veodl 1.8.3 → 1.9.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -0,0 +1,251 @@
1
+ import { spawn } from 'node:child_process';
2
+ import { mkdtemp, readFile, rm } from 'node:fs/promises';
3
+ import { tmpdir } from 'node:os';
4
+ import path from 'node:path';
5
+ import { setTimeout as delay } from 'node:timers/promises';
6
+ import { cleanText, validateUrl } from './utils.js';
7
+
8
+ const MEDIA_URL = /\.(?:m3u8|mpd|mp4|webm|mov)(?:[?#]|$)/i;
9
+ const MEDIA_TYPE = /(?:mpegurl|dash\+xml|video\/(?:mp4|webm|quicktime))/i;
10
+
11
+ function abortable(promise, signal) {
12
+ if (!signal) return promise;
13
+ if (signal.aborted) {
14
+ Promise.resolve(promise).catch(() => {});
15
+ throw signal.reason;
16
+ }
17
+ return new Promise((resolve, reject) => {
18
+ const onAbort = () => reject(signal.reason);
19
+ signal.addEventListener('abort', onAbort, { once: true });
20
+ Promise.resolve(promise).then(resolve, reject).finally(() => signal.removeEventListener('abort', onAbort));
21
+ });
22
+ }
23
+
24
+ export function parseSourceTimeout(value) {
25
+ const match = /^(\d+)(ms|s|m)?$/i.exec(String(value || '').trim());
26
+ if (!match) throw new Error('--timeout expects seconds, e.g. 30, 30s or 2m.');
27
+ const milliseconds = Number(match[1]) * ({ ms: 1, s: 1000, m: 60000 }[match[2]?.toLowerCase() || 's']);
28
+ if (!Number.isSafeInteger(milliseconds) || milliseconds < 5000 || milliseconds > 600000) {
29
+ throw new Error('--timeout must be between 5 seconds and 10 minutes.');
30
+ }
31
+ return milliseconds;
32
+ }
33
+
34
+ export function shouldOfferSourceDiscovery(error) {
35
+ if (error?.name === 'AbortError') return false;
36
+ return !/(?:HTTP (?:Error )?(?:401|403|404|429|5\d\d)|\b(?:DRM|live stream|rate limit|timed out|timeout|connection refused|login required|sign in|private video|geo.?block)\b|This URL is a collection)/i
37
+ .test(String(error?.message || error));
38
+ }
39
+
40
+ export function browserCandidates() {
41
+ const custom = process.env.VEO_BROWSER_PATH;
42
+ if (custom) return [custom];
43
+ if (process.platform === 'win32') {
44
+ const roots = [process.env.PROGRAMFILES, process.env['PROGRAMFILES(X86)'], process.env.LOCALAPPDATA].filter(Boolean);
45
+ return roots.flatMap(root => [
46
+ path.join(root, 'Google', 'Chrome', 'Application', 'chrome.exe'),
47
+ path.join(root, 'Microsoft', 'Edge', 'Application', 'msedge.exe'),
48
+ path.join(root, 'BraveSoftware', 'Brave-Browser', 'Application', 'brave.exe'),
49
+ ]);
50
+ }
51
+ if (process.platform === 'darwin') return [
52
+ '/Applications/Google Chrome.app/Contents/MacOS/Google Chrome',
53
+ '/Applications/Microsoft Edge.app/Contents/MacOS/Microsoft Edge',
54
+ '/Applications/Brave Browser.app/Contents/MacOS/Brave Browser',
55
+ ];
56
+ return ['chromium', 'chromium-browser', 'google-chrome', 'microsoft-edge', 'brave-browser'];
57
+ }
58
+
59
+ export function mediaUrlsFromText(body) {
60
+ if (typeof body !== 'string' || body.length > 524288) return [];
61
+ const normalized = body.replaceAll('\\/', '/').replaceAll('\\u0026', '&').replaceAll('&amp;', '&');
62
+ const urls = normalized.match(/https?:\/\/[^\s"'<>\\]+/gi) || [];
63
+ return [...new Set(urls.map(value => value.replace(/[),;]+$/, '')).filter(value => MEDIA_URL.test(value)).map(value => {
64
+ try { return validateUrl(value); } catch { return null; }
65
+ }).filter(Boolean))];
66
+ }
67
+
68
+ export function summarizeSource(metadata, url, index, pageUrl) {
69
+ const formats = (metadata.formats || []).filter(format => !format.has_drm);
70
+ const heights = [...new Set(formats.map(format => format.height).filter(Number.isFinite))].sort((a, b) => b - a);
71
+ const bytes = metadata.filesize || metadata.filesize_approx || null;
72
+ const bitrate = Number.isFinite(metadata.tbr) ? metadata.tbr : Math.max(0, ...formats.map(format => format.tbr).filter(Number.isFinite));
73
+ return {
74
+ index, title: cleanText(metadata.title || metadata.id || 'Video'),
75
+ source: new URL(url).hostname, type: /\.m3u8(?:[?#]|$)/i.test(url) ? 'HLS' : /\.mpd(?:[?#]|$)/i.test(url) ? 'DASH' : 'Video',
76
+ quality: heights.length ? heights.map(height => `${height}p`) : [],
77
+ estimatedMiB: bytes ? Math.round(bytes / 1048576 * 10) / 10 : null,
78
+ bitrateMbps: bitrate > 0 ? Math.round(bitrate / 100) / 10 : null,
79
+ pageUrl,
80
+ };
81
+ }
82
+
83
+ function createCdp(socket) {
84
+ let nextId = 0;
85
+ const pending = new Map();
86
+ const listeners = new Set();
87
+ socket.addEventListener('message', event => {
88
+ let message;
89
+ try { message = JSON.parse(event.data); } catch { return; }
90
+ if (message.id) {
91
+ const entry = pending.get(message.id);
92
+ if (!entry) return;
93
+ pending.delete(message.id);
94
+ if (message.error) entry.reject(new Error(message.error.message));
95
+ else entry.resolve(message.result || {});
96
+ } else for (const listener of listeners) listener(message);
97
+ });
98
+ socket.addEventListener('close', () => {
99
+ for (const entry of pending.values()) entry.reject(new Error('Browser connection closed.'));
100
+ pending.clear();
101
+ });
102
+ return {
103
+ on(listener) { listeners.add(listener); return () => listeners.delete(listener); },
104
+ send(method, params = {}, sessionId) {
105
+ const id = ++nextId;
106
+ const promise = new Promise((resolve, reject) => pending.set(id, { resolve, reject }));
107
+ socket.send(JSON.stringify({ id, method, params, ...(sessionId ? { sessionId } : {}) }));
108
+ return promise;
109
+ },
110
+ };
111
+ }
112
+
113
+ async function findDebuggerPort(directory, child, signal) {
114
+ for (let attempt = 0; attempt < 80; attempt++) {
115
+ signal?.throwIfAborted();
116
+ if (child.exitCode !== null) throw new Error('The browser exited before source discovery started.');
117
+ const text = await readFile(path.join(directory, 'DevToolsActivePort'), 'utf8').catch(() => null);
118
+ if (text) return Number(text.split(/\r?\n/)[0]);
119
+ await delay(100, undefined, { signal });
120
+ }
121
+ throw new Error('The browser did not start its debugging connection.');
122
+ }
123
+
124
+ /** Observe media requests in a temporary browser profile; no browser cookies are copied. */
125
+ export async function captureBrowserMedia(pageUrl, { signal, browserPaths = browserCandidates(), observeMs = 8000, deepScan = false } = {}) {
126
+ const directory = await mkdtemp(path.join(tmpdir(), 'veo-discover-'));
127
+ let child, socket;
128
+ const urls = new Set();
129
+ try {
130
+ let started = false;
131
+ for (const executable of browserPaths) {
132
+ try {
133
+ child = spawn(executable, ['--headless=new', '--remote-debugging-port=0', `--user-data-dir=${directory}`,
134
+ '--no-first-run', '--no-default-browser-check', '--autoplay-policy=no-user-gesture-required', 'about:blank'],
135
+ { stdio: 'ignore', windowsHide: true });
136
+ await new Promise((resolve, reject) => {
137
+ child.once('spawn', resolve);
138
+ child.once('error', reject);
139
+ });
140
+ started = true;
141
+ break;
142
+ } catch { child = null; }
143
+ }
144
+ if (!started) throw new Error('No Chromium-based browser found. Set VEO_BROWSER_PATH to Chrome, Edge or Chromium.');
145
+ const port = await findDebuggerPort(directory, child, signal);
146
+ const endpoint = await fetch(`http://127.0.0.1:${port}/json/version`, { signal }).then(response => response.json());
147
+ socket = new WebSocket(endpoint.webSocketDebuggerUrl);
148
+ await abortable(new Promise((resolve, reject) => {
149
+ socket.addEventListener('open', resolve, { once: true });
150
+ socket.addEventListener('error', reject, { once: true });
151
+ }), signal);
152
+ const cdp = createCdp(socket);
153
+ const responses = new Map();
154
+ const bodyReads = [];
155
+ const collect = (value, mime = '') => {
156
+ if (!MEDIA_URL.test(value) && !MEDIA_TYPE.test(mime)) return;
157
+ try { urls.add(validateUrl(value)); } catch {}
158
+ };
159
+ cdp.on(message => {
160
+ if (message.method === 'Target.attachedToTarget') {
161
+ const attached = message.params.sessionId;
162
+ cdp.send('Network.enable', {}, attached).catch(() => {});
163
+ return;
164
+ }
165
+ if (message.method === 'Network.requestWillBeSent') collect(message.params.request?.url);
166
+ if (message.method === 'Network.responseReceived') {
167
+ const response = message.params.response;
168
+ collect(response.url, response.mimeType);
169
+ if (/(?:json|javascript|text\/plain)/i.test(response.mimeType || '')) {
170
+ responses.set(`${message.sessionId || ''}:${message.params.requestId}`, { sessionId: message.sessionId, requestId: message.params.requestId });
171
+ }
172
+ }
173
+ if (message.method === 'Network.loadingFinished') {
174
+ const key = `${message.sessionId || ''}:${message.params.requestId}`;
175
+ const response = responses.get(key);
176
+ if (!response || message.params.encodedDataLength > 524288) return;
177
+ responses.delete(key);
178
+ if (!deepScan && bodyReads.length >= 100) return;
179
+ bodyReads.push(cdp.send('Network.getResponseBody', { requestId: response.requestId }, response.sessionId)
180
+ .then(result => { if (!result.base64Encoded) for (const url of mediaUrlsFromText(result.body)) urls.add(url); })
181
+ .catch(() => {}));
182
+ }
183
+ });
184
+ await abortable(cdp.send('Target.setAutoAttach', { autoAttach: true, waitForDebuggerOnStart: false, flatten: true }), signal);
185
+ const target = await abortable(cdp.send('Target.createTarget', { url: 'about:blank' }), signal);
186
+ const attached = await abortable(cdp.send('Target.attachToTarget', { targetId: target.targetId, flatten: true }), signal);
187
+ await abortable(cdp.send('Network.enable', {}, attached.sessionId), signal);
188
+ await abortable(cdp.send('Page.enable', {}, attached.sessionId), signal);
189
+ await abortable(cdp.send('Page.navigate', { url: pageUrl }, attached.sessionId), signal);
190
+ await delay(Math.min(observeMs / 2, 4000), undefined, { signal });
191
+ await abortable(cdp.send('Runtime.evaluate', { expression: `(() => {
192
+ for (const video of document.querySelectorAll('video')) video.play().catch(() => {});
193
+ const buttons = [...document.querySelectorAll('button,[role="button"]')];
194
+ const play = buttons.find(node => /^(play|watch|abspielen|ansehen|start)(?:\\s+(?:video|movie|film))?$/i.test((node.getAttribute('aria-label') || node.textContent || '').trim()));
195
+ if (play) play.click();
196
+ })()`, returnByValue: true }, attached.sessionId).catch(() => {}), signal);
197
+ await delay(Math.max(1000, observeMs / 2), undefined, { signal });
198
+ await Promise.race([Promise.allSettled(bodyReads), delay(1000, undefined, { signal }).catch(() => {})]);
199
+ return [...urls];
200
+ } catch (error) {
201
+ if (signal?.aborted && ['AbortError', 'TimeoutError'].includes(error?.name)) return [...urls];
202
+ throw error;
203
+ } finally {
204
+ socket?.close();
205
+ child?.kill();
206
+ await rm(directory, { recursive: true, force: true, maxRetries: 5, retryDelay: 100 }).catch(() => {});
207
+ }
208
+ }
209
+
210
+ /** Verify candidates with yt-dlp before offering them to the user. */
211
+ export async function discoverSources(pageUrl, { signal, capture = captureBrowserMedia, inspect, limit = 30,
212
+ deepScan = false, timeoutMs = deepScan ? 120000 : 45000 } = {}) {
213
+ signal?.throwIfAborted();
214
+ const timeout = new AbortController();
215
+ const timer = setTimeout(() => timeout.abort(new DOMException('Source scan timed out.', 'TimeoutError')), timeoutMs);
216
+ const scanSignal = signal ? AbortSignal.any([signal, timeout.signal]) : timeout.signal;
217
+ const sources = [];
218
+ const seen = new Set();
219
+ try {
220
+ const observeMs = deepScan ? Math.min(30000, Math.floor(timeoutMs * 0.6)) : Math.min(8000, Math.floor(timeoutMs * 0.5));
221
+ let urls;
222
+ try { urls = await capture(pageUrl, { signal: scanSignal, observeMs, deepScan }); }
223
+ catch (error) { if (!timeout.signal.aborted) throw error; urls = []; }
224
+ signal?.throwIfAborted();
225
+ for (const url of deepScan ? urls : urls.slice(0, limit)) {
226
+ if (timeout.signal.aborted) break;
227
+ signal?.throwIfAborted();
228
+ try {
229
+ const metadata = await abortable(inspect(url, pageUrl, { signal: scanSignal }), scanSignal);
230
+ if (metadata.has_drm || metadata.is_live || metadata.formats?.length && metadata.formats.every(format => format.has_drm)
231
+ || !metadata.formats?.length && !metadata.url) continue;
232
+ const address = new URL(url);
233
+ const key = `${address.origin}${address.pathname}:${metadata.id || ''}`;
234
+ if (seen.has(key)) continue;
235
+ seen.add(key);
236
+ sources.push({ ...summarizeSource(metadata, url, sources.length + 1, pageUrl), url });
237
+ } catch { /* A request alone is not proof of downloadable media. */ }
238
+ }
239
+ signal?.throwIfAborted();
240
+ Object.defineProperty(sources, 'timedOut', { value: timeout.signal.aborted });
241
+ return sources;
242
+ } finally {
243
+ clearTimeout(timer);
244
+ }
245
+ }
246
+
247
+ export function formatSource(source) {
248
+ const size = source.estimatedMiB === null ? 'Size unknown' : `about ${source.estimatedMiB} MiB`;
249
+ const bitrate = source.bitrateMbps === null ? '' : ` | ${source.bitrateMbps} Mbit/s`;
250
+ return `${source.index}. ${source.title} | ${source.source} | ${source.type} | ${source.quality.join(', ') || 'Quality unknown'} | ${size}${bitrate}`;
251
+ }
package/src/state.js CHANGED
@@ -1,27 +1,27 @@
1
- import { createHash, randomUUID } from 'node:crypto';
2
- import { mkdir, readFile, rename, rm, writeFile } from 'node:fs/promises';
3
- import path from 'node:path';
4
-
5
- export function digest(value) {
6
- return createHash('sha256').update(JSON.stringify(value)).digest('hex').slice(0, 24);
7
- }
8
-
9
- export async function readJson(file, fallback) {
10
- try { return JSON.parse(await readFile(file, 'utf8')); }
11
- catch (error) { if (error.code === 'ENOENT') return fallback; throw error; }
12
- }
13
-
14
- export async function writeJson(file, value) {
15
- await mkdir(path.dirname(file), { recursive: true });
16
- const temporary = `${file}.${randomUUID()}.tmp`;
17
- try {
18
- await writeFile(temporary, `${JSON.stringify(value, null, 2)}\n`, { mode: 0o600, flag: 'wx' });
19
- await rename(temporary, file);
20
- } finally { await rm(temporary, { force: true }); }
21
- }
22
-
23
- // Do not persist browser sessions or cookie paths in retry jobs.
24
- export function publicOptions(options) {
25
- const { cookies, cookiesFromBrowser, profile, urls, url, retryFailed, batchFile, ...safe } = options;
26
- return { ...safe, output: path.resolve(options.output) };
27
- }
1
+ import { createHash, randomUUID } from 'node:crypto';
2
+ import { mkdir, readFile, rename, rm, writeFile } from 'node:fs/promises';
3
+ import path from 'node:path';
4
+
5
+ export function digest(value) {
6
+ return createHash('sha256').update(JSON.stringify(value)).digest('hex').slice(0, 24);
7
+ }
8
+
9
+ export async function readJson(file, fallback) {
10
+ try { return JSON.parse(await readFile(file, 'utf8')); }
11
+ catch (error) { if (error.code === 'ENOENT') return fallback; throw error; }
12
+ }
13
+
14
+ export async function writeJson(file, value) {
15
+ await mkdir(path.dirname(file), { recursive: true });
16
+ const temporary = `${file}.${randomUUID()}.tmp`;
17
+ try {
18
+ await writeFile(temporary, `${JSON.stringify(value, null, 2)}\n`, { mode: 0o600, flag: 'wx' });
19
+ await rename(temporary, file);
20
+ } finally { await rm(temporary, { force: true }); }
21
+ }
22
+
23
+ // Do not persist browser sessions or cookie paths in retry jobs.
24
+ export function publicOptions(options) {
25
+ const { cookies, cookiesFromBrowser, profile, urls, url, retryFailed, batchFile, mediaUrl, ...safe } = options;
26
+ return { ...safe, output: path.resolve(options.output) };
27
+ }