@storyteller-platform/epub 0.7.0 → 0.7.1

package/dist/upgrade.d.ts

@@ -339,10 +339,6 @@ declare class Epub {
     static assertEpub3(epub: Epub): Promise<void>;
     copy(path?: string): Promise<Epub>;
     private removeEntry;
-    /**
-     * Read raw bytes (or utf-8 text) from the underlying adapter
-     */
-    private getFileData;
     /**
      * Length of the underlying archive entry for a manifest item, in bytes
      * Necessary to compute the readium page count which is for COMPRESSED content
@@ -391,7 +387,7 @@ declare class Epub {
      */
     findMetadataItem(predicate: (entry: MetadataEntry) => boolean): Promise<{
         id: string | undefined;
-        type: `?${string}` | `a${string}` | `b${string}` | `c${string}` | `d${string}` | `e${string}` | `f${string}` | `g${string}` | `h${string}` | `i${string}` | `j${string}` | `k${string}` | `l${string}` | `m${string}` | `n${string}` | `o${string}` | `p${string}` | `q${string}` | `r${string}` | `s${string}` | `t${string}` | `u${string}` | `v${string}` | `w${string}` | `x${string}` | `y${string}` | `z${string}` | `A${string}` | `B${string}` | `C${string}` | `D${string}` | `E${string}` | `F${string}` | `G${string}` | `H${string}` | `I${string}` | `J${string}` | `K${string}` | `L${string}` | `M${string}` | `N${string}` | `O${string}` | `P${string}` | `Q${string}` | `R${string}` | `S${string}` | `T${string}` | `U${string}` | `V${string}` | `W${string}` | `X${string}` | `Y${string}` | `Z${string}`;
+        type: `a${string}` | `b${string}` | `c${string}` | `d${string}` | `e${string}` | `f${string}` | `g${string}` | `h${string}` | `i${string}` | `j${string}` | `k${string}` | `l${string}` | `m${string}` | `n${string}` | `o${string}` | `p${string}` | `q${string}` | `r${string}` | `s${string}` | `t${string}` | `u${string}` | `v${string}` | `w${string}` | `x${string}` | `y${string}` | `z${string}` | `A${string}` | `B${string}` | `C${string}` | `D${string}` | `E${string}` | `F${string}` | `G${string}` | `H${string}` | `I${string}` | `J${string}` | `K${string}` | `L${string}` | `M${string}` | `N${string}` | `O${string}` | `P${string}` | `Q${string}` | `R${string}` | `S${string}` | `T${string}` | `U${string}` | `V${string}` | `W${string}` | `X${string}` | `Y${string}` | `Z${string}` | `?${string}`;
         properties: {
             [k: string]: string;
         };
@@ -403,7 +399,7 @@ declare class Epub {
      */
     findAllMetadataItems(predicate: (entry: MetadataEntry) => boolean): Promise<{
         id: string | undefined;
-        type: `?${string}` | `a${string}` | `b${string}` | `c${string}` | `d${string}` | `e${string}` | `f${string}` | `g${string}` | `h${string}` | `i${string}` | `j${string}` | `k${string}` | `l${string}` | `m${string}` | `n${string}` | `o${string}` | `p${string}` | `q${string}` | `r${string}` | `s${string}` | `t${string}` | `u${string}` | `v${string}` | `w${string}` | `x${string}` | `y${string}` | `z${string}` | `A${string}` | `B${string}` | `C${string}` | `D${string}` | `E${string}` | `F${string}` | `G${string}` | `H${string}` | `I${string}` | `J${string}` | `K${string}` | `L${string}` | `M${string}` | `N${string}` | `O${string}` | `P${string}` | `Q${string}` | `R${string}` | `S${string}` | `T${string}` | `U${string}` | `V${string}` | `W${string}` | `X${string}` | `Y${string}` | `Z${string}`;
+        type: `a${string}` | `b${string}` | `c${string}` | `d${string}` | `e${string}` | `f${string}` | `g${string}` | `h${string}` | `i${string}` | `j${string}` | `k${string}` | `l${string}` | `m${string}` | `n${string}` | `o${string}` | `p${string}` | `q${string}` | `r${string}` | `s${string}` | `t${string}` | `u${string}` | `v${string}` | `w${string}` | `x${string}` | `y${string}` | `z${string}` | `A${string}` | `B${string}` | `C${string}` | `D${string}` | `E${string}` | `F${string}` | `G${string}` | `H${string}` | `I${string}` | `J${string}` | `K${string}` | `L${string}` | `M${string}` | `N${string}` | `O${string}` | `P${string}` | `Q${string}` | `R${string}` | `S${string}` | `T${string}` | `U${string}` | `V${string}` | `W${string}` | `X${string}` | `Y${string}` | `Z${string}` | `?${string}`;
         properties: {
             [k: string]: string;
         };
@@ -922,13 +918,31 @@ declare class Epub {
         resolveToRoot?: boolean;
     }): Promise<Navigation | null>;
     /**
-     * Returns a Zip Entry path for an HREF
+     * Name the entry a resolved URL addresses.
+     *
+     * A correctly authored publication percent-encodes its hrefs, so the
+     * decoded reading is the one the spec calls for and the one we prefer.
+     * Some publications instead write the entry name verbatim, so when the two
+     * readings differ we ask the archive which one it holds. A verbatim name
+     * carrying a bare `%`, as `100%.xhtml` does, is not a valid encoding at all
+     * and can only be its own answer.
      */
-    private resolveInternalHref;
+    private entryForUrl;
+    /**
+     * Resolve an href to the name of the archive entry it addresses.
+     *
+     * @param from The entry the href appears in; it resolves against that
+     *   entry's directory
+     * @throws when the href addresses something outside the publication
+     */
+    private resolveEntry;
     /**
      * Returns a path-relative-scheme-less URL, relative to the
      * container root.
      *
+     * An href that points outside the publication, such as an external link in
+     * a navigation document, is handed back unchanged.
+     *
      * @param href The href to resolve
      * @param [relativeTo] Optional - The href to resolve this href relative to.
          Use if resolving a relative href from a file other than the package document.

package/dist/upgrade.js

@@ -1,5 +1,6 @@
 import "./chunk-BIEQXUOY.js";
 import { nanoid } from "nanoid";
+import { MediaType } from "@storyteller-platform/media-types";
 import {
   Epub
 } from "./index.js";
@@ -49,31 +50,7 @@ const GUIDE_TO_EPUBTYPE = {
   "other.volume": "volume",
   "other.warning": "warning"
 };
-const XHTML_MEDIA_TYPES = /* @__PURE__ */ new Set([
-  "application/xhtml+xml",
-  "application/vnd.adobe-page-template+xml",
-  "text/html"
-]);
-const OEB_FONTS = /* @__PURE__ */ new Set([
-  "application/vnd.ms-opentype",
-  "application/x-font-ttf",
-  "application/x-font-otf",
-  "application/x-font-truetype",
-  "application/font-sfnt",
-  "application/font-woff",
-  "application/font-woff2",
-  "font/woff",
-  "font/woff2",
-  "font/otf",
-  "font/ttf",
-  "font/sfnt"
-]);
-const FONT_MIME_BY_EXT = {
-  ".ttf": "application/font-sfnt",
-  ".otf": "application/font-sfnt",
-  ".woff": "application/font-woff",
-  ".woff2": "font/woff2"
-};
+const XHTML_MEDIA_TYPES = /* @__PURE__ */ new Set([MediaType.XHTML, MediaType.HTML]);
 function getMetadataElement(pkg) {
   return Epub.findXmlChildByName("metadata", Epub.getXmlChildren(pkg));
 }
@@ -350,11 +327,10 @@ function removeSpineTocRef(pkg) {
   delete spine[":@"]["@_toc"];
 }
 async function collectManifestProperties(epub) {
-  var _a;
   const manifest = await epub.getManifest();
   for (const item of Object.values(manifest)) {
-    const mediaType = ((_a = item.mediaType) == null ? void 0 : _a.toLowerCase()) ?? "";
-    if (!XHTML_MEDIA_TYPES.has(mediaType)) continue;
+    const mediaType = MediaType.fromMime(item.mediaType ?? "");
+    if (!mediaType || !XHTML_MEDIA_TYPES.has(mediaType)) continue;
     let xml;
     try {
       xml = await epub.readXhtmlItemContents(item.id);
@@ -375,21 +351,16 @@ async function collectManifestProperties(epub) {
   }
 }
 function fixFontMimeTypes(pkg) {
-  var _a, _b;
+  var _a, _b, _c;
   const manifest = getManifestElement(pkg);
   if (!manifest) return;
   for (const item of findAllByName("item", manifest.manifest)) {
-    const mt = (((_a = item[":@"]) == null ? void 0 : _a["@_media-type"]) ?? "").toLowerCase();
-    if (!OEB_FONTS.has(mt)) continue;
-    const href = ((_b = item[":@"]) == null ? void 0 : _b["@_href"]) ?? "";
-    const dotIdx = href.lastIndexOf(".");
-    if (dotIdx === -1) continue;
-    const ext = href.slice(dotIdx).toLowerCase();
-    const corrected = FONT_MIME_BY_EXT[ext];
-    if (corrected && corrected !== mt) {
-      item[":@"] ??= {};
-      item[":@"]["@_media-type"] = corrected;
-    }
+    const declared = ((_a = item[":@"]) == null ? void 0 : _a["@_media-type"]) ?? "";
+    if (!((_b = MediaType.fromMime(declared)) == null ? void 0 : _b.is("font"))) continue;
+    const corrected = MediaType.fromPath(((_c = item[":@"]) == null ? void 0 : _c["@_href"]) ?? "");
+    if (!(corrected == null ? void 0 : corrected.is("font")) || corrected.mime === declared) continue;
+    item[":@"] ??= {};
+    item[":@"]["@_media-type"] = corrected.mime;
   }
 }
 function buildTocOl(entries) {
@@ -482,10 +453,7 @@ async function chooseNavHref(epub) {
 async function removeNcx(epub) {
   const manifest = await epub.getManifest();
   const ncxItem = Object.values(manifest).find(
-    (item) => {
-      var _a;
-      return ((_a = item.mediaType) == null ? void 0 : _a.toLowerCase()) === "application/x-dtbncx+xml";
-    }
+    (item) => MediaType.fromMime(item.mediaType ?? "") === MediaType.NCX
   );
   if (ncxItem) {
     await epub.removeManifestItem(ncxItem.id);

package/dist/zip.cjs

@@ -0,0 +1,142 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var zip_exports = {};
+__export(zip_exports, {
+  openZip: () => openZip
+});
+module.exports = __toCommonJS(zip_exports);
+var import_node_fs = require("node:fs");
+var import_promises = require("node:fs/promises");
+var import_yauzl_promise = require("yauzl-promise");
+const EOCDR_SIGNATURE = Buffer.from([80, 75, 5, 6]);
+const EOCDR_SIZE = 22;
+const MAX_TAIL_SCAN = 1024 * 1024;
+function toBuffer(source) {
+  return Buffer.isBuffer(source) ? source : Buffer.from(source.buffer, source.byteOffset, source.byteLength);
+}
+function findArchiveEnd(tail, tailStart, fileSize) {
+  let searchFrom = tail.length - EOCDR_SIZE;
+  while (searchFrom >= 0) {
+    const index = tail.lastIndexOf(EOCDR_SIGNATURE, searchFrom);
+    if (index === -1) return null;
+    const commentLength = tail.readUInt16LE(index + 20);
+    const end = tailStart + index + EOCDR_SIZE + commentLength;
+    if (end < fileSize) return end;
+    searchFrom = index - 1;
+  }
+  return null;
+}
+async function readExactly(handle, length, position) {
+  const buffer = Buffer.allocUnsafe(length);
+  let offset = 0;
+  while (offset < length) {
+    const { bytesRead } = await handle.read(
+      buffer,
+      offset,
+      length - offset,
+      position + offset
+    );
+    if (bytesRead === 0) return null;
+    offset += bytesRead;
+  }
+  return buffer;
+}
+async function findFileArchiveEnd(path) {
+  let handle;
+  try {
+    handle = await (0, import_promises.open)(path, "r");
+  } catch {
+    return null;
+  }
+  try {
+    const { size } = await handle.stat();
+    const tailLength = Math.min(size, MAX_TAIL_SCAN);
+    if (tailLength < EOCDR_SIZE) return null;
+    const tailStart = size - tailLength;
+    const tail = await readExactly(handle, tailLength, tailStart);
+    if (!tail) return null;
+    return findArchiveEnd(tail, tailStart, size);
+  } finally {
+    await handle.close();
+  }
+}
+class TruncatedFileReader extends import_yauzl_promise.Reader {
+  constructor(path) {
+    super();
+    this.path = path;
+  }
+  handle = null;
+  async _open() {
+    this.handle = await (0, import_promises.open)(this.path, "r");
+    return {};
+  }
+  async _close() {
+    const handle = this.handle;
+    this.handle = null;
+    if (handle) await handle.close();
+    return {};
+  }
+  async _read(start, length) {
+    if (!this.handle) throw new Error(`reader for ${this.path} is not open`);
+    const buffer = await readExactly(this.handle, length, start);
+    if (!buffer) {
+      throw new Error(`unexpected end of file reading ${this.path}`);
+    }
+    return buffer;
+  }
+  _createReadStream(start, length) {
+    return (0, import_node_fs.createReadStream)(this.path, { start, end: start + length - 1 });
+  }
+}
+async function openBufferZip(data) {
+  try {
+    return await (0, import_yauzl_promise.fromBuffer)(data);
+  } catch (error) {
+    const tailStart = Math.max(0, data.length - MAX_TAIL_SCAN);
+    const end = findArchiveEnd(data.subarray(tailStart), tailStart, data.length);
+    if (end === null) throw error;
+    try {
+      return await (0, import_yauzl_promise.fromBuffer)(data.subarray(0, end));
+    } catch {
+      throw error;
+    }
+  }
+}
+async function openFileZip(path) {
+  try {
+    return await (0, import_yauzl_promise.open)(path);
+  } catch (error) {
+    const end = await findFileArchiveEnd(path);
+    if (end === null) throw error;
+    const reader = new TruncatedFileReader(path);
+    try {
+      return await (0, import_yauzl_promise.fromReader)(reader, end);
+    } catch {
+      await reader._close().catch(() => void 0);
+      throw error;
+    }
+  }
+}
+async function openZip(source) {
+  return typeof source === "string" ? openFileZip(source) : openBufferZip(toBuffer(source));
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  openZip
+});

package/dist/zip.d.cts

@@ -0,0 +1,5 @@
+import { ZipFile } from 'yauzl-promise';
+
+declare function openZip(source: string | Uint8Array): Promise<ZipFile>;
+
+export { openZip };

package/dist/zip.d.ts

@@ -0,0 +1,5 @@
+import { ZipFile } from 'yauzl-promise';
+
+declare function openZip(source: string | Uint8Array): Promise<ZipFile>;
+
+export { openZip };

package/dist/zip.js

@@ -0,0 +1,124 @@
+import "./chunk-BIEQXUOY.js";
+import { createReadStream } from "node:fs";
+import { open as openFile } from "node:fs/promises";
+import {
+  Reader,
+  fromBuffer,
+  fromReader,
+  open
+} from "yauzl-promise";
+const EOCDR_SIGNATURE = Buffer.from([80, 75, 5, 6]);
+const EOCDR_SIZE = 22;
+const MAX_TAIL_SCAN = 1024 * 1024;
+function toBuffer(source) {
+  return Buffer.isBuffer(source) ? source : Buffer.from(source.buffer, source.byteOffset, source.byteLength);
+}
+function findArchiveEnd(tail, tailStart, fileSize) {
+  let searchFrom = tail.length - EOCDR_SIZE;
+  while (searchFrom >= 0) {
+    const index = tail.lastIndexOf(EOCDR_SIGNATURE, searchFrom);
+    if (index === -1) return null;
+    const commentLength = tail.readUInt16LE(index + 20);
+    const end = tailStart + index + EOCDR_SIZE + commentLength;
+    if (end < fileSize) return end;
+    searchFrom = index - 1;
+  }
+  return null;
+}
+async function readExactly(handle, length, position) {
+  const buffer = Buffer.allocUnsafe(length);
+  let offset = 0;
+  while (offset < length) {
+    const { bytesRead } = await handle.read(
+      buffer,
+      offset,
+      length - offset,
+      position + offset
+    );
+    if (bytesRead === 0) return null;
+    offset += bytesRead;
+  }
+  return buffer;
+}
+async function findFileArchiveEnd(path) {
+  let handle;
+  try {
+    handle = await openFile(path, "r");
+  } catch {
+    return null;
+  }
+  try {
+    const { size } = await handle.stat();
+    const tailLength = Math.min(size, MAX_TAIL_SCAN);
+    if (tailLength < EOCDR_SIZE) return null;
+    const tailStart = size - tailLength;
+    const tail = await readExactly(handle, tailLength, tailStart);
+    if (!tail) return null;
+    return findArchiveEnd(tail, tailStart, size);
+  } finally {
+    await handle.close();
+  }
+}
+class TruncatedFileReader extends Reader {
+  constructor(path) {
+    super();
+    this.path = path;
+  }
+  handle = null;
+  async _open() {
+    this.handle = await openFile(this.path, "r");
+    return {};
+  }
+  async _close() {
+    const handle = this.handle;
+    this.handle = null;
+    if (handle) await handle.close();
+    return {};
+  }
+  async _read(start, length) {
+    if (!this.handle) throw new Error(`reader for ${this.path} is not open`);
+    const buffer = await readExactly(this.handle, length, start);
+    if (!buffer) {
+      throw new Error(`unexpected end of file reading ${this.path}`);
+    }
+    return buffer;
+  }
+  _createReadStream(start, length) {
+    return createReadStream(this.path, { start, end: start + length - 1 });
+  }
+}
+async function openBufferZip(data) {
+  try {
+    return await fromBuffer(data);
+  } catch (error) {
+    const tailStart = Math.max(0, data.length - MAX_TAIL_SCAN);
+    const end = findArchiveEnd(data.subarray(tailStart), tailStart, data.length);
+    if (end === null) throw error;
+    try {
+      return await fromBuffer(data.subarray(0, end));
+    } catch {
+      throw error;
+    }
+  }
+}
+async function openFileZip(path) {
+  try {
+    return await open(path);
+  } catch (error) {
+    const end = await findFileArchiveEnd(path);
+    if (end === null) throw error;
+    const reader = new TruncatedFileReader(path);
+    try {
+      return await fromReader(reader, end);
+    } catch {
+      await reader._close().catch(() => void 0);
+      throw error;
+    }
+  }
+}
+async function openZip(source) {
+  return typeof source === "string" ? openFileZip(source) : openBufferZip(toBuffer(source));
+}
+export {
+  openZip
+};

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@storyteller-platform/epub",
-  "version": "0.7.0",
+  "version": "0.7.1",
   "type": "module",
   "module": "dist/index.js",
   "types": "dist/index.d.ts",
@@ -32,7 +32,6 @@
     "@storyteller-platform/eslint": "0.1.0",
     "@storyteller-platform/tsup": "^0.1.0",
     "@tsconfig/strictest": "^2.0.5",
-    "@types/mime-types": "^2",
     "@types/node": "^24.0.0",
     "@types/yauzl-promise": "^4",
     "@types/yazl": "^3",
@@ -48,13 +47,13 @@
   },
   "dependencies": {
     "@readium/shared": "^2.2.0",
-    "@storyteller-platform/fs": "^0.1.4",
-    "@storyteller-platform/path": "^0.1.3",
+    "@storyteller-platform/fs": "^0.2.0",
+    "@storyteller-platform/media-types": "^0.1.0",
+    "@storyteller-platform/path": "^0.2.0",
     "@zip.js/zip.js": "^2.0.0",
     "async-mutex": "^0.5.0",
     "fast-xml-parser": "^4.0.1",
     "mem": "^8.0.0",
-    "mime-types": "^3.0.1",
     "nanoid": "^5.1.5",
     "yauzl-promise": "^4.0.0",
     "yazl": "^3.3.1"