XFA -- Load fonts permanently from the pdf

- Different fonts can be used in xfa and some of them are embedded in the pdf.
  - Load all the fonts in window.document.

Update src/core/document.js

Co-authored-by: Jonas Jenwald <jonas.jenwald@gmail.com>

Update src/core/worker.js

Co-authored-by: Jonas Jenwald <jonas.jenwald@gmail.com>

src/core/core_utils.js

@@ -19,6 +19,7 @@ import {
   bytesToString,
   objectSize,
   stringToPDFString,
+  warn,
 } from "../shared/util.js";
 import { Dict, isName, isRef, isStream, RefSet } from "./primitives.js";
 
@@ -376,6 +377,70 @@ function encodeToXmlString(str) {
   return buffer.join("");
 }
 
+function validateCSSFont(cssFontInfo) {
+  // See https://developer.mozilla.org/en-US/docs/Web/CSS/font-style.
+  const DEFAULT_CSS_FONT_OBLIQUE = "14";
+  // See https://developer.mozilla.org/en-US/docs/Web/CSS/font-weight.
+  const DEFAULT_CSS_FONT_WEIGHT = "400";
+  const CSS_FONT_WEIGHT_VALUES = new Set([
+    "100",
+    "200",
+    "300",
+    "400",
+    "500",
+    "600",
+    "700",
+    "800",
+    "900",
+    "1000",
+    "normal",
+    "bold",
+    "bolder",
+    "lighter",
+  ]);
+
+  const { fontFamily, fontWeight, italicAngle } = cssFontInfo;
+
+  // See https://developer.mozilla.org/en-US/docs/Web/CSS/string.
+  if (/^".*"$/.test(fontFamily)) {
+    if (/[^\\]"/.test(fontFamily.slice(1, fontFamily.length - 1))) {
+      warn(`XFA - FontFamily contains some unescaped ": ${fontFamily}.`);
+      return false;
+    }
+  } else if (/^'.*'$/.test(fontFamily)) {
+    if (/[^\\]'/.test(fontFamily.slice(1, fontFamily.length - 1))) {
+      warn(`XFA - FontFamily contains some unescaped ': ${fontFamily}.`);
+      return false;
+    }
+  } else {
+    // See https://developer.mozilla.org/en-US/docs/Web/CSS/custom-ident.
+    for (const ident of fontFamily.split(/[ \t]+/)) {
+      if (
+        /^([0-9]|(-([0-9]|-)))/.test(ident) ||
+        !/^[a-zA-Z0-9\-_\\]+$/.test(ident)
+      ) {
+        warn(
+          `XFA - FontFamily contains some invalid <custom-ident>: ${fontFamily}.`
+        );
+        return false;
+      }
+    }
+  }
+
+  const weight = fontWeight ? fontWeight.toString() : "";
+  cssFontInfo.fontWeight = CSS_FONT_WEIGHT_VALUES.has(weight)
+    ? weight
+    : DEFAULT_CSS_FONT_WEIGHT;
+
+  const angle = parseFloat(italicAngle);
+  cssFontInfo.italicAngle =
+    isNaN(angle) || angle < -90 || angle > 90
+      ? DEFAULT_CSS_FONT_OBLIQUE
+      : italicAngle.toString();
+
+  return true;
+}
+
 export {
   collectActions,
   encodeToXmlString,
@@ -391,6 +456,7 @@ export {
   readUint16,
   readUint32,
   toRomanNumerals,
+  validateCSSFont,
   XRefEntryException,
   XRefParseException,
 };

src/core/document.js

@@ -42,6 +42,7 @@ import {
   isName,
   isRef,
   isStream,
+  Name,
   Ref,
 } from "./primitives.js";
 import {
@@ -49,6 +50,7 @@ import {
   getInheritableProperty,
   isWhiteSpace,
   MissingDataException,
+  validateCSSFont,
   XRefEntryException,
   XRefParseException,
 } from "./core_utils.js";
@@ -854,6 +856,71 @@ class PDFDocument {
     return this.xfaFactory !== null;
   }
 
+  async loadXfaFonts(handler, task) {
+    const acroForm = await this.pdfManager.ensureCatalog("acroForm");
+
+    const resources = await acroForm.getAsync("DR");
+    if (!(resources instanceof Dict)) {
+      return;
+    }
+    const objectLoader = new ObjectLoader(resources, ["Font"], this.xref);
+    await objectLoader.load();
+
+    const fontRes = resources.get("Font");
+    if (!(fontRes instanceof Dict)) {
+      return;
+    }
+
+    const partialEvaluator = new PartialEvaluator({
+      xref: this.xref,
+      handler,
+      pageIndex: -1,
+      idFactory: this._globalIdFactory,
+      fontCache: this.catalog.fontCache,
+      builtInCMapCache: this.catalog.builtInCMapCache,
+    });
+    const operatorList = new OperatorList();
+    const initialState = {
+      font: null,
+      clone() {
+        return this;
+      },
+    };
+
+    const fonts = new Map();
+    fontRes.forEach((fontName, font) => {
+      fonts.set(fontName, font);
+    });
+    const promises = [];
+
+    for (const [fontName, font] of fonts) {
+      const descriptor = font.get("FontDescriptor");
+      if (descriptor instanceof Dict) {
+        const fontFamily = descriptor.get("FontFamily");
+        const fontWeight = descriptor.get("FontWeight");
+        const italicAngle = descriptor.get("ItalicAngle");
+        const cssFontInfo = { fontFamily, fontWeight, italicAngle };
+
+        if (!validateCSSFont(cssFontInfo)) {
+          continue;
+        }
+
+        const promise = partialEvaluator.handleSetFont(
+          resources,
+          [Name.get(fontName), 1],
+          /* fontRef = */ null,
+          operatorList,
+          task,
+          initialState,
+          /* fallbackFontDict = */ null,
+          /* cssFontInfo = */ cssFontInfo
+        );
+        promises.push(promise.catch(() => {}));
+      }
+    }
+    await Promise.all(promises);
+  }
+
   get formInfo() {
     const formInfo = {
       hasFields: false,

src/core/evaluator.js

@@ -792,12 +792,19 @@ class PartialEvaluator {
     operatorList,
     task,
     state,
-    fallbackFontDict = null
+    fallbackFontDict = null,
+    cssFontInfo = null
   ) {
     const fontName =
       fontArgs && fontArgs[0] instanceof Name ? fontArgs[0].name : null;
 
-    return this.loadFont(fontName, fontRef, resources, fallbackFontDict)
+    return this.loadFont(
+      fontName,
+      fontRef,
+      resources,
+      fallbackFontDict,
+      cssFontInfo
+    )
       .then(translated => {
         if (!translated.font.isType3Font) {
           return translated;
@@ -986,7 +993,13 @@ class PartialEvaluator {
     });
   }
 
-  loadFont(fontName, font, resources, fallbackFontDict = null) {
+  loadFont(
+    fontName,
+    font,
+    resources,
+    fallbackFontDict = null,
+    cssFontInfo = null
+  ) {
     const errorFont = async () => {
       return new TranslatedFont({
         loadedName: "g_font_error",
@@ -1055,6 +1068,7 @@ class PartialEvaluator {
     let preEvaluatedFont;
     try {
       preEvaluatedFont = this.preEvaluateFont(font);
+      preEvaluatedFont.cssFontInfo = cssFontInfo;
     } catch (reason) {
       warn(`loadFont - preEvaluateFont failed: "${reason}".`);
       return errorFont();
@@ -3529,6 +3543,7 @@ class PartialEvaluator {
       flags: descriptor.get("Flags"),
       italicAngle: descriptor.get("ItalicAngle"),
       isType3Font: false,
+      cssFontInfo: preEvaluatedFont.cssFontInfo,
     };
 
     if (composite) {

src/core/fonts.js

@@ -95,6 +95,7 @@ const EXPORT_DATA_PROPERTIES = [
   "bold",
   "charProcOperatorList",
   "composite",
+  "cssFontInfo",
   "data",
   "defaultVMetrics",
   "defaultWidth",
@@ -565,6 +566,7 @@ var Font = (function FontClosure() {
     this.loadedName = properties.loadedName;
     this.isType3Font = properties.isType3Font;
     this.missingFile = false;
+    this.cssFontInfo = properties.cssFontInfo;
 
     this.glyphCache = Object.create(null);
 
@@ -2963,8 +2965,15 @@ var Font = (function FontClosure() {
         glyphZeroId = 0;
       }
 
+      // When `cssFontInfo` is set, the font is used to render text in the HTML
+      // view (e.g. with Xfa) so nothing must be moved in the private area use.
+      if (!properties.cssFontInfo) {
         // Converting glyphs and ids into font's cmap table
-      var newMapping = adjustMapping(charCodeToGlyphId, hasGlyph, glyphZeroId);
+        var newMapping = adjustMapping(
+          charCodeToGlyphId,
+          hasGlyph,
+          glyphZeroId
+        );
         this.toFontChar = newMapping.toFontChar;
         tables.cmap = {
           tag: "cmap",
@@ -2981,6 +2990,7 @@ var Font = (function FontClosure() {
             ),
           };
         }
+      }
 
       if (!isTrueType) {
         try {

src/core/pdf_manager.js

@@ -73,6 +73,10 @@ class BasePdfManager {
     return this.pdfDocument.fontFallback(id, handler);
   }
 
+  loadXfaFonts(handler, task) {
+    return this.pdfDocument.loadXfaFonts(handler, task);
+  }
+
   cleanup(manuallyTriggered = false) {
     return this.pdfDocument.cleanup(manuallyTriggered);
   }

src/core/worker.js

@@ -193,6 +193,17 @@ class WorkerMessageHandler {
         pdfManager.ensureDoc("fingerprint"),
         pdfManager.ensureDoc("isPureXfa"),
       ]);
+
+      if (isPureXfa) {
+        const task = new WorkerTask("Load fonts for Xfa");
+        startWorkerTask(task);
+        await pdfManager
+          .loadXfaFonts(handler, task)
+          .catch(reason => {
+            // Ignore errors, to allow the document to load.
+          })
+          .then(() => finishWorkerTask(task));
+      }
       return { numPages, fingerprint, isPureXfa };
     }
 

src/display/font_loader.js

@@ -370,7 +370,22 @@ class FontFaceObject {
     if (!this.data || this.disableFontFace) {
       return null;
     }
-    const nativeFontFace = new FontFace(this.loadedName, this.data, {});
+    let nativeFontFace;
+    if (!this.cssFontInfo) {
+      nativeFontFace = new FontFace(this.loadedName, this.data, {});
+    } else {
+      const css = {
+        weight: this.cssFontInfo.fontWeight,
+      };
+      if (this.cssFontInfo.italicAngle) {
+        css.style = `oblique ${this.cssFontInfo.italicAngle}deg`;
+      }
+      nativeFontFace = new FontFace(
+        this.cssFontInfo.fontFamily,
+        this.data,
+        css
+      );
+    }
 
     if (this.fontRegistry) {
       this.fontRegistry.registerFont(this);
@@ -385,7 +400,16 @@ class FontFaceObject {
     const data = bytesToString(new Uint8Array(this.data));
     // Add the @font-face rule to the document.
     const url = `url(data:${this.mimetype};base64,${btoa(data)});`;
-    const rule = `@font-face {font-family:"${this.loadedName}";src:${url}}`;
+    let rule;
+    if (!this.cssFontInfo) {
+      rule = `@font-face {font-family:"${this.loadedName}";src:${url}}`;
+    } else {
+      let css = `font-weight: ${this.cssFontInfo.fontWeight};`;
+      if (this.cssFontInfo.italicAngle) {
+        css += `font-style: oblique ${this.cssFontInfo.italicAngle}deg;`;
+      }
+      rule = `@font-face {font-family:"${this.cssFontInfo.fontFamily}";${css}src:${url}}`;
+    }
 
     if (this.fontRegistry) {
       this.fontRegistry.registerFont(this, url);

test/unit/core_utils_spec.js

@@ -22,6 +22,7 @@ import {
   log2,
   parseXFAPath,
   toRomanNumerals,
+  validateCSSFont,
 } from "../../src/core/core_utils.js";
 import { XRefMock } from "./test_utils.js";
 
@@ -233,4 +234,103 @@ describe("core_utils", function () {
       expect(encodeToXmlString(str)).toEqual(str);
     });
   });
+
+  describe("validateCSSFont", function () {
+    it("Check font family", function () {
+      const cssFontInfo = {
+        fontFamily: `"blah blah " blah blah"`,
+        fontWeight: 0,
+        italicAngle: 0,
+      };
+
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+
+      cssFontInfo.fontFamily = `"blah blah \\" blah blah"`;
+      expect(validateCSSFont(cssFontInfo)).toEqual(true);
+
+      cssFontInfo.fontFamily = `'blah blah ' blah blah'`;
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+
+      cssFontInfo.fontFamily = `'blah blah \\' blah blah'`;
+      expect(validateCSSFont(cssFontInfo)).toEqual(true);
+
+      cssFontInfo.fontFamily = `"blah blah `;
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+
+      cssFontInfo.fontFamily = `blah blah"`;
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+
+      cssFontInfo.fontFamily = `'blah blah `;
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+
+      cssFontInfo.fontFamily = `blah blah'`;
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+
+      cssFontInfo.fontFamily = "blah blah blah";
+      expect(validateCSSFont(cssFontInfo)).toEqual(true);
+
+      cssFontInfo.fontFamily = "blah 0blah blah";
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+
+      cssFontInfo.fontFamily = "blah blah -0blah";
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+
+      cssFontInfo.fontFamily = "blah blah --blah";
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+
+      cssFontInfo.fontFamily = "blah blah -blah";
+      expect(validateCSSFont(cssFontInfo)).toEqual(true);
+
+      cssFontInfo.fontFamily = "blah fdqAJqjHJK23kl23__--Kj blah";
+      expect(validateCSSFont(cssFontInfo)).toEqual(true);
+
+      cssFontInfo.fontFamily = "blah fdqAJqjH$JK23kl23__--Kj blah";
+      expect(validateCSSFont(cssFontInfo)).toEqual(false);
+    });
+
+    it("Check font weight", function () {
+      const cssFontInfo = {
+        fontFamily: "blah",
+        fontWeight: 100,
+        italicAngle: 0,
+      };
+
+      validateCSSFont(cssFontInfo);
+      expect(cssFontInfo.fontWeight).toEqual("100");
+
+      cssFontInfo.fontWeight = "700";
+      validateCSSFont(cssFontInfo);
+      expect(cssFontInfo.fontWeight).toEqual("700");
+
+      cssFontInfo.fontWeight = "normal";
+      validateCSSFont(cssFontInfo);
+      expect(cssFontInfo.fontWeight).toEqual("normal");
+
+      cssFontInfo.fontWeight = 314;
+      validateCSSFont(cssFontInfo);
+      expect(cssFontInfo.fontWeight).toEqual("400");
+    });
+
+    it("Check italic angle", function () {
+      const cssFontInfo = {
+        fontFamily: "blah",
+        fontWeight: 100,
+        italicAngle: 10,
+      };
+      validateCSSFont(cssFontInfo);
+      expect(cssFontInfo.italicAngle).toEqual("10");
+
+      cssFontInfo.italicAngle = -123;
+      validateCSSFont(cssFontInfo);
+      expect(cssFontInfo.italicAngle).toEqual("14");
+
+      cssFontInfo.italicAngle = "91";
+      validateCSSFont(cssFontInfo);
+      expect(cssFontInfo.italicAngle).toEqual("14");
+
+      cssFontInfo.italicAngle = 2.718;
+      validateCSSFont(cssFontInfo);
+      expect(cssFontInfo.italicAngle).toEqual("2.718");
+    });
+  });
 });