From df352b375bbb43e6deca1b167125f82e047a7bab Mon Sep 17 00:00:00 2001
From: Yury Delendik <ydelendik@mozilla.com>
Date: Mon, 24 Jun 2013 15:33:50 -0500
Subject: [PATCH 1/4] Refactoring TTF repair logic
---
src/fonts.js | 210 +++++++++++++++++++--------------------------------
1 file changed, 77 insertions(+), 133 deletions(-)
diff --git a/src/fonts.js b/src/fonts.js
index 289a2dd9c..5b73b9ea7 100644
--- a/src/fonts.js
+++ b/src/fonts.js
@@ -3528,30 +3528,6 @@ var Font = (function FontClosure() {
glyf.data = newGlyfData.subarray(0, writeOffset);
}
- function findEmptyGlyphs(locaTable, isGlyphLocationsLong, emptyGlyphIds) {
- var itemSize, itemDecode;
- if (isGlyphLocationsLong) {
- itemSize = 4;
- itemDecode = function fontItemDecodeLong(data, offset) {
- return (data[offset] << 24) | (data[offset + 1] << 16) |
- (data[offset + 2] << 8) | data[offset + 3];
- };
- } else {
- itemSize = 2;
- itemDecode = function fontItemDecode(data, offset) {
- return (data[offset] << 9) | (data[offset + 1] << 1);
- };
- }
- var data = locaTable.data, length = data.length;
- var lastOffset = itemDecode(data, 0);
- for (var i = itemSize, j = 0; i < length; i += itemSize, j++) {
- var offset = itemDecode(data, i);
- if (offset == lastOffset)
- emptyGlyphIds[j] = true;
- lastOffset = offset;
- }
- }
-
function readPostScriptTable(post, properties, maxpNumGlyphs) {
var start = (font.start ? font.start : 0) + post.offset;
font.pos = start;
@@ -3886,75 +3862,59 @@ var Font = (function FontClosure() {
// The following steps modify the original font data, making copy
font = new Stream(new Uint8Array(font.getBytes()));
- // Check that required tables are present
- var requiredTables = ['OS/2', 'cmap', 'head', 'hhea',
- 'hmtx', 'maxp', 'name', 'post'];
+ var VALID_TABLES = ['OS/2', 'cmap', 'head', 'hhea', 'hmtx', 'maxp',
+ 'name', 'post', 'loca', 'glyf', 'fpgm', 'prep', 'cvt ', 'CFF '];
var header = readOpenTypeHeader(font);
var numTables = header.numTables;
- var cmap, post, maxp, hhea, hmtx, head, os2;
- var glyf, fpgm, loca, prep, cvt;
- var tables = [];
+ var tables = { 'OS/2': null, cmap: null, head: null, hhea: null,
+ hmtx: null, maxp: null, name: null, post: null};
for (var i = 0; i < numTables; i++) {
var table = readTableEntry(font);
- var index = requiredTables.indexOf(table.tag);
- if (index != -1) {
- if (table.tag == 'cmap')
- cmap = table;
- else if (table.tag == 'post')
- post = table;
- else if (table.tag == 'maxp')
- maxp = table;
- else if (table.tag == 'hhea')
- hhea = table;
- else if (table.tag == 'hmtx')
- hmtx = table;
- else if (table.tag == 'head')
- head = table;
- else if (table.tag == 'OS/2')
- os2 = table;
-
- requiredTables.splice(index, 1);
- } else {
- if (table.tag == 'loca')
- loca = table;
- else if (table.tag == 'glyf')
- glyf = table;
- else if (table.tag == 'fpgm')
- fpgm = table;
- else if (table.tag == 'prep')
- prep = table;
- else if (table.tag == 'cvt ')
- cvt = table;
- else if (table.tag == 'CFF ')
- return null; // XXX: OpenType font is found, stopping
- else // skipping table if it's not a required or optional table
- continue;
+ if (VALID_TABLES.indexOf(table.tag) < 0) {
+ continue; // skipping table if it's not a required or optional table
}
- tables.push(table);
+ tables[table.tag] = table;
}
- // Ensure the hmtx table contains the advance width and
- // sidebearings information for numGlyphs in the maxp table
- font.pos = (font.start || 0) + maxp.offset;
+ var isTrueType = !tables['CFF '];
+ if (!isTrueType) {
+ // OpenType font
+ delete tables.glyf;
+ delete tables.loca;
+ delete tables.fpgm;
+ delete tables.prep;
+ delete tables['cvt '];
+
+ return; // TODO: implement OpenType support
+ } else {
+ if (!tables.glyf || !tables.loca) {
+ error('Required "glyf" or "loca" tables are not found');
+ }
+ }
+
+ font.pos = (font.start || 0) + tables.maxp.offset;
var version = int32(font.getBytes(4));
var numGlyphs = int16(font.getBytes(2));
var maxFunctionDefs = 0;
- if (version >= 0x00010000 && maxp.length >= 22) {
+ if (version >= 0x00010000 && tables.maxp.length >= 22) {
font.pos += 14;
var maxFunctionDefs = int16(font.getBytes(2));
}
- var hintsValid = sanitizeTTPrograms(fpgm, prep, maxFunctionDefs);
+ var hintsValid = sanitizeTTPrograms(tables.fpgm, tables.prep,
+ maxFunctionDefs);
if (!hintsValid) {
- tables.splice(tables.indexOf(fpgm), 1);
- fpgm = null;
- tables.splice(tables.indexOf(prep), 1);
- prep = null;
+ delete tables.fpgm;
+ delete tables.prep;
}
- var numTables = tables.length + requiredTables.length;
+ // Tables needs to be written by ascendant alphabetic order
+ var tablesNames = Object.keys(tables);
+ tablesNames.sort();
+
+ numTables = tablesNames.length;
// header and new offsets. Table entry information is appended to the
// end of file. The virtualOffset represents where to put the actual
@@ -3968,35 +3928,36 @@ var Font = (function FontClosure() {
// of missing tables
createOpenTypeHeader(header.version, ttf, numTables);
- sanitizeMetrics(font, hhea, hmtx, numGlyphs);
+ // Ensure the hmtx table contains the advance width and
+ // sidebearings information for numGlyphs in the maxp table
+ sanitizeMetrics(font, tables.hhea, tables.hmtx, numGlyphs);
- if (head) {
- sanitizeHead(head, numGlyphs, loca.length);
+ if (!tables.head) {
+ error('Required "head" table is not found');
}
- var isGlyphLocationsLong = int16([head.data[50], head.data[51]]);
- if (head && loca && glyf) {
- sanitizeGlyphLocations(loca, glyf, numGlyphs, isGlyphLocationsLong,
- hintsValid);
- }
+ sanitizeHead(tables.head, numGlyphs, isTrueType ? tables.loca.length : 0);
+
+ if (isTrueType) {
+ var isGlyphLocationsLong = int16([tables.head.data[50],
+ tables.head.data[51]]);
- var emptyGlyphIds = [];
- if (glyf)
- findEmptyGlyphs(loca, isGlyphLocationsLong, emptyGlyphIds);
+ sanitizeGlyphLocations(tables.loca, tables.glyf, numGlyphs,
+ isGlyphLocationsLong, hintsValid);
+ }
// Sanitizer reduces the glyph advanceWidth to the maxAdvanceWidth
// Sometimes it's 0. That needs to be fixed
- if (hhea.data[10] === 0 && hhea.data[11] === 0) {
- hhea.data[10] = 0xFF;
- hhea.data[11] = 0xFF;
+ if (tables.hhea.data[10] === 0 && tables.hhea.data[11] === 0) {
+ tables.hhea.data[10] = 0xFF;
+ tables.hhea.data[11] = 0xFF;
}
// The 'post' table has glyphs names.
- if (post) {
- var valid = readPostScriptTable(post, properties, numGlyphs);
+ if (tables.post) {
+ var valid = readPostScriptTable(tables.post, properties, numGlyphs);
if (!valid) {
- tables.splice(tables.indexOf(post), 1);
- post = null;
+ tables.post = null;
}
}
@@ -4010,12 +3971,11 @@ var Font = (function FontClosure() {
// U+00AD (soft hyphen) is not drawn.
// So, offset all the glyphs by 0xFF to avoid these cases and use
// the encoding to map incoming characters to the new glyph positions
- if (!cmap) {
- cmap = {
+ if (!tables.cmap) {
+ tables.cmap = {
tag: 'cmap',
data: null
};
- tables.push(cmap);
}
var cidToGidMap = properties.cidToGidMap || [];
@@ -4083,7 +4043,7 @@ var Font = (function FontClosure() {
// but this.hasEncoding is currently true for any encodings on the
// Encodings object (e.g. MacExpertEncoding). So should consider using
// better check for this.
- var cmapTable = readCmapTable(cmap, font, this.hasEncoding,
+ var cmapTable = readCmapTable(tables.cmap, font, this.hasEncoding,
this.isSymbolicFont);
// TODO(mack): If the (3, 0) cmap table used, then the font is
@@ -4210,69 +4170,53 @@ var Font = (function FontClosure() {
}
// Converting glyphs and ids into font's cmap table
- cmap.data = createCmapTable(glyphs, ids);
+ tables.cmap.data = createCmapTable(glyphs, ids);
var unicodeIsEnabled = [];
for (var i = 0, ii = glyphs.length; i < ii; i++) {
unicodeIsEnabled[glyphs[i].unicode] = true;
}
this.unicodeIsEnabled = unicodeIsEnabled;
- if (os2 && !validateOS2Table(os2)) {
- tables.splice(tables.indexOf(os2), 1);
- os2 = null;
- }
-
- if (!os2) {
+ if (!tables['OS/2'] || !validateOS2Table(tables['OS/2'])) {
// extract some more font properties from the OpenType head and
// hhea tables; yMin and descent value are always negative
var override = {
- unitsPerEm: int16([head.data[18], head.data[19]]),
- yMax: int16([head.data[42], head.data[43]]),
- yMin: int16([head.data[38], head.data[39]]) - 0x10000,
- ascent: int16([hhea.data[4], hhea.data[5]]),
- descent: int16([hhea.data[6], hhea.data[7]]) - 0x10000
+ unitsPerEm: int16([tables.head.data[18], tables.head.data[19]]),
+ yMax: int16([tables.head.data[42], tables.head.data[43]]),
+ yMin: int16([tables.head.data[38], tables.head.data[39]]) - 0x10000,
+ ascent: int16([tables.hhea.data[4], tables.hhea.data[5]]),
+ descent: int16([tables.hhea.data[6], tables.hhea.data[7]]) - 0x10000
};
- tables.push({
+ tables['OS/2'] = {
tag: 'OS/2',
data: stringToArray(createOS2Table(properties, glyphs, override))
- });
+ };
}
// Rewrite the 'post' table if needed
- if (!post) {
- tables.push({
+ if (!tables.post) {
+ tables.post = {
tag: 'post',
data: stringToArray(createPostTable(properties))
- });
+ };
}
// Re-creating 'name' table
- if (requiredTables.indexOf('name') != -1) {
- tables.push({
+ if (!tables.name) {
+ tables.name = {
tag: 'name',
data: stringToArray(createNameTable(this.name))
- });
+ };
} else {
// ... using existing 'name' table as prototype
- for (var i = 0, ii = tables.length; i < ii; i++) {
- var table = tables[i];
- if (table.tag === 'name') {
- var namePrototype = readNameTable(table);
- table.data = stringToArray(createNameTable(name, namePrototype));
- break;
- }
- }
+ var namePrototype = readNameTable(tables.name);
+ tables.name.data = stringToArray(createNameTable(name, namePrototype));
}
- // Tables needs to be written by ascendant alphabetic order
- tables.sort(function tables_sort(a, b) {
- return (a.tag > b.tag) - (a.tag < b.tag);
- });
-
// rewrite the tables but tweak offsets
- for (var i = 0, ii = tables.length; i < ii; i++) {
- var table = tables[i];
+ for (var i = 0; i < numTables; i++) {
+ var table = tables[tablesNames[i]];
var data = [];
var tableData = table.data;
@@ -4282,8 +4226,8 @@ var Font = (function FontClosure() {
}
// Add the table datas
- for (var i = 0, ii = tables.length; i < ii; i++) {
- var table = tables[i];
+ for (var i = 0; i < numTables; i++) {
+ var table = tables[tablesNames[i]];
var tableData = table.data;
ttf.file += arrayToString(tableData);
From 6dc4b10b4094688566d91ea5799eb32a4f2fdb74 Mon Sep 17 00:00:00 2001
From: Yury Delendik <ydelendik@mozilla.com>
Date: Mon, 24 Jun 2013 18:45:31 -0500
Subject: [PATCH 2/4] Implements reparing of CFF table
---
src/fonts.js | 38 +++++++++++++++++++++++++++++++-------
1 file changed, 31 insertions(+), 7 deletions(-)
diff --git a/src/fonts.js b/src/fonts.js
index 5b73b9ea7..7c341aec7 100644
--- a/src/fonts.js
+++ b/src/fonts.js
@@ -2483,6 +2483,7 @@ var Font = (function FontClosure() {
data = this.convert(name, cff, properties);
break;
+ case 'OpenType':
case 'TrueType':
case 'CIDFontType2':
this.mimetype = 'font/opentype';
@@ -2490,10 +2491,6 @@ var Font = (function FontClosure() {
// Repair the TrueType file. It is can be damaged in the point of
// view of the sanitizer
data = this.checkAndRepair(name, file, properties);
- if (!data) {
- // TrueType data is not found, e.g. when the font is an OpenType font
- warn('Font is not a TrueType font');
- }
break;
default:
@@ -3881,26 +3878,36 @@ var Font = (function FontClosure() {
var isTrueType = !tables['CFF '];
if (!isTrueType) {
// OpenType font
+ if (!tables.head || !tables.hhea || !tables.maxp || !tables.post) {
+ // no major tables: throwing everything at CFFFont
+ var cffFile = new Stream(tables['CFF '].data);
+ var cff = new CFFFont(cffFile, properties);
+
+ return this.convert(name, cff, properties);
+ }
+
delete tables.glyf;
delete tables.loca;
delete tables.fpgm;
delete tables.prep;
delete tables['cvt '];
-
- return; // TODO: implement OpenType support
} else {
if (!tables.glyf || !tables.loca) {
error('Required "glyf" or "loca" tables are not found');
}
}
+ if (!tables.maxp) {
+ error('Required "maxp" table is not found');
+ }
+
font.pos = (font.start || 0) + tables.maxp.offset;
var version = int32(font.getBytes(4));
var numGlyphs = int16(font.getBytes(2));
var maxFunctionDefs = 0;
if (version >= 0x00010000 && tables.maxp.length >= 22) {
font.pos += 14;
- var maxFunctionDefs = int16(font.getBytes(2));
+ maxFunctionDefs = int16(font.getBytes(2));
}
var hintsValid = sanitizeTTPrograms(tables.fpgm, tables.prep,
@@ -3946,6 +3953,10 @@ var Font = (function FontClosure() {
isGlyphLocationsLong, hintsValid);
}
+ if (!tables.hhea) {
+ error('Required "hhea" table is not found');
+ }
+
// Sanitizer reduces the glyph advanceWidth to the maxAdvanceWidth
// Sometimes it's 0. That needs to be fixed
if (tables.hhea.data[10] === 0 && tables.hhea.data[11] === 0) {
@@ -4202,6 +4213,19 @@ var Font = (function FontClosure() {
};
}
+ if (!isTrueType) {
+ try {
+ // Trying to repair CFF file
+ var cffFile = new Stream(tables['CFF '].data);
+ var parser = new CFFParser(cffFile, properties);
+ var cff = parser.parse();
+ var compiler = new CFFCompiler(cff);
+ tables['CFF '].data = compiler.compile();
+ } catch (e) {
+ warn('Failed to compile font ' + properties.loadedName);
+ }
+ }
+
// Re-creating 'name' table
if (!tables.name) {
tables.name = {
From 121c183db1d8150319c133b65c54f730f17ad767 Mon Sep 17 00:00:00 2001
From: Yury Delendik <ydelendik@mozilla.com>
Date: Thu, 27 Jun 2013 14:38:52 -0500
Subject: [PATCH 3/4] Fail if other font types are found
---
src/fonts.js | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/fonts.js b/src/fonts.js
index 7c341aec7..e103a0893 100644
--- a/src/fonts.js
+++ b/src/fonts.js
@@ -2494,7 +2494,7 @@ var Font = (function FontClosure() {
break;
default:
- warn('Font ' + type + ' is not supported');
+ error('Font ' + type + ' is not supported');
break;
}
From 48146745b9b6085a25d0b17b1484068fc513311b Mon Sep 17 00:00:00 2001
From: Yury Delendik <ydelendik@mozilla.com>
Date: Mon, 8 Jul 2013 21:03:59 -0500
Subject: [PATCH 4/4] Tests for OpenType fonts
---
test/pdfs/issue2442.pdf.link | 1 +
test/pdfs/issue3062.pdf.link | 1 +
test/test_manifest.json | 16 ++++++++++++++++
3 files changed, 18 insertions(+)
create mode 100644 test/pdfs/issue2442.pdf.link
create mode 100644 test/pdfs/issue3062.pdf.link
diff --git a/test/pdfs/issue2442.pdf.link b/test/pdfs/issue2442.pdf.link
new file mode 100644
index 000000000..195d71f3d
--- /dev/null
+++ b/test/pdfs/issue2442.pdf.link
@@ -0,0 +1 @@
+https://401k.fidelity.com/static/dcl/shared/documents/WPS_PublicSite/Documents/403b_DistributionForm_441802.pdf
diff --git a/test/pdfs/issue3062.pdf.link b/test/pdfs/issue3062.pdf.link
new file mode 100644
index 000000000..499aa7bd1
--- /dev/null
+++ b/test/pdfs/issue3062.pdf.link
@@ -0,0 +1 @@
+http://www3.weforum.org/docs/TTCR/2013/TTCR_DataTables12_2013.pdf
diff --git a/test/test_manifest.json b/test/test_manifest.json
index c3072484c..63461af77 100644
--- a/test/test_manifest.json
+++ b/test/test_manifest.json
@@ -959,6 +959,22 @@
"link": true,
"type": "eq"
},
+ { "id": "issue3062",
+ "file": "pdfs/issue3062.pdf",
+ "md5": "206715f1258f0e117df4180d98dd4d68",
+ "rounds": 1,
+ "lastPage": 1,
+ "link": true,
+ "type": "eq"
+ },
+ { "id": "issue2442",
+ "file": "pdfs/issue2442.pdf",
+ "md5": "4656e36c44f44c71a499f02ce6c781d3",
+ "rounds": 1,
+ "lastPage": 1,
+ "link": true,
+ "type": "eq"
+ },
{ "id": "issue2337",
"file": "pdfs/issue2337.pdf",
"md5": "ea10f4131202b9b8f2a6cb7770d3f185",