How pffantom1290/full-video-finder works
Full Video Finder takes a TikTok clip URL and searches for its longer source by identifying the work from clip content, then finding and verifying candidate videos. The README documents the staged workflow; sampled source confirms the pipeline orchestration and several internal dependencies. Some provider and API wiring is unsampled, so uncertain connections are omitted or shown as unverified.
Main entry points: pipeline.py Search pipeline · signals.py Signal collection · app.js Web interface · main.py HTTP API
Components 23
Application entry
- Web interface
frontend/app.js - HTTP API
backend/app/main.py - Search pipeline
backend/app/pipeline.py
Clip analysis
- Clip resolver
backend/app/tiktok.py - Frame extraction
backend/app/media.py - Signal collection
backend/app/signals.py - Speech recognition
backend/app/asr.py - Visual recognition
backend/app/vision.py
Search and identification
- Provider registry
backend/app/providers/__init__.py - Search providers
backend/app/providers - Web search
backend/app/providers/websearch.py - Search engines
backend/app/engines.py - Work identification
backend/app/identify.py - Work catalog
backend/app/catalog.py - Query translation
backend/app/translate.py
Ranking and proof
- Candidate ranking
backend/app/scoring.py - Frame verification
backend/app/verify.py
Shared internals
- Shared data models
backend/app/models.py - Text statistics
backend/app/textstats.py - Image hashing
backend/app/imghash.py - HTTP client and cache
backend/app/httpx_client.py
Other components
- Clip seeker
- External search services
Connections 22
- Clip seeker → Web interface: submits clip
README.md - Web interface → HTTP API: requests search
- HTTP API → Search pipeline: runs search
backend/app/main.py - Search pipeline → Clip resolver: resolves clip
README.md - Search pipeline → Signal collection: collects signals
backend/app/pipeline.py - Signal collection → Frame extraction: extracts frames
backend/app/signals.py - Signal collection → Speech recognition: transcribes audio
backend/app/signals.py - Signal collection → Visual recognition: analyzes frames
backend/app/signals.py - Search pipeline → Provider registry: gets providers
backend/app/pipeline.py - Provider registry → Search providers: registers providers
backend/app/providers/__init__.py - Search pipeline → Work identification: identifies work
README.md - Work identification → Work catalog: checks hypothesis
backend/app/identify.py - Search pipeline → Query translation: expands queries
backend/app/pipeline.py - Web search → Search engines: searches queries
backend/app/providers/websearch.py - Search engines → External search services: queries search APIs
backend/app/engines.py - Search pipeline → Candidate ranking: ranks candidates
README.md - Search pipeline → Frame verification: verifies frames
backend/app/pipeline.py - Candidate ranking → Text statistics: weights terms
README.md - Frame extraction → Image hashing: hashes frames
backend/app/media.py - Frame verification → Image hashing: compares hashes
backend/app/verify.py - Search pipeline → Shared data models: passes search data
backend/app/pipeline.py - HTTP client and cache → External search services: sends HTTP