Merge pull request #5095 from Snuffleupagus/issue-5070

Adjust the heuristics to recognize more cases of unknown glyphs for |toUnicode| (issue 5070)

src/core/fonts.js

@@ -4299,6 +4299,7 @@ var Font = (function FontClosure() {
       if (!properties.composite /* is simple font */) {
         toUnicode = [];
         var encoding = properties.defaultEncoding.slice();
+        var baseEncodingName = properties.baseEncodingName;
         // Merge in the differences array.
         var differences = properties.differences;
         for (charcode in differences) {
@@ -4309,26 +4310,43 @@ var Font = (function FontClosure() {
           var glyphName = encoding[charcode];
           // b) Look up the character name in the Adobe Glyph List (see the
           //    Bibliography) to obtain the corresponding Unicode value.
-          if (glyphName === '' || !(glyphName in GlyphsUnicode)) {
+          if (glyphName === '') {
+            continue;
+          } else if (GlyphsUnicode[glyphName] === undefined) {
             // (undocumented) c) Few heuristics to recognize unknown glyphs
             // NOTE: Adobe Reader does not do this step, but OSX Preview does
-            var code;
-            // Gxx glyph
-            if (glyphName.length === 3 &&
-                glyphName[0] === 'G' &&
-                (code = parseInt(glyphName.substr(1), 16))) {
-              toUnicode[charcode] = String.fromCharCode(code);
+            var code = 0;
+            switch (glyphName[0]) {
+              case 'G': // Gxx glyph
+                if (glyphName.length === 3) {
+                  code = parseInt(glyphName.substr(1), 16);
+                }
+                break;
+              case 'g': // g00xx glyph
+                if (glyphName.length === 5) {
+                  code = parseInt(glyphName.substr(1), 16);
+                }
+                break;
+              case 'C': // Cddd glyph
+              case 'c': // cddd glyph
+                if (glyphName.length >= 3) {
+                  code = +glyphName.substr(1);
+                }
+                break;
             }
-            // g00xx glyph
-            if (glyphName.length === 5 &&
-                glyphName[0] === 'g' &&
-                (code = parseInt(glyphName.substr(1), 16))) {
-              toUnicode[charcode] = String.fromCharCode(code);
-            }
-            // Cddd glyph
-            if (glyphName.length >= 3 &&
-                glyphName[0] === 'C' &&
-                (code = +glyphName.substr(1))) {
+            if (code) {
+              // If |baseEncodingName| is one the predefined encodings,
+              // and |code| equals |charcode|, using the glyph defined in the
+              // baseEncoding seems to yield a better |toUnicode| mapping
+              // (fixes issue 5070).
+              if (baseEncodingName && code === +charcode) {
+                var baseEncoding = Encodings[baseEncodingName];
+                if (baseEncoding && (glyphName = baseEncoding[charcode])) {
+                  toUnicode[charcode] =
+                    String.fromCharCode(GlyphsUnicode[glyphName]);
+                  continue;
+                }
+              }
               toUnicode[charcode] = String.fromCharCode(code);
             }
             continue;

test/pdfs/.gitignore

@@ -58,6 +58,7 @@
 !issue925.pdf
 !issue4668.pdf
 !issue5039.pdf
+!issue5070.pdf
 !gradientfill.pdf
 !bug903856.pdf
 !bug850854.pdf

test/test_manifest.json

@@ -1081,6 +1081,13 @@
        "rounds": 1,
        "type": "eq"
     },
+    {  "id": "issue5070",
+       "file": "pdfs/issue5070.pdf",
+       "md5": "ec2ca0b4954c8390a5b3b0ffd79a8e92",
+       "link": false,
+       "rounds": 1,
+       "type": "eq"
+    },
     {  "id": "issue1257",
       "file": "pdfs/issue1257.pdf",
       "md5": "9111533826bc21ed774e8e01603a2f54",