Merge pull request #21717 from calixteman/fix/pdf-editor-image-only-metadata

Handle image-only PDF exports

src/core/editor/pdf_editor.js

@@ -997,13 +997,6 @@ class PDFEditor {
       // async.
       this.oldPages[newPageIndex] = null;
     };
-    // Image entries don't carry document identity, so ignore them when
-    // deciding whether we're operating on a single source PDF.
-    const docPageInfos = pageInfos.filter(info => !!info.document);
-    this.isSingleFile =
-      docPageInfos.length === 1 ||
-      (docPageInfos.length > 0 &&
-        docPageInfos.every(info => info.document === docPageInfos[0].document));
     const allDocumentData = [];
 
     if (annotationStorage) {
@@ -1099,7 +1092,11 @@ class PDFEditor {
       }
     }
     await Promise.all(promises);
+    if (this.oldPages.length === 0) {
+      throw new Error("extractPages: nothing to extract.");
+    }
     const copyCounts = new Map();
+    const documents = new Set();
     for (let i = 0, ii = this.oldPages.length; i < ii; i++) {
       const pageData = this.oldPages[i];
       if (pageData === undefined) {
@@ -1110,8 +1107,10 @@ class PDFEditor {
         const copyLevel = copyCounts.get(page) ?? 0;
         copyCounts.set(page, copyLevel + 1);
         pageData.copyLevel = copyLevel;
+        documents.add(pageData.documentData.document);
       }
     }
+    this.isSingleFile = documents.size === 1;
     promises.length = 0;
 
     this.#collectValidDestinations(allDocumentData);

test/unit/api_spec.js

@@ -7757,6 +7757,44 @@ small scripts as well as for`);
         await loadingTask.destroy();
       });
 
+      it("extracts only an image when all source pages are excluded", async function () {
+        if (isNodeJS) {
+          pending("Cannot create a bitmap from Node.js.");
+        }
+        let loadingTask = getDocument(buildGetDocumentParams("empty.pdf"));
+        let pdfDoc = await loadingTask.promise;
+        const bitmap = await getImageBitmap("firefox_logo.png");
+
+        const data = await pdfDoc.extractPages([
+          { document: null, excludePages: [0] },
+          { image: bitmap },
+        ]);
+        expect(data).not.toBeNull();
+        await loadingTask.destroy();
+
+        loadingTask = getDocument({ data });
+        pdfDoc = await loadingTask.promise;
+        expect(pdfDoc.numPages).toEqual(1);
+
+        const pdfPage = await pdfDoc.getPage(1);
+        const { fnArray } = await pdfPage.getOperatorList();
+        expect(fnArray).toContain(OPS.paintImageXObject);
+
+        await loadingTask.destroy();
+      });
+
+      it("returns null when every source page is excluded", async function () {
+        const loadingTask = getDocument(buildGetDocumentParams("basicapi.pdf"));
+        const pdfDoc = await loadingTask.promise;
+
+        const data = await pdfDoc.extractPages([
+          { document: null, excludePages: [[0, 2]] },
+        ]);
+        expect(data).toBeNull();
+
+        await loadingTask.destroy();
+      });
+
       it("preserves EmbeddedFiles (attachments) when extracting pages", async function () {
         let loadingTask = getDocument(buildGetDocumentParams("attachment.pdf"));
         let pdfDoc = await loadingTask.promise;