Files
KVideo/lib/utils/m3u-parser.ts
2026-04-02 22:14:46 +08:00

432 lines
12 KiB
TypeScript

/**
* M3U Playlist Parser
* Parses M3U/M3U8 IPTV playlist format and JSON channel lists
*/
export interface M3UChannel {
name: string;
url: string;
logo?: string;
group?: string;
tvgId?: string;
tvgName?: string;
routes?: string[];
sourceId?: string;
sourceName?: string;
httpUserAgent?: string;
httpReferrer?: string;
}
export interface M3UPlaylist {
channels: M3UChannel[];
groups: string[];
}
export interface PlaylistReference {
kind: 'playlist' | 'config';
name: string;
url: string;
httpUserAgent?: string;
httpReferrer?: string;
}
const STREAM_URL_RE = /^(https?:\/\/|rtmp:\/\/|rtsp:\/\/|udp:\/\/|rtp:\/\/|mms:\/\/|ftp:\/\/|file:\/\/|\/\/)/i;
function isRecord(value: unknown): value is Record<string, unknown> {
return typeof value === 'object' && value !== null;
}
function firstString(...values: unknown[]): string | undefined {
for (const value of values) {
if (typeof value === 'string') {
const trimmed = value.trim();
if (trimmed) return trimmed;
}
}
return undefined;
}
function unwrapProxyUrl(target: string): string {
const trimmed = target.trim();
if (!trimmed.startsWith('proxy://')) {
return trimmed;
}
const match = trimmed.match(/(?:^|[?&])(?:ext|url)=([^&]+)/i);
if (!match) {
return trimmed;
}
try {
return decodeURIComponent(match[1]);
} catch {
return match[1];
}
}
function looksLikeRelativeMediaPath(value: string): boolean {
return value.startsWith('/') ||
value.startsWith('./') ||
value.startsWith('../') ||
value.endsWith('.m3u8') ||
value.endsWith('.m3u') ||
value.includes('.m3u8?') ||
value.includes('.mp4');
}
function resolveReferenceUrl(baseUrl: string | undefined, target: string): string {
const normalizedTarget = unwrapProxyUrl(target);
if (!baseUrl) return normalizedTarget;
try {
return new URL(normalizedTarget, baseUrl).toString();
} catch {
return normalizedTarget;
}
}
function countStreamMatches(value: string): number {
const matches = value.match(/(?:https?:\/\/|rtmp:\/\/|rtsp:\/\/|udp:\/\/|rtp:\/\/|mms:\/\/|ftp:\/\/|file:\/\/|\/\/)/gi);
return matches ? matches.length : 0;
}
function normalizeRouteCandidate(rawValue: string): string | null {
const trimmed = unwrapProxyUrl(rawValue);
if (!trimmed) {
return null;
}
const fromDollar = trimmed.includes('$') ? trimmed.slice(trimmed.lastIndexOf('$') + 1).trim() : trimmed;
if (STREAM_URL_RE.test(fromDollar) || looksLikeRelativeMediaPath(fromDollar)) {
return fromDollar;
}
const fromComma = fromDollar.includes(',') ? fromDollar.slice(fromDollar.lastIndexOf(',') + 1).trim() : fromDollar;
if (STREAM_URL_RE.test(fromComma) || looksLikeRelativeMediaPath(fromComma)) {
return fromComma;
}
return STREAM_URL_RE.test(trimmed) || looksLikeRelativeMediaPath(trimmed) ? trimmed : null;
}
function extractRoutes(value: unknown, baseUrl?: string): string[] {
if (Array.isArray(value)) {
return Array.from(new Set(
value.flatMap((item) => extractRoutes(item, baseUrl))
));
}
if (typeof value !== 'string') {
if (isRecord(value)) {
return extractRoutes(firstString(value.url, value.src, value.link), baseUrl);
}
return [];
}
const segments = value
.split(/\r?\n/)
.map((part) => part.trim())
.filter(Boolean)
.flatMap((segment) => (
segment.includes('#') && countStreamMatches(segment) > 1
? segment.split('#').map((part) => part.trim()).filter(Boolean)
: [segment]
));
return Array.from(new Set(
segments
.map(normalizeRouteCandidate)
.filter((route): route is string => !!route)
.map((route) => resolveReferenceUrl(baseUrl, route))
));
}
function buildChannel(entry: Record<string, unknown>, baseUrl?: string, inheritedGroup?: string): M3UChannel | null {
const name = firstString(
entry.name,
entry.title,
entry.channel_name,
entry.channel,
entry.tvg_name,
entry.tvgName
);
const routes = extractRoutes(
entry.urls ?? entry.url ?? entry.stream_url ?? entry.src ?? entry.link ?? entry.stream ?? entry.playUrl ?? entry.play_url,
baseUrl
);
if (!name || routes.length === 0) {
return null;
}
const group = firstString(entry.group, entry.group_title, entry.groupName, entry.category, inheritedGroup);
const channel: M3UChannel = {
name,
url: routes[0],
logo: firstString(entry.logo, entry.icon, entry.tvg_logo),
group,
tvgId: firstString(entry.tvg_id, entry.tvgId),
tvgName: firstString(entry.tvg_name, entry.tvgName),
httpUserAgent: firstString(entry.http_user_agent, entry.httpUserAgent, entry.user_agent, entry.userAgent, entry.ua),
httpReferrer: firstString(entry.http_referrer, entry.httpReferrer, entry.referer, entry.referrer),
};
if (routes.length > 1) {
channel.routes = routes;
}
return channel;
}
function getJsonChannelEntries(data: unknown): Array<{ entry: Record<string, unknown>; inheritedGroup?: string }> {
if (Array.isArray(data)) {
return data.filter(isRecord).map((entry) => ({ entry }));
}
if (!isRecord(data)) {
return [];
}
for (const candidate of [data.channels, data.list, data.items, data.data]) {
if (Array.isArray(candidate)) {
return candidate.filter(isRecord).map((entry) => ({ entry }));
}
}
if (Array.isArray(data.lives)) {
const nestedChannels: Array<{ entry: Record<string, unknown>; inheritedGroup?: string }> = [];
for (const liveEntry of data.lives) {
if (!isRecord(liveEntry) || !Array.isArray(liveEntry.channels)) {
continue;
}
const inheritedGroup = firstString(
liveEntry.group,
liveEntry.group_title,
liveEntry.groupName,
liveEntry.name,
liveEntry.title
);
for (const channelEntry of liveEntry.channels) {
if (isRecord(channelEntry)) {
nestedChannels.push({ entry: channelEntry, inheritedGroup });
}
}
}
if (nestedChannels.length > 0) {
return nestedChannels;
}
}
return [];
}
function getReferenceTargets(entry: Record<string, unknown>): string[] {
const directUrl = firstString(entry.url);
if (directUrl) {
return [directUrl];
}
if (Array.isArray(entry.urls)) {
return entry.urls.filter((value): value is string => typeof value === 'string' && value.trim().length > 0);
}
return [];
}
export function extractPlaylistReferences(content: string, baseUrl?: string): PlaylistReference[] {
try {
const data = JSON.parse(content);
if (!isRecord(data)) return [];
const parsedData = data as Record<string, unknown>;
const references: PlaylistReference[] = [];
if (Array.isArray(parsedData.lives)) {
for (const entry of parsedData.lives) {
if (!isRecord(entry) || Array.isArray(entry.channels)) continue;
const targets = getReferenceTargets(entry);
for (const target of targets) {
references.push({
kind: 'playlist',
name: firstString(entry.name, entry.title) || '直播源',
url: resolveReferenceUrl(baseUrl, target),
httpUserAgent: firstString(entry.ua, entry.userAgent, entry.http_user_agent, entry.httpUserAgent),
httpReferrer: firstString(entry.referer, entry.referrer, entry.http_referrer, entry.httpReferrer),
});
}
}
}
if (Array.isArray(parsedData.urls)) {
for (const entry of parsedData.urls) {
if (!isRecord(entry)) continue;
const targets = getReferenceTargets(entry);
for (const target of targets) {
references.push({
kind: 'config',
name: firstString(entry.name, entry.title) || '配置源',
url: resolveReferenceUrl(baseUrl, target),
});
}
}
}
return references;
} catch {
return [];
}
}
/**
* Try to parse content as JSON channel list.
* Supports formats:
* - Array of channel objects: [{ name, url, group?, logo?, ... }]
* - Object with channels/list field: { channels: [...] } or { list: [...] }
* - TVBox/OK-style lives groups with nested channels arrays
*/
function tryParseJSON(content: string, baseUrl?: string): M3UPlaylist | null {
try {
const data = JSON.parse(content);
const entries = getJsonChannelEntries(data);
const groupSet = new Set<string>();
const parsed: M3UChannel[] = [];
for (const { entry, inheritedGroup } of entries) {
const channel = buildChannel(entry, baseUrl, inheritedGroup);
if (!channel) continue;
if (channel.group) {
groupSet.add(channel.group);
}
parsed.push(channel);
}
if (parsed.length === 0) return null;
return {
channels: parsed,
groups: Array.from(groupSet).sort(),
};
} catch {
return null;
}
}
/**
* Parse M3U playlist content into structured data.
* Also supports JSON format channel lists.
*/
export function parseM3U(content: string, baseUrl?: string): M3UPlaylist {
const trimmed = content.trim();
// Try JSON first if it looks like JSON
if (trimmed.startsWith('{') || trimmed.startsWith('[')) {
const jsonResult = tryParseJSON(trimmed, baseUrl);
if (jsonResult) return jsonResult;
}
const lines = content.split('\n').map(l => l.trim()).filter(l => l.length > 0);
const channels: M3UChannel[] = [];
const groupSet = new Set<string>();
for (let i = 0; i < lines.length; i++) {
const line = lines[i];
if (line.startsWith('#EXTINF:')) {
// Parse EXTINF line
const channel: M3UChannel = { name: '', url: '' };
// Extract attributes from EXTINF
const tvgNameMatch = line.match(/tvg-name="([^"]*)"/i);
const tvgLogoMatch = line.match(/tvg-logo="([^"]*)"/i);
const groupTitleMatch = line.match(/group-title="([^"]*)"/i);
const tvgIdMatch = line.match(/tvg-id="([^"]*)"/i);
const httpUserAgentMatch = line.match(/http-user-agent="([^"]*)"/i);
const httpReferrerMatch = line.match(/http-referrer="([^"]*)"/i);
if (tvgNameMatch) channel.tvgName = tvgNameMatch[1];
if (tvgLogoMatch) channel.logo = tvgLogoMatch[1];
if (groupTitleMatch) {
channel.group = groupTitleMatch[1];
if (channel.group) groupSet.add(channel.group);
}
if (tvgIdMatch) channel.tvgId = tvgIdMatch[1];
if (httpUserAgentMatch) channel.httpUserAgent = httpUserAgentMatch[1];
if (httpReferrerMatch) channel.httpReferrer = httpReferrerMatch[1];
// Extract channel name (after last comma)
const commaIndex = line.lastIndexOf(',');
if (commaIndex !== -1) {
channel.name = line.substring(commaIndex + 1).trim();
}
// Next non-comment line should be the URL
for (let j = i + 1; j < lines.length; j++) {
if (!lines[j].startsWith('#')) {
channel.url = resolveReferenceUrl(baseUrl, lines[j]);
i = j; // Skip to after URL
break;
}
}
if (channel.name && channel.url) {
// Use tvgName as fallback for name
if (!channel.name && channel.tvgName) {
channel.name = channel.tvgName;
}
channels.push(channel);
}
}
}
// If no EXTINF entries were found, also try JSON as a fallback
if (channels.length === 0) {
const jsonResult = tryParseJSON(content, baseUrl);
if (jsonResult) return jsonResult;
}
return {
channels,
groups: Array.from(groupSet).sort(),
};
}
/**
* Group channels with the same name into single entries with multiple routes.
* This merges duplicate channel names (common in M3U playlists with multiple streams).
*/
export function groupChannelsByName(channels: M3UChannel[]): M3UChannel[] {
const groups = new Map<string, M3UChannel>();
for (const ch of channels) {
const key = `${ch.sourceId || ''}::${ch.name.toLowerCase().trim()}`;
const existing = groups.get(key);
if (existing) {
if (!existing.routes) {
existing.routes = [existing.url];
}
if (!existing.routes.includes(ch.url)) {
existing.routes.push(ch.url);
}
// Use first logo found
if (!existing.logo && ch.logo) existing.logo = ch.logo;
} else {
groups.set(key, { ...ch });
}
}
// Only add routes array when there are multiple routes
const result = Array.from(groups.values());
for (const ch of result) {
if (ch.routes && ch.routes.length <= 1) {
delete ch.routes;
}
}
return result;
}