Repository navigation
Add opt-in incremental bounded HEIC decoding - #12
Merged
laurenspriem merged 1 commit intoOct 1, 2026
Merged
Conversation
laurens-pilot
force-pushed
the
feat/incremental-bounded-decode
branch
from
October 1, 2026 05:40
d26ac40 to
37cf033
Compare
Author
|
@codex review |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Add feature-gated incremental reconstruction for supported opaque 8-bit 4:2:0 HEVC still images, including non-grid images and grids with oversized tiles. Complete prediction and filtering in rolling bands before conversion and downsampling, so source raster storage grows with width rather than total pixels. Existing APIs and production callers retain their behavior.
A 199.8 MP photograph re-encoded as a single HEIC item produces 6000×4500 output with 103.3 MiB measured peak heap under a 128 MiB budget; all 299,635,200 reconstructed YUV samples match FFmpeg. This is host evidence for the supported subset, not native camera compatibility or a universal RSS guarantee. WPP, alpha, higher bit depths and other coding structures remain follow-up work.
Keep native reference samples exact and allow at most one RGB8 code value for display rounding. Tests use the existing external libheif/stress corpus and locally generated regression variants in ignored directories, with live strict libheif references and allocation ceilings. No new binary fixtures or golden PNGs are tracked. The existing exact normal-decoder comparison policy stays intact.
An additional high-frequency odd-crop probe has one green sample differing by two values despite exact native YUV agreement. The strict one-value RGB gate remains unchanged; that display-conversion discrepancy is recorded for the compatibility follow-up and is not counted as a passing comparison.
Source fork PR: laurens-pilot#2
@laurenspriem