pipu-music/services/api/internal/genius/metadata.go

package genius

import (
	"context"
	"encoding/json"
	"errors"
	"fmt"
	"net/url"
	"strings"
	"unicode"

	"pipu/api/internal/api/gen"
)

func normalize(s string) string {
	return strings.Join(strings.FieldsFunc(strings.ToLower(s), func(r rune) bool { return !unicode.IsLetter(r) && !unicode.IsNumber(r) }), " ")
}

// sameArtist сравнивает имя на Genius с нашим: у Genius часто «имя (транслитерация)»,
// поэтому скобочный хвост допустим.
func sameArtist(geniusName, name string) bool {
	if normalize(geniusName) == normalize(name) {
		return true
	}
	if i := strings.Index(geniusName, "("); i > 0 && strings.HasSuffix(strings.TrimSpace(geniusName), ")") {
		return normalize(geniusName[:i]) == normalize(name)
	}
	return false
}

// Exact normalized identity; retain version qualifiers such as live/remix.
func matchSong(songs []Song, artist, title string) (Song, bool) {
	var matched Song
	for _, s := range songs {
		if normalize(s.Title) == normalize(title) && sameArtist(s.Artist.Name, artist) {
			if matched.ID != 0 && matched.ID != s.ID {
				return Song{}, false
			}
			matched = s
		}
	}
	return matched, matched.ID != 0
}

func imageURL(raw string) string {
	u, err := url.Parse(strings.TrimSpace(raw))
	if err != nil || u.Scheme != "https" || u.User != nil || strings.Contains(strings.ToLower(u.Path), "default") {
		return ""
	}
	switch u.Host {
	case "images.genius.com", "images.rapgenius.com", "t2.genius.com":
		return u.String()
	default:
		return ""
	}
}

func firstImage(urls ...string) string {
	for _, raw := range urls {
		if u := imageURL(raw); u != "" {
			return u
		}
	}
	return ""
}

// songArtwork prefers the full song resource: search hits often omit album art and the artist image.
func songArtwork(song Song) (songImage, artistImage, albumImage string) {
	return firstImage(song.ImageURL, song.ThumbURL, song.HeaderURL), imageURL(song.Artist.ImageURL), firstImage(song.Album.ImageURL, song.Album.ThumbURL)
}

func urlPath(raw string) string {
	u, err := url.Parse(raw)
	if err != nil {
		return ""
	}
	return u.Path
}

func sourceURL(raw string) string {
	u, err := url.Parse(raw)
	if err != nil || u.Scheme != "https" || u.Host != "genius.com" || u.User != nil {
		return ""
	}
	return u.String()
}

func (c *Client) Song(ctx context.Context, id int64) (Song, error) {
	if id <= 0 {
		return Song{}, errors.New("genius: song id required")
	}
	body, err := c.get(ctx, fmt.Sprintf("https://api.genius.com/songs/%d?text_format=plain", id), true)
	if err != nil {
		return Song{}, err
	}
	var data struct {
		Response struct {
			Song Song `json:"song"`
		} `json:"response"`
	}
	if err = json.Unmarshal(body, &data); err != nil {
		return Song{}, fmt.Errorf("genius: decode song: %w", err)
	}
	if data.Response.Song.ID == 0 {
		return Song{}, errors.New("genius: song missing")
	}
	return data.Response.Song, nil
}

func (c *Client) Artist(ctx context.Context, id int64) (Artist, error) {
	if id <= 0 {
		return Artist{}, errors.New("genius: artist id required")
	}
	body, err := c.get(ctx, fmt.Sprintf("https://api.genius.com/artists/%d", id), true)
	if err != nil {
		return Artist{}, err
	}
	var data struct {
		Response struct {
			Artist Artist `json:"artist"`
		} `json:"response"`
	}
	if err = json.Unmarshal(body, &data); err != nil {
		return Artist{}, fmt.Errorf("genius: decode artist: %w", err)
	}
	if data.Response.Artist.ID == 0 {
		return Artist{}, errors.New("genius: artist missing")
	}
	return data.Response.Artist, nil
}

func (c *Client) Annotations(ctx context.Context, id int64) ([]gen.GeniusAnnotation, error) {
	out := []gen.GeniusAnnotation{}
	var seen int64
	haveSeen := false
	for page := 1; ; page++ {
		body, err := c.get(ctx, fmt.Sprintf("https://api.genius.com/referents?song_id=%d&text_format=plain&per_page=50&page=%d", id, page), true)
		if err != nil {
			return nil, err
		}
		var data struct {
			Response struct {
				Referents []struct {
					ID          int64  `json:"id"`
					Fragment    string `json:"fragment"`
					Annotations []struct {
						Body struct {
							Plain string `json:"plain"`
						} `json:"body"`
						URL      string `json:"url"`
						Verified bool   `json:"verified"`
						Authors  []struct {
							User struct {
								Name string `json:"name"`
							} `json:"user"`
						} `json:"authors"`
					} `json:"annotations"`
				} `json:"referents"`
			} `json:"response"`
		}
		if err := json.Unmarshal(body, &data); err != nil {
			return nil, fmt.Errorf("genius: decode annotations: %w", err)
		}
		refs := data.Response.Referents
		if len(refs) == 0 || (haveSeen && refs[0].ID == seen) {
			return out, nil
		}
		if refs[0].ID != 0 {
			seen = refs[0].ID
			haveSeen = true
		}
		for _, r := range refs {
			for _, a := range r.Annotations {
				if a.Body.Plain == "" {
					continue
				}
				names := []string{}
				for _, author := range a.Authors {
					if author.User.Name != "" {
						names = append(names, author.User.Name)
					}
				}
				out = append(out, gen.GeniusAnnotation{Fragment: r.Fragment, Text: a.Body.Plain, Url: sourceURL(a.URL), Verified: a.Verified, Authors: strings.Join(names, ", ")})
			}
		}
		if len(refs) < 50 {
			return out, nil
		}
		if page == 20 {
			return nil, fmt.Errorf("genius: annotation pagination limit")
		}
	}
}