@torrent-tv/proxy 2.87.4 → 2.87.6

package/services/torrent/worker/handle-state.js

@@ -1,23 +1,23 @@
-/**
- * @file Whether a torrent handle can still be read from.
- *
- * Its own module so it can be tested: importing the worker starts a torrent
- * client and a piece pool, which a unit test has no business doing.
- */
-
-/**
- * Whether a torrent handle can still be read from.
- *
- * `destroyed` is WebTorrent's own flag and the earliest signal. The empty file
- * list is the symptom that actually reaches a reader — it is what produced
- * `File 0 not found` in the field — and it is also true of a handle whose
- * metadata has not arrived yet, in which case adding the source again is
- * equally right: the add de-duplicates and yields the same torrent once it is
- * ready.
- *
- * @param {{ destroyed?: boolean, files?: unknown[] } | null | undefined} torrent
- * @returns {boolean}
- */
-export function isUsableTorrentHandle(torrent) {
-  return Boolean(torrent) && torrent.destroyed !== true && (torrent.files?.length ?? 0) > 0;
-}
+/**
+ * @file Whether a torrent handle can still be read from.
+ *
+ * Its own module so it can be tested: importing the worker starts a torrent
+ * client and a piece pool, which a unit test has no business doing.
+ */
+
+/**
+ * Whether a torrent handle can still be read from.
+ *
+ * `destroyed` is WebTorrent's own flag and the earliest signal. The empty file
+ * list is the symptom that actually reaches a reader — it is what produced
+ * `File 0 not found` in the field — and it is also true of a handle whose
+ * metadata has not arrived yet, in which case adding the source again is
+ * equally right: the add de-duplicates and yields the same torrent once it is
+ * ready.
+ *
+ * @param {{ destroyed?: boolean, files?: unknown[] } | null | undefined} torrent
+ * @returns {boolean}
+ */
+export function isUsableTorrentHandle(torrent) {
+  return Boolean(torrent) && torrent.destroyed !== true && (torrent.files?.length ?? 0) > 0;
+}

package/services/torrent/worker/held-bytes.js

@@ -1,155 +1,155 @@
-/**
- * @file What of a file the torrent already HOLDS, and reading it without
- * asking the swarm for anything.
- *
- * Two questions, and both are the torrent's own: which pieces have arrived is
- * its bitfield, and reading a piece it has is its store. They are here because
- * of that and for no other reason — everything else that used to keep them
- * company, the subtitle plan and the cue walk and the cursor a browser follows,
- * is what a FILE says about itself and moved to the thread the sessions are on
- * (2026-09-15).
- *
- * **Why the subtitle walk cannot use the ordinary range read.** That one
- * declares demand: it registers a window, steers the swarm toward the piece it
- * is stopped on, and waits. Right for a viewer waiting on a segment, wrong
- * here — switching subtitles on must never pull bytes the viewer is not waiting
- * for, which is the rule the walk has kept since it was written. So these two
- * read the store and say "not here" rather than fetching.
- *
- * **The held ranges are answered as a LIST, not asked piece by piece.** The
- * walk asks "is this cluster downloaded" for every cluster of the file — on the
- * field files that is hundreds of questions per pass, and a pass runs every
- * three seconds. One list per pass costs one message; the same answers asked
- * one at a time cost hundreds of round trips for a walk that is meant to be
- * free when there is nothing new to find.
- */
-
-/**
- * The byte ranges of one file that are downloaded WHOLE.
- *
- * Whole because a piece is the unit the swarm delivers and the unit the
- * bitfield counts: a range that is half a piece short cannot be read, so
- * reporting it would be reporting bytes that are not there.
- *
- * @param {object} torrent
- * @param {number} fileIndex
- * @returns {Array<[number, number]>} Ascending, non-overlapping, inclusive
- *   offsets WITHIN the file. Empty when nothing of it has arrived, or when the
- *   torrent cannot say — which reads the same way: nothing may be walked yet.
- */
-export function heldRangesOf(torrent, fileIndex) {
-  const file = torrent?.files?.[fileIndex];
-  const pieceLength = Number(torrent?.pieceLength);
-  if (!file || !Number.isFinite(pieceLength) || pieceLength <= 0 || !torrent?.bitfield) {
-    return [];
-  }
-  const offset = Number(file.offset) || 0;
-  const length = Number(file.length) || 0;
-  if (!(length > 0)) {
-    return [];
-  }
-  const first = Math.floor(offset / pieceLength);
-  const last = Math.floor((offset + length - 1) / pieceLength);
-  /** @type {Array<[number, number]>} */
-  const ranges = [];
-  let runStart = -1;
-  for (let index = first; index <= last; index += 1) {
-    if (torrent.bitfield.get(index)) {
-      if (runStart < 0) {
-        runStart = index;
-      }
-      continue;
-    }
-    if (runStart >= 0) {
-      ranges.push(rangeWithin(runStart, index - 1, pieceLength, offset, length));
-      runStart = -1;
-    }
-  }
-  if (runStart >= 0) {
-    ranges.push(rangeWithin(runStart, last, pieceLength, offset, length));
-  }
-  return ranges;
-}
-
-/**
- * A run of whole pieces, as offsets within the file.
- *
- * @param {number} firstPiece
- * @param {number} lastPiece
- * @param {number} pieceLength
- * @param {number} offset - Where the file begins in the torrent.
- * @param {number} length - How long the file is.
- * @returns {[number, number]}
- */
-function rangeWithin(firstPiece, lastPiece, pieceLength, offset, length) {
-  const start = Math.max(0, firstPiece * pieceLength - offset);
-  const end = Math.min(length - 1, (lastPiece + 1) * pieceLength - 1 - offset);
-  return [start, end];
-}
-
-/**
- * How long a read of bytes the torrent already holds may take before it is
- * given up.
- *
- * Not a measurement and nothing is derived from it: such a read either answers
- * or it does not, and this is the point past which it is presumed lost — so
- * that one stream which never ends cannot hold a file's walk, and with it the
- * browser's own request for its subtitles, for the rest of the session.
- */
-const READ_ABANDON_MS = 30_000;
-
-/**
- * Read a byte range of a file straight from the store, without asking the swarm
- * for anything.
- *
- * @param {object} torrent
- * @param {number} fileIndex
- * @param {number} start
- * @param {number} end - Inclusive.
- * @param {{ info: Function }} [logger]
- * @returns {Promise<Buffer | null>}
- */
-export function readHeldBytes(torrent, fileIndex, start, end, logger = null) {
-  const file = torrent?.files?.[fileIndex];
-  if (!file || !(end >= start) || !(start >= 0)) {
-    return Promise.resolve(null);
-  }
-  const last = Math.min(end, Number(file.length) - 1);
-  return new Promise((resolve) => {
-    const chunks = [];
-    let stream;
-    try {
-      stream = file.createReadStream({ start, end: last });
-    } catch {
-      resolve(null);
-      return;
-    }
-    let settled = false;
-    /** @type {ReturnType<typeof setTimeout> | null} */
-    let abandon = null;
-    const settle = (value) => {
-      if (settled) {
-        return;
-      }
-      settled = true;
-      if (abandon !== null) {
-        clearTimeout(abandon);
-      }
-      if (value === null) {
-        stream.destroy?.();
-      }
-      resolve(value);
-    };
-    abandon = setTimeout(() => {
-      logger?.info(
-        `subtitles: a read of ${start}-${last} in "${String(file.name).slice(0, 40)}" ` +
-        `did not finish in ${READ_ABANDON_MS / 1000}s and was given up`
-      );
-      settle(null);
-    }, READ_ABANDON_MS);
-    abandon.unref?.();
-    stream.on("data", (chunk) => chunks.push(chunk));
-    stream.on("end", () => settle(Buffer.concat(chunks)));
-    stream.on("error", () => settle(null));
-  });
-}
+/**
+ * @file What of a file the torrent already HOLDS, and reading it without
+ * asking the swarm for anything.
+ *
+ * Two questions, and both are the torrent's own: which pieces have arrived is
+ * its bitfield, and reading a piece it has is its store. They are here because
+ * of that and for no other reason — everything else that used to keep them
+ * company, the subtitle plan and the cue walk and the cursor a browser follows,
+ * is what a FILE says about itself and moved to the thread the sessions are on
+ * (2026-09-15).
+ *
+ * **Why the subtitle walk cannot use the ordinary range read.** That one
+ * declares demand: it registers a window, steers the swarm toward the piece it
+ * is stopped on, and waits. Right for a viewer waiting on a segment, wrong
+ * here — switching subtitles on must never pull bytes the viewer is not waiting
+ * for, which is the rule the walk has kept since it was written. So these two
+ * read the store and say "not here" rather than fetching.
+ *
+ * **The held ranges are answered as a LIST, not asked piece by piece.** The
+ * walk asks "is this cluster downloaded" for every cluster of the file — on the
+ * field files that is hundreds of questions per pass, and a pass runs every
+ * three seconds. One list per pass costs one message; the same answers asked
+ * one at a time cost hundreds of round trips for a walk that is meant to be
+ * free when there is nothing new to find.
+ */
+
+/**
+ * The byte ranges of one file that are downloaded WHOLE.
+ *
+ * Whole because a piece is the unit the swarm delivers and the unit the
+ * bitfield counts: a range that is half a piece short cannot be read, so
+ * reporting it would be reporting bytes that are not there.
+ *
+ * @param {object} torrent
+ * @param {number} fileIndex
+ * @returns {Array<[number, number]>} Ascending, non-overlapping, inclusive
+ *   offsets WITHIN the file. Empty when nothing of it has arrived, or when the
+ *   torrent cannot say — which reads the same way: nothing may be walked yet.
+ */
+export function heldRangesOf(torrent, fileIndex) {
+  const file = torrent?.files?.[fileIndex];
+  const pieceLength = Number(torrent?.pieceLength);
+  if (!file || !Number.isFinite(pieceLength) || pieceLength <= 0 || !torrent?.bitfield) {
+    return [];
+  }
+  const offset = Number(file.offset) || 0;
+  const length = Number(file.length) || 0;
+  if (!(length > 0)) {
+    return [];
+  }
+  const first = Math.floor(offset / pieceLength);
+  const last = Math.floor((offset + length - 1) / pieceLength);
+  /** @type {Array<[number, number]>} */
+  const ranges = [];
+  let runStart = -1;
+  for (let index = first; index <= last; index += 1) {
+    if (torrent.bitfield.get(index)) {
+      if (runStart < 0) {
+        runStart = index;
+      }
+      continue;
+    }
+    if (runStart >= 0) {
+      ranges.push(rangeWithin(runStart, index - 1, pieceLength, offset, length));
+      runStart = -1;
+    }
+  }
+  if (runStart >= 0) {
+    ranges.push(rangeWithin(runStart, last, pieceLength, offset, length));
+  }
+  return ranges;
+}
+
+/**
+ * A run of whole pieces, as offsets within the file.
+ *
+ * @param {number} firstPiece
+ * @param {number} lastPiece
+ * @param {number} pieceLength
+ * @param {number} offset - Where the file begins in the torrent.
+ * @param {number} length - How long the file is.
+ * @returns {[number, number]}
+ */
+function rangeWithin(firstPiece, lastPiece, pieceLength, offset, length) {
+  const start = Math.max(0, firstPiece * pieceLength - offset);
+  const end = Math.min(length - 1, (lastPiece + 1) * pieceLength - 1 - offset);
+  return [start, end];
+}
+
+/**
+ * How long a read of bytes the torrent already holds may take before it is
+ * given up.
+ *
+ * Not a measurement and nothing is derived from it: such a read either answers
+ * or it does not, and this is the point past which it is presumed lost — so
+ * that one stream which never ends cannot hold a file's walk, and with it the
+ * browser's own request for its subtitles, for the rest of the session.
+ */
+const READ_ABANDON_MS = 30_000;
+
+/**
+ * Read a byte range of a file straight from the store, without asking the swarm
+ * for anything.
+ *
+ * @param {object} torrent
+ * @param {number} fileIndex
+ * @param {number} start
+ * @param {number} end - Inclusive.
+ * @param {{ info: Function }} [logger]
+ * @returns {Promise<Buffer | null>}
+ */
+export function readHeldBytes(torrent, fileIndex, start, end, logger = null) {
+  const file = torrent?.files?.[fileIndex];
+  if (!file || !(end >= start) || !(start >= 0)) {
+    return Promise.resolve(null);
+  }
+  const last = Math.min(end, Number(file.length) - 1);
+  return new Promise((resolve) => {
+    const chunks = [];
+    let stream;
+    try {
+      stream = file.createReadStream({ start, end: last });
+    } catch {
+      resolve(null);
+      return;
+    }
+    let settled = false;
+    /** @type {ReturnType<typeof setTimeout> | null} */
+    let abandon = null;
+    const settle = (value) => {
+      if (settled) {
+        return;
+      }
+      settled = true;
+      if (abandon !== null) {
+        clearTimeout(abandon);
+      }
+      if (value === null) {
+        stream.destroy?.();
+      }
+      resolve(value);
+    };
+    abandon = setTimeout(() => {
+      logger?.info(
+        `subtitles: a read of ${start}-${last} in "${String(file.name).slice(0, 40)}" ` +
+        `did not finish in ${READ_ABANDON_MS / 1000}s and was given up`
+      );
+      settle(null);
+    }, READ_ABANDON_MS);
+    abandon.unref?.();
+    stream.on("data", (chunk) => chunks.push(chunk));
+    stream.on("end", () => settle(Buffer.concat(chunks)));
+    stream.on("error", () => settle(null));
+  });
+}

package/services/torrent/worker/install-webrtc-shim.js

@@ -1,32 +1,32 @@
-/**
- * @file Point `webrtc-polyfill` at the JavaScript WebRTC stack, in this thread only.
- *
- * Imported for its side effect, and imported FIRST by `worker.js` — ES module
- * bodies run in import order, so registering the hook here happens before
- * `torrent/torrent-pool.js` pulls in WebTorrent, which is what reaches
- * `@thaunknown/simple-peer` and, through it, `webrtc-polyfill`.
- *
- * Scope is deliberately narrow. The hook lives in the worker's isolate, so the
- * main thread keeps using node-datachannel directly for the browser's video
- * channel — see `webrtc-shim.js` for why the two cannot share one process
- * isolate at all.
- */
-
-import { registerHooks } from "node:module";
-
-const SHIM_URL = new URL("./webrtc-shim.js", import.meta.url).href;
-
-registerHooks({
-  /**
-   * @param {string} specifier
-   * @param {object} context
-   * @param {(specifier: string, context: object) => { url: string }} nextResolve
-   * @returns {{ url: string, shortCircuit?: boolean }}
-   */
-  resolve(specifier, context, nextResolve) {
-    if (specifier === "webrtc-polyfill") {
-      return { url: SHIM_URL, shortCircuit: true };
-    }
-    return nextResolve(specifier, context);
-  }
-});
+/**
+ * @file Point `webrtc-polyfill` at the JavaScript WebRTC stack, in this thread only.
+ *
+ * Imported for its side effect, and imported FIRST by `worker.js` — ES module
+ * bodies run in import order, so registering the hook here happens before
+ * `torrent/torrent-pool.js` pulls in WebTorrent, which is what reaches
+ * `@thaunknown/simple-peer` and, through it, `webrtc-polyfill`.
+ *
+ * Scope is deliberately narrow. The hook lives in the worker's isolate, so the
+ * main thread keeps using node-datachannel directly for the browser's video
+ * channel — see `webrtc-shim.js` for why the two cannot share one process
+ * isolate at all.
+ */
+
+import { registerHooks } from "node:module";
+
+const SHIM_URL = new URL("./webrtc-shim.js", import.meta.url).href;
+
+registerHooks({
+  /**
+   * @param {string} specifier
+   * @param {object} context
+   * @param {(specifier: string, context: object) => { url: string }} nextResolve
+   * @returns {{ url: string, shortCircuit?: boolean }}
+   */
+  resolve(specifier, context, nextResolve) {
+    if (specifier === "webrtc-polyfill") {
+      return { url: SHIM_URL, shortCircuit: true };
+    }
+    return nextResolve(specifier, context);
+  }
+});