Repository navigation
feat(analysis): places and dates from anytopdf as hints for the graph - #229
Merged
Merged
Conversation
Settings › Analysis › Find places and dates with anytopdf (analysis.anytopdf, off by default). Analysis then asks anytopdf, run here offline with no plugins, for the places a recording's text names, the dates it gives (with their ISO value) and where a photo was taken (its GPS fix, named offline). They join the extractor's names per line, a bare term anytopdf knows as a place becomes one, and the entity setup still decides what is kept. What was found is kept on the recording and shown on its Details tab to members only. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018GMFApXZSk7XERM3J5gNTy
…-kgtpxt # Conflicts: # CHANGELOG.md # fastapi_backend/app/domain/anytopdf.py
6 tasks done
21 of 23 tasks
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Requested by Adeel · project thread
Description
Before: Lens's rules extractor found "Berlin" only as a term, gave dates no value, and knew nothing about where a photo was taken.
After: with Settings › Analysis › Find places and dates with anytopdf on (
analysis.anytopdf, off by default), analysis asks anytopdf, run here offline with no plugins, for:exiftoolbeside anytopdf). Read once per file.They join the extractor's names per line; a bare TERM anytopdf knows becomes a PLACE or DATE; the entity setup still decides what is kept. What was found is kept on the recording as
enrichmentand shown on the Details tab ("Places and dates") to namespace members only. Visitors, embeds and signed visitor links don't get it. A failing anytopdf never stops analysis.Motivation and Context
Phase 3 of the anytopdf integration plan (enrichment: location, dates, entity hints). Builds on #221. CLIP photo search follows as its own PR.
How: new
enrich.py;anytopdf.graph()runsconvert --dump-graphwith--no-plugins(0.4.0 has no bundle yet, and the PDF's embedded chunks drop locations);analyze_recordingmerges the hints; the anytopdf component is fetched foranalyzewhen the switch is on.Types of changes
Checklist:
tests/api/test_enrich.py, plus a real-binary test underLENS_TEST_ANYTOPDF, passed locally with 0.4.0 and exiftool).docs/accordingly.## UnreleasedinCHANGELOG.md.🤖 Generated with Claude Code
https://claude.ai/code/session_018GMFApXZSk7XERM3J5gNTy
Generated by Claude Code