How adminfarhanaftab/long_video_analysis works
This is a video-processing application: a user uploads a video, starts a job, and views timestamp detections as runner chunks complete. The Worker signs object-storage requests, dispatches processing, tracks progress and timestamps, and deletes the source video after the configured completion count. The runner samples video frames for motion and posts chunk results. The HTML entry point is present but unsampled, so its interface and connections are not inferred; relationships shown are limited to those documented in the Worker and runner excerpts.
Main entry points: worker.js Request Worker · main.py Motion runner · index.html Video interface
Components 7
User interface
- Video interface
index.html
Job coordination
- Request Worker
cloudflare/worker.js
Video analysis
- Motion runner
runner/main.py
External services
- Video object storage
- Job state
- GitHub dispatch
Other components
- User
Connections 8
- User → Video interface: uses
- Video interface → Request Worker: requests
- Request Worker → Video object storage: signs requests
cloudflare/worker.js - Request Worker → Job state: reads and writes
cloudflare/worker.js - Request Worker → GitHub dispatch: dispatches job
cloudflare/worker.js - GitHub dispatch → Motion runner: runs chunks
- Motion runner → Video object storage: reads video
- Motion runner → Request Worker: posts detections
runner/main.py