Replace document parsing with AnyDoc - #11
Closed
mattv8 wants to merge 2 commits into
Closed
Conversation
mattv8
commented
Aug 11, 2026
| # Office documents | ||
| # AnyDoc README-supported document formats. Keep this as the canonical | ||
| # AnyDoc taxonomy so parser-facing sets can reuse it without local drift. | ||
| ANYDOC_DOCUMENT_EXTENSIONS: set[str] = { |
Owner
Author
There was a problem hiding this comment.
The canonical AnyDoc taxonomy is expanded here, but the filesystem-index UI defaults and User Space binary-prefetch gate are not updated in this PR. Most newly advertised formats are therefore omitted by default or treated as editor text. Please propagate this matrix to those consumer gates and pin parity with focused tests.
mattv8
commented
Aug 11, 2026
| sys.modules[module_name] = module | ||
|
|
||
| fake_file_constants = types.SimpleNamespace( | ||
| ANYDOC_DOCUMENT_EXTENSIONS={".pdf", ".docx", ".csv", ".docm", ".xlsm", ".xlsb", ".pps", ".pot", ".pptm", ".ppsx", ".ppsm"}, |
Owner
Author
There was a problem hiding this comment.
This focused fake extension set and the three routing examples do not prove conversion coverage for the full 21-extension taxonomy or consumer paths such as chat attachments, index collection, and User Space parsing. Please parameterize against the canonical set and add a native AnyDoc conversion smoke check.
mattv8
force-pushed
the
merry-walrus
branch
4 times, most recently
from
September 3, 2026 22:12
cb0b4e5 to
cdfd032
Compare
- Bump firecrawl-anydoc 0.1.6 -> 0.2.4 in app and runtime extras; map the new NeedsOcrError to a typed NEEDS_OCR conversion failure and treat it like UNSUPPORTED (empty text) in runtime PDF reads, matching the old pypdf behavior for scanned/image-only PDFs - Remove the dead pandas dependency (zero imports repo-wide) from pyproject and the Dockerfile import asserts; dev image assert now verifies anydoc instead - Prefill all AnyDoc-supported document extensions in the filesystem indexer default include patterns (both ToolWizard defaults) - Deny-list the new binary office variants (.docm, .pptm, .pps, .ppsx, .ppsm, .pot, .rtf, .epub) in the userspace prefetch heuristic; .csv stays prefetchable since it is text-editable - Revert Optional[str] style churn in document_parser to pre-PR unions - Cover NEEDS_OCR in document parser and runtime PDF read tests
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.
Summary
Verification
docker build --target python-pytest -f docker/Dockerfile .