Prefer /Resources from the /Contents stream-dict, if available

In rare cases /Resources are also found in the /Contents stream-dict, in addition to in the /Page dict, hence we need to prefer those when available; see `issue18894.pdf`.

Author: Snuffleupagus

src/core/annotation.js

@@ -49,6 +49,8 @@ import {
   lookupNormalRect,
   lookupRect,
   numberToString,
+  RESOURCES_KEYS_OPERATOR_LIST,
+  RESOURCES_KEYS_TEXT_CONTENT,
   stringToAsciiOrUTF16BE,
   stringToUTF16String,
 } from "./core_utils.js";
@@ -1196,7 +1198,7 @@ class Annotation {
 
     const appearanceDict = appearance.dict;
     const resources = await this.loadResources(
-      ["ExtGState", "ColorSpace", "Pattern", "Shading", "XObject", "Font"],
+      RESOURCES_KEYS_OPERATOR_LIST,
       appearance
     );
     const bbox = lookupRect(appearanceDict.getArray("BBox"), [0, 0, 1, 1]);
@@ -1257,7 +1259,7 @@ class Annotation {
     }
 
     const resources = await this.loadResources(
-      ["ExtGState", "Font", "Properties", "XObject"],
+      RESOURCES_KEYS_TEXT_CONTENT,
       this.appearance
     );
 

src/core/core_utils.js

@@ -32,6 +32,23 @@ const MIN_INT_32 = -(2 ** 31);
 
 const IDENTITY_MATRIX = [1, 0, 0, 1, 0, 0];
 
+const RESOURCES_KEYS_OPERATOR_LIST = [
+  "ColorSpace",
+  "ExtGState",
+  "Font",
+  "Pattern",
+  "Properties",
+  "Shading",
+  "XObject",
+];
+
+const RESOURCES_KEYS_TEXT_CONTENT = [
+  "ExtGState",
+  "Font",
+  "Properties",
+  "XObject",
+];
+
 function getLookupTableFactory(initializer) {
   let lookup;
   return function () {
@@ -745,6 +762,8 @@ export {
   readUint16,
   readUint32,
   recoverJsURL,
+  RESOURCES_KEYS_OPERATOR_LIST,
+  RESOURCES_KEYS_TEXT_CONTENT,
   stringToAsciiOrUTF16BE,
   stringToUTF16HexString,
   stringToUTF16String,

src/core/document.js

@@ -45,6 +45,8 @@ import {
   lookupNormalRect,
   MissingDataException,
   PDF_VERSION_REGEXP,
+  RESOURCES_KEYS_OPERATOR_LIST,
+  RESOURCES_KEYS_TEXT_CONTENT,
   validateCSSFont,
   XRefEntryException,
   XRefParseException,
@@ -419,6 +421,25 @@ class Page {
     await objectLoader.load();
   }
 
+  async #getMergedResources(streamDict, keys) {
+    // In rare cases /Resources are also found in the /Contents stream-dict,
+    // in addition to in the /Page dict, hence we need to prefer those when
+    // available (see issue18894.pdf).
+    const localResources = streamDict?.get("Resources");
+
+    if (!(localResources instanceof Dict)) {
+      return this.resources;
+    }
+    const objectLoader = new ObjectLoader(localResources, keys, this.xref);
+    await objectLoader.load();
+
+    return Dict.merge({
+      xref: this.xref,
+      dictArray: [localResources, this.resources],
+      mergeSubDicts: true,
+    });
+  }
+
   async getOperatorList({
     handler,
     sink,
@@ -429,15 +450,7 @@ class Page {
     modifiedIds = null,
   }) {
     const contentStreamPromise = this.getContentStream();
-    const resourcesPromise = this.loadResources([
-      "ColorSpace",
-      "ExtGState",
-      "Font",
-      "Pattern",
-      "Properties",
-      "Shading",
-      "XObject",
-    ]);
+    const resourcesPromise = this.loadResources(RESOURCES_KEYS_OPERATOR_LIST);
 
     const partialEvaluator = new PartialEvaluator({
       xref: this.xref,
@@ -525,11 +538,15 @@ class Page {
       contentStreamPromise,
       resourcesPromise,
     ]).then(async ([contentStream]) => {
+      const resources = await this.#getMergedResources(
+        contentStream.dict,
+        RESOURCES_KEYS_OPERATOR_LIST
+      );
       const opList = new OperatorList(intent, sink);
 
       handler.send("StartRenderPage", {
         transparency: partialEvaluator.hasBlendModes(
-          this.resources,
+          resources,
           this.nonBlendModesSet
         ),
         pageIndex: this.pageIndex,
@@ -539,7 +556,7 @@ class Page {
       await partialEvaluator.getOperatorList({
         stream: contentStream,
         task,
-        resources: this.resources,
+        resources,
         operatorList: opList,
       });
       return opList;
@@ -642,19 +659,19 @@ class Page {
     sink,
   }) {
     const contentStreamPromise = this.getContentStream();
-    const resourcesPromise = this.loadResources([
-      "ExtGState",
-      "Font",
-      "Properties",
-      "XObject",
-    ]);
+    const resourcesPromise = this.loadResources(RESOURCES_KEYS_TEXT_CONTENT);
     const langPromise = this.pdfManager.ensureCatalog("lang");
 
     const [contentStream, , lang] = await Promise.all([
       contentStreamPromise,
       resourcesPromise,
       langPromise,
     ]);
+    const resources = await this.#getMergedResources(
+      contentStream.dict,
+      RESOURCES_KEYS_TEXT_CONTENT
+    );
+
     const partialEvaluator = new PartialEvaluator({
       xref: this.xref,
       handler,
@@ -672,7 +689,7 @@ class Page {
     return partialEvaluator.getTextContent({
       stream: contentStream,
       task,
-      resources: this.resources,
+      resources,
       includeMarkedContent,
       disableNormalization,
       sink,