fresh" uses a new parser per parse |
 | simdjson On-Demand | none: forward-only, lazy | no | no | only traverse and select |
@@ -8,8 +8,9 @@

 // Same-feature-set benchmark: read-only JSON documents with random access.
 //
-//   json_view      nlohmann/json_view.hpp (fresh document per parse / reused)
-//   yyjson         yyjson_read(): immutable document, random access
+//   json_view      nlohmann/json_view.hpp (fresh document per parse / reused;
+//                  "(editable)": json_editable_document, same read API)
+//   yyjson        yyjson_read(): immutable document, random access
 //   simdjson DOM   dom::parser (reused, as recommended; "fresh": a new parser
 //                  per parse): immutable, random access
 // references (different feature sets):
@@ -19,7 +20,11 @@
 //
 // Workloads: parse (build + free), traverse (visit everything, convert every
 // number, touch every string and key), select (a few fields per document),
-// dump (compact serialization of the parsed document).
+// dump (compact serialization of the parsed document), and parse+dump (a fresh
+// document, then its dump, in one timed call). json_view converts floating-point
+// numbers lazily (when read or dumped), yyjson and simdjson DOM while parsing:
+// for float-heavy documents, traverse (which includes the parse) and parse+dump
+// are the fair reading of the parse and dump rows.
 // All engines run interleaved in every round, each timed call after an untimed
 // one of the same engine; the best round is reported.
 #include <nlohmann/json_view.hpp>
@@ -43,6 +48,8 @@

 using nlohmann::json;
 using nlohmann::json_document;
+using nlohmann::json_editable_document;
+using nlohmann::json_editable_view;
 using nlohmann::json_view;

 static volatile double g_sink;
@@ -55,7 +62,8 @@ struct stats

 // ---------------- traversal ----------------

-static void walk(json_view v, stats& st)
+template<class V>
+static void walk_view(V v, stats& st)
 {
    ++st.nodes;
    switch (v.type())
@@ -64,32 +72,42 @@ static void walk(json_view v, stats& st)
            for (auto it = v.begin(); it != v.end(); ++it)
            {
                st.str += it.key().size();
-                walk(*it, st);
+                walk_view<V>(*it, st);
            }
            break;
        case json::value_t::array:
-            for (const json_view e : v)
+            for (const V e : v)
            {
-                walk(e, st);
+                walk_view<V>(e, st);
            }
            break;
        case json::value_t::string:
            st.str += v.get_string().size();
            break;
        case json::value_t::number_integer:
-            st.num += static_cast<double>(v.get<std::int64_t>());
+            st.num += static_cast<double>(v.template get<std::int64_t>());
            break;
        case json::value_t::number_unsigned:
-            st.num += static_cast<double>(v.get<std::uint64_t>());
+            st.num += static_cast<double>(v.template get<std::uint64_t>());
            break;
        case json::value_t::number_float:
-            st.num += v.get<double>();
+            st.num += v.template get<double>();
            break;
        default:
            break;
    }
 }

+static void walk(json_view v, stats& st)
+{
+    walk_view<json_view>(v, st);
+}
+
+static void walk(json_editable_view v, stats& st)
+{
+    walk_view<json_editable_view>(v, st);
+}
+
 static void walk(const json& j, stats& st)
 {
    ++st.nodes;
@@ -300,20 +318,21 @@ static void walk(const boost::json::value& v, stats& st)
 // status (one tweet): id, user.screen_name, retweet_count
 // rpc: method, params.minuend, id

-static double pick(const std::string& name, json_view r)
+template<class V>
+static double pick_view(const std::string& name, V r)
 {
    double acc = 0;
    if (name == "twitter" || name == "status")
    {
-        auto one = [&](json_view s)
+        auto one = [&](V s)
        {
-            acc += static_cast<double>(s["id"].get<std::uint64_t>());
+            acc += static_cast<double>(s["id"].template get<std::uint64_t>());
            acc += static_cast<double>(s["user"]["screen_name"].get_string().size());
-            acc += static_cast<double>(s["retweet_count"].get<std::int64_t>());
+            acc += static_cast<double>(s["retweet_count"].template get<std::int64_t>());
        };
        if (name == "twitter")
        {
-            for (const json_view s : r["statuses"])
+            for (const V s : r["statuses"])
            {
                one(s);
            }
@@ -325,20 +344,20 @@ static double pick(const std::string& name, json_view r)
    }
    else if (name == "citm_catalog")
    {
-        for (const json_view p : r["performances"])
+        for (const V p : r["performances"])
        {
-            acc += static_cast<double>(p["id"].get<std::uint64_t>() + p["eventId"].get<std::uint64_t>() + p["seatCategories"].size());
+            acc += static_cast<double>(p["id"].template get<std::uint64_t>() + p["eventId"].template get<std::uint64_t>() + p["seatCategories"].size());
        }
        acc += static_cast<double>(r["events"].size());
    }
    else if (name == "canada")
    {
-        const json_view g = r["features"][0]["geometry"];
+        const V g = r["features"][0]["geometry"];
        acc += static_cast<double>(r["type"].get_string().size() + g["type"].get_string().size() + g["coordinates"].size());
    }
    else if (name == "jeopardy")
    {
-        for (const json_view q : r)
+        for (const V q : r)
        {
            acc += q["round"].get_string() == "Final Jeopardy!" ? 1 : 0;
            acc += static_cast<double>(q["category"].get_string().size());
@@ -347,11 +366,21 @@ static double pick(const std::string& name, json_view r)
    else if (name == "rpc")
    {
        acc += static_cast<double>(r["method"].get_string().size());
-        acc += static_cast<double>(r["params"]["minuend"].get<std::int64_t>() + r["id"].get<std::int64_t>());
+        acc += static_cast<double>(r["params"]["minuend"].template get<std::int64_t>() + r["id"].template get<std::int64_t>());
    }
    return acc;
 }

+static double pick(const std::string& name, json_view r)
+{
+    return pick_view<json_view>(name, r);
+}
+
+static double pick(const std::string& name, json_editable_view r)
+{
+    return pick_view<json_editable_view>(name, r);
+}
+
 static double pick(const std::string& name, const json& r)
 {
    double acc = 0;
@@ -634,15 +663,35 @@ int main(int argc, char** argv)
        walk(json::parse(dc.text), a);
        auto d = json_document::parse(dc.text);
        walk(d.root(), b);
+        auto ed = json_editable_document::parse(dc.text);
+        stats e;
+        walk(ed.root(), e);
        yyjson_doc* y = yyjson_read(dc.text.data(), dc.text.size(), 0);
        walk(yyjson_doc_get_root(y), c);
        simdjson::dom::parser p;
        walk(p.parse(dc.text).value(), dd);
-        const bool ok = a.nodes == b.nodes && a.nodes == c.nodes && a.nodes == dd.nodes && a.str == b.str && a.str == c.str && a.str == dd.str
+        const bool ok = a.nodes == b.nodes && a.nodes == e.nodes && a.str == e.str && std::fabs(a.num - e.num) <= 1e-9 * std::fabs(a.num) && a.nodes == c.nodes && a.nodes == dd.nodes && a.str == b.str && a.str == c.str && a.str == dd.str
                        && std::fabs(a.num - b.num) <= 1e-9 * std::fabs(a.num) && std::fabs(a.num - c.num) <= 1e-9 * std::fabs(a.num);
-        const double pa = pick(dc.name, json::parse(dc.text)), pb = pick(dc.name, d.root()), pc = pick(dc.name, yyjson_doc_get_root(y)), pd = pick(dc.name, p.parse(dc.text).value());
-        std::printf("check %-13s traverse %s select %s\n", dc.name.c_str(), ok ? "OK" : "MISMATCH", (pa == pb && pa == pc && pa == pd) ? "OK" : "MISMATCH");
+        const double pa = pick(dc.name, json::parse(dc.text)), pb = pick(dc.name, d.root()), pe = pick(dc.name, ed.root()), pc = pick(dc.name, yyjson_doc_get_root(y)), pd = pick(dc.name, p.parse(dc.text).value());
+        std::printf("check %-13s traverse %s select %s\n", dc.name.c_str(), ok ? "OK" : "MISMATCH", (pa == pb && pa == pe && pa == pc && pa == pd) ? "OK" : "MISMATCH");
        yyjson_doc_free(y);
+
+        // the dumps of the engines may format floats differently, so only: every dump is non-empty, and the
+        // dump of json_view (also editable) and of json::parse parse back to the same number of nodes
+        const std::string dj = json::parse(dc.text).dump();
+        const std::string dv = d.root().dump();
+        const std::string de = ed.root().dump();
+        std::size_t yn = 0;
+        yyjson_doc* y2 = yyjson_read(dc.text.data(), dc.text.size(), 0);
+        char* yo = yyjson_write(y2, 0, &yn);
+        const bool dumps_ok = !dj.empty() && !dv.empty() && !de.empty() && yo != nullptr && yn > 0 && !simdjson::to_string(p.parse(dc.text).value()).empty();
+        std::free(yo);
+        yyjson_doc_free(y2);
+        stats sj_nodes, sv_nodes, se_nodes;
+        walk(json::parse(dj), sj_nodes);
+        walk(json::parse(dv), sv_nodes);
+        walk(json::parse(de), se_nodes);
+        std::printf("check %-13s dump %s\n", dc.name.c_str(), (dumps_ok && sj_nodes.nodes == a.nodes && sv_nodes.nodes == a.nodes && se_nodes.nodes == a.nodes) ? "OK" : "MISMATCH");
    }

    std::FILE* csv = std::fopen("bench_view.csv", "w");
@@ -669,6 +718,7 @@ int main(int argc, char** argv)
        workloads.push_back({"parse", {
                {"json_view", [&] { auto d = json_document::parse(s); g_sink = static_cast<double>(d.node_count()); }},
                {"json_view (reused)", [&] { reused.read(s); g_sink = static_cast<double>(reused.node_count()); }},
+                {"json_view (editable)", [&] { auto d = json_editable_document::parse(s); g_sink = static_cast<double>(d.root().size()); }},
                {"yyjson", [&] { yyjson_doc* d = yyjson_read(s.data(), s.size(), 0); g_sink = static_cast<double>(yyjson_doc_get_val_count(d)); yyjson_doc_free(d); }},
                {"simdjson DOM", [&] { auto e = sj.parse(ps).value_unsafe(); g_sink = e.is_object(); }},
                {"simdjson DOM (fresh)", [&] { simdjson::dom::parser p; auto e = p.parse(ps).value_unsafe(); g_sink = e.is_object(); }},
@@ -680,6 +730,7 @@ int main(int argc, char** argv)
        workloads.push_back({"traverse", {
                {"json_view", [&] { auto d = json_document::parse(s); stats st; walk(d.root(), st); g_sink = st.num; }},
                {"json_view (reused)", [&] { reused.read(s); stats st; walk(reused.root(), st); g_sink = st.num; }},
+                {"json_view (editable)", [&] { auto d = json_editable_document::parse(s); stats st; walk(d.root(), st); g_sink = st.num; }},
                {"yyjson", [&] { yyjson_doc* d = yyjson_read(s.data(), s.size(), 0); stats st; walk(yyjson_doc_get_root(d), st); g_sink = st.num; yyjson_doc_free(d); }},
                {"simdjson DOM", [&] { stats st; walk(sj.parse(ps).value_unsafe(), st); g_sink = st.num; }},
                {"simdjson OD", [&] { auto d = od.iterate(ps).value_unsafe(); stats st; walk_od(d.get_value().value_unsafe(), st); g_sink = st.num; }},
@@ -691,15 +742,31 @@ int main(int argc, char** argv)
        workloads.push_back({"select", {
                {"json_view", [&] { auto d = json_document::parse(s); g_sink = pick(name, d.root()); }},
                {"json_view (reused)", [&] { reused.read(s); g_sink = pick(name, reused.root()); }},
+                {"json_view (editable)", [&] { auto d = json_editable_document::parse(s); g_sink = pick(name, d.root()); }},
                {"yyjson", [&] { yyjson_doc* d = yyjson_read(s.data(), s.size(), 0); g_sink = pick(name, yyjson_doc_get_root(d)); yyjson_doc_free(d); }},
                {"simdjson DOM", [&] { g_sink = pick(name, sj.parse(ps).value_unsafe()); }},
                {"simdjson OD", [&] { auto d = od.iterate(ps).value_unsafe(); g_sink = pick_od(name, d); }},
                {"json::parse", [&] { json j = json::parse(s); g_sink = pick(name, j); }},
            }});
+        // pipeline: a fresh document, then a dump of it, in one timed call
+        // (simdjson DOM reuses its parser, as in the parse row); the traverse
+        // workload above is already parse+traverse
+        workloads.push_back({"parse+dump", {
+                {"json_view", [&] { auto d = json_document::parse(s); std::string o = d.root().dump(); g_sink = static_cast<double>(o.size()); }},
+                {"json_view (editable)", [&] { auto d = json_editable_document::parse(s); std::string o = d.root().dump(); g_sink = static_cast<double>(o.size()); }},
+                {"yyjson", [&] { yyjson_doc* d = yyjson_read(s.data(), s.size(), 0); std::size_t n = 0; char* o = yyjson_write(d, 0, &n); g_sink = static_cast<double>(n); std::free(o); yyjson_doc_free(d); }},
+                {"simdjson DOM", [&] { std::string o = simdjson::to_string(sj.parse(ps).value_unsafe()); g_sink = static_cast<double>(o.size()); }},
+#if JSON_VIEW_BENCH_BOOST
+                {"Boost.JSON", [&] { boost::json::monotonic_resource mr; auto v = boost::json::parse(s, &mr); std::string o = boost::json::serialize(v); g_sink = static_cast<double>(o.size()); }},
+#endif
+                {"json::parse", [&] { json j = json::parse(s); std::string o = j.dump(); g_sink = static_cast<double>(o.size()); }},
+            }});
        {
            // serialization of an already parsed document
            static json_document vd;
            vd.read(s);
+            static json_editable_document ed = json_editable_document::parse(s);
+            ed = json_editable_document::parse(s);
            static yyjson_doc* yd = nullptr;
            if (yd)
            {
@@ -713,6 +780,7 @@ int main(int argc, char** argv)
            jd = json::parse(s);
            workloads.push_back({"dump", {
                    {"json_view", [&] { std::string o = vd.root().dump(); g_sink = static_cast<double>(o.size()); }},
+                    {"json_view (editable)", [&] { std::string o = ed.root().dump(); g_sink = static_cast<double>(o.size()); }},
                    {"yyjson", [&] { std::size_t n = 0; char* o = yyjson_write(yd, 0, &n); g_sink = static_cast<double>(n); std::free(o); }},
                    {"simdjson DOM", [&] { std::string o = simdjson::to_string(se); g_sink = static_cast<double>(o.size()); }},
                    {"json::parse", [&] { std::string o = jd.dump(); g_sink = static_cast<double>(o.size()); }},
@@ -740,7 +808,7 @@ int main(int argc, char** argv)
                }
            }
            const double ref = best[0];
-            std::printf("%-13s %-9s", dc.name.c_str(), wl.first.c_str());
+            std::printf("%-13s %-14s", dc.name.c_str(), wl.first.c_str());
            for (std::size_t k = 0; k < wl.second.size(); ++k)
            {
                const double us = best[k] / 1e3;
@@ -239,6 +239,8 @@ def main():
    ap.add_argument('--download', action='store_true', help='use pinned downloads instead of system libraries')
    ap.add_argument('--no-boost', action='store_true', help='skip Boost.JSON')
    ap.add_argument('--native', action='store_true', help='compile for this CPU (-march=native / -mcpu=native)')
+    ap.add_argument('--aligned', action='store_true',
+                    help='also build and run with all functions (and loops) aligned, to make code layout artifacts visible')
    ap.add_argument('--rounds', type=int, default=30, help='rounds of bench_view (default: 30)')
    ap.add_argument('--corpus', nargs='*', default=[], help='more files for bench_corpus')
    ap.add_argument('--build-dir', default=os.path.join(HERE, 'build'), help='where to build (default: build/ next to this script)')
@@ -282,32 +284,46 @@ def main():
                + ['-I' + d for d in lib.include])
            objects.append(obj)

-    binaries = {}
-    for bench in ['bench_view', 'bench_corpus', 'bench_edit', 'bench_decode']:
-        exe = os.path.join(args.build_dir, bench)
-        run(cxx + flags + include + [os.path.join(HERE, bench + '.cpp')] + objects + link + ['-o', exe])
-        binaries[bench] = exe
+    cxx_version = output(cxx + ['--version'])
+    is_clang = 'clang' in cxx_version.lower()
+    # alignment flags: functions and (gcc) loops on 64-byte boundaries
+    align_flags = ['-mllvm', '-align-all-functions=6'] if is_clang else ['-falign-functions=64', '-falign-loops=64']
+
+    # the builds: the default one and, with --aligned, one with aligned code
+    variants = [('default build', args.build_dir, [])]
+    if args.aligned:
+        variants.append(('aligned build', os.path.join(args.build_dir, 'aligned'), align_flags))

-    # run: bench_view on its documents, bench_corpus on those and the given files
    corpus = [os.path.join(args.data, f) for f in DEFAULT_CORPUS] + args.corpus
-    outputs = {}
-    outputs['bench_view'] = run([binaries['bench_view'], args.data, str(args.rounds)], cwd=args.build_dir,
-                                capture_output=True, text=True).stdout
-    outputs['bench_corpus'] = run([binaries['bench_corpus']] + corpus, cwd=args.build_dir,
-                                  capture_output=True, text=True).stdout
-    outputs['bench_edit'] = run([binaries['bench_edit'], args.data, str(max(1, args.rounds // 2))], cwd=args.build_dir,
-                                capture_output=True, text=True).stdout
-    outputs['bench_decode'] = run([binaries['bench_decode'], args.data, str(max(1, args.rounds // 2))], cwd=args.build_dir,
-                                  capture_output=True, text=True).stdout
-    for name, text in outputs.items():
-        print(text)
+    # label -> {benchmark: output}
+    all_outputs = {}
+    for label, build_dir, extra in variants:
+        os.makedirs(build_dir, exist_ok=True)
+        binaries = {}
+        for bench in ['bench_view', 'bench_corpus', 'bench_edit', 'bench_decode']:
+            exe = os.path.join(build_dir, bench)
+            run(cxx + flags + extra + include + [os.path.join(HERE, bench + '.cpp')] + objects + link + ['-o', exe])
+            binaries[bench] = exe
+
+        # run: bench_view on its documents, bench_corpus on those and the given files
+        outputs = {}
+        outputs['bench_view'] = run([binaries['bench_view'], args.data, str(args.rounds)], cwd=build_dir,
+                                    capture_output=True, text=True).stdout
+        outputs['bench_corpus'] = run([binaries['bench_corpus']] + corpus, cwd=build_dir,
+                                      capture_output=True, text=True).stdout
+        outputs['bench_edit'] = run([binaries['bench_edit'], args.data, str(max(1, args.rounds // 2))], cwd=build_dir,
+                                    capture_output=True, text=True).stdout
+        outputs['bench_decode'] = run([binaries['bench_decode'], args.data, str(max(1, args.rounds // 2))],
+                                      cwd=build_dir, capture_output=True, text=True).stdout
+        for name, text in outputs.items():
+            print(text)
+        all_outputs[label] = outputs

    # results with their metadata
    now = datetime.datetime.now()
    host = re.sub(r'[^A-Za-z0-9-]+', '-', platform.node().split('.')[0]) or 'host'
    stem = os.path.join(HERE, 'results', f'{now:%Y-%m-%d}-{host}')
    os.makedirs(os.path.dirname(stem), exist_ok=True)
-    cxx_version = output(cxx + ['--version'])
    compiler_version = cxx_version.splitlines()[0] if cxx_version else ' '.join(cxx)
    meta = [
        ('date', f'{now:%Y-%m-%d %H:%M}'),
@@ -316,6 +332,7 @@ def main():
        ('OS', f'{platform.system()} {platform.release()} ({platform.machine()})'),
        ('compiler', compiler_version),
        ('flags', ' '.join(flags)),
+        ('aligned results', ('included (flags: ' + ' '.join(align_flags) + ')') if args.aligned else 'not included'),
        ('yyjson', libs['yyjson'].version),
        ('simdjson', libs['simdjson'].version),
        ('Boost.JSON', libs['boost'].version if with_boost else 'skipped (not found)'),
@@ -328,15 +345,20 @@ def main():
        f.write('| | |\n|---|---|\n')
        for key, value in meta:
            f.write(f'| {key} | {value} |\n')
-        for name, text in outputs.items():
-            f.write(f'\n## {name}\n\n```\n{text.rstrip()}\n```\n')
+        for label, outputs in all_outputs.items():
+            # (without --aligned there is only the default build, and no extra heading level)
+            prefix = f'{label}: ' if args.aligned else ''
+            for name, text in outputs.items():
+                f.write(f'\n## {prefix}{name}\n\n```\n{text.rstrip()}\n```\n')
    with open(stem + '.csv', 'w', encoding='utf-8') as out:
        out.write(''.join(f'# {key}: {value}\n' for key, value in meta))
-        for name in ['bench_view', 'bench_corpus', 'bench_edit', 'bench_decode']:
-            path = os.path.join(args.build_dir, name + '.csv')
-            if os.path.isfile(path):
-                with open(path, encoding='utf-8') as f:
-                    out.write(f'# {name}\n' + f.read())
+        for label, build_dir, _ in variants:
+            for name in ['bench_view', 'bench_corpus', 'bench_edit', 'bench_decode']:
+                path = os.path.join(build_dir, name + '.csv')
+                if os.path.isfile(path):
+                    with open(path, encoding='utf-8') as f:
+                        # (the default build keeps its plain section names)
+                        out.write(f'# {name}' + (f' ({label})' if args.aligned else '') + '\n' + f.read())
    print(f'results: {stem}.md, {stem}.csv')


@@ -523,6 +523,47 @@ 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 + "]");
+            check_same("[" + number + "," + number + "]");
+            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
@@ -0,0 +1,364 @@
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  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;
+
+#include "json_view_test_helpers.hpp"
+using json_view_test::generator;
+
+#include <array>
+#include <cstddef>
+#include <string>
+#include <vector>
+
+// These tests were split off unit-json_view_dump.cpp, whose object file got too
+// large for the MinGW assembler ("file too big"). 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("objects whose first keys differ or repeat")
+    {
+        const std::array<const char*, 17> texts =
+        {
+            {
+                R"({"a":1,"b":2})", R"({"a":1,"c":2})", R"({"b":1,"a":2})", R"({"a":1,"a":2,"b":3})", R"({"a":2,"b":3})",
+                R"({"a":1,"b":1,"a":2})", R"({"a":2,"b":1})", R"({"a":1,"a":2})", R"({"b":1})", R"({"a":1,"b":2,"c":3,"d":4,"e":5})",
+                R"({"a":1,"b":2,"c":3,"d":4,"f":5})", R"({"a":1,"b":2,"c":3,"a":4,"e":5})", R"({"a":4,"b":2,"c":3,"e":5})",
+                R"({"\u0061":1,"b":2})", R"({})", R"([{"a":1,"b":2}])", R"([{"a":1,"c":2}])"
+            }
+        };
+        for (const char* const x : texts)
+        {
+            for (const char* const y : texts)
+            {
+                CAPTURE(x)
+                CAPTURE(y)
+                const ordered_json_document dx = ordered_json_document::parse(x);
+                const ordered_json_document dy = ordered_json_document::parse(y);
+                const ordered_json_editable_document ex = ordered_json_editable_document::parse(x);
+                const ordered_json jx = ordered_json::parse(x);
+                const ordered_json jy = ordered_json::parse(y);
+                CHECK((dx.root() == dy.root()) == (jx == jy));
+                CHECK((dx.root() == jy) == (jx == jy));
+                CHECK((ex.root() == dy.root()) == (jx == jy));
+            }
+        }
+    }
+
+    SECTION("objects with duplicate keys, small and large")
+    {
+        // n members with keys in descending order; with duplicates, some keys
+        // repeat later (the first after unique ones in key order); in reverse,
+        // the members are in ascending key order
+        const auto object = [](int n, bool duplicates, bool reverse, int value)
+        {
+            std::string s = "{";
+            for (int i = 0; i < n; ++i)
+            {
+                const int k = reverse ? i : n - 1 - i;
+                s += (i != 0 ? ",\"k" : "\"k") + std::to_string(100 + k) + "\":" + std::to_string(value + k);
+            }
+            if (duplicates)
+            {
+                for (int k = n / 2; k < n; k += 3)
+                {
+                    s += ",\"k" + std::to_string(100 + k) + "\":" + std::to_string(value + k);
+                }
+            }
+            return s + "}";
+        };
+        std::vector<std::string> texts;
+        for (const int n :
+                {
+                    3, 16, 17, 40
+                })
+        {
+            for (const int flags :
+                    {
+                        0, 1, 2, 3
+                    })
+            {
+                texts.push_back(object(n, (flags & 1) != 0, (flags & 2) != 0, 0));
+            }
+            texts.push_back(object(n, true, false, 1)); // other values
+        }
+        // many objects in one comparison (working memory reused across objects)
+        std::string array = "[";
+        std::string array_other = "[";
+        for (std::size_t i = 0; i < texts.size(); ++i)
+        {
+            array += (i != 0 ? "," : "") + texts[i];
+            array_other += (i != 0 ? "," : "") + texts[texts.size() - 1 - i];
+        }
+        texts.push_back(array + "]");
+        texts.push_back(array_other + "]");
+        for (const std::string& x : texts)
+        {
+            for (const std::string& y : texts)
+            {
+                CAPTURE(x)
+                CAPTURE(y)
+                const ordered_json_document dx = ordered_json_document::parse(x);
+                const ordered_json_document dy = ordered_json_document::parse(y);
+                const ordered_json jx = ordered_json::parse(x);
+                const ordered_json jy = ordered_json::parse(y);
+                CHECK((dx.root() == dy.root()) == (jx == jy));
+                CHECK((dx.root() == jy) == (jx == jy));
+                const json_document ux = json_document::parse(x);
+                const json uy = json::parse(y);
+                CHECK((ux.root() == uy) == (json::parse(x) == uy));
+            }
+        }
+    }
+
+    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("values that are the same text, or the same value written differently, at any depth")
+    {
+        // nested arrays and objects around depths where the walk over equal text gives up
+        for (const int depth :
+                {
+                    1, 5, 47, 48, 49, 50, 60, 300
+                })
+        {
+            CAPTURE(depth)
+            std::string head;
+            std::string tail;
+            for (int i = 0; i < depth; ++i)
+            {
+                head += i % 2 == 0 ? R"({"k":[1,"s",{"x":true},)" : R"([null,2.5,{"y":)";
+                tail.insert(0, i % 2 == 0 ? "]}" : "}]");
+            }
+            // the same text, one number changed, and one number written differently
+            const auto with_number = [&](const char* number)
+            {
+                std::string result = head;
+                result += number;
+                result += tail;
+                return result;
+            };
+            const std::string text = with_number("7");
+            const std::string changed = with_number("8");
+            const std::string spelled = with_number("7.0");
+            const json_document a = json_document::parse(text);
+            const json_document b = json_document::parse(text);
+            const json_document c = json_document::parse(changed);
+            const json_document d = json_document::parse(spelled);
+            CHECK(a.root() == b.root());
+            CHECK(a.root() != c.root());
+            CHECK(c.root() != a.root());
+            CHECK(a.root() == d.root());
+            CHECK(d.root() == a.root());
+            CHECK(a.root() == json::parse(text));
+            CHECK(a.root() != json::parse(changed));
+            CHECK(json::parse(spelled) == a.root());
+            const nlohmann::json_editable_document e = nlohmann::json_editable_document::parse(text);
+            CHECK(a.root() == e.root());
+            CHECK(e.root() == a.root());
+            CHECK(c.root() != e.root());
+            CHECK(e.root() == e.root());
+            const ordered_json_document oa = ordered_json_document::parse(text);
+            const ordered_json_document oc = ordered_json_document::parse(changed);
+            CHECK(oa.root() != oc.root());
+            CHECK(oa.root() == ordered_json::parse(text));
+            CHECK(oc.root() != ordered_json::parse(text));
+        }
+
+        // not the same text, but the same value
+        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);
+            const bool equal = dx.root() == dy.root();
+            CHECK(equal == (json::parse(x) == json::parse(y)));
+            CHECK(equal == (dx.root() == json::parse(y)));
+            return equal;
+        };
+        CHECK(same(R"({"a":1,"a":2})", R"({"a":3,"a":2})"));
+        CHECK(same(R"({"a":1,"b":[1,{"c":2,"c":3}]})", R"({"b":[1,{"c":9,"c":3}],"a":1})"));
+        CHECK(!same(R"({"a":1,"a":2})", R"({"a":2,"a":3})"));
+        CHECK(same("[1.0,-0,0,1e2]", "[1.00,0,-0,100.0]"));
+        CHECK(same(R"(["\u00e9","\n"])", "[\"\xc3\xa9\",\"\\n\"]"));
+        CHECK(!same(R"({"a":[1,2,{"b":null}]})", R"({"a":[1,2,{"b":false}]})"));
+        CHECK(!same(R"({"a":[1,2,{"b":null}]})", R"({"a":[1,2,{"c":null}]})"));
+        CHECK(!same(R"({"a":[1,2,{"b":null}]})", R"({"a":[1,2,{"b":null,"c":1}]})"));
+
+        // edited documents: moved sequences, links, and values of new nodes
+        nlohmann::json_editable_document edited = nlohmann::json_editable_document::parse(R"({"a":[1,2,3],"b":{"c":"d"},"e":0.5})");
+        const json_document plain = json_document::parse(R"({"a":[1,2,3,4],"b":{"c":"x","y":null},"e":0.5,"z":[]})");
+        edited.push_back(edited.root()["a"], 4);
+        edited.set(edited.root()["b"], "c", "x");
+        edited.set(edited.root()["b"], "y", nullptr);
+        edited.set(edited.root(), "z", json::array());
+        CHECK(edited.root() == plain.root());
+        CHECK(plain.root() == edited.root());
+        edited.erase(edited.root()["a"], 0);
+        CHECK(edited.root() != plain.root());
+        CHECK(edited.root()["a"] == json::parse("[2,3,4]"));
+        const json_document tail = json_document::parse("[2,3,4]");
+        CHECK(edited.root()["a"] == tail.root());
+    }
+
+    SECTION("large documents, with a difference at any place")
+    {
+        const auto make = [](std::size_t count, std::size_t changed, bool swap_members)
+        {
+            std::string text = R"({"b":1,"list":[)";
+            for (std::size_t i = 0; i < count; ++i)
+            {
+                text += (i != 0 ? "," : "");
+                const std::string id = std::to_string(i == changed ? i + 1 : i);
+                if (swap_members && i % 3 == 0)
+                {
+                    text += R"({"v":[1.5,"s",{"k":)";
+                    text += id;
+                    text += R"(}],"id":)";
+                    text += id;
+                    text += "}";
+                }
+                else
+                {
+                    text += R"({"id":)";
+                    text += id;
+                    text += R"(,"v":[1.5,"s",{"k":)";
+                    text += id;
+                    text += "}]}";
+                }
+            }
+            return text + R"(],"a":{"y":2,"x":3}})";
+        };
+        const std::size_t count = 3000;
+        for (const std::size_t changed :
+                {
+                    count, std::size_t{0}, std::size_t{1}, count / 2, count - 1
+                })
+        {
+            for (const bool swap :
+                    {
+                        false, true
+                    })
+            {
+                CAPTURE(changed)
+                CAPTURE(swap)
+                const std::string text_a = make(count, count, false);
+                const std::string text_b = make(count, changed, swap);
+                const json_document a = json_document::parse(text_a);
+                const json_document b = json_document::parse(text_b);
+                const bool equal = changed == count;
+                CHECK((a.root() == b.root()) == equal);
+                CHECK((b.root() == a.root()) == equal);
+                CHECK((a.root() == json::parse(text_b)) == equal);
+                CHECK((json::parse(text_b) == a.root()) == equal);
+                CHECK((a.root()["list"] == b.root()["list"]) == equal);
+                const ordered_json_document oa = ordered_json_document::parse(text_a);
+                const ordered_json_document ob = ordered_json_document::parse(text_b);
+                CHECK((oa.root() == ob.root()) == (equal && !swap));
+                CHECK((oa.root() == ordered_json::parse(text_b)) == (equal && !swap));
+            }
+        }
+    }
+
+    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());
+    }
+}
@@ -432,101 +432,6 @@ TEST_CASE("json_view dump")
    }
 }

-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
@@ -941,3 +941,123 @@ TEST_CASE("json_view edits: duplicate keys")
        CHECK(d.root()["o"]["d"].get<int>() == 0);
    }
 }
+
+#if !defined(JSON_NOEXCEPTION)
+TEST_CASE("json_view edits: check_utf8 agrees with the byte-wise check")
+{
+    namespace dv = nlohmann::detail::view;
+
+    // "ok" or the exception message
+    const auto run = [](bool fast, const std::string & s) -> std::string
+    {
+        try
+        {
+            if (fast)
+            {
+                dv::check_utf8(s.data(), s.size());
+            }
+            else
+            {
+                dv::check_utf8_bytewise(reinterpret_cast<const unsigned char*>(s.data()), s.size(), 0);
+            }
+        }
+        catch (const json::type_error& e)
+        {
+            return {e.what()};
+        }
+        return {"ok"};
+    };
+    std::size_t errors = 0;
+    const auto same = [&](const std::string & s)
+    {
+        const std::string a = run(true, s);
+        const std::string b = run(false, s);
+        if (a != b)
+        {
+            INFO("fast: " << a << ", byte-wise: " << b << ", length " << s.size());
+            CHECK(a == b);
+        }
+        errors += b != "ok" ? 1u : 0u;
+    };
+
+    const std::vector<std::string> pieces =
+    {
+        "a", "Z", " ", "\"", "\\", "\n", "\x01", "\x7f", "\xc3\xa4", "\xe4\xb8\xad", "\xf0\x9f\x98\x80",
+        "\x80", "\xbf", "\xc0\x80", "\xc1\xbf", "\xc2", "\xe0\x80\x80", "\xe0\x9f\xbf", "\xed\xa0\x80", "\xed\xbf\xbf",
+        "\xf0\x80\x80\x80", "\xf0\x8f\xbf\xbf", "\xf4\x90\x80\x80", "\xf4\x8f\xbf\xbf", "\xf5", "\xf8", "\xff", "\xfe",
+        "\xe4\xb8", "\xf0\x9f\x98", "\xf0\x9f", "\xf0"
+    };
+
+    // every piece behind prefixes of every length, and followed by more bytes
+    for (std::size_t prefix = 0; prefix <= 64; ++prefix)
+    {
+        for (const auto& piece : pieces)
+        {
+            same(std::string(prefix, 'x') + piece);
+            same(std::string(prefix, 'x') + piece + "yz");
+            same(std::string(prefix, '"') + piece);
+            same(std::string(prefix, '\xc3') + piece);
+            same(std::string(prefix / 2, '\xc3').append(prefix / 2, '\xa4') + piece);
+            same(std::string(prefix, 'x') + "\xe4\xb8\xad" + piece);
+        }
+    }
+    // an invalid byte at every position of valid strings of every length
+    for (std::size_t len = 0; len <= 64; ++len)
+    {
+        const std::string ascii(len, 'a');
+        std::string mixed;
+        while (mixed.size() < len)
+        {
+            if (mixed.size() % 7 == 0)
+            {
+                mixed += "\xc3\xa4";
+            }
+            else if (mixed.size() % 11 == 0)
+            {
+                mixed += "\xf0\x9f\x98\x80";
+            }
+            else
+            {
+                mixed += "a";
+            }
+        }
+        same(ascii);
+        same(mixed);
+        for (std::size_t pos = 0; pos < len; ++pos)
+        {
+            for (const char bad :
+                    {'\x80', '\xc0', '\xc2', '\xe0', '\xf4', '\xff', '"', '\\', '\x1f'
+                    })
+            {
+                std::string s = ascii;
+                s[pos] = bad;
+                same(s);
+                s = mixed;
+                s[pos] = bad;
+                same(s);
+            }
+        }
+    }
+    // random bytes, biased to the interesting ones
+    std::mt19937 gen(5869); // NOLINT(cert-msc32-c,cert-msc51-cpp,bugprone-random-generator-seed): reproducible
+    static const std::array<unsigned char, 23> interesting = {{0x00, 0x1f, 0x22, 0x5c, 0x7f, 0x80, 0x8f, 0x90, 0x9f, 0xa0, 0xbf, 0xc0, 0xc2, 0xdf, 0xe0, 0xe1, 0xed, 0xef, 0xf0, 0xf1, 0xf4, 0xf5, 0xff}};
+    for (int i = 0; i < 100000; ++i)
+    {
+        std::string s(gen() % 80, 'a');
+        for (auto& c : s)
+        {
+            const unsigned r = gen() % 8;
+            if (r == 0)
+            {
+                c = static_cast<char>(gen() & 0xff);
+            }
+            else if (r == 1)
+            {
+                c = static_cast<char>(interesting[gen() % interesting.size()]);
+            }
+        }
+        same(s);
+    }
+    CHECK(errors > 1000);
+}
+#endif
@@ -701,6 +701,64 @@ TEST_CASE("json_view images: ownership")
 #if !defined(JSON_NOEXCEPTION)
 TEST_CASE("json_view images: errors")
 {
+    SECTION("float tokens: a changed byte is accepted if the text is still a number with the digits the node records")
+    {
+        for (const char* const token :
+                {
+                    "-12.375", "0.5", "123456.0", "-0.0", "9999999999999999999.5", "0.000000000000000000000001", "1.25e10", "5E-3"
+                })
+        {
+            CAPTURE(token)
+            const std::string text = std::string("[") + token + "]";
+            const json_document d = json_document::parse(text);
+            const std::vector<std::uint8_t> image = d.save();
+            const std::size_t at = text_at(image) + 1; // the first byte of the token
+            const std::uint16_t recorded = node_at(image, 1).extra;
+            CHECK(loaded_dump(image) == json::parse(text).dump());
+            for (std::size_t pos = 0; token[pos] != '\0'; ++pos)
+            {
+                for (unsigned int byte = 0; byte < 256; ++byte)
+                {
+                    std::vector<std::uint8_t> changed = image;
+                    changed[at + pos] = static_cast<std::uint8_t>(byte);
+                    std::string t = token;
+                    t[pos] = static_cast<char>(byte);
+                    // the digits of the integer part and of the fraction of the changed token
+                    std::size_t integer = 0;
+                    std::size_t fraction = 0;
+                    std::size_t k = (t[0] == '-') ? 1 : 0;
+                    for (; k < t.size() && t[k] >= '0' && t[k] <= '9'; ++k)
+                    {
+                        ++integer;
+                    }
+                    if (k < t.size() && t[k] == '.')
+                    {
+                        for (++k; k < t.size() && t[k] >= '0' && t[k] <= '9'; ++k)
+                        {
+                            ++fraction;
+                        }
+                    }
+                    // (as a JSON text, the token may have white space around it)
+                    const bool float_token = json::accept(t) && t.find_first_of(std::string(" \t\r\n\0", 5)) == std::string::npos && (t.find('.') != std::string::npos || t.find_first_of("eE") != std::string::npos) // NOLINT(abseil-string-find-str-contains)
+                                             && ((integer | (fraction << 8U)) == recorded);
+                    CAPTURE(pos)
+                    CAPTURE(byte)
+                    std::string result;
+                    try
+                    {
+                        result = loaded_dump(changed);
+                    }
+                    catch (const json::parse_error& e)
+                    {
+                        CHECK(std::string(e.what()) == check_failed);
+                        result = "rejected";
+                    }
+                    CHECK((result != "rejected") == float_token);
+                }
+            }
+        }
+    }
+
    SECTION("invalid UTF-8 of a decoded string: the compact dump writes it, the others throw")
    {
        const json_document d = json_document::parse(R"({"k\n":"\u00e9x"})");
@@ -23,7 +23,6 @@ using nlohmann::json;
 #include <cstdint>
 #include <sstream>
 #include <iomanip>
-#include <limits>
 #include <string>
 #include <vector>

@@ -249,51 +248,6 @@ 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});
Author	SHA1	Message	Date
Niels Lohmann	6491563cbb	json_view: skip the digits of a fraction 16 bytes at a time (SSE2) The digits after the decimal point were skipped by skip_digits(), 16 compare-and-branch steps in a row: about 3.5 instructions per digit, and an exit branch that is mispredicted when the lengths of the fractions vary. On x86 with SSE2, the fraction is now skipped by skip_fraction_digits(): 16 bytes are compared at once and the end of the run is taken from the bit mask, which is about a third of the instructions for a fraction of 12 digits. The last bytes of the input are skipped one at a time as before. Where there is no SSE2 (AArch64, other targets, JSON_NO_SIMD) skip_fraction_digits() is skip_digits(): nothing changes there. Integer digits keep skip_digits(), whose exit is predictable. GCC unrolls the loop over the last bytes in every inlined copy of the new function, which made the parse loop 4 KiB larger and parse() of documents with hardly any fractions about 4 % slower. For GCC that loop is therefore out of line (skip_digits_tail()); Clang is faster with it inline. json_document::parse(), A/B in one binary, 6 code layouts, x86-64 Haswell, time after / before (GCC 16.2 / Clang 23): - numbers.json (10,001 floats): 0.715 / 0.680 - canada.json: 0.778 / 0.790 - mesh.json: 0.908 / 0.883 - marine_ik.json: 0.937 / 0.926 - mesh.pretty.json: 0.923 / 0.960 Files with few or no fractions and files with fractions of a fixed length (20,000 prices with two decimals, 60,000 coordinates with six): between 0.98 and 1.014 on both compilers, except citm_catalog.json with GCC (1.05: the instruction count is +0.3 %, and that row varies by +-7 % between builds of unrelated variants on this machine). New test: 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, in several positions, and cut in every place near the end of the text, against json::parse() (the length of the number is part of what is compared), for all option combinations and with and without a terminating NUL. Checked on x86-64 with GCC 16.2, Clang 23 and Apple clang under Rosetta, and with a deliberate off-by-one at the 16th byte. Signed-off-by: Niels Lohmann <mail@nlohmann.me>	2026-10-30 22:17:00 +01:00
Niels Lohmann	755bb16d7a	Fix the red CI on develop: remove the duplicate is_editable in basic_json_view (#5966 ) #5919 replaced the bool template parameter Editable of basic_json_view by view_mode Mode and defined is_editable from Mode. #5890 was written before that and added a second is_editable = Editable to the private section, which no longer compiles ("duplicate member", "undeclared identifier Editable"). It broke every job that builds the json_view tests (macOS, MSVC, clang-cl, clang on Windows, the C++20 module job, CodeQL). Signed-off-by: Niels Lohmann <mail@nlohmann.me>	2026-10-30 16:48:31 +01:00
Niels Lohmann	d20c4dcd89	json_view benchmarks: compare fairly and state what each row means (#5914 ) - bench_view: add "(editable)" rows for json_editable_document and a parse+dump workload, so the cost of lazy float conversion shows next to the parse row (the traverse workload already includes the parse). Check that all engines' dumps agree where outputs are comparable. - compare.py: add --aligned, which also builds and runs with aligned code and reports both tables, so code-layout artifacts become visible. Off by default; the default output is unchanged. Not run end to end yet. - README: say which engines are like-for-like peers (yyjson, simdjson DOM) and which are mutable reference points (Boost.JSON, json::parse, the editable json_view); explain the lazy float conversion and the eager prototype that was measured and dropped. Signed-off-by: Niels Lohmann <mail@nlohmann.me>	2026-10-30 15:20:12 +01:00
Niels Lohmann	ea03d601af	json_view: comparison without temporaries and a faster image float check (#5890 ) * Compare json_view values without a vector per container and a basic_json per number operator== of views (view/view, view/json, and editable documents) built two vectors of members for every object it entered, sorted them with stable_sort for every object, and converted each number to a basic_json value. It now: - keeps the members of all open objects in one vector (reserved with geometric growth, so that nested objects do not reallocate), and orders them by insertion sort for up to 16 members, or not at all when the keys ascend already; - compares numbers of one type directly, and builds a basic_json only for numbers of different types; - compares short texts (keys, numbers, most strings) with two overlapping loads (bytes_equal(), in node.hpp); - first walks both documents in lockstep, in the order of their text, which finds most pairs of equal values without the work of the comparison. The walk only enters an object if the comparison would visit its members in the same order (the object type keeps the order, the keys ascend, or all values are scalars), so a difference is never found later than before, and it records the containers it found different, so that the comparison does not walk them again; - rejects values of different types and arrays of different sizes before any of this, and, for an object type that keeps the order, objects whose first (up to four) members have different keys, which ends a comparison that differs near the start without preparing the members. Results are unchanged: numbers compare by value across types, duplicate keys resolve as parse() resolves them, and objects compare in document order for ordered_json and by key otherwise. Instructions of one comparison of a document with an equal copy, macOS arm64 (develop -> now), json_document / ordered_json_document / editable: twitter 16.1M -> 7.0M 17.7M -> 1.5M 16.6M -> 7.3M citm 45.0M -> 3.4M 47.0M -> 3.2M 45.7M -> 4.3M canada 111.4M -> 13.7M 111.4M -> 13.7M 113.3M -> 15.0M jeopardy 26.2M -> 3.7M 27.3M -> 3.7M 26.6M -> 12.1M and of a view with a basic_json value: twitter 11.4M -> 5.9M 11.2M -> 7.3M 11.6M -> 5.9M citm 33.9M -> 9.7M 32.6M -> 9.1M 34.6M -> 9.8M canada 92.9M -> 32.1M 92.9M -> 32.1M 93.9M -> 32.8M and Linux instruction counts (GCC 13 / Clang 18) of the same comparisons: document with an equal copy json_document ordered_json_document editable twitter -49 % / -52 % -89 % / -89 % -48 % / -50 % citm -90 % / -89 % -91 % / -90 % -88 % / -87 % jeopardy (4,000 questions) -85 % / -85 % -86 % / -86 % -46 % / -49 % view with a basic_json value twitter -41 % / -45 % -28 % / -35 % -45 % / -44 % citm -63 % / -65 % -63 % / -66 % -65 % / -64 % jeopardy -39 % / -43 % -14 % / -24 % -44 % / -43 % A sweep that places one difference at about 2,100 positions of eight documents (a token at an even spacing, the first and last tokens, the first, middle and last member of the root removed, one appended, the document wrapped) in five kinds of comparison (view/view of read-only and of editable documents, view/json, json_document, editable/view) finds the same results as before and no placement slower on the Mac (Clang); on Linux no placement is slower than 3 % (Clang and GCC). Signed-off-by: Niels Lohmann <mail@nlohmann.me> * Check the float tokens of a json_document image without a sort and a scan of every character load(image_check::full) checked every float token of an image by sorting the token ranges, scanning the token character by character, and comparing the digit layout with the node's. Now: - the ranges are sorted only if they are not sorted already (save() writes them in the order of the nodes, so they are); - a token of the shape [-]digits.digits whose digit counts are the ones the node records (nearly all of them) is accepted by checking its digits eight at a time, without the general scan; - the general scan skips runs of digits eight at a time. Damaged tokens are still rejected as before: the check accepts exactly the tokens that the general scan accepts. Instructions of load(image_check::full) (develop -> now): macOS arm64 canada 44.6M -> 30.9M (-31 %), jeopardy 11.3M -> 10.7M, citm 10.4M -> 10.0M, twitter 5.06M -> 4.94M Linux GCC 13 canada -50 %, jeopardy -43 %, twitter -31 %, citm -28 % Linux Clang 18 canada -42 %, jeopardy -48 %, twitter -35 %, citm -32 % image_check::bounds is unchanged (within 0.4 %). Signed-off-by: Niels Lohmann <mail@nlohmann.me> * Rename names that Flawfinder takes for the file function open() The comparison cursors' open() is now enter(), and a test's open and close strings are head and tail. Signed-off-by: Niels Lohmann <mail@nlohmann.me> * Fix the compiler and lint failures of the json_view comparison - all_digits(): the lambda declares its constant itself; MSVC does not let a lambda without captures read a constexpr local of the function - equal(): the frame's "in document order" index is a namespace-scope constant (document_order); the local constexpr npos in the default member initializers of the local struct crashed GCC 9 and MinGW GCC and was not found by MSVC - keys_differ_early() keeps the first keys in std::arrays (cpplint: variable-length arrays; clang-tidy: C arrays), and the basic_json side's first_keys() is not noexcept (key() may throw) - the numbers of one type are compared in an if chain instead of a nested conditional operator - the dump test avoids C arrays and temporary string concatenations View comparison instructions are unchanged (Apple clang, twitter, citm, canada: -0.1 % to +0.1 %). Signed-off-by: Niels Lohmann <mail@nlohmann.me> * Keep the byte comparison helpers in compare.hpp Address review comments: bytes_equal() and bytes_less() move from node.hpp to compare.hpp, their only user; key_less() no longer splits the comment of resolve_indices() from its function; and the loop counters of bytes_equal() and all_digits() are scoped to their loops. Signed-off-by: Niels Lohmann <mail@nlohmann.me> * Dispatch on the compile-time conditions of the json_view comparison MSVC (/W4 /WX) reported C4127 (conditional expression is constant) for the template parameter EA and the object type's ordering in an if condition; both are now tag-dispatched. Signed-off-by: Niels Lohmann <mail@nlohmann.me> * Remove a nested duplicate guard from the json_view image test clang-tidy reported the inner #if !defined(JSON_NOEXCEPTION) as redundant (the test case is in the same #if) and a find() != npos, which gets the NOLINT the other tests use. Signed-off-by: Niels Lohmann <mail@nlohmann.me> * Mark Infer's null dereference in materialize() as a false positive Infer (ci_infer, also red on develop since #5882) reports a null dereference in json_sax_dom_parser::null() called from materialize(). The member pointer the parser stores a value of an object in is set by the key() before it, so it is not null there. Signed-off-by: Niels Lohmann <mail@nlohmann.me> * Move the json_view comparison tests into their own file With the tests this branch adds, the object file of unit-json_view_dump.cpp got too large for the MinGW x64 assembler ("file too big"). The test case "json_view comparison" is now unit-json_view_compare.cpp, unchanged and also built with C++17. Signed-off-by: Niels Lohmann <mail@nlohmann.me> * Build the DOM parser with Visual Studio 2015 again Since #5882, the three VS 2015 jobs of AppVeyor fail on develop: - the detection of keys that can be indexed instantiated std::hash for every string type, which VS 2015 rejects with a static_assert for a type it does not hash (alt_string): std::hash is now looked at only for objects in a vector - the frame stack copied its frames with std::copy to a raw pointer, which VS 2015 reports as unsafe (C4996, an error with /WX): a loop now Signed-off-by: Niels Lohmann <mail@nlohmann.me> * json_view: no per-object allocations in ordered_json comparisons For an object type that keeps the order of its members (ordered_json), resolve_indices() allocated a vector per compared object, and resolve_members() two. They now use one working vector owned by equal(), whose capacity is reused, and resolve_indices() needs none when the keys are unique. resolve_members() sorts small objects by insertion, as the other paths do. Instructions per comparison (callgrind, GCC 13 and clang 18, x86-64): twitter, ordered_json view == ordered_json: 0.939 / 0.941; 20000 small objects with a duplicate key: 0.741-0.768; all other rows (json, citm_catalog, canada, view == view) within 0.988-1.003. Signed-off-by: Niels Lohmann <mail@nlohmann.me> --------- Signed-off-by: Niels Lohmann <mail@nlohmann.me>	2026-10-30 15:19:42 +01:00
dependabot[bot]	e2cb15c479	Bump the codeql-action group across 1 directory with 4 updates (#5961 ) Bumps the codeql-action group with 4 updates in the / directory: [github/codeql-action/init](https://github.com/github/codeql-action), [github/codeql-action/autobuild](https://github.com/github/codeql-action), [github/codeql-action/analyze](https://github.com/github/codeql-action) and [github/codeql-action/upload-sarif](https://github.com/github/codeql-action). Updates `github/codeql-action/init` from 4.39.0 to 4.39.1 - [Release notes](https://github.com/github/codeql-action/releases) - [Changelog](https://github.com/github/codeql-action/blob/main/CHANGELOG.md) - [Commits](https://github.com/github/codeql-action/compare/6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee...9b3254085494b55a0f7c4c2b3d2d52cc712ce370) Updates `github/codeql-action/autobuild` from 4.39.0 to 4.39.1 - [Release notes](https://github.com/github/codeql-action/releases) - [Changelog](https://github.com/github/codeql-action/blob/main/CHANGELOG.md) - [Commits](https://github.com/github/codeql-action/compare/6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee...9b3254085494b55a0f7c4c2b3d2d52cc712ce370) Updates `github/codeql-action/analyze` from 4.39.0 to 4.39.1 - [Release notes](https://github.com/github/codeql-action/releases) - [Changelog](https://github.com/github/codeql-action/blob/main/CHANGELOG.md) - [Commits](https://github.com/github/codeql-action/compare/6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee...9b3254085494b55a0f7c4c2b3d2d52cc712ce370) Updates `github/codeql-action/upload-sarif` from 4.39.0 to 4.39.1 - [Release notes](https://github.com/github/codeql-action/releases) - [Changelog](https://github.com/github/codeql-action/blob/main/CHANGELOG.md) - [Commits](https://github.com/github/codeql-action/compare/6f2ac81f5cd81fe351a197ba0b00d91d03cd2aee...9b3254085494b55a0f7c4c2b3d2d52cc712ce370) --- updated-dependencies: - dependency-name: github/codeql-action/analyze dependency-version: 4.39.1 dependency-type: direct:production update-type: version-update:semver-patch dependency-group: codeql-action - dependency-name: github/codeql-action/autobuild dependency-version: 4.39.1 dependency-type: direct:production update-type: version-update:semver-patch dependency-group: codeql-action - dependency-name: github/codeql-action/init dependency-version: 4.39.1 dependency-type: direct:production update-type: version-update:semver-patch dependency-group: codeql-action - dependency-name: github/codeql-action/upload-sarif dependency-version: 4.39.1 dependency-type: direct:production update-type: version-update:semver-patch dependency-group: codeql-action ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>	2026-10-30 15:19:08 +01:00
dependabot[bot]	c451fac1bc	Bump step-security/harden-runner from 2.22.1 to 2.22.2 (#5962 ) Bumps [step-security/harden-runner](https://github.com/step-security/harden-runner) from 2.22.1 to 2.22.2. - [Release notes](https://github.com/step-security/harden-runner/releases) - [Commits](https://github.com/step-security/harden-runner/compare/ccd8616d44fd3846e67624a50d5aad6d37bf2d25...dc2769abcc0fcfbadfc1df54081c1608579ef6b2) --- updated-dependencies: - dependency-name: step-security/harden-runner dependency-version: 2.22.2 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>	2026-10-30 15:18:56 +01:00
dependabot[bot]	28f71914b5	Bump egor-tensin/setup-mingw (#5963 ) Bumps [egor-tensin/setup-mingw](https://github.com/egor-tensin/setup-mingw) from 41b837e47d7f85214629d255b9c4bc3fcbe9fd63 to d54f5ef0ce585bd3fbb1f8a71b5b254f59dbf931. - [Release notes](https://github.com/egor-tensin/setup-mingw/releases) - [Changelog](https://github.com/egor-tensin/setup-mingw/blob/master/CHANGELOG.md) - [Commits](https://github.com/egor-tensin/setup-mingw/compare/41b837e47d7f85214629d255b9c4bc3fcbe9fd63...d54f5ef0ce585bd3fbb1f8a71b5b254f59dbf931) --- updated-dependencies: - dependency-name: egor-tensin/setup-mingw dependency-version: d54f5ef0ce585bd3fbb1f8a71b5b254f59dbf931 dependency-type: direct:production ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>	2026-10-30 15:18:44 +01:00
Niels Lohmann	0b91f9ddd8	json_view: check_utf8 skips well-formed runs with the parser's string scan (#5926 ) * json_view: check_utf8 skips well-formed runs with the parser's string scan Strings of 32 bytes or more that are written into an editable document are validated with scan_string_run(), which skips well-formed UTF-8 16 bytes at a time; the byte-wise check runs only from where the scan found ill-formed UTF-8, to report the same type_error.316. Shorter strings keep the byte-wise check (inlined with a constant start, so that this path is the code it was), where the vector scan costs more than it saves. set()/push_back() of strings of 32 bytes or more take 0.29-0.76 of the time (x86-64 GCC, averaged over code layouts) and 0.25-0.80 (AArch64 clang); shorter strings take 1.5-4 % fewer instructions; the json_view rows of bench_edit execute at most 0.04 % more instructions. Signed-off-by: Niels Lohmann <mail@nlohmann.me> * tests: avoid sign conversion in json_view_edit UTF-8 check Signed-off-by: Niels Lohmann <mail@nlohmann.me> * json_view: use while in check_utf8_bytewise Signed-off-by: Niels Lohmann <mail@nlohmann.me> --------- Signed-off-by: Niels Lohmann <mail@nlohmann.me>	2026-10-30 15:17:55 +01:00
Niels Lohmann	0115bb47b3	Name the source and license of adapted code in the file comments (#5965 ) The files with code adapted from fast_float, Zmij and simdjson carried the authors' SPDX copyright lines, but their file comments did not say what was taken or under which license: number_parse.hpp, to_chars.hpp, simd_scan.hpp and view/simd.hpp now do, as zmij.hpp and the yyjson-based view files already did. README and the license page now also name fast_float's eight-digit parsing and the lexer's use of simdjson's UTF-8 check. Comments only. Signed-off-by: Niels Lohmann <mail@nlohmann.me>	2026-10-30 15:17:24 +01:00