feat(uploads): match release groups with MusicBrainz and AcoustID

services/api/internal/musicbrainz/client.go

@@ -0,0 +1,353 @@
+// Package musicbrainz adapts MusicBrainz and Cover Art Archive without receiving audio files.
+package musicbrainz
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"sort"
+	"strings"
+	"sync"
+	"time"
+
+	"github.com/google/uuid"
+)
+
+type Track struct {
+	RecordingID string `json:"recording_id"`
+	Title       string `json:"title"`
+	Artist      string `json:"artist"`
+	Disc        int    `json:"disc"`
+	Position    int    `json:"position"`
+	DurationMs  int    `json:"duration_ms"`
+}
+type Release struct {
+	ID       string  `json:"id"`
+	GroupID  string  `json:"group_id"`
+	Title    string  `json:"title"`
+	Artist   string  `json:"artist"`
+	Type     string  `json:"type"`
+	Date     string  `json:"date"`
+	Country  string  `json:"country"`
+	Comment  string  `json:"comment"`
+	HasCover bool    `json:"has_cover"`
+	Tracks   []Track `json:"tracks"`
+}
+type Query struct {
+	GroupID      string   `json:"group_id,omitempty"`
+	Artist       string   `json:"artist"`
+	Album        string   `json:"album"`
+	ReleaseID    string   `json:"release_id"`
+	Audio        bool     `json:"audio,omitempty"`
+	RecordingIDs []string `json:"-"`
+}
+type Cover struct {
+	Data  []byte
+	MIME  string
+	URL   string
+	Exact bool
+}
+type Cache interface {
+	GetMetadataCache(context.Context, string) ([]byte, bool, error)
+	PutMetadataCache(context.Context, string, []byte, time.Duration) error
+}
+type Client struct {
+	http      *http.Client
+	cache     Cache
+	mu        sync.Mutex
+	next      time.Time
+	base      string
+	coverBase string
+	interval  time.Duration
+	userAgent string
+}
+
+func New(cache Cache, userAgent string) *Client {
+	return &Client{cache: cache, base: "https://musicbrainz.org/ws/2", coverBase: "https://coverartarchive.org", interval: 1100 * time.Millisecond,
+		userAgent: userAgent, http: &http.Client{Timeout: 20 * time.Second, CheckRedirect: func(req *http.Request, via []*http.Request) error {
+			h := req.URL.Hostname()
+			if len(via) > 5 || req.URL.Scheme != "https" || req.URL.Port() != "" || req.URL.User != nil ||
+				!(h == "musicbrainz.org" || h == "coverartarchive.org" || h == "archive.org" || strings.HasSuffix(h, ".archive.org")) {
+				return errors.New("metadata: redirect outside provider")
+			}
+			return nil
+		}}}
+}
+func (c *Client) wait(ctx context.Context) error {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	if delay := time.Until(c.next); delay > 0 {
+		t := time.NewTimer(delay)
+		defer t.Stop()
+		select {
+		case <-ctx.Done():
+			return ctx.Err()
+		case <-t.C:
+		}
+	}
+	c.next = time.Now().Add(c.interval)
+	return nil
+}
+
+var ErrNotFound = errors.New("metadata: not found")
+
+func (c *Client) get(ctx context.Context, address string, limited bool, maxBytes int64) ([]byte, error) {
+	if data, ok, err := c.cache.GetMetadataCache(ctx, address); err == nil && ok {
+		return data, nil
+	}
+	if limited {
+		if err := c.wait(ctx); err != nil {
+			return nil, err
+		}
+	}
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, address, nil)
+	if err != nil {
+		return nil, err
+	}
+	req.Header.Set("User-Agent", c.userAgent)
+	req.Header.Set("Accept", "application/json, image/jpeg, image/png, image/webp")
+	res, err := c.http.Do(req)
+	if err != nil {
+		return nil, err
+	}
+	defer res.Body.Close()
+	if res.StatusCode == http.StatusNotFound {
+		return nil, ErrNotFound
+	}
+	if res.StatusCode != http.StatusOK {
+		return nil, fmt.Errorf("metadata: HTTP %d", res.StatusCode)
+	}
+	data, err := io.ReadAll(io.LimitReader(res.Body, maxBytes+1))
+	if err != nil {
+		return nil, err
+	}
+	if int64(len(data)) > maxBytes {
+		return nil, errors.New("metadata: response too large")
+	}
+	if limited && !json.Valid(data) {
+		return nil, errors.New("metadata: invalid JSON response")
+	}
+	if !limited {
+		mime := http.DetectContentType(data)
+		if mime != "image/jpeg" && mime != "image/png" && mime != "image/webp" {
+			return nil, errors.New("metadata: unsupported cover")
+		}
+	}
+	_ = c.cache.PutMetadataCache(ctx, address, data, 7*24*time.Hour)
+	return data, nil
+}
+
+type credit struct {
+	Name   string `json:"name"`
+	Join   string `json:"joinphrase"`
+	Artist struct {
+		Name string `json:"name"`
+	} `json:"artist"`
+}
+
+func artist(credits []credit) string {
+	var s strings.Builder
+	for _, a := range credits {
+		name := a.Name
+		if name == "" {
+			name = a.Artist.Name
+		}
+		s.WriteString(name)
+		s.WriteString(a.Join)
+	}
+	return s.String()
+}
+
+type releaseJSON struct {
+	ID      string   `json:"id"`
+	Title   string   `json:"title"`
+	Date    string   `json:"date"`
+	Country string   `json:"country"`
+	Comment string   `json:"disambiguation"`
+	Credits []credit `json:"artist-credit"`
+	Group   struct {
+		ID   string `json:"id"`
+		Type string `json:"primary-type"`
+	} `json:"release-group"`
+	Cover struct {
+		Front bool `json:"front"`
+	} `json:"cover-art-archive"`
+	Media []struct {
+		Position int `json:"position"`
+		Tracks   []struct {
+			Position  int      `json:"position"`
+			Title     string   `json:"title"`
+			Length    int      `json:"length"`
+			Credits   []credit `json:"artist-credit"`
+			Recording struct {
+				ID      string   `json:"id"`
+				Title   string   `json:"title"`
+				Length  int      `json:"length"`
+				Credits []credit `json:"artist-credit"`
+			} `json:"recording"`
+		} `json:"tracks"`
+	} `json:"media"`
+}
+
+func (c *Client) release(ctx context.Context, id string) (Release, error) {
+	if _, err := uuid.Parse(id); err != nil {
+		return Release{}, errors.New("metadata: invalid release ID")
+	}
+	data, err := c.get(ctx, c.base+"/release/"+id+"?fmt=json&inc=recordings%2Bartist-credits%2Brelease-groups", true, 8<<20)
+	if err != nil {
+		return Release{}, err
+	}
+	var raw releaseJSON
+	if err = json.Unmarshal(data, &raw); err != nil {
+		return Release{}, err
+	}
+	r := Release{ID: raw.ID, GroupID: raw.Group.ID, Title: raw.Title, Artist: artist(raw.Credits), Type: raw.Group.Type, Date: raw.Date, Country: raw.Country, Comment: raw.Comment, HasCover: raw.Cover.Front, Tracks: []Track{}}
+	if _, err = uuid.Parse(r.ID); err != nil {
+		return Release{}, errors.New("metadata: invalid release response")
+	}
+	for _, m := range raw.Media {
+		for _, t := range m.Tracks {
+			title := t.Title
+			if title == "" {
+				title = t.Recording.Title
+			}
+			a := artist(t.Credits)
+			if a == "" {
+				a = artist(t.Recording.Credits)
+			}
+			if a == "" {
+				a = r.Artist
+			}
+			length := t.Length
+			if length == 0 {
+				length = t.Recording.Length
+			}
+			r.Tracks = append(r.Tracks, Track{RecordingID: t.Recording.ID, Title: title, Artist: a, Disc: m.Position, Position: t.Position, DurationMs: length})
+		}
+	}
+	if len(r.Tracks) == 0 || len(r.Tracks) > 1000 {
+		return Release{}, errors.New("metadata: invalid track list")
+	}
+	return r, nil
+}
+func quote(s string) string { return `"` + strings.NewReplacer(`\`, `\\`, `"`, `\"`).Replace(s) + `"` }
+func (c *Client) Search(ctx context.Context, q Query) ([]Release, error) {
+	if len(q.RecordingIDs) > 0 {
+		return c.byRecordings(ctx, q.RecordingIDs)
+	}
+	if q.ReleaseID != "" {
+		r, err := c.release(ctx, q.ReleaseID)
+		if errors.Is(err, ErrNotFound) {
+			return []Release{}, nil
+		}
+		return []Release{r}, err
+	}
+	params := url.Values{"fmt": {"json"}, "limit": {"5"}, "query": {"release:" + quote(q.Album) + " AND artist:" + quote(q.Artist)}}
+	data, err := c.get(ctx, c.base+"/release?"+params.Encode(), true, 4<<20)
+	if err != nil {
+		return nil, err
+	}
+	var found struct {
+		Releases []struct {
+			ID string `json:"id"`
+		} `json:"releases"`
+	}
+	if err = json.Unmarshal(data, &found); err != nil {
+		return nil, err
+	}
+	out := []Release{}
+	for _, candidate := range found.Releases {
+		r, err := c.release(ctx, candidate.ID)
+		if err != nil {
+			return nil, err
+		}
+		out = append(out, r)
+	}
+	return out, nil
+}
+
+// byRecordings discovers editions from a bounded sample; all batch files are matched afterwards.
+func (c *Client) byRecordings(ctx context.Context, ids []string) ([]Release, error) {
+	votes := map[string]int{}
+	seen := map[string]bool{}
+	for _, id := range ids {
+		if _, err := uuid.Parse(id); err != nil {
+			return nil, errors.New("metadata: invalid recording ID")
+		}
+		if seen[id] {
+			continue
+		}
+		if len(seen) == 8 {
+			break
+		}
+		seen[id] = true
+		params := url.Values{"fmt": {"json"}, "recording": {id}, "limit": {"100"}}
+		data, err := c.get(ctx, c.base+"/release?"+params.Encode(), true, 4<<20)
+		if err != nil {
+			return nil, err
+		}
+		var found struct {
+			Releases []struct {
+				ID string `json:"id"`
+			} `json:"releases"`
+		}
+		if err = json.Unmarshal(data, &found); err != nil {
+			return nil, err
+		}
+		perRecording := map[string]bool{}
+		for _, r := range found.Releases {
+			if _, err := uuid.Parse(r.ID); err == nil && !perRecording[r.ID] {
+				votes[r.ID]++
+				perRecording[r.ID] = true
+			}
+		}
+	}
+	candidates := []string{}
+	for id := range votes {
+		candidates = append(candidates, id)
+	}
+	sort.Slice(candidates, func(i, j int) bool {
+		if votes[candidates[i]] == votes[candidates[j]] {
+			return candidates[i] < candidates[j]
+		}
+		return votes[candidates[i]] > votes[candidates[j]]
+	})
+	out := []Release{}
+	for _, id := range candidates[:min(5, len(candidates))] {
+		r, err := c.release(ctx, id)
+		if err != nil {
+			return nil, err
+		}
+		out = append(out, r)
+	}
+	return out, nil
+}
+func (c *Client) Cover(ctx context.Context, r Release) (Cover, error) {
+	for i, id := range []string{r.ID, r.GroupID} {
+		if _, err := uuid.Parse(id); err != nil {
+			continue
+		}
+		kind := "release"
+		if i == 1 {
+			kind = "release-group"
+		}
+		address := c.coverBase + "/" + kind + "/" + id + "/front-1200"
+		data, err := c.get(ctx, address, false, 8<<20)
+		if errors.Is(err, ErrNotFound) {
+			continue
+		}
+		if err != nil {
+			return Cover{}, err
+		}
+		mime := http.DetectContentType(data)
+		if mime != "image/jpeg" && mime != "image/png" && mime != "image/webp" {
+			return Cover{}, errors.New("metadata: unsupported cover")
+		}
+		return Cover{Data: data, MIME: mime, URL: address, Exact: i == 0}, nil
+	}
+	return Cover{}, ErrNotFound
+}