XFA - Move the fake HTML representation of XFA from the worker to the…

… main thread

  - the only goal of this patch is to be able to get synchronously the fake html when printing from firefox:
    - in order to print we need to inject some html in beforeprint callback but we cannot block in waiting for all the pages.
  - from a memory point of view: it doesn't change anything since the fake HTML is deleted in the worker;
  - this way we don't break any assumptions.

src/core/document.js

@@ -146,8 +146,7 @@ class Page {
 
   _getBoundingBox(name) {
     if (this.xfaData) {
-      const { width, height } = this.xfaData.attributes.style;
-      return [0, 0, parseInt(width), parseInt(height)];
+      return this.xfaData.bbox;
     }
 
     const box = this._getInheritableProperty(name, /* getArray = */ true);
@@ -241,7 +240,9 @@ class Page {
 
   get xfaData() {
     if (this.xfaFactory) {
-      return shadow(this, "xfaData", this.xfaFactory.getPage(this.pageIndex));
+      return shadow(this, "xfaData", {
+        bbox: this.xfaFactory.getBoundingBox(this.pageIndex),
+      });
     }
     return shadow(this, "xfaData", null);
   }
@@ -851,8 +852,11 @@ class PDFDocument {
     return shadow(this, "xfaFaxtory", null);
   }
 
-  get isPureXfa() {
-    return this.xfaFactory !== null;
+  get htmlForXfa() {
+    if (this.xfaFactory) {
+      return this.xfaFactory.getPages();
+    }
+    return null;
   }
 
   async loadXfaFonts(handler, task) {

src/core/worker.js

@@ -187,13 +187,13 @@ class WorkerMessageHandler {
         await pdfManager.ensureDoc("checkFirstPage");
       }
 
-      const [numPages, fingerprint, isPureXfa] = await Promise.all([
+      const [numPages, fingerprint, htmlForXfa] = await Promise.all([
         pdfManager.ensureDoc("numPages"),
         pdfManager.ensureDoc("fingerprint"),
-        pdfManager.ensureDoc("isPureXfa"),
+        pdfManager.ensureDoc("htmlForXfa"),
       ]);
 
-      if (isPureXfa) {
+      if (htmlForXfa) {
         const task = new WorkerTask("loadXfaFonts");
         startWorkerTask(task);
         await pdfManager
@@ -203,7 +203,7 @@ class WorkerMessageHandler {
           })
           .then(() => finishWorkerTask(task));
       }
-      return { numPages, fingerprint, isPureXfa };
+      return { numPages, fingerprint, htmlForXfa };
     }
 
     function getPdfManager(data, evaluatorOptions, enableXfa) {
@@ -501,12 +501,6 @@ class WorkerMessageHandler {
       });
     });
 
-    handler.on("GetPageXfa", function wphSetupGetXfa({ pageIndex }) {
-      return pdfManager.getPage(pageIndex).then(function (page) {
-        return pdfManager.ensure(page, "xfaData");
-      });
-    });
-
     handler.on("GetOutline", function wphSetupGetOutline(data) {
       return pdfManager.ensureCatalog("documentOutline");
     });

src/core/xfa/factory.js

@@ -22,18 +22,35 @@ class XFAFactory {
     try {
       this.root = new XFAParser().parse(XFAFactory._createDocument(data));
       this.form = new Binder(this.root).bind();
-      this.pages = this.form[$toHTML]();
+      this._createPages();
     } catch (e) {
       console.log(e);
     }
   }
 
-  getPage(pageIndex) {
-    return this.pages.children[pageIndex];
+  _createPages() {
+    this.pages = this.form[$toHTML]();
+    this.dims = this.pages.children.map(c => {
+      const { width, height } = c.attributes.style;
+      return [0, 0, parseInt(width), parseInt(height)];
+    });
+  }
+
+  getBoundingBox(pageIndex) {
+    return this.dims[pageIndex];
   }
 
   get numberPages() {
-    return this.pages.children.length;
+    return this.dims.length;
+  }
+
+  getPages() {
+    if (!this.pages) {
+      this._createPages();
+    }
+    const pages = this.pages;
+    this.pages = null;
+    return pages;
   }
 
   static _createDocument(data) {

src/display/api.js

@@ -695,7 +695,15 @@ class PDFDocumentProxy {
    * @type {boolean} True if only XFA form.
    */
   get isPureXfa() {
-    return this._pdfInfo.isPureXfa;
+    return !!this._transport._htmlForXfa;
+  }
+
+  /**
+   * @type {Object | null} An object representing a HTML tree structure
+   * to render the XFA, or `null` when no XFA form exists.
+   */
+  get allXfaHtml() {
+    return this._transport._htmlForXfa;
   }
 
   /**
@@ -1255,8 +1263,8 @@ class PDFPageProxy {
    *   are {Object} with a name, attributes (class, style, ...), value and
    *   children, very similar to a HTML DOM tree), or `null` if no XFA exists.
    */
-  getXfa() {
-    return (this._xfaPromise ||= this._transport.getPageXfa(this._pageIndex));
+  async getXfa() {
+    return this._transport._htmlForXfa?.children[this._pageIndex] || null;
   }
 
   /**
@@ -1552,7 +1560,6 @@ class PDFPageProxy {
     this.objs.clear();
     this._annotationsPromise = null;
     this._jsActionsPromise = null;
-    this._xfaPromise = null;
     this._structTreePromise = null;
     this.pendingCleanup = false;
     return Promise.all(waitOn);
@@ -1588,7 +1595,6 @@ class PDFPageProxy {
     this.objs.clear();
     this._annotationsPromise = null;
     this._jsActionsPromise = null;
-    this._xfaPromise = null;
     this._structTreePromise = null;
     if (resetStats && this._stats) {
       this._stats = new StatTimer();
@@ -2456,6 +2462,8 @@ class WorkerTransport {
 
     messageHandler.on("GetDoc", ({ pdfInfo }) => {
       this._numPages = pdfInfo.numPages;
+      this._htmlForXfa = pdfInfo.htmlForXfa;
+      delete pdfInfo.htmlForXfa;
       loadingTask._capability.resolve(new PDFDocumentProxy(pdfInfo, this));
     });
 
@@ -2812,12 +2820,6 @@ class WorkerTransport {
     });
   }
 
-  getPageXfa(pageIndex) {
-    return this.messageHandler.sendWithPromise("GetPageXfa", {
-      pageIndex,
-    });
-  }
-
   getStructTree(pageIndex) {
     return this.messageHandler.sendWithPromise("GetStructTree", {
       pageIndex,

test/unit/xfa_tohtml_spec.js

@@ -57,7 +57,8 @@ describe("XFAFactory", function () {
 
       expect(factory.numberPages).toEqual(2);
 
-      const page1 = factory.getPage(0);
+      const pages = factory.getPages();
+      const page1 = pages.children[0];
       expect(page1.attributes.style).toEqual({
         height: "789px",
         width: "456px",
@@ -99,7 +100,7 @@ describe("XFAFactory", function () {
 
       // draw element must be on each page.
       expect(draw.attributes.style).toEqual(
-        factory.getPage(1).children[1].children[0].attributes.style
+        pages.children[1].children[1].children[0].attributes.style
       );
     });
   });