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

services/api/internal/acoustid/client.go

@@ -0,0 +1,126 @@
+// Package acoustid identifies recordings using Chromaprint, never audio files.
+package acoustid
+
+import (
+	"context"
+	"crypto/sha256"
+	"encoding/hex"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"strings"
+	"sync"
+	"time"
+
+	"github.com/google/uuid"
+)
+
+type Match struct {
+	RecordingID string
+	Score       float64
+}
+type Cache interface {
+	GetMetadataCache(context.Context, string) ([]byte, bool, error)
+	PutMetadataCache(context.Context, string, []byte, time.Duration) error
+}
+type Client struct {
+	key, endpoint string
+	cache         Cache
+	http          *http.Client
+	mu            sync.Mutex
+	next          time.Time
+	interval      time.Duration
+}
+
+func New(key string, cache Cache) *Client {
+	return &Client{key: key, endpoint: "https://api.acoustid.org/v2/lookup", cache: cache, interval: 350 * time.Millisecond,
+		http: &http.Client{Timeout: 20 * time.Second, CheckRedirect: func(*http.Request, []*http.Request) error { return http.ErrUseLastResponse }}}
+}
+func (c *Client) wait(ctx context.Context) error {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	t := time.NewTimer(max(time.Duration(0), time.Until(c.next)))
+	defer t.Stop()
+	select {
+	case <-ctx.Done():
+		return ctx.Err()
+	case <-t.C:
+	}
+	c.next = time.Now().Add(c.interval)
+	return nil
+}
+func (c *Client) Lookup(ctx context.Context, fingerprint string, duration int) ([]Match, error) {
+	if c.key == "" || duration <= 0 || len(fingerprint) == 0 || len(fingerprint) > 65536 {
+		return nil, errors.New("acoustid: invalid lookup input")
+	}
+	hash := sha256.Sum256([]byte(fmt.Sprint(duration) + ":" + fingerprint))
+	key := "acoustid:" + hex.EncodeToString(hash[:])
+	if data, ok, err := c.cache.GetMetadataCache(ctx, key); err == nil && ok {
+		return parse(data)
+	}
+	if err := c.wait(ctx); err != nil {
+		return nil, err
+	}
+	body := url.Values{"client": {c.key}, "duration": {fmt.Sprint(duration)}, "fingerprint": {fingerprint}, "meta": {"recordingids"}, "format": {"json"}}
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, c.endpoint, strings.NewReader(body.Encode()))
+	if err != nil {
+		return nil, err
+	}
+	req.Header.Set("Content-Type", "application/x-www-form-urlencoded")
+	req.Header.Set("User-Agent", "PipuMusic (https://gitea.hsrv.site/mrleo1nid/pipu-music)")
+	res, err := c.http.Do(req)
+	if err != nil {
+		return nil, errors.New("acoustid: request failed")
+	}
+	defer res.Body.Close()
+	if res.StatusCode != http.StatusOK {
+		return nil, fmt.Errorf("acoustid: HTTP %d", res.StatusCode)
+	}
+	data, err := io.ReadAll(io.LimitReader(res.Body, (2<<20)+1))
+	if err != nil || len(data) > 2<<20 {
+		return nil, errors.New("acoustid: invalid response size")
+	}
+	matches, err := parse(data)
+	if err != nil {
+		return nil, err
+	}
+	_ = c.cache.PutMetadataCache(ctx, key, data, 7*24*time.Hour)
+	return matches, nil
+}
+func parse(data []byte) ([]Match, error) {
+	var response struct {
+		Status  string `json:"status"`
+		Results []struct {
+			Score      float64 `json:"score"`
+			Recordings []struct {
+				ID string `json:"id"`
+			} `json:"recordings"`
+		} `json:"results"`
+	}
+	if err := json.Unmarshal(data, &response); err != nil {
+		return nil, errors.New("acoustid: invalid response")
+	}
+	if response.Status != "ok" {
+		return nil, errors.New("acoustid: lookup rejected; check application key")
+	}
+	best := map[string]float64{}
+	for _, result := range response.Results {
+		if result.Score <= 0 || result.Score > 1 {
+			continue
+		}
+		for _, recording := range result.Recordings {
+			id, err := uuid.Parse(recording.ID)
+			if err == nil && id != uuid.Nil && result.Score > best[id.String()] {
+				best[id.String()] = result.Score
+			}
+		}
+	}
+	out := []Match{}
+	for id, score := range best {
+		out = append(out, Match{RecordingID: id, Score: score})
+	}
+	return out, nil
+}