feat(backend): perceptual hashing for images and video

Adds a 64-bit dHash perceptual hash (internal/imagehash, built on the existing
disintegration/imaging — no new dependency) and starts populating the long-unused
data.files.phash column:

- Upload sets phash inline for images (cheap, from the in-memory bytes).
- Replace recomputes it from new content for images and clears it for anything
  else, so a stale hash never survives a content swap.
- FileRepo.SetPHash sets/clears the hash (used by Replace and, later, the dedup
  backfill).
- DiskStorage.VideoFrameMiddle extracts a frame from the middle of a clip
  (ffprobe duration -> ffmpeg -ss duration/2), avoiding the shared-intro collision
  a fixed early offset causes. It is a concrete method, not part of the storage
  port: only the dedup CLI needs it, keeping ffmpeg off the upload path. Video
  phashes are therefore computed by that CLI, not at upload time.
- DUPLICATE_HASH_THRESHOLD config (default 10/64) for the later pair rescan.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
2026-06-16 12:20:52 +03:00
parent 58cea88f52
commit 88849cc16b
7 changed files with 278 additions and 7 deletions
+8
View File
@@ -63,6 +63,12 @@ type Config struct {
// Import
ImportPath string
// DuplicateHashThreshold is the maximum Hamming distance (out of 64) between
// two perceptual hashes for the files to be treated as duplicate candidates.
// Lower = stricter (fewer, more confident matches); higher = looser. Used only
// by the dedup rescan that (re)builds data.duplicate_pairs.
DuplicateHashThreshold int
// Static SPA. When set, the server serves the built frontend (and falls
// back to index.html for client routes) on the same port as the API. Empty
// in local development, where the Vite dev server serves the UI separately.
@@ -176,6 +182,8 @@ func Load() (*Config, error) {
ImportPath: requireStr("IMPORT_PATH"),
DuplicateHashThreshold: parseInt("DUPLICATE_HASH_THRESHOLD", 10),
StaticDir: defaultStr("STATIC_DIR", ""),
}