|
| 1 | +import { type LyricLine, parseLrc } from "@applemusic-like-lyrics/lyric"; |
| 2 | + |
| 3 | +/** |
| 4 | + * LRC 格式类型 |
| 5 | + */ |
| 6 | +export enum LrcFormat { |
| 7 | + /** 普通逐行 LRC */ |
| 8 | + Line = "line", |
| 9 | + /** 逐字 LRC：[00:28.850]曲[00:32.455]：[00:36.060]钱 */ |
| 10 | + WordByWord = "word-by-word", |
| 11 | + /** 增强型 LRC (ESLyric)：[01:37.305]<01:37.624>怕<01:37.943>你 */ |
| 12 | + Enhanced = "enhanced", |
| 13 | +} |
| 14 | + |
| 15 | +/** LyricWord 类型 */ |
| 16 | +type LyricWord = { word: string; startTime: number; endTime: number; romanWord: string }; |
| 17 | + |
| 18 | +// 预编译正则表达式 |
| 19 | +const META_TAG_REGEX = /^\[[a-z]+:/i; |
| 20 | +const TIME_TAG_REGEX = /\[(\d{2}):(\d{2})\.(\d{2,3})\]/g; |
| 21 | +const ENHANCED_TIME_TAG_REGEX = /<(\d{2}):(\d{2})\.(\d{2,3})>/; |
| 22 | +const WORD_BY_WORD_REGEX = /\[(\d{2}):(\d{2})\.(\d{2,3})\]([^\[\]]*)/g; |
| 23 | +const ENHANCED_WORD_REGEX = /<(\d{2}):(\d{2})\.(\d{2,3})>([^<]*)/g; |
| 24 | +const LINE_TIME_REGEX = /^\[(\d{2}):(\d{2})\.(\d{2,3})\]/; |
| 25 | + |
| 26 | +/** |
| 27 | + * 解析时间戳为毫秒 |
| 28 | + */ |
| 29 | +const parseTimeToMs = (min: string, sec: string, ms: string): number => { |
| 30 | + const minutes = parseInt(min, 10); |
| 31 | + const seconds = parseInt(sec, 10); |
| 32 | + const milliseconds = ms.length === 2 ? parseInt(ms, 10) * 10 : parseInt(ms, 10); |
| 33 | + return minutes * 60 * 1000 + seconds * 1000 + milliseconds; |
| 34 | +}; |
| 35 | + |
| 36 | +/** |
| 37 | + * 创建 LyricWord 对象 |
| 38 | + */ |
| 39 | +const createWord = (word: string, startTime: number, endTime: number = startTime): LyricWord => ({ |
| 40 | + word, |
| 41 | + startTime, |
| 42 | + endTime, |
| 43 | + romanWord: "", |
| 44 | +}); |
| 45 | + |
| 46 | +/** |
| 47 | + * 创建 LyricLine 对象 |
| 48 | + */ |
| 49 | +const createLine = (words: LyricWord[], startTime: number, endTime: number = 0): LyricLine => ({ |
| 50 | + words, |
| 51 | + startTime, |
| 52 | + endTime, |
| 53 | + translatedLyric: "", |
| 54 | + romanLyric: "", |
| 55 | + isBG: false, |
| 56 | + isDuet: false, |
| 57 | +}); |
| 58 | + |
| 59 | +/** |
| 60 | + * 修正歌词行的结束时间 |
| 61 | + * 每行最后一个字的结束时间 = 下一行的开始时间 |
| 62 | + */ |
| 63 | +const fixLineEndTimes = (lines: LyricLine[]): void => { |
| 64 | + const len = lines.length; |
| 65 | + for (let i = 0; i < len; i++) { |
| 66 | + const line = lines[i]; |
| 67 | + const lastWord = line.words[line.words.length - 1]; |
| 68 | + const nextLineStart = lines[i + 1]?.startTime; |
| 69 | + // 如果有下一行，使用下一行的开始时间；否则使用最后一个字开始时间 + 1s |
| 70 | + lastWord.endTime = nextLineStart ?? lastWord.startTime + 1000; |
| 71 | + line.endTime = lastWord.endTime; |
| 72 | + } |
| 73 | +}; |
| 74 | + |
| 75 | +/** |
| 76 | + * 检测 LRC 格式类型 |
| 77 | + */ |
| 78 | +export const detectLrcFormat = (content: string): LrcFormat => { |
| 79 | + const lines = content.split(/\r?\n/); |
| 80 | + for (const rawLine of lines) { |
| 81 | + const line = rawLine.trim(); |
| 82 | + if (!line || META_TAG_REGEX.test(line)) continue; |
| 83 | + // 检查增强型LRC |
| 84 | + if (ENHANCED_TIME_TAG_REGEX.test(line)) { |
| 85 | + return LrcFormat.Enhanced; |
| 86 | + } |
| 87 | + // 检查逐字LRC |
| 88 | + const matches = line.match(TIME_TAG_REGEX); |
| 89 | + if (matches && matches.length > 1) { |
| 90 | + return LrcFormat.WordByWord; |
| 91 | + } |
| 92 | + } |
| 93 | + return LrcFormat.Line; |
| 94 | +}; |
| 95 | + |
| 96 | +/** |
| 97 | + * 解析逐字 LRC 格式 |
| 98 | + */ |
| 99 | +export const parseWordByWordLrc = (content: string): LyricLine[] => { |
| 100 | + const result: LyricLine[] = []; |
| 101 | + |
| 102 | + for (const rawLine of content.split(/\r?\n/)) { |
| 103 | + const line = rawLine.trim(); |
| 104 | + if (!line || META_TAG_REGEX.test(line)) continue; |
| 105 | + |
| 106 | + const words: LyricWord[] = []; |
| 107 | + let lineStartTime = Infinity; |
| 108 | + let match: RegExpExecArray | null; |
| 109 | + |
| 110 | + // 重置正则状态 |
| 111 | + WORD_BY_WORD_REGEX.lastIndex = 0; |
| 112 | + |
| 113 | + while ((match = WORD_BY_WORD_REGEX.exec(line)) !== null) { |
| 114 | + const startTime = parseTimeToMs(match[1], match[2], match[3]); |
| 115 | + const word = match[4]; |
| 116 | + |
| 117 | + if (!word && words.length === 0) continue; |
| 118 | + |
| 119 | + lineStartTime = Math.min(lineStartTime, startTime); |
| 120 | + |
| 121 | + // 上一个字的结束时间 = 当前字的开始时间 |
| 122 | + if (words.length > 0) { |
| 123 | + words[words.length - 1].endTime = startTime; |
| 124 | + } |
| 125 | + |
| 126 | + if (word) { |
| 127 | + words.push(createWord(word, startTime)); |
| 128 | + } |
| 129 | + } |
| 130 | + |
| 131 | + if (words.length > 0) { |
| 132 | + result.push(createLine(words, lineStartTime === Infinity ? 0 : lineStartTime)); |
| 133 | + } |
| 134 | + } |
| 135 | + |
| 136 | + fixLineEndTimes(result); |
| 137 | + return result; |
| 138 | +}; |
| 139 | + |
| 140 | +/** |
| 141 | + * 解析增强型 LRC 格式 (ESLyric) |
| 142 | + */ |
| 143 | +export const parseEnhancedLrc = (content: string): LyricLine[] => { |
| 144 | + const result: LyricLine[] = []; |
| 145 | + |
| 146 | + for (const rawLine of content.split(/\r?\n/)) { |
| 147 | + const line = rawLine.trim(); |
| 148 | + if (!line || META_TAG_REGEX.test(line)) continue; |
| 149 | + |
| 150 | + const lineTimeMatch = LINE_TIME_REGEX.exec(line); |
| 151 | + if (!lineTimeMatch) continue; |
| 152 | + |
| 153 | + const lineStartTime = parseTimeToMs(lineTimeMatch[1], lineTimeMatch[2], lineTimeMatch[3]); |
| 154 | + const contentAfterTime = line.slice(lineTimeMatch[0].length); |
| 155 | + |
| 156 | + const words: LyricWord[] = []; |
| 157 | + |
| 158 | + // 检查是否有增强型标记 |
| 159 | + if (ENHANCED_TIME_TAG_REGEX.test(contentAfterTime)) { |
| 160 | + let match: RegExpExecArray | null; |
| 161 | + ENHANCED_WORD_REGEX.lastIndex = 0; |
| 162 | + |
| 163 | + while ((match = ENHANCED_WORD_REGEX.exec(contentAfterTime)) !== null) { |
| 164 | + const startTime = parseTimeToMs(match[1], match[2], match[3]); |
| 165 | + const word = match[4]; |
| 166 | + |
| 167 | + if (words.length > 0) { |
| 168 | + words[words.length - 1].endTime = startTime; |
| 169 | + } |
| 170 | + |
| 171 | + if (word) { |
| 172 | + words.push(createWord(word, startTime)); |
| 173 | + } |
| 174 | + } |
| 175 | + } else { |
| 176 | + // 无增强型标记，作为整行处理 |
| 177 | + const text = contentAfterTime.trim(); |
| 178 | + if (text) { |
| 179 | + words.push(createWord(text, lineStartTime)); |
| 180 | + } |
| 181 | + } |
| 182 | + |
| 183 | + if (words.length > 0) { |
| 184 | + result.push(createLine(words, lineStartTime)); |
| 185 | + } |
| 186 | + } |
| 187 | + |
| 188 | + fixLineEndTimes(result); |
| 189 | + return result; |
| 190 | +}; |
| 191 | + |
| 192 | +/** |
| 193 | + * 智能解析 LRC 歌词 |
| 194 | + */ |
| 195 | +export const parseSmartLrc = (content: string): { format: LrcFormat; lines: LyricLine[] } => { |
| 196 | + const format = detectLrcFormat(content); |
| 197 | + |
| 198 | + let lines: LyricLine[]; |
| 199 | + switch (format) { |
| 200 | + case LrcFormat.WordByWord: |
| 201 | + lines = parseWordByWordLrc(content); |
| 202 | + break; |
| 203 | + case LrcFormat.Enhanced: |
| 204 | + lines = parseEnhancedLrc(content); |
| 205 | + break; |
| 206 | + default: |
| 207 | + lines = parseLrc(content) || []; |
| 208 | + } |
| 209 | + |
| 210 | + console.log(`[LyricParser] 检测到歌词格式: ${format}, 共 ${lines.length} 行`); |
| 211 | + return { format, lines }; |
| 212 | +}; |
| 213 | + |
| 214 | +/** |
| 215 | + * 判断解析结果是否为逐字格式 |
| 216 | + */ |
| 217 | +export const isWordLevelFormat = (format: LrcFormat): boolean => |
| 218 | + format === LrcFormat.WordByWord || format === LrcFormat.Enhanced; |
0 commit comments