PDF stage 4.8: image masks, stencil masks, soft masks#572
Merged
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1299bea6ef
ℹ️ About Codex in GitHub
Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".
Render the three PDF image-transparency mechanisms (ISO 32000-1 8.9.6 / 11.6.5.2): - `/ImageMask true` stencils (image XObject + inline `BI…ID…EI`) are painted in the *current fill colour*, which is known only at `Do` time — so the decoded 1-bpc bitmap is stashed on the XObject and the page extractor recolours it to an RGBA PNG (`encode_stencil_png`) with the graphics-state fill colour. - `/SMask` (alpha) and `/Mask` (a stencil sub-image or a colour-key array) are composited into an RGBA PNG on the raster path at parse time: `resolve_mask_alpha` decodes a sub-image into a base-sized coverage plane (`decode_mask_alpha`, nearest-neighbour resampled), and `encode_image` / `encode_image_png` emit RGBA when alpha or a colour key is present. Adds `write_png_rgba`. A mask on a JPEG (`DCTDecode`) base is ignored — decoding the JPEG to composite is out of scope. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The two PNG writers differed only in the stride multiplier (3 vs 4) and the IHDR colour-type byte (2 vs 6). Collapse them into a single `write_png(pixels, width, height, channels)` taking the channel count, removing the near-duplicated container-emission code. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Uqx1dUpDUykRxui8FYwxvM
Align pdf_page_extractor and pdf_document_parser with the module's prevailing convention (const on by-value scalar definition params). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Uqx1dUpDUykRxui8FYwxvM
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
🤖 Generated with Claude Code
Stage 4.8 of the PDF graphics work (
src/odr/internal/pdf/STAGE4_PLAN.md): render the three PDF image-transparency mechanisms (ISO 32000-1 8.9.6 / 11.6.5.2).What
/ImageMask truestencils (image XObject + inlineBI…ID…EI) are painted in the current fill colour, which is known only atDotime. The decoded 1-bpc bitmap + geometry are stashed on theXObjectat parse time (parse_stencil_mask); the page extractor recolours it to an RGBA PNG (encode_stencil_png) using the graphics-state fill colour./SMask(alpha) and/Mask(a stencil sub-image or a colour-key array) are composited into an RGBA PNG on the raster path at parse time.resolve_mask_alphadecodes a sub-image into a base-sized coverage plane (decode_mask_alpha, nearest-neighbour resampled if the mask differs in size);encode_image/encode_image_pngemit RGBA when alpha or a colour key is present.write_png_rgba(PNG colour type 6) alongside the existing RGB writer.Limitations (documented)
A
/SMaskor/Maskon a JPEG (DCTDecode) base is ignored — decoding the JPEG to composite alpha is out of scope; the opaque JPEG still renders.Tests
pdf_image:write_png_rgbaround-trip; alpha-plane and colour-key → RGBA;encode_stencil_png(paint colour +/Decodeinversion);decode_mask_alpha(soft-mask grey→alpha, stencil set-bits, resampling).pdf_page_extractor: stencil image XObject and inline image mask both emit a recoloured RGBA PNGImageElement.All 186 PDF unit tests pass. Reference-output regeneration is left to the maintainer.