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

package genius

import (
	"bytes"
	"encoding/json"
	"regexp"
	"strings"

	"golang.org/x/net/html"
)

// Parse only designated lyrics containers, never the entire page (navigation, ads, annotations).
func parseLyrics(body []byte) (string, error) {
	doc, err := html.Parse(bytes.NewReader(body))
	if err != nil {
		return "", err
	}
	var blocks []string
	var walk func(*html.Node)
	walk = func(n *html.Node) {
		if excluded(n) {
			return
		}
		if attr(n, "data-lyrics-container") == "true" {
			if text := nodeText(n); text != "" {
				blocks = append(blocks, text)
			}
			return
		}
		for child := n.FirstChild; child != nil; child = child.NextSibling {
			walk(child)
		}
	}
	walk(doc)
	if len(blocks) == 0 {
		if text := embeddedLyrics(doc); text != "" {
			return text, nil
		}
		return "", ErrLyricsUnavailable
	}
	return strings.Join(blocks, "\n\n"), nil
}

func attr(n *html.Node, name string) string {
	for _, a := range n.Attr {
		if a.Key == name {
			return a.Val
		}
	}
	return ""
}

func excluded(n *html.Node) bool {
	return n.Type == html.ElementNode && (n.Data == "script" || n.Data == "style" || n.Data == "button" ||
		attr(n, "data-exclude-from-selection") == "true" || attr(n, "aria-hidden") == "true" ||
		strings.Contains(attr(n, "class"), "LyricsHeader"))
}

var statePattern = regexp.MustCompile(`(?s)window\.__PRELOADED_STATE__\s*=\s*JSON\.parse\('((?:[^'\\]|\\.)*)'\)`)

// Read a JSON string literal only; never evaluate JavaScript from a remote page.
func embeddedLyrics(doc *html.Node) string {
	var result string
	var walk func(*html.Node)
	walk = func(n *html.Node) {
		if result != "" {
			return
		}
		if n.Type == html.ElementNode && n.Data == "script" && n.FirstChild != nil {
			match := statePattern.FindStringSubmatch(n.FirstChild.Data)
			if len(match) == 2 {
				decoded, ok := decodeJSString(match[1])
				if ok {
					var state struct {
						SongPage struct {
							LyricsData struct {
								Body struct {
									HTML string `json:"html"`
								} `json:"body"`
							} `json:"lyricsData"`
						} `json:"songPage"`
					}
					if json.Unmarshal([]byte(decoded), &state) == nil {
						fragment, err := html.Parse(strings.NewReader(strings.ReplaceAll(state.SongPage.LyricsData.Body.HTML, "\n", "")))
						if err == nil {
							result = nodeText(fragment)
						}
					}
				}
			}
			return
		}
		for child := n.FirstChild; child != nil; child = child.NextSibling {
			walk(child)
		}
	}
	walk(doc)
	return result
}

func decodeJSString(raw string) (string, bool) {
	var quoted strings.Builder
	quoted.WriteByte('"')
	for i := 0; i < len(raw); i++ {
		switch raw[i] {
		case '"':
			quoted.WriteString(`\"`)
		case '\\':
			i++
			if i >= len(raw) {
				return "", false
			}
			switch raw[i] {
			case '\'':
				quoted.WriteByte('\'')
			case 'x':
				if i+2 >= len(raw) {
					return "", false
				}
				quoted.WriteString(`\u00`)
				quoted.WriteString(raw[i+1 : i+3])
				i += 2
			default:
				quoted.WriteByte('\\')
				quoted.WriteByte(raw[i])
			}
		default:
			quoted.WriteByte(raw[i])
		}
	}
	quoted.WriteByte('"')
	var decoded string
	err := json.Unmarshal([]byte(quoted.String()), &decoded)
	return decoded, err == nil
}

func nodeText(root *html.Node) string {
	var out strings.Builder
	var walk func(*html.Node)
	walk = func(n *html.Node) {
		if excluded(n) {
			return
		}
		if n.Type == html.TextNode {
			out.WriteString(n.Data)
			return
		}
		if n.Type == html.ElementNode && n.Data == "br" {
			out.WriteByte('\n')
			return
		}
		block := n.Type == html.ElementNode && (n.Data == "p" || n.Data == "div") && n != root
		if block {
			out.WriteByte('\n')
		}
		for child := n.FirstChild; child != nil; child = child.NextSibling {
			walk(child)
		}
		if block {
			out.WriteByte('\n')
		}
	}
	walk(root)
	lines := strings.Split(strings.ReplaceAll(out.String(), "\r", ""), "\n")
	clean := make([]string, 0, len(lines))
	for _, line := range lines {
		line = strings.Join(strings.Fields(line), " ")
		if line == "" && (len(clean) == 0 || clean[len(clean)-1] == "") {
			continue
		}
		clean = append(clean, line)
	}
	return strings.TrimSpace(strings.Join(clean, "\n"))
}