feat: initial implementation of metadata aggregator

- gRPC service with MusicBrainz provider
- PostgreSQL schema with migrations
- Service layer with database-first caching
- Repository pattern for data access
- YAML configuration support
- Research documentation for 17 music metadata projects
This commit is contained in:
Alexander
2026-04-28 16:27:14 +02:00
commit a1f6701bac
163 changed files with 95884 additions and 0 deletions
+68
View File
@@ -0,0 +1,68 @@
package config
import (
"fmt"
"os"
"gopkg.in/yaml.v3"
)
type Config struct {
Server ServerConfig `yaml:"server"`
Database DatabaseConfig `yaml:"database"`
}
type ServerConfig struct {
Port int `yaml:"port"`
}
type DatabaseConfig struct {
Host string `yaml:"host"`
Port int `yaml:"port"`
User string `yaml:"user"`
Password string `yaml:"password"`
Name string `yaml:"name"`
SSLMode string `yaml:"sslmode"`
}
func Load(path string) (*Config, error) {
cfg := &Config{
Server: ServerConfig{
Port: 50051,
},
Database: DatabaseConfig{
Host: "localhost",
Port: 5432,
SSLMode: "disable",
},
}
if path == "" {
return cfg, nil
}
data, err := os.ReadFile(path)
if err != nil {
if os.IsNotExist(err) {
return cfg, nil
}
return nil, fmt.Errorf("read config: %w", err)
}
if err := yaml.Unmarshal(data, cfg); err != nil {
return nil, fmt.Errorf("parse config: %w", err)
}
return cfg, nil
}
func (d *DatabaseConfig) DSN() string {
if d.Host == "" || d.User == "" || d.Name == "" {
return ""
}
return fmt.Sprintf(
"postgres://%s:%s@%s:%d/%s?sslmode=%s",
d.User, d.Password, d.Host, d.Port, d.Name, d.SSLMode,
)
}
+84
View File
@@ -0,0 +1,84 @@
package domain
import "time"
type Artist struct {
ID string
Name string
SortName string
Type string
Country string
FormedDate *time.Time
DisbandedDate *time.Time
Description string
ImageURL string
Genres []Genre
ExternalIDs []ExternalID
}
type Album struct {
ID string
Title string
Type string
ReleaseDate *time.Time
UPC string
TotalTracks int
TotalDiscs int
CoverURL string
Artists []ArtistCredit
Label *Label
Genres []Genre
ExternalIDs []ExternalID
}
type Track struct {
ID string
Title string
DurationMs int
ISRC string
Explicit bool
DiscNumber int
TrackNumber int
Artists []ArtistCredit
Work *Work
ExternalIDs []ExternalID
}
type Work struct {
ID string
Title string
Type string
Language string
Composers []ArtistCredit
}
type Label struct {
ID string
Name string
Country string
}
type Genre struct {
ID string
Name string
}
type ArtistCredit struct {
Artist Artist
Role string
Position int
JoinPhrase string
}
type ExternalID struct {
Source string
SourceID string
URL string
}
type SearchResult[T any] struct {
Items []T
Total int
Limit int
Offset int
}
+127
View File
@@ -0,0 +1,127 @@
package musicbrainz
import (
"context"
"encoding/json"
"fmt"
"io"
"net/http"
"net/url"
"time"
"golang.org/x/time/rate"
)
const (
baseURL = "https://musicbrainz.org/ws/2"
userAgent = "MetadataAggregator/0.1.0 (https://github.com/metadata-agregator)"
)
type client struct {
http *http.Client
limiter *rate.Limiter
}
func newClient() *client {
return &client{
http: &http.Client{
Timeout: 30 * time.Second,
},
limiter: rate.NewLimiter(rate.Every(time.Second), 1),
}
}
func (c *client) get(ctx context.Context, endpoint string, params url.Values) ([]byte, error) {
if err := c.limiter.Wait(ctx); err != nil {
return nil, fmt.Errorf("rate limiter: %w", err)
}
if params == nil {
params = url.Values{}
}
params.Set("fmt", "json")
reqURL := fmt.Sprintf("%s/%s?%s", baseURL, endpoint, params.Encode())
req, err := http.NewRequestWithContext(ctx, http.MethodGet, reqURL, nil)
if err != nil {
return nil, fmt.Errorf("create request: %w", err)
}
req.Header.Set("User-Agent", userAgent)
req.Header.Set("Accept", "application/json")
resp, err := c.http.Do(req)
if err != nil {
return nil, fmt.Errorf("do request: %w", err)
}
defer resp.Body.Close()
if resp.StatusCode == http.StatusNotFound {
return nil, ErrNotFound
}
if resp.StatusCode == http.StatusServiceUnavailable {
return nil, ErrRateLimited
}
if resp.StatusCode != http.StatusOK {
body, _ := io.ReadAll(resp.Body)
return nil, fmt.Errorf("unexpected status %d: %s", resp.StatusCode, string(body))
}
return io.ReadAll(resp.Body)
}
func (c *client) lookup(ctx context.Context, entity, id string, inc []string) ([]byte, error) {
params := url.Values{}
if len(inc) > 0 {
incStr := ""
for i, v := range inc {
if i > 0 {
incStr += "+"
}
incStr += v
}
params.Set("inc", incStr)
}
return c.get(ctx, fmt.Sprintf("%s/%s", entity, id), params)
}
func (c *client) browse(ctx context.Context, entity, linkedEntity, linkedID string, limit, offset int, inc []string) ([]byte, error) {
params := url.Values{}
params.Set(linkedEntity, linkedID)
params.Set("limit", fmt.Sprintf("%d", limit))
params.Set("offset", fmt.Sprintf("%d", offset))
if len(inc) > 0 {
incStr := ""
for i, v := range inc {
if i > 0 {
incStr += "+"
}
incStr += v
}
params.Set("inc", incStr)
}
return c.get(ctx, entity, params)
}
func (c *client) search(ctx context.Context, entity, query string, limit, offset int) ([]byte, error) {
params := url.Values{}
params.Set("query", query)
params.Set("limit", fmt.Sprintf("%d", limit))
params.Set("offset", fmt.Sprintf("%d", offset))
return c.get(ctx, entity, params)
}
func decode[T any](data []byte) (*T, error) {
var result T
if err := json.Unmarshal(data, &result); err != nil {
return nil, fmt.Errorf("decode: %w", err)
}
return &result, nil
}
+8
View File
@@ -0,0 +1,8 @@
package musicbrainz
import "errors"
var (
ErrNotFound = errors.New("not found")
ErrRateLimited = errors.New("rate limited")
)
+212
View File
@@ -0,0 +1,212 @@
package musicbrainz
import (
"fmt"
"time"
"github.com/metadata-agregator/internal/domain"
)
func mapArtist(mb *mbArtist) *domain.Artist {
if mb == nil {
return nil
}
artist := &domain.Artist{
ID: mb.ID,
Name: mb.Name,
SortName: mb.SortName,
Type: mb.Type,
Country: mb.Country,
Description: mb.Disambiguation,
ExternalIDs: []domain.ExternalID{{
Source: "musicbrainz",
SourceID: mb.ID,
URL: fmt.Sprintf("https://musicbrainz.org/artist/%s", mb.ID),
}},
}
if mb.LifeSpan.Begin != "" {
if t := parseDate(mb.LifeSpan.Begin); t != nil {
artist.FormedDate = t
}
}
if mb.LifeSpan.End != "" {
if t := parseDate(mb.LifeSpan.End); t != nil {
artist.DisbandedDate = t
}
}
for _, g := range mb.Genres {
artist.Genres = append(artist.Genres, domain.Genre{
ID: g.ID,
Name: g.Name,
})
}
for _, rel := range mb.Relations {
if rel.Type == "image" && rel.URL != nil {
artist.ImageURL = rel.URL.Resource
break
}
}
return artist
}
func mapAlbum(mb *mbReleaseGroup, release *mbRelease) *domain.Album {
if mb == nil {
return nil
}
album := &domain.Album{
ID: mb.ID,
Title: mb.Title,
Type: mb.PrimaryType,
ExternalIDs: []domain.ExternalID{{
Source: "musicbrainz",
SourceID: mb.ID,
URL: fmt.Sprintf("https://musicbrainz.org/release-group/%s", mb.ID),
}},
}
if mb.FirstReleaseDate != "" {
album.ReleaseDate = parseDate(mb.FirstReleaseDate)
}
for _, ac := range mb.ArtistCredit {
album.Artists = append(album.Artists, mapArtistCredit(&ac, "primary"))
}
for _, g := range mb.Genres {
album.Genres = append(album.Genres, domain.Genre{
ID: g.ID,
Name: g.Name,
})
}
if release != nil {
album.UPC = release.Barcode
if len(release.LabelInfo) > 0 && release.LabelInfo[0].Label != nil {
album.Label = mapLabel(release.LabelInfo[0].Label)
}
for _, m := range release.Media {
album.TotalTracks += m.TrackCount
}
album.TotalDiscs = len(release.Media)
if release.CoverArtArchive.Front {
album.CoverURL = fmt.Sprintf("https://coverartarchive.org/release/%s/front", release.ID)
}
}
return album
}
func mapTrack(mb *mbRecording, discNum, trackNum int) *domain.Track {
if mb == nil {
return nil
}
track := &domain.Track{
ID: mb.ID,
Title: mb.Title,
DurationMs: mb.Length,
DiscNumber: discNum,
TrackNumber: trackNum,
ExternalIDs: []domain.ExternalID{{
Source: "musicbrainz",
SourceID: mb.ID,
URL: fmt.Sprintf("https://musicbrainz.org/recording/%s", mb.ID),
}},
}
if len(mb.ISRCs) > 0 {
track.ISRC = mb.ISRCs[0]
}
for _, ac := range mb.ArtistCredit {
track.Artists = append(track.Artists, mapArtistCredit(&ac, "primary"))
}
for _, rel := range mb.Relations {
if rel.TargetType == "work" && rel.Work != nil {
track.Work = mapWork(rel.Work)
break
}
}
return track
}
func mapWork(mb *mbWork) *domain.Work {
if mb == nil {
return nil
}
work := &domain.Work{
ID: mb.ID,
Title: mb.Title,
Type: mb.Type,
Language: mb.Language,
}
for _, rel := range mb.Relations {
if rel.TargetType == "artist" && rel.Artist != nil {
role := "writer"
if rel.Type == "composer" || rel.Type == "lyricist" || rel.Type == "writer" {
role = rel.Type
}
work.Composers = append(work.Composers, domain.ArtistCredit{
Artist: *mapArtist(rel.Artist),
Role: role,
})
}
}
return work
}
func mapLabel(mb *mbLabel) *domain.Label {
if mb == nil {
return nil
}
return &domain.Label{
ID: mb.ID,
Name: mb.Name,
Country: mb.Country,
}
}
func mapArtistCredit(ac *mbArtistCredit, defaultRole string) domain.ArtistCredit {
credit := domain.ArtistCredit{
Role: defaultRole,
JoinPhrase: ac.JoinPhrase,
}
if ac.Artist != nil {
credit.Artist = *mapArtist(ac.Artist)
}
return credit
}
func parseDate(s string) *time.Time {
formats := []string{
"2006-01-02",
"2006-01",
"2006",
}
for _, f := range formats {
if t, err := time.Parse(f, s); err == nil {
return &t
}
}
return nil
}
+282
View File
@@ -0,0 +1,282 @@
package musicbrainz
import (
"context"
"encoding/json"
"fmt"
"strings"
"github.com/metadata-agregator/internal/domain"
)
type Provider struct {
client *client
}
func New() *Provider {
return &Provider{
client: newClient(),
}
}
func (p *Provider) Name() string {
return "musicbrainz"
}
func (p *Provider) GetArtist(ctx context.Context, id string) (*domain.Artist, error) {
data, err := p.client.lookup(ctx, "artist", id, []string{"genres", "url-rels"})
if err != nil {
return nil, fmt.Errorf("lookup artist: %w", err)
}
mb, err := decode[mbArtist](data)
if err != nil {
return nil, err
}
return mapArtist(mb), nil
}
func (p *Provider) SearchArtists(ctx context.Context, query string, limit, offset int) (*domain.SearchResult[domain.Artist], error) {
if limit <= 0 || limit > 100 {
limit = 25
}
escapedQuery := escapeQuery(query)
data, err := p.client.search(ctx, "artist", fmt.Sprintf("artist:%s", escapedQuery), limit, offset)
if err != nil {
return nil, fmt.Errorf("search artists: %w", err)
}
var resp struct {
Count int `json:"count"`
Offset int `json:"offset"`
Artists []*mbArtist `json:"artists"`
}
if err := decodeInto(data, &resp); err != nil {
return nil, err
}
result := &domain.SearchResult[domain.Artist]{
Total: resp.Count,
Limit: limit,
Offset: offset,
}
for _, mb := range resp.Artists {
if artist := mapArtist(mb); artist != nil {
result.Items = append(result.Items, *artist)
}
}
return result, nil
}
func (p *Provider) GetAlbum(ctx context.Context, id string) (*domain.Album, error) {
data, err := p.client.lookup(ctx, "release-group", id, []string{"releases", "artist-credits", "genres"})
if err != nil {
return nil, fmt.Errorf("lookup release-group: %w", err)
}
mb, err := decode[mbReleaseGroup](data)
if err != nil {
return nil, err
}
var release *mbRelease
if len(mb.Releases) > 0 {
release = selectCanonicalRelease(mb.Releases)
}
return mapAlbum(mb, release), nil
}
func (p *Provider) GetArtistAlbums(ctx context.Context, artistID string, limit, offset int) (*domain.SearchResult[domain.Album], error) {
if limit <= 0 || limit > 100 {
limit = 25
}
data, err := p.client.browse(ctx, "release-group", "artist", artistID, limit, offset, []string{"artist-credits"})
if err != nil {
return nil, fmt.Errorf("browse release-groups: %w", err)
}
var resp struct {
ReleaseGroupCount int `json:"release-group-count"`
ReleaseGroupOffset int `json:"release-group-offset"`
ReleaseGroups []*mbReleaseGroup `json:"release-groups"`
}
if err := decodeInto(data, &resp); err != nil {
return nil, err
}
result := &domain.SearchResult[domain.Album]{
Total: resp.ReleaseGroupCount,
Limit: limit,
Offset: resp.ReleaseGroupOffset,
}
for _, mb := range resp.ReleaseGroups {
if album := mapAlbum(mb, nil); album != nil {
result.Items = append(result.Items, *album)
}
}
return result, nil
}
func (p *Provider) GetTrack(ctx context.Context, id string) (*domain.Track, error) {
data, err := p.client.lookup(ctx, "recording", id, []string{"artist-credits", "isrcs", "work-rels"})
if err != nil {
return nil, fmt.Errorf("lookup recording: %w", err)
}
mb, err := decode[mbRecording](data)
if err != nil {
return nil, err
}
return mapTrack(mb, 0, 0), nil
}
func (p *Provider) GetAlbumTracks(ctx context.Context, albumID string) ([]domain.Track, error) {
data, err := p.client.browse(ctx, "release", "release-group", albumID, 100, 0, nil)
if err != nil {
return nil, fmt.Errorf("browse releases: %w", err)
}
var resp struct {
Releases []*mbRelease `json:"releases"`
}
if err := decodeInto(data, &resp); err != nil {
return nil, err
}
if len(resp.Releases) == 0 {
return nil, ErrNotFound
}
release := selectCanonicalRelease(resp.Releases)
releaseData, err := p.client.lookup(ctx, "release", release.ID, []string{"recordings", "artist-credits", "isrcs"})
if err != nil {
return nil, fmt.Errorf("lookup release: %w", err)
}
fullRelease, err := decode[mbRelease](releaseData)
if err != nil {
return nil, err
}
var tracks []domain.Track
for _, medium := range fullRelease.Media {
for _, t := range medium.Tracks {
if track := mapTrack(&t.Recording, medium.Position, t.Position); track != nil {
tracks = append(tracks, *track)
}
}
}
return tracks, nil
}
func (p *Provider) GetTrackByISRC(ctx context.Context, isrc string) (*domain.Track, error) {
data, err := p.client.get(ctx, fmt.Sprintf("isrc/%s", isrc), nil)
if err != nil {
return nil, fmt.Errorf("lookup isrc: %w", err)
}
var resp struct {
Recordings []*mbRecording `json:"recordings"`
}
if err := decodeInto(data, &resp); err != nil {
return nil, err
}
if len(resp.Recordings) == 0 {
return nil, ErrNotFound
}
return p.GetTrack(ctx, resp.Recordings[0].ID)
}
func (p *Provider) GetLabel(ctx context.Context, id string) (*domain.Label, error) {
data, err := p.client.lookup(ctx, "label", id, nil)
if err != nil {
return nil, fmt.Errorf("lookup label: %w", err)
}
mb, err := decode[mbLabel](data)
if err != nil {
return nil, err
}
return mapLabel(mb), nil
}
func (p *Provider) GetWork(ctx context.Context, id string) (*domain.Work, error) {
data, err := p.client.lookup(ctx, "work", id, []string{"artist-rels"})
if err != nil {
return nil, fmt.Errorf("lookup work: %w", err)
}
mb, err := decode[mbWork](data)
if err != nil {
return nil, err
}
return mapWork(mb), nil
}
func selectCanonicalRelease(releases []*mbRelease) *mbRelease {
if len(releases) == 0 {
return nil
}
var best *mbRelease
bestScore := -1
for _, r := range releases {
score := 0
switch r.Status {
case "Official":
score += 100
case "Promotion":
score += 50
}
if len(r.Media) > 0 {
switch r.Media[0].Format {
case "Digital Media":
score += 20
case "CD":
score += 15
}
}
if r.Barcode != "" {
score += 5
}
if score > bestScore {
bestScore = score
best = r
}
}
return best
}
func escapeQuery(s string) string {
special := []string{`+`, `-`, `&`, `|`, `!`, `(`, `)`, `{`, `}`, `[`, `]`, `^`, `"`, `~`, `*`, `?`, `:`, `/`, `\`}
result := s
for _, char := range special {
result = strings.ReplaceAll(result, char, `\`+char)
}
return `"` + result + `"`
}
func decodeInto(data []byte, v any) error {
return json.Unmarshal(data, v)
}
+138
View File
@@ -0,0 +1,138 @@
package musicbrainz
type mbArtist struct {
ID string `json:"id"`
Name string `json:"name"`
SortName string `json:"sort-name"`
Type string `json:"type"`
Country string `json:"country"`
Disambiguation string `json:"disambiguation"`
LifeSpan mbLifeSpan `json:"life-span"`
Genres []mbGenre `json:"genres"`
Relations []mbRelation `json:"relations"`
}
type mbLifeSpan struct {
Begin string `json:"begin"`
End string `json:"end"`
Ended bool `json:"ended"`
}
type mbReleaseGroup struct {
ID string `json:"id"`
Title string `json:"title"`
PrimaryType string `json:"primary-type"`
FirstReleaseDate string `json:"first-release-date"`
ArtistCredit []mbArtistCredit `json:"artist-credit"`
Genres []mbGenre `json:"genres"`
Releases []*mbRelease `json:"releases"`
}
type mbRelease struct {
ID string `json:"id"`
Title string `json:"title"`
Status string `json:"status"`
Date string `json:"date"`
Country string `json:"country"`
Barcode string `json:"barcode"`
LabelInfo []mbLabelInfo `json:"label-info"`
Media []mbMedium `json:"media"`
ReleaseGroup *mbReleaseGroup `json:"release-group"`
ArtistCredit []mbArtistCredit `json:"artist-credit"`
CoverArtArchive mbCoverArtArchive `json:"cover-art-archive"`
}
type mbCoverArtArchive struct {
Artwork bool `json:"artwork"`
Front bool `json:"front"`
Back bool `json:"back"`
}
type mbLabelInfo struct {
CatalogNumber string `json:"catalog-number"`
Label *mbLabel `json:"label"`
}
type mbLabel struct {
ID string `json:"id"`
Name string `json:"name"`
Country string `json:"country"`
}
type mbMedium struct {
Position int `json:"position"`
Format string `json:"format"`
TrackCount int `json:"track-count"`
Tracks []mbTrack `json:"tracks"`
}
type mbTrack struct {
ID string `json:"id"`
Number string `json:"number"`
Title string `json:"title"`
Length int `json:"length"`
Position int `json:"position"`
Recording mbRecording `json:"recording"`
}
type mbRecording struct {
ID string `json:"id"`
Title string `json:"title"`
Length int `json:"length"`
ISRCs []string `json:"isrcs"`
ArtistCredit []mbArtistCredit `json:"artist-credit"`
Relations []mbRelation `json:"relations"`
}
type mbWork struct {
ID string `json:"id"`
Title string `json:"title"`
Type string `json:"type"`
Language string `json:"language"`
ISWCs []string `json:"iswcs"`
Relations []mbRelation `json:"relations"`
}
type mbArtistCredit struct {
Name string `json:"name"`
JoinPhrase string `json:"joinphrase"`
Artist *mbArtist `json:"artist"`
}
type mbGenre struct {
ID string `json:"id"`
Name string `json:"name"`
Count int `json:"count"`
}
type mbRelation struct {
Type string `json:"type"`
TypeID string `json:"type-id"`
Direction string `json:"direction"`
TargetType string `json:"target-type"`
URL *mbURL `json:"url"`
Artist *mbArtist `json:"artist"`
Work *mbWork `json:"work"`
Attributes []string `json:"attributes"`
}
type mbURL struct {
ID string `json:"id"`
Resource string `json:"resource"`
}
type mbSearchResponse[T any] struct {
Created string `json:"created"`
Count int `json:"count"`
Offset int `json:"offset"`
Artists []T `json:"artists,omitempty"`
}
type mbBrowseResponse[T any] struct {
ReleaseGroupCount int `json:"release-group-count"`
ReleaseGroupOffset int `json:"release-group-offset"`
ReleaseGroups []T `json:"release-groups,omitempty"`
ReleaseCount int `json:"release-count"`
ReleaseOffset int `json:"release-offset"`
Releases []T `json:"releases,omitempty"`
}
+25
View File
@@ -0,0 +1,25 @@
package provider
import (
"context"
"github.com/metadata-agregator/internal/domain"
)
type Provider interface {
Name() string
GetArtist(ctx context.Context, id string) (*domain.Artist, error)
SearchArtists(ctx context.Context, query string, limit, offset int) (*domain.SearchResult[domain.Artist], error)
GetAlbum(ctx context.Context, id string) (*domain.Album, error)
GetArtistAlbums(ctx context.Context, artistID string, limit, offset int) (*domain.SearchResult[domain.Album], error)
GetTrack(ctx context.Context, id string) (*domain.Track, error)
GetAlbumTracks(ctx context.Context, albumID string) ([]domain.Track, error)
GetTrackByISRC(ctx context.Context, isrc string) (*domain.Track, error)
GetLabel(ctx context.Context, id string) (*domain.Label, error)
GetWork(ctx context.Context, id string) (*domain.Work, error)
}
+5
View File
@@ -0,0 +1,5 @@
package repository
import "errors"
var ErrNotFound = errors.New("not found")
+238
View File
@@ -0,0 +1,238 @@
package postgres
import (
"context"
"errors"
"time"
"github.com/jackc/pgx/v5"
"github.com/jackc/pgx/v5/pgxpool"
"github.com/metadata-agregator/internal/domain"
"github.com/metadata-agregator/internal/repository"
)
type AlbumRepository struct {
pool *pgxpool.Pool
}
func NewAlbumRepository(pool *pgxpool.Pool) *AlbumRepository {
return &AlbumRepository{pool: pool}
}
func (r *AlbumRepository) GetByID(ctx context.Context, id string) (*domain.Album, error) {
query := `
SELECT id, title, album_type, release_date, upc, total_tracks, total_discs,
cover_url, source, source_id
FROM albums
WHERE id = $1`
album, err := r.scanAlbum(ctx, query, id)
if err != nil {
return nil, err
}
if err := r.loadRelations(ctx, album); err != nil {
return nil, err
}
return album, nil
}
func (r *AlbumRepository) GetByExternalID(ctx context.Context, source, sourceID string) (*domain.Album, error) {
query := `
SELECT a.id, a.title, a.album_type, a.release_date, a.upc, a.total_tracks,
a.total_discs, a.cover_url, a.source, a.source_id
FROM albums a
JOIN album_external_ids e ON a.id = e.album_id
WHERE e.source = $1 AND e.source_id = $2`
album, err := r.scanAlbum(ctx, query, source, sourceID)
if err != nil {
return nil, err
}
if err := r.loadRelations(ctx, album); err != nil {
return nil, err
}
return album, nil
}
func (r *AlbumRepository) GetByArtistID(ctx context.Context, artistID string, limit, offset int) (*domain.SearchResult[domain.Album], error) {
countQuery := `
SELECT COUNT(DISTINCT a.id)
FROM albums a
JOIN album_artists aa ON a.id = aa.album_id
JOIN artist_external_ids ae ON aa.artist_id = ae.artist_id
WHERE ae.source_id = $1`
searchQuery := `
SELECT DISTINCT a.id, a.title, a.album_type, a.release_date, a.upc,
a.total_tracks, a.total_discs, a.cover_url, a.source, a.source_id
FROM albums a
JOIN album_artists aa ON a.id = aa.album_id
JOIN artist_external_ids ae ON aa.artist_id = ae.artist_id
WHERE ae.source_id = $1
ORDER BY a.release_date DESC NULLS LAST
LIMIT $2 OFFSET $3`
var total int
if err := r.pool.QueryRow(ctx, countQuery, artistID).Scan(&total); err != nil {
return nil, err
}
rows, err := r.pool.Query(ctx, searchQuery, artistID, limit, offset)
if err != nil {
return nil, err
}
defer rows.Close()
var albums []domain.Album
for rows.Next() {
album, err := r.scanAlbumFromRow(rows)
if err != nil {
return nil, err
}
albums = append(albums, *album)
}
return &domain.SearchResult[domain.Album]{
Items: albums,
Total: total,
Limit: limit,
Offset: offset,
}, nil
}
func (r *AlbumRepository) Save(ctx context.Context, album *domain.Album) error {
tx, err := r.pool.Begin(ctx)
if err != nil {
return err
}
defer tx.Rollback(ctx)
var source, sourceID string
if len(album.ExternalIDs) > 0 {
source = album.ExternalIDs[0].Source
sourceID = album.ExternalIDs[0].SourceID
}
query := `
INSERT INTO albums (id, title, album_type, release_date, upc, total_tracks,
total_discs, cover_url, source, source_id)
VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10)
ON CONFLICT (id) DO UPDATE SET
title = EXCLUDED.title,
album_type = EXCLUDED.album_type,
release_date = EXCLUDED.release_date,
upc = EXCLUDED.upc,
total_tracks = EXCLUDED.total_tracks,
total_discs = EXCLUDED.total_discs,
cover_url = EXCLUDED.cover_url,
updated_at = now()`
_, err = tx.Exec(ctx, query,
album.ID, album.Title, nullString(album.Type), album.ReleaseDate,
nullString(album.UPC), album.TotalTracks, album.TotalDiscs,
nullString(album.CoverURL), source, sourceID)
if err != nil {
return err
}
for _, ext := range album.ExternalIDs {
extQuery := `
INSERT INTO album_external_ids (album_id, source, source_id, url)
VALUES ($1, $2, $3, $4)
ON CONFLICT (album_id, source, source_id) DO UPDATE SET
url = EXCLUDED.url,
fetched_at = now()`
_, err = tx.Exec(ctx, extQuery, album.ID, ext.Source, ext.SourceID, nullString(ext.URL))
if err != nil {
return err
}
}
for _, ac := range album.Artists {
artistQuery := `
INSERT INTO album_artists (album_id, artist_id, role, position)
VALUES ($1, $2, $3, $4)
ON CONFLICT (album_id, artist_id, role) DO NOTHING`
_, err = tx.Exec(ctx, artistQuery, album.ID, ac.Artist.ID, ac.Role, ac.Position)
if err != nil {
return err
}
}
return tx.Commit(ctx)
}
func (r *AlbumRepository) scanAlbum(ctx context.Context, query string, args ...any) (*domain.Album, error) {
row := r.pool.QueryRow(ctx, query, args...)
return r.scanAlbumRow(row)
}
func (r *AlbumRepository) scanAlbumFromRow(row pgx.Row) (*domain.Album, error) {
return r.scanAlbumRow(row)
}
func (r *AlbumRepository) scanAlbumRow(row pgx.Row) (*domain.Album, error) {
var (
album domain.Album
albumType *string
releaseDate *time.Time
upc *string
totalTracks *int
totalDiscs *int
coverURL *string
source string
sourceID *string
)
err := row.Scan(
&album.ID, &album.Title, &albumType, &releaseDate, &upc,
&totalTracks, &totalDiscs, &coverURL, &source, &sourceID,
)
if errors.Is(err, pgx.ErrNoRows) {
return nil, repository.ErrNotFound
}
if err != nil {
return nil, err
}
album.Type = derefString(albumType)
album.ReleaseDate = releaseDate
album.UPC = derefString(upc)
if totalTracks != nil {
album.TotalTracks = *totalTracks
}
if totalDiscs != nil {
album.TotalDiscs = *totalDiscs
}
album.CoverURL = derefString(coverURL)
return &album, nil
}
func (r *AlbumRepository) loadRelations(ctx context.Context, album *domain.Album) error {
extQuery := `SELECT source, source_id, url FROM album_external_ids WHERE album_id = $1`
rows, err := r.pool.Query(ctx, extQuery, album.ID)
if err != nil {
return err
}
defer rows.Close()
for rows.Next() {
var ext domain.ExternalID
var url *string
if err := rows.Scan(&ext.Source, &ext.SourceID, &url); err != nil {
return err
}
ext.URL = derefString(url)
album.ExternalIDs = append(album.ExternalIDs, ext)
}
return rows.Err()
}
+260
View File
@@ -0,0 +1,260 @@
package postgres
import (
"context"
"errors"
"time"
"github.com/jackc/pgx/v5"
"github.com/jackc/pgx/v5/pgxpool"
"github.com/metadata-agregator/internal/domain"
"github.com/metadata-agregator/internal/repository"
)
type ArtistRepository struct {
pool *pgxpool.Pool
}
func NewArtistRepository(pool *pgxpool.Pool) *ArtistRepository {
return &ArtistRepository{pool: pool}
}
func (r *ArtistRepository) GetByID(ctx context.Context, id string) (*domain.Artist, error) {
query := `
SELECT id, name, sort_name, artist_type, country, formed_date, disbanded_date,
description, image_url, source, source_id
FROM artists
WHERE id = $1`
artist, err := r.scanArtist(ctx, query, id)
if err != nil {
return nil, err
}
if err := r.loadExternalIDs(ctx, artist); err != nil {
return nil, err
}
return artist, nil
}
func (r *ArtistRepository) GetByExternalID(ctx context.Context, source, sourceID string) (*domain.Artist, error) {
query := `
SELECT a.id, a.name, a.sort_name, a.artist_type, a.country, a.formed_date,
a.disbanded_date, a.description, a.image_url, a.source, a.source_id
FROM artists a
JOIN artist_external_ids e ON a.id = e.artist_id
WHERE e.source = $1 AND e.source_id = $2`
artist, err := r.scanArtist(ctx, query, source, sourceID)
if err != nil {
return nil, err
}
if err := r.loadExternalIDs(ctx, artist); err != nil {
return nil, err
}
return artist, nil
}
func (r *ArtistRepository) Search(ctx context.Context, query string, limit, offset int) (*domain.SearchResult[domain.Artist], error) {
countQuery := `SELECT COUNT(*) FROM artists WHERE name ILIKE $1`
searchQuery := `
SELECT id, name, sort_name, artist_type, country, formed_date, disbanded_date,
description, image_url, source, source_id
FROM artists
WHERE name ILIKE $1
ORDER BY name
LIMIT $2 OFFSET $3`
pattern := "%" + query + "%"
var total int
if err := r.pool.QueryRow(ctx, countQuery, pattern).Scan(&total); err != nil {
return nil, err
}
rows, err := r.pool.Query(ctx, searchQuery, pattern, limit, offset)
if err != nil {
return nil, err
}
defer rows.Close()
var artists []domain.Artist
for rows.Next() {
artist, err := r.scanArtistFromRow(rows)
if err != nil {
return nil, err
}
artists = append(artists, *artist)
}
return &domain.SearchResult[domain.Artist]{
Items: artists,
Total: total,
Limit: limit,
Offset: offset,
}, nil
}
func (r *ArtistRepository) Save(ctx context.Context, artist *domain.Artist) error {
tx, err := r.pool.Begin(ctx)
if err != nil {
return err
}
defer tx.Rollback(ctx)
var source, sourceID string
if len(artist.ExternalIDs) > 0 {
source = artist.ExternalIDs[0].Source
sourceID = artist.ExternalIDs[0].SourceID
}
query := `
INSERT INTO artists (id, name, sort_name, artist_type, country, formed_date,
disbanded_date, description, image_url, source, source_id)
VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10, $11)
ON CONFLICT (id) DO UPDATE SET
name = EXCLUDED.name,
sort_name = EXCLUDED.sort_name,
artist_type = EXCLUDED.artist_type,
country = EXCLUDED.country,
formed_date = EXCLUDED.formed_date,
disbanded_date = EXCLUDED.disbanded_date,
description = EXCLUDED.description,
image_url = EXCLUDED.image_url,
updated_at = now()`
_, err = tx.Exec(ctx, query,
artist.ID, artist.Name, nullString(artist.SortName), nullString(artist.Type),
nullString(artist.Country), artist.FormedDate, artist.DisbandedDate,
nullString(artist.Description), nullString(artist.ImageURL), source, sourceID)
if err != nil {
return err
}
for _, ext := range artist.ExternalIDs {
extQuery := `
INSERT INTO artist_external_ids (artist_id, source, source_id, url)
VALUES ($1, $2, $3, $4)
ON CONFLICT (artist_id, source, source_id) DO UPDATE SET
url = EXCLUDED.url,
fetched_at = now()`
_, err = tx.Exec(ctx, extQuery, artist.ID, ext.Source, ext.SourceID, nullString(ext.URL))
if err != nil {
return err
}
}
return tx.Commit(ctx)
}
func (r *ArtistRepository) scanArtist(ctx context.Context, query string, args ...any) (*domain.Artist, error) {
row := r.pool.QueryRow(ctx, query, args...)
var (
artist domain.Artist
sortName *string
artistType *string
country *string
formedDate *time.Time
disbandDate *time.Time
description *string
imageURL *string
source string
sourceID *string
)
err := row.Scan(
&artist.ID, &artist.Name, &sortName, &artistType, &country,
&formedDate, &disbandDate, &description, &imageURL, &source, &sourceID,
)
if errors.Is(err, pgx.ErrNoRows) {
return nil, repository.ErrNotFound
}
if err != nil {
return nil, err
}
artist.SortName = derefString(sortName)
artist.Type = derefString(artistType)
artist.Country = derefString(country)
artist.FormedDate = formedDate
artist.DisbandedDate = disbandDate
artist.Description = derefString(description)
artist.ImageURL = derefString(imageURL)
return &artist, nil
}
func (r *ArtistRepository) scanArtistFromRow(row pgx.Row) (*domain.Artist, error) {
var (
artist domain.Artist
sortName *string
artistType *string
country *string
formedDate *time.Time
disbandDate *time.Time
description *string
imageURL *string
source string
sourceID *string
)
err := row.Scan(
&artist.ID, &artist.Name, &sortName, &artistType, &country,
&formedDate, &disbandDate, &description, &imageURL, &source, &sourceID,
)
if err != nil {
return nil, err
}
artist.SortName = derefString(sortName)
artist.Type = derefString(artistType)
artist.Country = derefString(country)
artist.FormedDate = formedDate
artist.DisbandedDate = disbandDate
artist.Description = derefString(description)
artist.ImageURL = derefString(imageURL)
return &artist, nil
}
func (r *ArtistRepository) loadExternalIDs(ctx context.Context, artist *domain.Artist) error {
query := `SELECT source, source_id, url FROM artist_external_ids WHERE artist_id = $1`
rows, err := r.pool.Query(ctx, query, artist.ID)
if err != nil {
return err
}
defer rows.Close()
for rows.Next() {
var ext domain.ExternalID
var url *string
if err := rows.Scan(&ext.Source, &ext.SourceID, &url); err != nil {
return err
}
ext.URL = derefString(url)
artist.ExternalIDs = append(artist.ExternalIDs, ext)
}
return rows.Err()
}
func nullString(s string) *string {
if s == "" {
return nil
}
return &s
}
func derefString(s *string) string {
if s == nil {
return ""
}
return *s
}
+226
View File
@@ -0,0 +1,226 @@
package postgres
import (
"context"
"errors"
"github.com/jackc/pgx/v5"
"github.com/jackc/pgx/v5/pgxpool"
"github.com/metadata-agregator/internal/domain"
"github.com/metadata-agregator/internal/repository"
)
type TrackRepository struct {
pool *pgxpool.Pool
}
func NewTrackRepository(pool *pgxpool.Pool) *TrackRepository {
return &TrackRepository{pool: pool}
}
func (r *TrackRepository) GetByID(ctx context.Context, id string) (*domain.Track, error) {
query := `
SELECT id, title, duration_ms, isrc, explicit, source, source_id
FROM tracks
WHERE id = $1`
track, err := r.scanTrack(ctx, query, id)
if err != nil {
return nil, err
}
if err := r.loadExternalIDs(ctx, track); err != nil {
return nil, err
}
return track, nil
}
func (r *TrackRepository) GetByExternalID(ctx context.Context, source, sourceID string) (*domain.Track, error) {
query := `
SELECT t.id, t.title, t.duration_ms, t.isrc, t.explicit, t.source, t.source_id
FROM tracks t
JOIN track_external_ids e ON t.id = e.track_id
WHERE e.source = $1 AND e.source_id = $2`
track, err := r.scanTrack(ctx, query, source, sourceID)
if err != nil {
return nil, err
}
if err := r.loadExternalIDs(ctx, track); err != nil {
return nil, err
}
return track, nil
}
func (r *TrackRepository) GetByISRC(ctx context.Context, isrc string) (*domain.Track, error) {
query := `
SELECT id, title, duration_ms, isrc, explicit, source, source_id
FROM tracks
WHERE isrc = $1`
track, err := r.scanTrack(ctx, query, isrc)
if err != nil {
return nil, err
}
if err := r.loadExternalIDs(ctx, track); err != nil {
return nil, err
}
return track, nil
}
func (r *TrackRepository) GetByAlbumID(ctx context.Context, albumID string) ([]domain.Track, error) {
query := `
SELECT t.id, t.title, t.duration_ms, t.isrc, t.explicit, t.source, t.source_id,
at.disc_number, at.track_number
FROM tracks t
JOIN album_tracks at ON t.id = at.track_id
JOIN album_external_ids ae ON at.album_id = ae.album_id
WHERE ae.source_id = $1
ORDER BY at.disc_number, at.track_number`
rows, err := r.pool.Query(ctx, query, albumID)
if err != nil {
return nil, err
}
defer rows.Close()
var tracks []domain.Track
for rows.Next() {
var (
track domain.Track
durationMs *int
isrc *string
explicit *bool
source string
sourceID *string
)
err := rows.Scan(
&track.ID, &track.Title, &durationMs, &isrc, &explicit,
&source, &sourceID, &track.DiscNumber, &track.TrackNumber,
)
if err != nil {
return nil, err
}
if durationMs != nil {
track.DurationMs = *durationMs
}
track.ISRC = derefString(isrc)
if explicit != nil {
track.Explicit = *explicit
}
tracks = append(tracks, track)
}
return tracks, rows.Err()
}
func (r *TrackRepository) Save(ctx context.Context, track *domain.Track) error {
tx, err := r.pool.Begin(ctx)
if err != nil {
return err
}
defer tx.Rollback(ctx)
var source, sourceID string
if len(track.ExternalIDs) > 0 {
source = track.ExternalIDs[0].Source
sourceID = track.ExternalIDs[0].SourceID
}
query := `
INSERT INTO tracks (id, title, duration_ms, isrc, explicit, source, source_id)
VALUES ($1, $2, $3, $4, $5, $6, $7)
ON CONFLICT (id) DO UPDATE SET
title = EXCLUDED.title,
duration_ms = EXCLUDED.duration_ms,
isrc = EXCLUDED.isrc,
explicit = EXCLUDED.explicit,
updated_at = now()`
_, err = tx.Exec(ctx, query,
track.ID, track.Title, track.DurationMs, nullString(track.ISRC),
track.Explicit, source, sourceID)
if err != nil {
return err
}
for _, ext := range track.ExternalIDs {
extQuery := `
INSERT INTO track_external_ids (track_id, source, source_id, url)
VALUES ($1, $2, $3, $4)
ON CONFLICT (track_id, source, source_id) DO UPDATE SET
url = EXCLUDED.url,
fetched_at = now()`
_, err = tx.Exec(ctx, extQuery, track.ID, ext.Source, ext.SourceID, nullString(ext.URL))
if err != nil {
return err
}
}
return tx.Commit(ctx)
}
func (r *TrackRepository) scanTrack(ctx context.Context, query string, args ...any) (*domain.Track, error) {
row := r.pool.QueryRow(ctx, query, args...)
var (
track domain.Track
durationMs *int
isrc *string
explicit *bool
source string
sourceID *string
)
err := row.Scan(
&track.ID, &track.Title, &durationMs, &isrc, &explicit, &source, &sourceID,
)
if errors.Is(err, pgx.ErrNoRows) {
return nil, repository.ErrNotFound
}
if err != nil {
return nil, err
}
if durationMs != nil {
track.DurationMs = *durationMs
}
track.ISRC = derefString(isrc)
if explicit != nil {
track.Explicit = *explicit
}
return &track, nil
}
func (r *TrackRepository) loadExternalIDs(ctx context.Context, track *domain.Track) error {
query := `SELECT source, source_id, url FROM track_external_ids WHERE track_id = $1`
rows, err := r.pool.Query(ctx, query, track.ID)
if err != nil {
return err
}
defer rows.Close()
for rows.Next() {
var ext domain.ExternalID
var url *string
if err := rows.Scan(&ext.Source, &ext.SourceID, &url); err != nil {
return err
}
ext.URL = derefString(url)
track.ExternalIDs = append(track.ExternalIDs, ext)
}
return rows.Err()
}
+29
View File
@@ -0,0 +1,29 @@
package repository
import (
"context"
"github.com/metadata-agregator/internal/domain"
)
type ArtistRepository interface {
GetByID(ctx context.Context, id string) (*domain.Artist, error)
GetByExternalID(ctx context.Context, source, sourceID string) (*domain.Artist, error)
Search(ctx context.Context, query string, limit, offset int) (*domain.SearchResult[domain.Artist], error)
Save(ctx context.Context, artist *domain.Artist) error
}
type AlbumRepository interface {
GetByID(ctx context.Context, id string) (*domain.Album, error)
GetByExternalID(ctx context.Context, source, sourceID string) (*domain.Album, error)
GetByArtistID(ctx context.Context, artistID string, limit, offset int) (*domain.SearchResult[domain.Album], error)
Save(ctx context.Context, album *domain.Album) error
}
type TrackRepository interface {
GetByID(ctx context.Context, id string) (*domain.Track, error)
GetByExternalID(ctx context.Context, source, sourceID string) (*domain.Track, error)
GetByISRC(ctx context.Context, isrc string) (*domain.Track, error)
GetByAlbumID(ctx context.Context, albumID string) ([]domain.Track, error)
Save(ctx context.Context, track *domain.Track) error
}
+142
View File
@@ -0,0 +1,142 @@
package service
import (
"context"
"errors"
"github.com/metadata-agregator/internal/domain"
"github.com/metadata-agregator/internal/provider"
"github.com/metadata-agregator/internal/repository"
)
type MetadataService struct {
artists repository.ArtistRepository
albums repository.AlbumRepository
tracks repository.TrackRepository
provider provider.Provider
}
func NewMetadataService(
artists repository.ArtistRepository,
albums repository.AlbumRepository,
tracks repository.TrackRepository,
prov provider.Provider,
) *MetadataService {
return &MetadataService{
artists: artists,
albums: albums,
tracks: tracks,
provider: prov,
}
}
func (s *MetadataService) GetArtist(ctx context.Context, id string) (*domain.Artist, error) {
artist, err := s.artists.GetByExternalID(ctx, s.provider.Name(), id)
if err == nil {
return artist, nil
}
if !errors.Is(err, repository.ErrNotFound) {
return nil, err
}
artist, err = s.provider.GetArtist(ctx, id)
if err != nil {
return nil, err
}
if saveErr := s.artists.Save(ctx, artist); saveErr != nil {
return artist, nil
}
return artist, nil
}
func (s *MetadataService) SearchArtists(ctx context.Context, query string, limit, offset int) (*domain.SearchResult[domain.Artist], error) {
result, err := s.artists.Search(ctx, query, limit, offset)
if err == nil && len(result.Items) > 0 {
return result, nil
}
return s.provider.SearchArtists(ctx, query, limit, offset)
}
func (s *MetadataService) GetAlbum(ctx context.Context, id string) (*domain.Album, error) {
album, err := s.albums.GetByExternalID(ctx, s.provider.Name(), id)
if err == nil {
return album, nil
}
if !errors.Is(err, repository.ErrNotFound) {
return nil, err
}
album, err = s.provider.GetAlbum(ctx, id)
if err != nil {
return nil, err
}
if saveErr := s.albums.Save(ctx, album); saveErr != nil {
return album, nil
}
return album, nil
}
func (s *MetadataService) GetArtistAlbums(ctx context.Context, artistID string, limit, offset int) (*domain.SearchResult[domain.Album], error) {
result, err := s.albums.GetByArtistID(ctx, artistID, limit, offset)
if err == nil && len(result.Items) > 0 {
return result, nil
}
return s.provider.GetArtistAlbums(ctx, artistID, limit, offset)
}
func (s *MetadataService) GetTrack(ctx context.Context, id string) (*domain.Track, error) {
track, err := s.tracks.GetByExternalID(ctx, s.provider.Name(), id)
if err == nil {
return track, nil
}
if !errors.Is(err, repository.ErrNotFound) {
return nil, err
}
track, err = s.provider.GetTrack(ctx, id)
if err != nil {
return nil, err
}
if saveErr := s.tracks.Save(ctx, track); saveErr != nil {
return track, nil
}
return track, nil
}
func (s *MetadataService) GetTrackByISRC(ctx context.Context, isrc string) (*domain.Track, error) {
track, err := s.tracks.GetByISRC(ctx, isrc)
if err == nil {
return track, nil
}
if !errors.Is(err, repository.ErrNotFound) {
return nil, err
}
track, err = s.provider.GetTrackByISRC(ctx, isrc)
if err != nil {
return nil, err
}
if saveErr := s.tracks.Save(ctx, track); saveErr != nil {
return track, nil
}
return track, nil
}
func (s *MetadataService) GetAlbumTracks(ctx context.Context, albumID string) ([]domain.Track, error) {
tracks, err := s.tracks.GetByAlbumID(ctx, albumID)
if err == nil && len(tracks) > 0 {
return tracks, nil
}
return s.provider.GetAlbumTracks(ctx, albumID)
}