XFA - Overwrite AcroForm dictionary when saving if no datasets in XFA (bug 1720179)

src/core/catalog.js

@@ -130,6 +130,11 @@ class Catalog {
     return shadow(this, "acroForm", acroForm);
   }
 
+  get acroFormRef() {
+    const value = this._catDict.getRaw("AcroForm");
+    return shadow(this, "acroFormRef", isRef(value) ? value : null);
+  }
+
   get metadata() {
     const streamRef = this._catDict.getRaw("Metadata");
     if (!isRef(streamRef)) {

src/core/worker.js

@@ -573,6 +573,7 @@ class WorkerMessageHandler {
         const promises = [
           pdfManager.onLoadedStream(),
           pdfManager.ensureCatalog("acroForm"),
+          pdfManager.ensureCatalog("acroFormRef"),
           pdfManager.ensureDoc("xref"),
           pdfManager.ensureDoc("startXRef"),
         ];
@@ -597,6 +598,7 @@ class WorkerMessageHandler {
         return Promise.all(promises).then(function ([
           stream,
           acroForm,
+          acroFormRef,
           xref,
           startXRef,
           ...refs
@@ -621,15 +623,22 @@ class WorkerMessageHandler {
             }
           }
 
-          const xfa = (acroForm instanceof Dict && acroForm.get("XFA")) || [];
+          const xfa = (acroForm instanceof Dict && acroForm.get("XFA")) || null;
           let xfaDatasets = null;
+          let hasDatasets = false;
           if (Array.isArray(xfa)) {
             for (let i = 0, ii = xfa.length; i < ii; i += 2) {
               if (xfa[i] === "datasets") {
                 xfaDatasets = xfa[i + 1];
+                acroFormRef = null;
+                hasDatasets = true;
               }
             }
+            if (xfaDatasets === null) {
+              xfaDatasets = xref.getNewRef();
+            }
           } else {
+            acroFormRef = null;
             // TODO: Support XFA streams.
             warn("Unsupported XFA type.");
           }
@@ -666,6 +675,9 @@ class WorkerMessageHandler {
             newRefs,
             xref,
             datasetsRef: xfaDatasets,
+            hasDatasets,
+            acroFormRef,
+            acroForm,
             xfaData,
           });
         });

src/core/writer.js

@@ -146,10 +146,54 @@ function writeXFADataForAcroform(str, newRefs) {
   return buffer.join("");
 }
 
-function updateXFA(xfaData, datasetsRef, newRefs, xref) {
-  if (datasetsRef === null || xref === null) {
+function updateXFA({
+  xfaData,
+  datasetsRef,
+  hasDatasets,
+  acroFormRef,
+  acroForm,
+  newRefs,
+  xref,
+  xrefInfo,
+}) {
+  if (xref === null) {
     return;
   }
+
+  if (!hasDatasets) {
+    if (!acroFormRef) {
+      warn("XFA - Cannot save it");
+      return;
+    }
+
+    // We've a XFA array which doesn't contain a datasets entry.
+    // So we'll update the AcroForm dictionary to have an XFA containing
+    // the datasets.
+    const oldXfa = acroForm.get("XFA");
+    const newXfa = oldXfa.slice();
+    newXfa.splice(2, 0, "datasets");
+    newXfa.splice(3, 0, datasetsRef);
+
+    acroForm.set("XFA", newXfa);
+
+    const encrypt = xref.encrypt;
+    let transform = null;
+    if (encrypt) {
+      transform = encrypt.createCipherTransform(
+        acroFormRef.num,
+        acroFormRef.gen
+      );
+    }
+
+    const buffer = [`${acroFormRef.num} ${acroFormRef.gen} obj\n`];
+    writeDict(acroForm, buffer, transform);
+    buffer.push("\n");
+
+    acroForm.set("XFA", oldXfa);
+
+    newRefs.push({ ref: acroFormRef, data: buffer.join("") });
+  }
+
   if (xfaData === null) {
     const datasets = xref.fetchIfRef(datasetsRef);
     xfaData = writeXFADataForAcroform(datasets.getString(), newRefs);
@@ -178,9 +222,21 @@ function incrementalUpdate({
   newRefs,
   xref = null,
   datasetsRef = null,
+  hasDatasets = false,
+  acroFormRef = null,
+  acroForm = null,
   xfaData = null,
 }) {
-  updateXFA(xfaData, datasetsRef, newRefs, xref);
+  updateXFA({
+    xfaData,
+    datasetsRef,
+    hasDatasets,
+    acroFormRef,
+    acroForm,
+    newRefs,
+    xref,
+    xrefInfo,
+  });
 
   const newXref = new Dict(null);
   const refForXrefTable = xrefInfo.newRef;

test/unit/writer_spec.js

@@ -142,4 +142,67 @@ describe("Writer", function () {
       expect(buffer.join("")).toEqual(expected);
     });
   });
+
+  describe("XFA", function () {
+    it("should update AcroForm when no datasets in XFA array", function () {
+      const originalData = new Uint8Array();
+      const newRefs = [];
+
+      const acroForm = new Dict(null);
+      acroForm.set("XFA", [
+        "preamble",
+        Ref.get(123, 0),
+        "postamble",
+        Ref.get(456, 0),
+      ]);
+      const acroFormRef = Ref.get(789, 0);
+      const datasetsRef = Ref.get(101112, 0);
+      const xfaData = "<hello>world</hello>";
+
+      const xrefInfo = {
+        newRef: Ref.get(131415, 0),
+        startXRef: 314,
+        fileIds: null,
+        rootRef: null,
+        infoRef: null,
+        encryptRef: null,
+        filename: "foo.pdf",
+        info: {},
+      };
+
+      let data = incrementalUpdate({
+        originalData,
+        xrefInfo,
+        newRefs,
+        datasetsRef,
+        hasDatasets: false,
+        acroFormRef,
+        acroForm,
+        xfaData,
+        xref: {},
+      });
+      data = bytesToString(data);
+
+      const expected =
+        "\n" +
+        "789 0 obj\n" +
+        "<< /XFA [(preamble) 123 0 R (datasets) 101112 0 R (postamble) 456 0 R]>>\n" +
+        "101112 0 obj\n" +
+        "<< /Type /EmbeddedFile /Length 20>>\n" +
+        "stream\n" +
+        "<hello>world</hello>\n" +
+        "endstream\n" +
+        "endobj\n" +
+        "131415 0 obj\n" +
+        "<< /Size 131416 /Prev 314 /Type /XRef /Index [0 1 789 1 101112 1 131415 1] /W [1 1 2] /Length 16>> stream\n" +
+        "\u0000\u0001ÿÿ\u0001\u0001\u0000\u0000\u0001T\u0000\u0000\u0001²\u0000\u0000\n" +
+        "endstream\n" +
+        "endobj\n" +
+        "startxref\n" +
+        "178\n" +
+        "%%EOF\n";
+
+      expect(data).toEqual(expected);
+    });
+  });
 });