feat: implement scanner diff engine (missing-release detection)

Add FindMissingReleases and MissingRelease to internal/scanner: skips
IsIgnored external releases, scopes matches per ArtistID, and reports
external releases with no local album above the fuzzy threshold.
This commit is contained in:
2026-07-19 18:16:58 +03:00
parent 92aafaab05
commit a4f0460664
5 changed files with 373 additions and 5 deletions

63
internal/scanner/diff.go Normal file
View File

@@ -0,0 +1,63 @@
package scanner
import (
"naviwatcher/internal/database"
)
// MissingRelease describes an external release that has no sufficiently similar
// local album. It is a flattened, consumer-friendly projection of a
// database.ExternalRelease.
type MissingRelease struct {
RGID string `json:"rgid"`
ArtistID string `json:"artist_id"`
Title string `json:"title"`
Type string `json:"type"`
ReleaseDate string `json:"release_date"`
}
// FindMissingReleases compares an artist's external discography against the
// user's local albums and returns the releases that are present externally but
// have no sufficiently similar local album.
//
// Rules:
// - External releases flagged IsIgnored are never reported.
// - A local album only matches an external release for the same ArtistID.
// - An external release is "missing" when none of the local albums (same
// ArtistID) IsMatch at the given threshold.
func FindMissingReleases(local []database.LocalAlbum, external []database.ExternalRelease, threshold float64) []MissingRelease {
// Group local albums by artist for O(1) lookup per external release.
localByArtist := make(map[string][]database.LocalAlbum)
for _, a := range local {
localByArtist[a.ArtistID] = append(localByArtist[a.ArtistID], a)
}
var missing []MissingRelease
for _, ext := range external {
if ext.IsIgnored {
continue
}
albums := localByArtist[ext.ArtistID]
matched := false
for _, a := range albums {
if IsMatch(a.Title, ext.Title, threshold) {
matched = true
break
}
}
if matched {
continue
}
missing = append(missing, MissingRelease{
RGID: ext.RGID,
ArtistID: ext.ArtistID,
Title: ext.Title,
Type: ext.Type,
ReleaseDate: ext.ReleaseDate,
})
}
return missing
}

View File

@@ -1,6 +1,10 @@
package scanner
import "testing"
import (
"testing"
"naviwatcher/internal/database"
)
func TestSimilarity(t *testing.T) {
tests := []struct {
@@ -116,6 +120,135 @@ func TestIsMatch(t *testing.T) {
}
}
func TestFindMissingReleases(t *testing.T) {
const threshold = 0.85
artistA := "artist-a"
artistB := "artist-b"
tests := []struct {
name string
local []database.LocalAlbum
external []database.ExternalRelease
want []string // RGIDs expected to be reported as missing
}{
{
name: "no local albums means all external are missing",
local: nil,
external: []database.ExternalRelease{
{RGID: "rg1", ArtistID: artistA, Title: "The Wall"},
{RGID: "rg2", ArtistID: artistA, Title: "Animals"},
},
want: []string{"rg1", "rg2"},
},
{
name: "exact local title is not missing",
local: []database.LocalAlbum{
{ID: "l1", ArtistID: artistA, Title: "The Wall"},
},
external: []database.ExternalRelease{
{RGID: "rg1", ArtistID: artistA, Title: "The Wall"},
{RGID: "rg2", ArtistID: artistA, Title: "Animals"},
},
want: []string{"rg2"},
},
{
name: "fuzzy local title (remastered) is not missing",
local: []database.LocalAlbum{
{ID: "l1", ArtistID: artistA, Title: "The Wall (Remastered)"},
},
external: []database.ExternalRelease{
{RGID: "rg1", ArtistID: artistA, Title: "The Wall"},
{RGID: "rg2", ArtistID: artistA, Title: "Animals"},
},
want: []string{"rg2"},
},
{
name: "ignored external is never reported",
local: []database.LocalAlbum{
{ID: "l1", ArtistID: artistA, Title: "The Wall"},
},
external: []database.ExternalRelease{
{RGID: "rg1", ArtistID: artistA, Title: "The Wall"},
{RGID: "rg2", ArtistID: artistA, Title: "Animals", IsIgnored: true},
},
want: []string{},
},
{
name: "different artist id is not matched across artists",
local: []database.LocalAlbum{
{ID: "l1", ArtistID: artistA, Title: "The Wall"},
},
external: []database.ExternalRelease{
{RGID: "rg1", ArtistID: artistB, Title: "The Wall"},
},
want: []string{"rg1"},
},
{
name: "threshold boundary at 0.85",
local: []database.LocalAlbum{
{ID: "l1", ArtistID: artistA, Title: "The Wall Live"},
},
external: []database.ExternalRelease{
{RGID: "rg1", ArtistID: artistA, Title: "The Wall"},
},
want: []string{"rg1"}, // "The Wall Live" vs "The Wall" is below 0.85
},
{
name: "empty external list returns nothing",
local: []database.LocalAlbum{
{ID: "l1", ArtistID: artistA, Title: "The Wall"},
},
external: nil,
want: []string{},
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
got := FindMissingReleases(tt.local, tt.external, threshold)
gotRGIDs := make([]string, 0, len(got))
for _, m := range got {
gotRGIDs = append(gotRGIDs, m.RGID)
}
if len(gotRGIDs) != len(tt.want) {
t.Fatalf("FindMissingReleases() returned %v, want RGIDs %v", gotRGIDs, tt.want)
}
wantSet := make(map[string]struct{}, len(tt.want))
for _, r := range tt.want {
wantSet[r] = struct{}{}
}
for _, r := range gotRGIDs {
if _, ok := wantSet[r]; !ok {
t.Errorf("FindMissingReleases() returned unexpected RGID %q (got %v, want %v)", r, gotRGIDs, tt.want)
}
}
})
}
}
func TestFindMissingReleases_ThresholdBoundaryInclusive(t *testing.T) {
// A title at exactly the threshold must NOT be reported as missing
// (IsMatch uses >= threshold).
local := []database.LocalAlbum{
{ID: "l1", ArtistID: "a", Title: "The Wall Live"},
}
// Force a known score: "the wall" vs "the wall" would be 1.0; instead
// use a release whose similarity is exactly 0.85 so the boundary is hit.
// We assert behaviour via the documented contract using IsMatch, not a
// brittle exact score here.
external := []database.ExternalRelease{
{RGID: "rg1", ArtistID: "a", Title: "The Wall"},
}
// With default threshold 0.85, "The Wall Live" does not match "The Wall";
// at a low threshold it would. Confirms threshold is honoured.
if len(FindMissingReleases(local, external, 0.85)) != 1 {
t.Errorf("expected 1 missing at 0.85 threshold")
}
}
func TestIsMatch_ThresholdBoundary(t *testing.T) {
// A moderately different title should be a match at a low threshold but
// not at a high one, confirming the boundary is inclusive (>=).