Merge pull request #19719 from Snuffleupagus/structTree-xref

Pass the `XRef`-instance explicitly to the `StructTreeRoot` class, and simplify the `StructElementNode.prototype.role` getter

extensions/chromium/preferences_schema.json

@@ -228,7 +228,7 @@
     "enableAutoLinking": {
       "description": "Enable creation of hyperlinks from text that look like URLs.",
       "type": "boolean",
-      "default": false
+      "default": true
     }
   }
 }

src/core/catalog.js

@@ -262,7 +262,7 @@ class Catalog {
   get structTreeRoot() {
     let structTree = null;
     try {
-      structTree = this._readStructTreeRoot();
+      structTree = this.#readStructTreeRoot();
     } catch (ex) {
       if (ex instanceof MissingDataException) {
         throw ex;
@@ -272,17 +272,14 @@ class Catalog {
     return shadow(this, "structTreeRoot", structTree);
   }
 
-  /**
+  #readStructTreeRoot() {
-   * @private
-   */
-  _readStructTreeRoot() {
     const rawObj = this._catDict.getRaw("StructTreeRoot");
     const obj = this.xref.fetchIfRef(rawObj);
     if (!(obj instanceof Dict)) {
       return null;
     }
 
-    const root = new StructTreeRoot(obj, rawObj);
+    const root = new StructTreeRoot(this.xref, obj, rawObj);
     root.init();
     return root;
   }

src/core/struct_tree.js

@@ -29,7 +29,8 @@ const StructElementType = {
 };
 
 class StructTreeRoot {
-  constructor(rootDict, rootRef) {
+  constructor(xref, rootDict, rootRef) {
+    this.xref = xref;
     this.dict = rootDict;
     this.ref = rootRef instanceof Ref ? rootRef : null;
     this.roleMap = new Map();
@@ -158,7 +159,7 @@ class StructTreeRoot {
     }
   }
 
-  async canUpdateStructTree({ pdfManager, xref, newAnnotationsByPage }) {
+  async canUpdateStructTree({ pdfManager, newAnnotationsByPage }) {
     if (!this.ref) {
       warn("Cannot update the struct tree: no root reference.");
       return false;
@@ -180,7 +181,7 @@ class StructTreeRoot {
       warn("Cannot update the struct tree: nums isn't an array.");
       return false;
     }
-    const numberTree = new NumberTree(parentTree, xref);
+    const numberTree = new NumberTree(parentTree, this.xref);
 
     for (const pageIndex of newAnnotationsByPage.keys()) {
       const { pageDict } = await pdfManager.getPage(pageIndex);
@@ -201,7 +202,7 @@ class StructTreeRoot {
       const { pageDict } = await pdfManager.getPage(pageIndex);
       StructTreeRoot.#collectParents({
         elements,
-        xref: this.dict.xref,
+        xref: this.xref,
         pageDict,
         numberTree,
       });
@@ -233,9 +234,8 @@ class StructTreeRoot {
   }
 
   async updateStructureTree({ newAnnotationsByPage, pdfManager, changes }) {
-    const xref = this.dict.xref;
+    const { ref: structTreeRootRef, xref } = this;
     const structTreeRoot = this.dict.clone();
-    const structTreeRootRef = this.ref;
     const cache = new RefSetCache();
     cache.put(structTreeRootRef, structTreeRoot);
 
@@ -541,6 +541,7 @@ class StructTreeRoot {
 class StructElementNode {
   constructor(tree, dict) {
     this.tree = tree;
+    this.xref = tree.xref;
     this.dict = dict;
     this.kids = [];
     this.parseKids();
@@ -550,10 +551,7 @@ class StructElementNode {
     const nameObj = this.dict.get("S");
     const name = nameObj instanceof Name ? nameObj.name : "";
     const { root } = this.tree;
-    if (root.roleMap.has(name)) {
+    return root.roleMap.get(name) ?? name;
-      return root.roleMap.get(name);
-    }
-    return name;
   }
 
   parseKids() {
@@ -565,7 +563,7 @@ class StructElementNode {
     const kids = this.dict.get("K");
     if (Array.isArray(kids)) {
       for (const kid of kids) {
-        const element = this.parseKid(pageObjId, kid);
+        const element = this.parseKid(pageObjId, this.xref.fetchIfRef(kid));
         if (element) {
           this.kids.push(element);
         }
@@ -592,33 +590,26 @@ class StructElementNode {
       });
     }
 
-    // Find the dictionary for the kid.
+    if (!(kid instanceof Dict)) {
-    let kidDict = null;
-    if (kid instanceof Ref) {
-      kidDict = this.dict.xref.fetch(kid);
-    } else if (kid instanceof Dict) {
-      kidDict = kid;
-    }
-    if (!kidDict) {
       return null;
     }
-    const pageRef = kidDict.getRaw("Pg");
+
+    const pageRef = kid.getRaw("Pg");
     if (pageRef instanceof Ref) {
       pageObjId = pageRef.toString();
     }
 
-    const type =
+    const type = kid.get("Type") instanceof Name ? kid.get("Type").name : null;
-      kidDict.get("Type") instanceof Name ? kidDict.get("Type").name : null;
     if (type === "MCR") {
       if (this.tree.pageDict.objId !== pageObjId) {
         return null;
       }
-      const kidRef = kidDict.getRaw("Stm");
+      const kidRef = kid.getRaw("Stm");
       return new StructElement({
         type: StructElementType.STREAM_CONTENT,
         refObjId: kidRef instanceof Ref ? kidRef.toString() : null,
         pageObjId,
-        mcid: kidDict.get("MCID"),
+        mcid: kid.get("MCID"),
       });
     }
 
@@ -626,7 +617,7 @@ class StructElementNode {
       if (this.tree.pageDict.objId !== pageObjId) {
         return null;
       }
-      const kidRef = kidDict.getRaw("Obj");
+      const kidRef = kid.getRaw("Obj");
       return new StructElement({
         type: StructElementType.OBJECT,
         refObjId: kidRef instanceof Ref ? kidRef.toString() : null,
@@ -636,7 +627,7 @@ class StructElementNode {
 
     return new StructElement({
       type: StructElementType.ELEMENT,
-      dict: kidDict,
+      dict: kid,
     });
   }
 }
@@ -661,7 +652,8 @@ class StructElement {
 class StructTreePage {
   constructor(structTreeRoot, pageDict) {
     this.root = structTreeRoot;
-    this.rootDict = structTreeRoot ? structTreeRoot.dict : null;
+    this.xref = structTreeRoot?.xref ?? null;
+    this.rootDict = structTreeRoot?.dict ?? null;
     this.pageDict = pageDict;
     this.nodes = [];
   }
@@ -687,7 +679,7 @@ class StructTreePage {
     }
 
     const map = new Map();
-    const numberTree = new NumberTree(parentTree, this.rootDict.xref);
+    const numberTree = new NumberTree(parentTree, this.xref);
 
     for (const [elemId] of ids) {
       const obj = numberTree.getRaw(elemId);
@@ -714,14 +706,14 @@ class StructTreePage {
     }
 
     const map = new Map();
-    const numberTree = new NumberTree(parentTree, this.rootDict.xref);
+    const numberTree = new NumberTree(parentTree, this.xref);
 
     if (Number.isInteger(id)) {
       const parentArray = numberTree.get(id);
       if (Array.isArray(parentArray)) {
         for (const ref of parentArray) {
           if (ref instanceof Ref) {
-            this.addNode(this.rootDict.xref.fetch(ref), map);
+            this.addNode(this.xref.fetch(ref), map);
           }
         }
       }
@@ -733,7 +725,7 @@ class StructTreePage {
     for (const [elemId, type] of ids) {
       const obj = numberTree.get(elemId);
       if (obj) {
-        const elem = this.addNode(this.rootDict.xref.fetchIfRef(obj), map);
+        const elem = this.addNode(this.xref.fetchIfRef(obj), map);
         if (
           elem?.kids?.length === 1 &&
           elem.kids[0].type === StructElementType.OBJECT

src/core/worker.js

@@ -558,7 +558,6 @@ class WorkerMessageHandler {
           } else if (
             await _structTreeRoot.canUpdateStructTree({
               pdfManager,
-              xref,
               newAnnotationsByPage,
             })
           ) {

src/core/xfa/template.js

@@ -5721,12 +5721,7 @@ class Text extends ContentObject {
     if (typeof this[$content] === "string") {
       return this[$content]
         .split(/[\u2029\u2028\n]/)
-        .reduce((acc, line) => {
+        .filter(line => !!line)
-          if (line) {
-            acc.push(line);
-          }
-          return acc;
-        }, [])
         .join("\n");
     }
     return this[$content][$text]();
@@ -5748,18 +5743,15 @@ class Text extends ContentObject {
           .map(para =>
             // Convert a paragraph into a set of <span> (for lines)
             // separated by <br>.
-            para.split(/[\u2028\n]/).reduce((acc, line) => {
+            para.split(/[\u2028\n]/).flatMap(line => [
-              acc.push(
+              {
-                {
+                name: "span",
-                  name: "span",
+                value: line,
-                  value: line,
+              },
-                },
+              {
-                {
+                name: "br",
-                  name: "br",
+              },
-                }
+            ])
-              );
-              return acc;
-            }, [])
           )
           .forEach(lines => {
             html.children.push({

web/app_options.js

@@ -202,7 +202,7 @@ const defaultOptions = {
   },
   enableAutoLinking: {
     /** @type {boolean} */
-    value: typeof PDFJSDev === "undefined" || PDFJSDev.test("MOZCENTRAL"),
+    value: true,
     kind: OptionKind.VIEWER + OptionKind.PREFERENCE,
   },
   enableDetailCanvas: {

web/pdf_page_view.js

@@ -95,7 +95,7 @@ import { XfaLayerBuilder } from "./xfa_layer_builder.js";
  * @property {boolean} [enableHWA] - Enables hardware acceleration for
  *   rendering. The default value is `false`.
  * @property {boolean} [enableAutoLinking] - Enable creation of hyperlinks from
- *   text that look like URLs. The default value is `false`.
+ *   text that look like URLs. The default value is `true`.
  */
 
 const DEFAULT_LAYER_PROPERTIES =
@@ -130,7 +130,7 @@ class PDFPageView extends BasePDFPageView {
 
   #canvasWrapper = null;
 
-  #enableAutoLinking = false;
+  #enableAutoLinking = true;
 
   #hasRestrictedScaling = false;
 
@@ -188,7 +188,7 @@ class PDFPageView extends BasePDFPageView {
     this.maxCanvasPixels =
       options.maxCanvasPixels ?? AppOptions.get("maxCanvasPixels");
     this.maxCanvasDim = options.maxCanvasDim || AppOptions.get("maxCanvasDim");
-    this.#enableAutoLinking = options.enableAutoLinking || false;
+    this.#enableAutoLinking = options.enableAutoLinking !== false;
 
     this.l10n = options.l10n;
     if (typeof PDFJSDev === "undefined" || PDFJSDev.test("GENERIC")) {

web/pdf_viewer.js

@@ -137,7 +137,7 @@ function isValidAnnotationEditorMode(mode) {
  * @property {boolean} [supportsPinchToZoom] - Enable zooming on pinch gesture.
  *   The default value is `true`.
  * @property {boolean} [enableAutoLinking] - Enable creation of hyperlinks from
- *   text that look like URLs. The default value is `false`.
+ *   text that look like URLs. The default value is `true`.
  */
 
 class PDFPageViewBuffer {
@@ -238,7 +238,7 @@ class PDFViewer {
 
   #enableNewAltTextWhenAddingImage = false;
 
-  #enableAutoLinking = false;
+  #enableAutoLinking = true;
 
   #eventAbortController = null;
 
@@ -340,7 +340,7 @@ class PDFViewer {
     this.#mlManager = options.mlManager || null;
     this.#enableHWA = options.enableHWA || false;
     this.#supportsPinchToZoom = options.supportsPinchToZoom !== false;
-    this.#enableAutoLinking = options.enableAutoLinking || false;
+    this.#enableAutoLinking = options.enableAutoLinking !== false;
 
     this.defaultRenderingQueue = !options.renderingQueue;
     if (