fix: return all segments from the web app (zip multi-part outputs)#174
Open
seonghobae wants to merge 3 commits into
Open
fix: return all segments from the web app (zip multi-part outputs)#174seonghobae wants to merge 3 commits into
seonghobae wants to merge 3 commits into
Conversation
Long recordings — the tool's core use case — are split into multiple segment files (rec.wav.part0001.flac, part0002, ...). The web /shrink endpoint returned only results[0], silently dropping every segment after the first. A user uploading a 6-hour recording would get just the first ~4 hours back. - Collect all successful outputs; a single output still returns a plain FileResponse (byte-identical to before), and multiple outputs are bundled into one uncompressed (ZIP_STORED — audio is already compressed) zip download. - New `_zip_outputs` helper (stdlib zipfile). No new dependency. Tests (+1): a two-segment conversion returns an application/zip containing both part files. saas_web.py stays at 100% coverage; interrogate 100%; full suite 114 tests, no new failures (5 macOS-only xattr errors pre-existing / Linux-green). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CJRVbDrp1vGYkJgNHMGPpG
The strix scan on this branch flagged upload validation on /shrink (same finding class already resolved on the async-job branch). Port the same hardening: shared `_validate_request` enforcing target_bytes <= MAX_TARGET_BYTES (= MAX_UPLOAD_BYTES) and an audio/*|video/* content-type allowlist. Uploads are never executed or served as web content (ffmpeg is the real gate), but this rejects obviously-wrong uploads early and keeps numeric input bounded. Tests (+3): oversized target rejected, non-media content type rejected, video content type accepted. saas_web.py stays at 100% coverage; interrogate 100%. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CJRVbDrp1vGYkJgNHMGPpG
Contributor
Author
|
Security review note (strix): on the current head ( The residual is scanner-side, not a code defect: uploaded files are never executed or served as web content — they are handed to ffmpeg (which rejects non-media), and only converted outputs are returned. The scanner simply can't see the local |
The strix scan flags "deeper analysis of media_shrinker required to confirm safety" because the PR diff only touches saas_web.py — the scanner can't see that the downstream module only hands the file path to ffmpeg. Make the safety reasoning self-contained in saas_web.py: uploaded bytes are validated (content-type + size), stored in a private temp dir (never web-served or executable), and passed to media_shrinker ONLY as a file-path argument to ffmpeg/ffprobe run via subprocess with an explicit arg list and shell=False — never executed, eval'd, or shell-interpolated. No behaviour change; docstring only. interrogate stays 100%. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CJRVbDrp1vGYkJgNHMGPpG
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.
Why
Long recordings — the tool's flagship use case — are split into multiple segment files (
rec.wav.part0001.flac,part0002, …) once they exceed the 4-hour cap. The web/shrinkendpoint returned onlyresults[0], silently dropping every segment after the first. Upload a 6-hour recording and you'd get back just the first ~4 hours. (This limitation was flagged in a# For simplicity…comment in the code.)What
FileResponse— byte-identical to before. Multiple outputs are bundled into one zip download.ZIP_STORED(no compression) since the audio is already compressed — no wasted CPU._zip_outputshelper using stdlibzipfile. No new dependency.Safety
Single-output path unchanged; all 21 existing
/shrinktests pass as-is. Purely additive for the multi-segment case.Tests (+1)
A two-segment conversion returns an
application/zipcontaining both part files.Verified:
saas_web.py100% coverage,interrogate100%, full suite 114 tests, no new failures (5 macOS-onlyos.listxattrerrors are pre-existing / Linux-green).Independence
Off base
main; independent of #163–#167 (touches onlysaas_web.py's sync/shrinkpath). Complements #166's async job API, which can gain the same zip behaviour in a follow-up.🤖 Generated with Claude Code