Rewrite OBJ/MTL parser to byte-stream based with StreamReader, add mmap support, and integrate Windows long path/Unicode support by Copilot · Pull Request #416 · tinyobjloader/tinyobjloader

Copilot · 2026-03-03T17:05:23Z

Confirm actionable scope for new comment (4009096646)
Check recent CI workflow runs and inspect failures/logs
Perform final stage-3 deep code/security review with maximum effort
Apply additional fixes and add regression tests
Run targeted tests and full make check
Run code review and CodeQL checks
Reply on PR comment with final stage-3 outcome and commit hash

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

…oading Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

syoyo · 2026-03-03T17:45:47Z

@copilot rewrite line based and raw token pointer scanning/parsing to byte-stream based. see https://raw.githubusercontent.com/lighttransport/tinyusdz/refs/heads/release/src/stream-reader.hh for the reference.

rewrite IS_NEW_LINE, IS_SPACE, parse***, strspn, etc. to use the byte-stream based token parsing(add eof guard for each token consumption)

Replace safeGetline + raw const char* token pointer scanning with a StreamReader class that wraps (buf, len, idx) with bounds-checked peek/advance/match operations. Every byte access is guarded by an EOF check. Key changes: - Add StreamReader class for bounds-checked text parsing - Add sr_* parse functions that use StreamReader instead of const char** - Add LoadMtlInternal, LoadObjInternal, LoadObjWithCallbackInternal using StreamReader; public APIs wrap these - mmap paths construct StreamReader directly from mapped memory - Remove safeGetline (no longer needed) - Keep old parse functions for backward compat with public ParseTextureNameAndOption API - Keep IS_SPACE, IS_NEW_LINE, IS_DIGIT macros for old parse functions Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

…ked parsing - Add StreamReader class wrapping (buf, length) with EOF-guarded peek/advance/match - Add sr_* parse functions using StreamReader instead of const char** pointers - Add LoadMtlInternal, LoadObjInternal, LoadObjWithCallbackInternal taking StreamReader - mmap paths construct StreamReader directly from mapped memory - Remove safeGetline (no longer needed) - Keep old parse functions + macros for ParseTextureNameAndOption backward compat Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot · 2026-03-03T18:09:28Z

@copilot rewrite line based and raw token pointer scanning/parsing to byte-stream based. see https://raw.githubusercontent.com/lighttransport/tinyusdz/refs/heads/release/src/stream-rea...

Implemented in f2c657b and c57ed2c. The parser is now byte-stream based using a StreamReader class that wraps (const char *buf, size_t length) with EOF-guarded peek()/advance()/match() on every byte access.

Key changes:

StreamReader class with eof(), peek(), advance(), skip_space(), skip_line(), at_line_end(), match(), peek_at(), remaining(), etc.
sr_* parse functions (sr_parseString, sr_parseInt, sr_parseReal, sr_parseTriple, sr_parseRawTriple, etc.) all using StreamReader instead of raw const char** + strspn/strcspn
Internal functions LoadMtlInternal, LoadObjInternal, LoadObjWithCallbackInternal take StreamReader& and do all parsing with byte-stream scanning
safeGetline removed — StreamReader handles line-by-line byte scanning internally
mmap paths construct StreamReader directly from mapped memory (no double-buffering)
Old parse functions + IS_SPACE/IS_NEW_LINE macros preserved for ParseTextureNameAndOption public API backward compatibility

…-mtl-parsing

Copilot

Pull request overview

This PR rewrites the OBJ/MTL parser from line-based (safeGetline + raw const char* token scanning) to a byte-stream-based approach using a new StreamReader class, and adds opt-in memory-mapped I/O support (TINYOBJLOADER_USE_MMAP).

Changes:

Introduces StreamReader class with bounds-checked byte access, and rewrites all parse functions as sr_* variants consuming StreamReader instead of const char** pointers.
Converts LoadObj, LoadMtl, and LoadObjWithCallback to thin wrappers over new LoadObjInternal, LoadMtlInternal, and LoadObjWithCallbackInternal helpers.
Adds mmap support for both POSIX and Windows for both .obj and .mtl files, with tests for the new paths.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 5 comments.

File	Description
`tiny_obj_loader.h`	Core rewrite: `StreamReader` class, all `sr_*` parse functions, mmap infrastructure for Windows/POSIX, internal helper functions, and `membuf`/`imemstream` dead code
`tests/tester.cc`	Two new tests: `test_mmap_and_standard_load_agree` (compares mmap vs stream results) and `test_load_from_memory_buffer` (verifies in-memory parsing consistency)

Comments suppressed due to low confidence (1)

tiny_obj_loader.h:3165

The mmap open/fstat/map/parse/unmap pattern is duplicated across six call sites: twice in MaterialFileReader::operator() (multi-path loop and single-path branch, each for both Windows and POSIX) and twice in LoadObj (Windows and POSIX). Extracting the common logic into a static helper (e.g., LoadMtlFromFile or LoadObjFromFile) would significantly reduce the code surface and the risk of the two copies diverging.

#ifdef TINYOBJLOADER_USE_MMAP
#if defined(_WIN32)
      {
        HANDLE hFile =
            CreateFileA(filepath.c_str(), GENERIC_READ, FILE_SHARE_READ, NULL,
                        OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL);
        if (hFile == INVALID_HANDLE_VALUE) continue;
        LARGE_INTEGER fileSize;
        if (!GetFileSizeEx(hFile, &fileSize)) {
          CloseHandle(hFile);
          continue;
        }
        const size_t file_size = static_cast<size_t>(fileSize.QuadPart);
        if (file_size == 0) {
          CloseHandle(hFile);
          StreamReader empty_sr("", 0);
          LoadMtlInternal(matMap, materials, empty_sr, warn, err);
          return true;
        }
        HANDLE hMapping =
            CreateFileMappingA(hFile, NULL, PAGE_READONLY, 0, 0, NULL);
        if (hMapping == NULL) {
          CloseHandle(hFile);
          continue;
        }
        const char *mmap_data = static_cast<const char *>(
            MapViewOfFile(hMapping, FILE_MAP_READ, 0, 0, 0));
        if (!mmap_data) {
          CloseHandle(hMapping);
          CloseHandle(hFile);
          continue;
        }
        {
          StreamReader sr(mmap_data, file_size);
          LoadMtlInternal(matMap, materials, sr, warn, err);
        }
        UnmapViewOfFile(mmap_data);
        CloseHandle(hMapping);
        CloseHandle(hFile);
        return true;
      }
#else  // POSIX mmap
      {
        int fd = open(filepath.c_str(), O_RDONLY);
        if (fd == -1) continue;
        struct stat sb;
        if (fstat(fd, &sb) != 0) {
          close(fd);
          continue;
        }
        const size_t file_size = static_cast<size_t>(sb.st_size);
        if (file_size == 0) {
          close(fd);
          StreamReader empty_sr("", 0);
          LoadMtlInternal(matMap, materials, empty_sr, warn, err);
          return true;
        }
        void *mapped = mmap(NULL, file_size, PROT_READ, MAP_PRIVATE, fd, 0);
        close(fd);
        if (mapped == MAP_FAILED) continue;
        const char *mmap_data = static_cast<const char *>(mapped);
        {
          StreamReader sr(mmap_data, file_size);
          LoadMtlInternal(matMap, materials, sr, warn, err);
        }
        munmap(mapped, file_size);
        return true;
      }
#endif  // _WIN32
#else   // !TINYOBJLOADER_USE_MMAP
      std::ifstream matIStream(filepath.c_str());
      if (matIStream) {
        LoadMtl(matMap, materials, &matIStream, warn, err);

        return true;
      }
#endif  // TINYOBJLOADER_USE_MMAP
    }

    std::stringstream ss;
    ss << "Material file [ " << matId
       << " ] not found in a path : " << m_mtlBaseDir << "\n";
    if (warn) {
      (*warn) += ss.str();
    }
    return false;

  } else {
    std::string filepath = matId;

#ifdef TINYOBJLOADER_USE_MMAP
#if defined(_WIN32)
    {
      HANDLE hFile =
          CreateFileA(filepath.c_str(), GENERIC_READ, FILE_SHARE_READ, NULL,
                      OPEN_EXISTING, FILE_ATTRIBUTE_NORMAL, NULL);
      if (hFile != INVALID_HANDLE_VALUE) {
        LARGE_INTEGER fileSize;
        if (GetFileSizeEx(hFile, &fileSize)) {
          const size_t file_size = static_cast<size_t>(fileSize.QuadPart);
          if (file_size == 0) {
            CloseHandle(hFile);
            StreamReader empty_sr("", 0);
            LoadMtlInternal(matMap, materials, empty_sr, warn, err);
            return true;
          }
          HANDLE hMapping =
              CreateFileMappingA(hFile, NULL, PAGE_READONLY, 0, 0, NULL);
          if (hMapping != NULL) {
            const char *mmap_data = static_cast<const char *>(
                MapViewOfFile(hMapping, FILE_MAP_READ, 0, 0, 0));
            if (mmap_data) {
              {
                StreamReader sr(mmap_data, file_size);
                LoadMtlInternal(matMap, materials, sr, warn, err);
              }
              UnmapViewOfFile(mmap_data);
              CloseHandle(hMapping);
              CloseHandle(hFile);
              return true;
            }
            CloseHandle(hMapping);
          }
        }
        CloseHandle(hFile);
      }
    }
#else  // POSIX mmap
    {
      int fd = open(filepath.c_str(), O_RDONLY);
      if (fd != -1) {
        struct stat sb;
        if (fstat(fd, &sb) == 0) {
          const size_t file_size = static_cast<size_t>(sb.st_size);
          if (file_size == 0) {
            close(fd);
            StreamReader empty_sr("", 0);
            LoadMtlInternal(matMap, materials, empty_sr, warn, err);
            return true;
          }
          void *mapped = mmap(NULL, file_size, PROT_READ, MAP_PRIVATE, fd, 0);
          close(fd);
          if (mapped != MAP_FAILED) {
            const char *mmap_data = static_cast<const char *>(mapped);
            {
              StreamReader sr(mmap_data, file_size);
              LoadMtlInternal(matMap, materials, sr, warn, err);
            }
            munmap(mapped, file_size);
            return true;
          }
        } else {
          close(fd);
        }
      }
    }
#endif  // _WIN32
#else   // !TINYOBJLOADER_USE_MMAP
    std::ifstream matIStream(filepath.c_str());
    if (matIStream) {
      LoadMtl(matMap, materials, &matIStream, warn, err);

      return true;
    }
#endif  // TINYOBJLOADER_USE_MMAP

    std::stringstream ss;
    ss << "Material file [ " << filepath
       << " ] not found in a path : " << m_mtlBaseDir << "\n";
    if (warn) {
      (*warn) += ss.str();
    }

    return false;
  }
}

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

tiny_obj_loader.h

tests/tester.cc

tiny_obj_loader.h

syoyo · 2026-03-04T00:24:34Z

@copilot resolve conflicts nicely. basically accept both changes.

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

…ent, BOM in callback, test fixes - Add UTF8ToWchar/LongPathW helpers for Windows long path and Unicode support - Add Windows long path support (#ifdef _WIN32) in non-mmap ifstream fallbacks - Add vt_w vector and optional w component parsing in LoadObjInternal - Fix texcoord_ws.swap(vt) -> texcoord_ws.swap(vt_w) corruption - Add BOM stripping to LoadObjWithCallbackInternal - Fix invalid-relative-texture-vertex-index.obj -> invalid-relative-texture-index.obj - Add test models: texcoord-w.obj, texcoord-w-mixed.obj, utf8-path-test.obj/.mtl - Add tests: texcoord_w, texcoord_w_mixed, utf8_path, long_path, callback_BOM Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot · 2026-03-04T00:37:02Z

@copilot resolve conflicts nicely. basically accept both changes.

All changes from both release commits (#414, #415) have been manually integrated into our StreamReader-based branch in 5ef0c48:

Windows long paths (Add Windows long path and Unicode file path support for .obj/.mtl loading #415): UTF8ToWchar/LongPathW helpers + #ifdef _WIN32 in all non-mmap std::ifstream fallback paths
texcoord w-component (Fix texcoord_ws corruption, alpha_texname double-write, missing BOM strip in callback, and wrong test fixture path #414): vt_w vector + sr_parseReal for optional w + texcoord_ws.swap(vt_w) fix
BOM in callback (Fix texcoord_ws corruption, alpha_texname double-write, missing BOM strip in callback, and wrong test fixture path #414): BOM check added to LoadObjWithCallbackInternal
alpha_texname double-write (Fix texcoord_ws corruption, alpha_texname double-write, missing BOM strip in callback, and wrong test fixture path #414): already correct in our StreamReader version
Test fix (Fix texcoord_ws corruption, alpha_texname double-write, missing BOM strip in callback, and wrong test fixture path #414): invalid-relative-texture-vertex-index.obj → invalid-relative-texture-index.obj
All new test models and test functions added

All 51 tests pass. However, since I can't do git merge + git push, the git-level conflicts still show on GitHub. You'll need to merge release into this branch to resolve them — the content should match on both sides now.

…lease Keep StreamReader-based parsing from the harden branch for all conflicting regions: map_d parsing, mmap material loading, LoadObjInternal, texcoord w-component, BOM handling, and main parse loop. Retain mmap/memory-buffer tests in tester.cc. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

- Add column tracking (col_num_) to StreamReader, updated in all position-advancing methods (get, advance, skip_space, etc.) - Add current_line_text(), format_error(filename, msg) for clang-style diagnostics (file:line:col with source line and caret) - Add error stack to StreamReader (push_error, has_errors, get_errors) - Add error-reporting sr_parse overloads that return bool for success/failure: sr_parseInt, sr_parseReal, sr_parseReal2/3, sr_parseVertexWithColor (returns -1 on error) - Wire error-reporting overloads into LoadMtlInternal (Ka, Kd, Ks, Ni, Ns, illum, d, Tr, PBR properties) with early return on parse failure; change LoadMtlInternal from void to bool - Wire error-reporting overloads into LoadObjInternal (v, vn, vt, vw) with early return on parse failure - Upgrade existing error messages (face, line, points, skin weight) to use sr.format_error() for consistent clang-style output - Pass filename through call chain: LoadObj -> LoadObjInternal, MaterialFileReader -> LoadMtlInternal - Add warning_context.filename for fixIndex warnings - Add 6 new tests: column tracking, format_error output, error stack, malformed vertex/mtl errors, backward compatibility Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

- Move warning_context initialization (warn, filename) before the main parse loop in LoadObjInternal; only update line_number per iteration. Avoids a std::string copy of filename on every face/line/points line. - Replace test_parse_error_backward_compat (was just TEST_CHECK(true)) with an actual round-trip test: parse valid OBJ, verify no errors and correct vertex count. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

syoyo · 2026-03-04T16:00:09Z

@copilot code review[agent] Review changes with GPT-5.3-Codex

…m naming, extract sr_skipTagField helper, refactor mmap to RAII MappedFile Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

…erals Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot · 2026-03-05T20:20:04Z

@copilot work on Low severity items.

Addressed in ea9bf10 and b62d3ba.

Changes made for each low severity item:

Cppcheck suggested performance improvement #7 (read_token undocumented): Added doc comment marking it as a utility/test method.
Some small additions... #9 (line_no vs line_num): Renamed line_no → line_num in LoadMtlInternal to match LoadObjInternal.
added clearing of unknown parameters map #6 (duplicate advance loop in sr_parseTagTriple): Extracted into sr_skipTagField() helper.
add missing includes #10 (mmap code duplication): Introduced RAII MappedFile struct with open()/close()/destructor, replacing ~250 lines of duplicated platform-specific mmap logic with ~5 lines per call site. Uses an is_mapped flag to safely distinguish empty-file string literals from actual mapped regions.
Added a Cmake build system configuration #8 (callback path non-error overload): Left as-is — the callback API is intentionally fire-and-forget per the existing contract, changing it could break user code that relies on the callback path not halting on malformed data.
set default material #11 (ParseTextureNameAndOption legacy): Noted for future refactoring, no change.

All 51 tests pass (both standard and TINYOBJLOADER_USE_MMAP modes). CodeQL: 0 alerts.

syoyo · 2026-03-05T21:13:56Z

@copilot run comprehensive review again, now GPT-5.3-Codex

Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot · 2026-03-05T21:18:24Z

@copilot run comprehensive review again, now GPT-5.3-Codex

Ran a comprehensive GPT-5.3-Codex review on 496cd28.

Findings:

Medium: StreamReader(std::istream&) still forces parsing from stream start (seekg(0, beg)), so non-zero stream offsets are ignored (tiny_obj_loader.h:786-809).
Medium: stream-based load paths buffer entire input into memory, which can amplify memory usage on very large/untrusted streams (tiny_obj_loader.h:786-809, used by LoadObj/callback/material stream paths).
Low: missing regression tests for non-zero stream offsets and very large stream behavior (tests/tester.cc:1850-2045).

Also checked recent Actions runs/logs for this branch; latest Unit Tests run is passing, with no failed jobs in recent action_required runs.

No UI changes in this PR (no screenshot).

syoyo · 2026-03-05T21:55:34Z

@copilot run comprehensive review again, now GPT-5.3-Codex

Ran a comprehensive GPT-5.3-Codex review on 496cd28.

Findings:

Medium: StreamReader(std::istream&) still forces parsing from stream start (seekg(0, beg)), so non-zero stream offsets are ignored (tiny_obj_loader.h:786-809).

Medium: stream-based load paths buffer entire input into memory, which can amplify memory usage on very large/untrusted streams (tiny_obj_loader.h:786-809, used by LoadObj/callback/material stream paths).

Low: missing regression tests for non-zero stream offsets and very large stream behavior (tests/tester.cc:1850-2045).

@copilot do all these

Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot · 2026-03-05T22:02:24Z

@copilot run comprehensive review again, now GPT-5.3-Codex
...

Implemented in d5aac19.

Done all three items:

StreamReader(std::istream&) now reads from the current stream position (no forced rewind to beginning).
Added bounded stream-read protection (TINYOBJLOADER_STREAM_READER_MAX_BYTES) and error propagation so oversized stream inputs fail cleanly instead of attempting unbounded buffering.
Added regression tests:
- test_stream_load_from_current_offset
- test_stream_load_rejects_oversized_input

Validation:

Targeted tests for the new cases passed.
Full make check passed.
CodeQL: 0 alerts.

No UI changes in this PR (no screenshot).

syoyo · 2026-03-06T01:40:57Z

@copilot deeply review. make this time final round.

Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot · 2026-03-06T01:48:37Z

@copilot deeply review. make this time final round.

Final deep review completed and fixes applied in 9b6d594.

I found and fixed a high-severity callback/material-loading issue:

MaterialFileReader / MaterialStreamReader now correctly propagate LoadMtlInternal failure instead of always returning success.
LoadObjWithCallbackInternal now guards mtllib_cb with !materials.empty() to avoid materials.at(0) out_of_range on failed/empty material loads.
Added regression test test_loadObjWithCallback_mtllib_failure_does_not_crash.

Validation:

Targeted tests passed.
Full make check passed.
Code review: no comments.
CodeQL: 0 alerts.

No UI changes in this PR (no screenshot).

syoyo · 2026-03-06T02:02:50Z

@copilot deeper review again. final stage 2. maximum effort. ultrathink.

… parity Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot · 2026-03-06T02:13:26Z

@copilot deeper review again. final stage 2. maximum effort. ultrathink.

Completed stage-2 deep review and applied fixes in 14e6a5a.

Key findings and fixes:

Fixed mtllib whitespace-only handling: empty tokens are now filtered before processing, so whitespace-only mtllib no longer triggers material loading attempts.
Added regression tests:
- test_mtllib_empty_filename_is_ignored_loadobj
- test_mtllib_empty_filename_is_ignored_callback
Enforced TINYOBJLOADER_STREAM_READER_MAX_BYTES on mmap-backed OBJ/MTL paths for parity with istream size limits.

Validation:

Targeted tests passed.
Full make check passed.
Code review run completed.
CodeQL: 0 alerts.
Recent Actions runs on this branch show action_required with no failed jobs/log failures returned.

No UI changes in this PR (no screenshot).

syoyo · 2026-03-06T02:15:57Z

@copilot further deep code review. final stage 3

…ze check Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot · 2026-03-06T02:23:45Z

@copilot further deep code review. final stage 3

Completed final stage-3 deep review and pushed fixes in 07d2c25.

Applied improvements:

Fixed SplitString so non-escape backslashes are preserved (prevents mtllib path corruption like subdir\file.mtl).
Hardened stream chunk-limit check against size_t overflow in StreamReader(std::istream&).
Added regression test: test_split_string_preserves_non_escape_backslash.

Validation:

Baseline and post-change full make check passed.
Targeted tests passed.
Code review: no comments.
CodeQL: 0 alerts.
Recent Actions runs on this branch still show action_required with no failed jobs/log failures returned.

No UI changes in this PR (no screenshot).

Initial plan

5f9f7f6

Copilot AI assigned Copilot and syoyo Mar 3, 2026

Copilot started work on behalf of syoyo March 3, 2026 17:05 View session

Add max_iter guards to parsing loops and mmap support for .obj/.mtl l…

5df0146

…oading Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot AI changed the title ~~[WIP] Make .obj/.mtl parsing robust and efficient~~ Harden OBJ/MTL parsing and add mmap support for large file I/O Mar 3, 2026

Copilot finished work on behalf of syoyo March 3, 2026 17:21

Copilot started work on behalf of syoyo March 3, 2026 17:46 View session

Copilot AI and others added 2 commits March 3, 2026 18:01

Copilot AI changed the title ~~Harden OBJ/MTL parsing and add mmap support for large file I/O~~ Rewrite OBJ/MTL parser to byte-stream based with StreamReader and add mmap support Mar 3, 2026

Copilot finished work on behalf of syoyo March 3, 2026 18:11

Merge remote-tracking branch 'origin/release' into copilot/harden-obj…

70ba4b3

…-mtl-parsing

syoyo marked this pull request as ready for review March 3, 2026 23:44

Copilot AI review requested due to automatic review settings March 3, 2026 23:44

Copilot started reviewing on behalf of syoyo March 3, 2026 23:44 View session

Copilot AI reviewed Mar 3, 2026

View reviewed changes

tiny_obj_loader.h Outdated Show resolved Hide resolved

tiny_obj_loader.h Outdated Show resolved Hide resolved

tests/tester.cc Outdated Show resolved Hide resolved

tiny_obj_loader.h Outdated Show resolved Hide resolved

tiny_obj_loader.h Show resolved Hide resolved

Copilot started work on behalf of syoyo March 4, 2026 00:24 View session

syoyo and others added 2 commits March 4, 2026 09:31

Update tiny_obj_loader.h

07ba329

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Copilot AI changed the title ~~Rewrite OBJ/MTL parser to byte-stream based with StreamReader and add mmap support~~ Rewrite OBJ/MTL parser to byte-stream based with StreamReader, add mmap support, and integrate Windows long path/Unicode support Mar 4, 2026

Copilot finished work on behalf of syoyo March 4, 2026 00:37

syoyo and others added 3 commits March 4, 2026 09:47

Copilot AI and others added 2 commits March 5, 2026 20:16

Address low severity review items: document read_token, unify line_nu…

ea9bf10

…m naming, extract sr_skipTagField helper, refactor mmap to RAII MappedFile Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Harden MappedFile with is_mapped flag to prevent unmapping string lit…

b62d3ba

…erals Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot finished work on behalf of syoyo March 5, 2026 20:21

Copilot started work on behalf of syoyo March 5, 2026 21:14 View session

Copilot AI and others added 2 commits March 5, 2026 21:17

Run GPT-5.3-Codex comprehensive review

7dde60a

Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Remove accidental CTest artifact and ignore Testing/Temporary

496cd28

Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot finished work on behalf of syoyo March 5, 2026 21:18

Copilot started work on behalf of syoyo March 5, 2026 21:55 View session

Fix stream offset handling and add oversized stream guards/tests

d5aac19

Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot finished work on behalf of syoyo March 5, 2026 22:02

Copilot started work on behalf of syoyo March 6, 2026 01:41 View session

Final deep-review fixes for mtllib callback failure path

9b6d594

Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot finished work on behalf of syoyo March 6, 2026 01:49

Copilot started work on behalf of syoyo March 6, 2026 02:03 View session

Stage-2 deep review fixes for mtllib empty tokens and mmap size guard…

14e6a5a

… parity Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot finished work on behalf of syoyo March 6, 2026 02:13

Copilot started work on behalf of syoyo March 6, 2026 02:16 View session

Stage-3 deep review: preserve mtllib backslashes and harden stream si…

07d2c25

…ze check Co-authored-by: syoyo <18676+syoyo@users.noreply.github.com>

Copilot finished work on behalf of syoyo March 6, 2026 02:24

Uh oh!

Conversation

Copilot AI commented Mar 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

syoyo commented Mar 3, 2026

Uh oh!

Copilot AI commented Mar 3, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

syoyo commented Mar 4, 2026

Uh oh!

Copilot AI commented Mar 4, 2026

Uh oh!

syoyo commented Mar 4, 2026

Uh oh!

Copilot AI commented Mar 5, 2026

Uh oh!

syoyo commented Mar 5, 2026

Uh oh!

Copilot AI commented Mar 5, 2026

Uh oh!

syoyo commented Mar 5, 2026

Uh oh!

Copilot AI commented Mar 5, 2026

Uh oh!

syoyo commented Mar 6, 2026

Uh oh!

Copilot AI commented Mar 6, 2026

Uh oh!

syoyo commented Mar 6, 2026

Uh oh!

Copilot AI commented Mar 6, 2026

Uh oh!

syoyo commented Mar 6, 2026

Uh oh!

Copilot AI commented Mar 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Copilot AI commented Mar 3, 2026 •

edited

Loading