Merge pull request #13105 from Snuffleupagus/BasePdfManager-parseDocBaseUrl

Improve memory usage around the `BasePdfManager.docBaseUrl` parameter (PR 7689 follow-up)

web/app.js

@@ -22,7 +22,6 @@ import {
   DEFAULT_SCALE_VALUE,
   EventBus,
   getActiveOrFocusedElement,
-  getPDFFileNameFromURL,
   isValidRotation,
   isValidScrollMode,
   isValidSpreadMode,
@@ -44,6 +43,7 @@ import {
   createPromiseCapability,
   getDocument,
   getFilenameFromUrl,
+  getPdfFilenameFromUrl,
   GlobalWorkerOptions,
   InvalidPDFException,
   isPdfFile,
@@ -749,7 +749,7 @@ const PDFViewerApplication = {
   setTitleUsingUrl(url = "") {
     this.url = url;
     this.baseUrl = url.split("#")[0];
-    let title = getPDFFileNameFromURL(url, "");
+    let title = getPdfFilenameFromUrl(url, "");
     if (!title) {
       try {
         title = decodeURIComponent(getFilenameFromUrl(url)) || url;
@@ -773,7 +773,7 @@ const PDFViewerApplication = {
   get _docFilename() {
     // Use `this.url` instead of `this.baseUrl` to perform filename detection
     // based on the reference fragment as ultimate fallback if needed.
-    return this._contentDispositionFilename || getPDFFileNameFromURL(this.url);
+    return this._contentDispositionFilename || getPdfFilenameFromUrl(this.url);
   },
 
   /**

web/generic_scripting.js

@@ -13,8 +13,7 @@
  * limitations under the License.
  */
 
-import { getPDFFileNameFromURL } from "./ui_utils.js";
-import { loadScript } from "pdfjs-lib";
+import { getPdfFilenameFromUrl, loadScript } from "pdfjs-lib";
 
 async function docPropertiesLookup(pdfDocument) {
   const url = "",
@@ -37,7 +36,7 @@ async function docPropertiesLookup(pdfDocument) {
     ...info,
     baseURL: baseUrl,
     filesize: contentLength,
-    filename: contentDispositionFilename || getPDFFileNameFromURL(url),
+    filename: contentDispositionFilename || getPdfFilenameFromUrl(url),
     metadata: metadata?.getRaw(),
     authors: metadata?.get("dc:creator"),
     numPages: pdfDocument.numPages,

web/pdf_document_properties.js

@@ -13,12 +13,12 @@
  * limitations under the License.
  */
 
-import { createPromiseCapability, PDFDateString } from "pdfjs-lib";
 import {
-  getPageSizeInches,
-  getPDFFileNameFromURL,
-  isPortraitOrientation,
-} from "./ui_utils.js";
+  createPromiseCapability,
+  getPdfFilenameFromUrl,
+  PDFDateString,
+} from "pdfjs-lib";
+import { getPageSizeInches, isPortraitOrientation } from "./ui_utils.js";
 
 const DEFAULT_FIELD_CONTENT = "-";
 
@@ -140,7 +140,7 @@ class PDFDocumentProperties {
       pageSize,
       isLinearized,
     ] = await Promise.all([
-      contentDispositionFilename || getPDFFileNameFromURL(this.url),
+      contentDispositionFilename || getPdfFilenameFromUrl(this.url),
       this._parseFileSize(contentLength),
       this._parseDate(info.CreationDate),
       this._parseDate(info.ModDate),

web/ui_utils.js

@@ -570,60 +570,6 @@ function noContextMenuHandler(evt) {
   evt.preventDefault();
 }
 
-function isDataSchema(url) {
-  let i = 0;
-  const ii = url.length;
-  while (i < ii && url[i].trim() === "") {
-    i++;
-  }
-  return url.substring(i, i + 5).toLowerCase() === "data:";
-}
-
-/**
- * Returns the filename or guessed filename from the url (see issue 3455).
- * @param {string} url - The original PDF location.
- * @param {string} defaultFilename - The value returned if the filename is
- *   unknown, or the protocol is unsupported.
- * @returns {string} Guessed PDF filename.
- */
-function getPDFFileNameFromURL(url, defaultFilename = "document.pdf") {
-  if (typeof url !== "string") {
-    return defaultFilename;
-  }
-  if (isDataSchema(url)) {
-    console.warn(
-      "getPDFFileNameFromURL: " +
-        'ignoring "data:" URL for performance reasons.'
-    );
-    return defaultFilename;
-  }
-  const reURI = /^(?:(?:[^:]+:)?\/\/[^/]+)?([^?#]*)(\?[^#]*)?(#.*)?$/;
-  //              SCHEME        HOST        1.PATH  2.QUERY   3.REF
-  // Pattern to get last matching NAME.pdf
-  const reFilename = /[^/?#=]+\.pdf\b(?!.*\.pdf\b)/i;
-  const splitURI = reURI.exec(url);
-  let suggestedFilename =
-    reFilename.exec(splitURI[1]) ||
-    reFilename.exec(splitURI[2]) ||
-    reFilename.exec(splitURI[3]);
-  if (suggestedFilename) {
-    suggestedFilename = suggestedFilename[0];
-    if (suggestedFilename.includes("%")) {
-      // URL-encoded %2Fpath%2Fto%2Ffile.pdf should be file.pdf
-      try {
-        suggestedFilename = reFilename.exec(
-          decodeURIComponent(suggestedFilename)
-        )[0];
-      } catch (ex) {
-        // Possible (extremely rare) errors:
-        // URIError "Malformed URI", e.g. for "%AA.pdf"
-        // TypeError "null has no properties", e.g. for "%2F.pdf"
-      }
-    }
-  }
-  return suggestedFilename || defaultFilename;
-}
-
 function normalizeWheelEventDirection(evt) {
   let delta = Math.hypot(evt.deltaX, evt.deltaY);
   const angle = Math.atan2(evt.deltaY, evt.deltaX);
@@ -1063,7 +1009,6 @@ export {
   getActiveOrFocusedElement,
   getOutputScale,
   getPageSizeInches,
-  getPDFFileNameFromURL,
   getVisibleElements,
   isPortraitOrientation,
   isValidRotation,