Merge pull request #7550 from Snuffleupagus/Type1-toUnicode-builtInEncoding-fallback

For embedded Type1 fonts without included `ToUnicode`/`Encoding` data, attempt to improve text selection by using the `builtInEncoding` to amend the `toUnicode` map (issue 6901, issue 7182, issue 7217, bug 917796, bug 1242142)

src/core/evaluator.js

@@ -1757,6 +1757,7 @@ var PartialEvaluator = (function PartialEvaluatorClosure() {
 
       properties.differences = differences;
       properties.baseEncodingName = baseEncodingName;
+      properties.hasEncoding = !!baseEncodingName || differences.length > 0;
       properties.dict = dict;
       return toUnicodePromise.then(function(toUnicode) {
         properties.toUnicode = toUnicode;
@@ -1774,8 +1775,10 @@ var PartialEvaluator = (function PartialEvaluatorClosure() {
      *   {ToUnicodeMap|IdentityToUnicodeMap} object.
      */
     buildToUnicode: function PartialEvaluator_buildToUnicode(properties) {
+      properties.hasIncludedToUnicodeMap =
+        !!properties.toUnicode && properties.toUnicode.length > 0;
       // Section 9.10.2 Mapping Character Codes to Unicode Values
-      if (properties.toUnicode && properties.toUnicode.length !== 0) {
+      if (properties.hasIncludedToUnicodeMap) {
         return Promise.resolve(properties.toUnicode);
       }
       // According to the spec if the font is a simple font we should only map

src/core/fonts.js

@@ -163,6 +163,30 @@ function adjustWidths(properties) {
   properties.defaultWidth *= scale;
 }
 
+function adjustToUnicode(properties, builtInEncoding) {
+  if (properties.hasIncludedToUnicodeMap) {
+    return; // The font dictionary has a `ToUnicode` entry.
+  }
+  if (properties.hasEncoding) {
+    return; // The font dictionary has an `Encoding` entry.
+  }
+  if (builtInEncoding === properties.defaultEncoding) {
+    return; // No point in trying to adjust `toUnicode` if the encodings match.
+  }
+  if (properties.toUnicode instanceof IdentityToUnicodeMap) {
+    return;
+  }
+  var toUnicode = [], glyphsUnicodeMap = getGlyphsUnicode();
+  for (var charCode in builtInEncoding) {
+    var glyphName = builtInEncoding[charCode];
+    var unicode = getUnicodeForGlyph(glyphName, glyphsUnicodeMap);
+    if (unicode !== -1) {
+      toUnicode[charCode] = String.fromCharCode(unicode);
+    }
+  }
+  properties.toUnicode.amend(toUnicode);
+}
+
 function getFontType(type, subtype) {
   switch (type) {
     case 'Type1':
@@ -261,7 +285,13 @@ var ToUnicodeMap = (function ToUnicodeMapClosure() {
 
     charCodeOf: function(v) {
       return this._map.indexOf(v);
-    }
+    },
+
+    amend: function (map) {
+      for (var charCode in map) {
+        this._map[charCode] = map[charCode];
+      }
+    },
   };
 
   return ToUnicodeMap;
@@ -297,7 +327,11 @@ var IdentityToUnicodeMap = (function IdentityToUnicodeMapClosure() {
 
     charCodeOf: function (v) {
       return (isInt(v) && v >= this.firstChar && v <= this.lastChar) ? v : -1;
-    }
+    },
+
+    amend: function (map) {
+      error('Should not call amend()');
+    },
   };
 
   return IdentityToUnicodeMap;
@@ -765,6 +799,7 @@ var Font = (function FontClosure() {
     this.fontMatrix = properties.fontMatrix;
     this.widths = properties.widths;
     this.defaultWidth = properties.defaultWidth;
+    this.toUnicode = properties.toUnicode;
     this.encoding = properties.baseEncoding;
     this.seacMap = properties.seacMap;
 
@@ -2386,10 +2421,8 @@ var Font = (function FontClosure() {
       } else {
         // Most of the following logic in this code branch is based on the
         // 9.6.6.4 of the PDF spec.
-        var hasEncoding =
-          properties.differences.length > 0 || !!properties.baseEncodingName;
-        var cmapTable =
-          readCmapTable(tables['cmap'], font, this.isSymbolicFont, hasEncoding);
+        var cmapTable = readCmapTable(tables['cmap'], font, this.isSymbolicFont,
+                                      properties.hasEncoding);
         var cmapPlatformId = cmapTable.platformId;
         var cmapEncodingId = cmapTable.encodingId;
         var cmapMappings = cmapTable.mappings;
@@ -2398,7 +2431,7 @@ var Font = (function FontClosure() {
         // The spec seems to imply that if the font is symbolic the encoding
         // should be ignored, this doesn't appear to work for 'preistabelle.pdf'
         // where the the font is symbolic and it has an encoding.
-        if (hasEncoding &&
+        if (properties.hasEncoding &&
             (cmapPlatformId === 3 && cmapEncodingId === 1 ||
              cmapPlatformId === 1 && cmapEncodingId === 0) ||
             (cmapPlatformId === -1 && cmapEncodingId === -1 && // Temporary hack
@@ -2562,6 +2595,12 @@ var Font = (function FontClosure() {
       // TODO: Check the charstring widths to determine this.
       properties.fixedPitch = false;
 
+      if (properties.builtInEncoding) {
+        // For Type1 fonts that do not include either `ToUnicode` or `Encoding`
+        // data, attempt to use the `builtInEncoding` to improve text selection.
+        adjustToUnicode(properties, properties.builtInEncoding);
+      }
+
       var mapping = font.getGlyphMapping(properties);
       var newMapping = adjustMapping(mapping, properties);
       this.toFontChar = newMapping.toFontChar;

test/pdfs/.gitignore

@@ -22,6 +22,7 @@
 !issue5808.pdf
 !issue6204.pdf
 !issue6782.pdf
+!issue6901.pdf
 !issue6961.pdf
 !issue6962.pdf
 !issue7020.pdf

test/test_manifest.json

@@ -1220,6 +1220,20 @@
        "link": false,
        "type": "text"
     },
+    {  "id": "issue6901-eq",
+       "file": "pdfs/issue6901.pdf",
+       "md5": "1a0604b1a7a3aaf2162b425a9a84230b",
+       "rounds": 1,
+       "link": false,
+       "type": "eq"
+    },
+    {  "id": "issue6901-text",
+       "file": "pdfs/issue6901.pdf",
+       "md5": "1a0604b1a7a3aaf2162b425a9a84230b",
+       "rounds": 1,
+       "link": false,
+       "type": "text"
+    },
     {  "id": "issue6962",
        "file": "pdfs/issue6962.pdf",
        "md5": "d40e871ecca68baf93114bd28c782148",