fix(jpeg2000): decode subsampled HTJ2K components correctly by lgritz · Pull Request #5643 · AcademySoftwareFoundation/OpenImageIO

lgritz · 2026-11-02T08:36:20Z

The OpenJPH (HTJ2K) path assumed every component is full resolution, pulling one full-width line per component for each image row. Subsampled components cause it to fall out of step, and OpenJPH spins indefinitely in codestream::pull(). Found by our nightly fuzzing with corrupt headers, but valid 4:2:0 files hang the same way.

Each component is now pulled at its own reconstructed size, and each sample is replicated over the full-resolution pixels it covers (the same nearest-sample mapping as the OpenJPEG path, but honoring the image offset). The image size now comes from the image extent rather than from component 0, which is smaller when it is subsampled.

Color-transformed files stay in OpenJPH's interleaved order, which needs equal subsampling, so unequal ones are rejected. Empty components, which also hung pull(), are rejected too.

Additionally, when tiles are smaller than a component's subsampling, the component has no samples in some tiles. OpenJPH 0.24 crashes pulling such a component (0.33 copes), so reject these files in the header check.

Assisted-by: Claude Code / Claude Opus 5.5

Copilot

🟡 Changes recommended

The decoder can corrupt color-transformed images and hang on zero-height components, while offset and extent calculations lack targeted regression coverage.

3 open issues

What changed in this PR

This PR updates OpenImageIO’s OpenJPH decoder to handle subsampled HTJ2K components without hanging and reconstruct them at full image resolution.

Changes:

Decode subsampled components in planar order and expand samples using offset-aware mapping.
Derive image dimensions from the image extent rather than component 0.
Add subsampling regression cases and document the decoder behavior.

File	Description
`testsuite/htj2k/run.py`	Adds valid subsampling cases and a corrupt-file hang regression.
`testsuite/htj2k/ref/out.txt`	Records expected dimensions, pixel hashes, and completion output.
`src/jpeg2000.imageio/jpeg2000input.cpp`	Implements per-component decoding, sample expansion, and canvas sizing.
`.agents/skills/oiio-input-security-audit/references/format-notes.md`	Documents the subsampling fix and OpenJPEG fallback limitations.
`.agents/skills/oiio-input-security-audit/references/bug-catalog.md`	Records the component-size mismatch and its decoding consequences.

_{Review effort: Balanced}

The OpenJPH (HTJ2K) path assumed every component is full resolution, pulling one full-width line per component for each image row. Subsampled components cause it to fall out of step, and OpenJPH spins indefinitely in codestream::pull(). Found by our nightly fuzzing with corrupt headers, but valid 4:2:0 files hang the same way. Each component is now pulled at its own reconstructed size, and each sample is replicated over the full-resolution pixels it covers (the same nearest-sample mapping as the OpenJPEG path, but honoring the image offset). The image size now comes from the image extent rather than from component 0, which is smaller when it is subsampled. Color-transformed files stay in OpenJPH's interleaved order, which needs equal subsampling, so unequal ones are rejected. Empty components, which also hung pull(), are rejected too. Assisted-by: Claude Code / Claude Opus 5.5 Signed-off-by: Larry Gritz <lg@larrygritz.com>

Keep OpenJPH's pull order (interleaved for color-transformed files), reject empty components, add offset/MCT/empty-component tests. Assisted-by: Claude Code / Claude Opus 5.5 Signed-off-by: Larry Gritz <lg@larrygritz.com>

lgritz · 2026-11-07T13:46:41Z

Addressed the copilot comments and pushed an amentment.

Copilot

🟡 Changes recommended

The column-map allocation can fail outside the scanline error handler, and the regression test can pass even when the decoder crashes.

2 open issues

3 resolved since last review

_{Review effort: Balanced}

Claude says: Report column-map allocation failure as a read error, and let the hang regression test fail if the decoder crashes. Assisted-by: Claude Code / Claude Opus 5.5 Signed-off-by: Larry Gritz <lg@larrygritz.com>

Claude says: When tiles are smaller than a component's subsampling, the component has no samples in some tiles. OpenJPH 0.24 crashes pulling such a component (0.33 copes), so reject these files in the header check. The hang regression test now checks the reader's error message rather than relying on a bare exit status, so a crash fails the test. Assisted-by: Claude Code / Claude Opus 5.5 Signed-off-by: Larry Gritz <lg@larrygritz.com>

Copilot

🟡 Changes recommended

The tile-layout check incorrectly rejects valid nonempty layouts with more than two tiles when the tile size is smaller than the subsampling factor.

1 open issue

Rejects valid layouts when tile size is smaller than subsampling · New

2 resolved since last review

_{Review effort: Balanced}

Claude says: The previous check rejected every layout with interior tiles smaller than the subsampling, though some have a sample in every tile. Such tiles hold at most one sample each, so compare sample and tile counts. A small valid file with such tiles is added to the htj2k test. Assisted-by: Claude Code / Claude Opus 5.5 Signed-off-by: Larry Gritz <lg@larrygritz.com>

…ures Claude says: Only a tile-component with no columns crashes OpenJPH, and only before 0.27.0, where tile::pull gained a zero-width guard. Tile rows without samples are skipped by every version. The per-tile check is now x-only and compiled only against older OpenJPH, with a message that no longer calls the component empty. Newer builds decode such files. All valid subsampled fixtures now come from a committed generator, make_subsampled_htj2k.cpp, coded losslessly from a known pattern so the expected pixels can be checked independently of the reader; the empty-component fixture is patched by make_malformed_htj2k.py. New fixtures cover 10-bit, tiles with image and tile offsets, non-power-of-2 factors, a color transform with tiles, four components, and tiles holding no rows or no columns of a component. The version-dependent outcomes have their own output file and refs. Assisted-by: Claude Code / Claude Fable 5.1 Signed-off-by: Larry Gritz <lg@larrygritz.com>

Copilot

🟡 Changes recommended

The regressions lack coverage for MCT files with five or more components, and the corrupt-fixture test can mask decoder crashes by hiding the exit status and forcing success.

2 open issues

1 resolved since last review

Rejects valid layouts when tile size is smaller than subsampling

_{Review effort: Balanced}

Claude says: Add a five-component color-transformed fixture, so the interleaved decode must pull a row of every component rather than just the four it keeps. On OpenJPH >= 0.27 the corrupt hang fixture is now decoded with the exit status enforced, with its log captured and filtered in a separate step, so a crash fails the test instead of leaving an empty output that matches the reference. Assisted-by: Claude Code / Claude Fable 5.1 Signed-off-by: Larry Gritz <lg@larrygritz.com>

Copilot

🟢 Approval recommended

No active review issues are reported for the subsampled HTJ2K decoding fix.

No open issues 🎉

2 resolved since last review

_{Review effort: Balanced}

lgritz force-pushed the lg-j2k-subsampling branch 2 times, most recently from 6dbae55 to b0b387c Compare November 6, 2026 06:04

lgritz requested a balanced review from Copilot November 7, 2026 08:04

Copilot started reviewing on behalf of lgritz November 7, 2026 08:04

View session

Copilot AI reviewed Nov 7, 2026

View reviewed changes

Comment thread src/jpeg2000.imageio/jpeg2000input.cpp Resolved

Comment thread src/jpeg2000.imageio/jpeg2000input.cpp Resolved

Comment thread src/jpeg2000.imageio/jpeg2000input.cpp Resolved

lgritz added 2 commits November 7, 2026 22:34

fix(jpeg2000): address review of subsampled HTJ2K decoding

7a70cbf

Keep OpenJPH's pull order (interleaved for color-transformed files), reject empty components, add offset/MCT/empty-component tests. Assisted-by: Claude Code / Claude Opus 5.5 Signed-off-by: Larry Gritz <lg@larrygritz.com>

lgritz force-pushed the lg-j2k-subsampling branch from b0b387c to 7a70cbf Compare November 7, 2026 13:35

lgritz requested a balanced review from Copilot November 7, 2026 13:47

Copilot started reviewing on behalf of lgritz November 7, 2026 13:48

View session

Copilot AI reviewed Nov 7, 2026

View reviewed changes

Comment thread src/jpeg2000.imageio/jpeg2000input.cpp Resolved

Comment thread testsuite/htj2k/run.py Resolved

lgritz added 2 commits November 8, 2026 00:21

fix(jpeg2000): address second review of subsampled HTJ2K decoding

a3b1309

Claude says: Report column-map allocation failure as a read error, and let the hang regression test fail if the decoder crashes. Assisted-by: Claude Code / Claude Opus 5.5 Signed-off-by: Larry Gritz <lg@larrygritz.com>

lgritz requested a balanced review from Copilot November 8, 2026 01:32

Copilot started reviewing on behalf of lgritz November 8, 2026 01:33

View session

Copilot AI reviewed Nov 8, 2026

View reviewed changes

Comment thread src/jpeg2000.imageio/jpeg2000input.cpp Resolved

lgritz added 2 commits November 9, 2026 17:25

lgritz requested a balanced review from Copilot November 9, 2026 09:31

Copilot started reviewing on behalf of lgritz November 9, 2026 09:32

View session

Copilot AI reviewed Nov 9, 2026

View reviewed changes

Comment thread src/jpeg2000.imageio/jpeg2000input.cpp Resolved

Comment thread testsuite/htj2k/run.py Resolved

lgritz requested a balanced review from Copilot November 9, 2026 11:58

Copilot started reviewing on behalf of lgritz November 9, 2026 11:59

View session

Copilot AI reviewed Nov 9, 2026

View reviewed changes

Navigation Menu

Repository navigation

fix(jpeg2000): decode subsampled HTJ2K components correctly - #5643

fix(jpeg2000): decode subsampled HTJ2K components correctly#5643
lgritz wants to merge 7 commits into
AcademySoftwareFoundation:mainfrom
lgritz:lg-j2k-subsampling

lgritz commented Nov 2, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

lgritz commented Nov 7, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

lgritz commented Nov 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟡 Changes recommended

Uh oh!

Uh oh!

Uh oh!

Uh oh!

lgritz commented Nov 7, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟡 Changes recommended

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟡 Changes recommended

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟡 Changes recommended

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟢 Approval recommended

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

lgritz commented Nov 2, 2026 •

edited

Loading