Sakurai/pdf.js
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Represent cid chars using integers, not strings.

Browse Source 
cid chars are 16-bit unsigned integers. Currently we convert them to
single-char strings when inserting them into the CMap, and then convert
them back to integers when extracting them from the CMap. This patch
changes CMap so that cid chars stay in integer format throughout, saving
both time and space.

When loading the PDF from issue #4580, this change reduces peak RSS from
~600 to ~370 MiB. It also improves overall speed on that PDF by ~26%,
going from 724 ms to 533 ms.
This commit is contained in:
Nicholas Nethercote 2014-07-31 23:46:37 -07:00
parent ad2ea78280
commit adf58ed687
3 changed files with 46 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

62
src/core/cmap.js
View File

@ -199,6 +199,10 @@ var CMap = (function CMapClosure() {
// where nBytePairs are ranges e.g. [low1, high1, low2, high2, ...] // where nBytePairs are ranges e.g. [low1, high1, low2, high2, ...]
this.codespaceRanges = [[], [], [], []]; this.codespaceRanges = [[], [], [], []];
this.numCodespaceRanges = 0; this.numCodespaceRanges = 0;
// Map entries have one of two forms.
// - cid chars are 16-bit unsigned integers, stored as integers.
// - bf chars are variable-length byte sequences, stored as strings, with
// one byte per character.
this._map = []; this._map = [];
this.vertical = false; this.vertical = false;
this.useCMap = null; this.useCMap = null;
@ -210,18 +214,23 @@ var CMap = (function CMapClosure() {
this.numCodespaceRanges++; this.numCodespaceRanges++;
}, },
mapRange: function(low, high, dstLow) { mapCidRange: function(low, high, dstLow) {
var lastByte = dstLow.length - 1;
while (low <= high) { while (low <= high) {
this._map[low] = dstLow; this._map[low++] = dstLow++;
// Only the last byte has to be incremented.
dstLow = dstLow.substr(0, lastByte) +
String.fromCharCode(dstLow.charCodeAt(lastByte) + 1);
++low;
} }
}, },
mapRangeToArray: function(low, high, array) { mapBfRange: function(low, high, dstLow) {
var lastByte = dstLow.length - 1;
while (low <= high) {
this._map[low++] = dstLow;
// Only the last byte has to be incremented.
dstLow = dstLow.substr(0, lastByte) +
String.fromCharCode(dstLow.charCodeAt(lastByte) + 1);
}
},
mapBfRangeToArray: function(low, high, array) {
var i = 0, ii = array.length; var i = 0, ii = array.length;
while (low <= high && i < ii) { while (low <= high && i < ii) {
this._map[low] = array[i++]; this._map[low] = array[i++];
@ -229,6 +238,7 @@ var CMap = (function CMapClosure() {
} }
}, },
// This is used for both bf and cid chars.
mapOne: function(src, dst) { mapOne: function(src, dst) {
this._map[src] = dst; this._map[src] = dst;
}, },
@ -302,7 +312,7 @@ var IdentityCMap = (function IdentityCMapClosure() {
CMap.call(this); CMap.call(this);
this.vertical = vertical; this.vertical = vertical;
this.addCodespaceRange(n, 0, 0xffff); this.addCodespaceRange(n, 0, 0xffff);
this.mapRange(0, 0xffff, '\u0000'); this.mapCidRange(0, 0xffff, 0);
} }
Util.inherit(IdentityCMap, CMap, {}); Util.inherit(IdentityCMap, CMap, {});
@ -522,7 +532,7 @@ var BinaryCMapReader = (function BinaryCMapReaderClosure() {
case 2: // cidchar case 2: // cidchar
stream.readHex(char, dataSize); stream.readHex(char, dataSize);
code = stream.readNumber(); code = stream.readNumber();
cMap.mapOne(hexToInt(char, dataSize), String.fromCharCode(code)); cMap.mapOne(hexToInt(char, dataSize), code);
for (i = 1; i < subitemsCount; i++) { for (i = 1; i < subitemsCount; i++) {
incHex(char, dataSize); incHex(char, dataSize);
if (!sequence) { if (!sequence) {
@ -530,7 +540,7 @@ var BinaryCMapReader = (function BinaryCMapReaderClosure() {
addHex(char, tmp, dataSize); addHex(char, tmp, dataSize);
} }
code = stream.readSigned() + (code + 1); code = stream.readSigned() + (code + 1);
cMap.mapOne(hexToInt(char, dataSize), String.fromCharCode(code)); cMap.mapOne(hexToInt(char, dataSize), code);
} }
break; break;
case 3: // cidrange case 3: // cidrange
@ -538,8 +548,8 @@ var BinaryCMapReader = (function BinaryCMapReaderClosure() {
stream.readHexNumber(end, dataSize); stream.readHexNumber(end, dataSize);
addHex(end, start, dataSize); addHex(end, start, dataSize);
code = stream.readNumber(); code = stream.readNumber();
cMap.mapRange(hexToInt(start, dataSize), hexToInt(end, dataSize), cMap.mapCidRange(hexToInt(start, dataSize), hexToInt(end, dataSize),
String.fromCharCode(code)); code);
for (i = 1; i < subitemsCount; i++) { for (i = 1; i < subitemsCount; i++) {
incHex(end, dataSize); incHex(end, dataSize);
if (!sequence) { if (!sequence) {
@ -551,8 +561,8 @@ var BinaryCMapReader = (function BinaryCMapReaderClosure() {
stream.readHexNumber(end, dataSize); stream.readHexNumber(end, dataSize);
addHex(end, start, dataSize); addHex(end, start, dataSize);
code = stream.readNumber(); code = stream.readNumber();
cMap.mapRange(hexToInt(start, dataSize), hexToInt(end, dataSize), cMap.mapCidRange(hexToInt(start, dataSize), hexToInt(end, dataSize),
String.fromCharCode(code)); code);
} }
break; break;
case 4: // bfchar case 4: // bfchar
@ -578,9 +588,9 @@ var BinaryCMapReader = (function BinaryCMapReaderClosure() {
stream.readHexNumber(end, ucs2DataSize); stream.readHexNumber(end, ucs2DataSize);
addHex(end, start, ucs2DataSize); addHex(end, start, ucs2DataSize);
stream.readHex(charCode, dataSize); stream.readHex(charCode, dataSize);
cMap.mapRange(hexToInt(start, ucs2DataSize), cMap.mapBfRange(hexToInt(start, ucs2DataSize),
hexToInt(end, ucs2DataSize), hexToInt(end, ucs2DataSize),
hexToStr(charCode, dataSize)); hexToStr(charCode, dataSize));
for (i = 1; i < subitemsCount; i++) { for (i = 1; i < subitemsCount; i++) {
incHex(end, ucs2DataSize); incHex(end, ucs2DataSize);
if (!sequence) { if (!sequence) {
@ -592,9 +602,9 @@ var BinaryCMapReader = (function BinaryCMapReaderClosure() {
stream.readHexNumber(end, ucs2DataSize); stream.readHexNumber(end, ucs2DataSize);
addHex(end, start, ucs2DataSize); addHex(end, start, ucs2DataSize);
stream.readHex(charCode, dataSize); stream.readHex(charCode, dataSize);
cMap.mapRange(hexToInt(start, ucs2DataSize), cMap.mapBfRange(hexToInt(start, ucs2DataSize),
hexToInt(end, ucs2DataSize), hexToInt(end, ucs2DataSize),
hexToStr(charCode, dataSize)); hexToStr(charCode, dataSize));
} }
break; break;
default: default:
@ -675,7 +685,7 @@ var CMapFactory = (function CMapFactoryClosure() {
obj = lexer.getObj(); obj = lexer.getObj();
if (isInt(obj) || isString(obj)) { if (isInt(obj) || isString(obj)) {
var dstLow = isInt(obj) ? String.fromCharCode(obj) : obj; var dstLow = isInt(obj) ? String.fromCharCode(obj) : obj;
cMap.mapRange(low, high, dstLow); cMap.mapBfRange(low, high, dstLow);
} else if (isCmd(obj, '[')) { } else if (isCmd(obj, '[')) {
obj = lexer.getObj(); obj = lexer.getObj();
var array = []; var array = [];
@ -683,7 +693,7 @@ var CMapFactory = (function CMapFactoryClosure() {
array.push(obj); array.push(obj);
obj = lexer.getObj(); obj = lexer.getObj();
} }
cMap.mapRangeToArray(low, high, array); cMap.mapBfRangeToArray(low, high, array);
} else { } else {
break; break;
} }
@ -704,7 +714,7 @@ var CMapFactory = (function CMapFactoryClosure() {
var src = strToInt(obj); var src = strToInt(obj);
obj = lexer.getObj(); obj = lexer.getObj();
expectInt(obj); expectInt(obj);
var dst = String.fromCharCode(obj); var dst = obj;
cMap.mapOne(src, dst); cMap.mapOne(src, dst);
} }
} }
@ -725,8 +735,8 @@ var CMapFactory = (function CMapFactoryClosure() {
var high = strToInt(obj); var high = strToInt(obj);
obj = lexer.getObj(); obj = lexer.getObj();
expectInt(obj); expectInt(obj);
var dstLow = String.fromCharCode(obj); var dstLow = obj;
cMap.mapRange(low, high, dstLow); cMap.mapCidRange(low, high, dstLow);
} }
} }

15
src/core/fonts.js
View File

@ -3899,8 +3899,7 @@ var Font = (function FontClosure() {
var cidToGidMap = properties.cidToGidMap || []; var cidToGidMap = properties.cidToGidMap || [];
var cidToGidMapLength = cidToGidMap.length; var cidToGidMapLength = cidToGidMap.length;
properties.cMap.forEach(function(charCode, cid) { properties.cMap.forEach(function(charCode, cid) {
assert(cid.length === 1, 'Max size of CID is 65,535'); assert(cid <= 0xffff, 'Max size of CID is 65,535');
cid = cid.charCodeAt(0);
var glyphId = -1; var glyphId = -1;
if (cidToGidMapLength === 0) { if (cidToGidMapLength === 0) {
glyphId = charCode; glyphId = charCode;
@ -4370,10 +4369,10 @@ var Font = (function FontClosure() {
var cMap = properties.cMap; var cMap = properties.cMap;
toUnicode = []; toUnicode = [];
cMap.forEach(function(charcode, cid) { cMap.forEach(function(charcode, cid) {
assert(cid.length === 1, 'Max size of CID is 65,535'); assert(cid <= 0xffff, 'Max size of CID is 65,535');
// e) Map the CID obtained in step (a) according to the CMap obtained // e) Map the CID obtained in step (a) according to the CMap obtained
// in step (d), producing a Unicode value. // in step (d), producing a Unicode value.
var ucs2 = ucs2CMap.lookup(cid.charCodeAt(0)); var ucs2 = ucs2CMap.lookup(cid);
if (ucs2) { if (ucs2) {
toUnicode[charcode] = toUnicode[charcode] =
String.fromCharCode((ucs2.charCodeAt(0) << 8) + String.fromCharCode((ucs2.charCodeAt(0) << 8) +
@ -4415,7 +4414,7 @@ var Font = (function FontClosure() {
var charcode = 0; var charcode = 0;
if (this.composite) { if (this.composite) {
if (this.cMap.contains(glyphUnicode)) { if (this.cMap.contains(glyphUnicode)) {
charcode = this.cMap.lookup(glyphUnicode).charCodeAt(0); charcode = this.cMap.lookup(glyphUnicode);
} }
} }
// ... via toUnicode map // ... via toUnicode map
@ -4444,7 +4443,7 @@ var Font = (function FontClosure() {
var widthCode = charcode; var widthCode = charcode;
if (this.cMap && this.cMap.contains(charcode)) { if (this.cMap && this.cMap.contains(charcode)) {
widthCode = this.cMap.lookup(charcode).charCodeAt(0); widthCode = this.cMap.lookup(charcode);
} }
width = this.widths[widthCode]; width = this.widths[widthCode];
width = isNum(width) ? width : this.defaultWidth; width = isNum(width) ? width : this.defaultWidth;
@ -5626,8 +5625,8 @@ var CFFFont = (function CFFFontClosure() {
// If the font is actually a CID font then we should use the charset // If the font is actually a CID font then we should use the charset
// to map CIDs to GIDs. // to map CIDs to GIDs.
for (glyphId = 0; glyphId < charsets.length; glyphId++) { for (glyphId = 0; glyphId < charsets.length; glyphId++) {
var cidString = String.fromCharCode(charsets[glyphId]); var cid = charsets[glyphId];
var charCode = properties.cMap.charCodeOf(cidString); var charCode = properties.cMap.charCodeOf(cid);
charCodeToGlyphId[charCode] = glyphId; charCodeToGlyphId[charCode] = glyphId;
} }
} else { } else {

6
test/unit/cmap_spec.js
View File

@ -44,7 +44,7 @@ describe('cmap', function() {
'endcidchar\n'; 'endcidchar\n';
var stream = new StringStream(str); var stream = new StringStream(str);
var cmap = CMapFactory.create(stream); var cmap = CMapFactory.create(stream);
expect(cmap.lookup(0x14)).toEqual(String.fromCharCode(0x00)); expect(cmap.lookup(0x14)).toEqual(0x00);
expect(cmap.lookup(0x15)).toBeUndefined(); expect(cmap.lookup(0x15)).toBeUndefined();
}); });
it('parses begincidrange', function() { it('parses begincidrange', function() {
@ -54,8 +54,8 @@ describe('cmap', function() {
var stream = new StringStream(str); var stream = new StringStream(str);
var cmap = CMapFactory.create(stream); var cmap = CMapFactory.create(stream);
expect(cmap.lookup(0x15)).toBeUndefined(); expect(cmap.lookup(0x15)).toBeUndefined();
expect(cmap.lookup(0x16)).toEqual(String.fromCharCode(0x00)); expect(cmap.lookup(0x16)).toEqual(0x00);
expect(cmap.lookup(0x1B)).toEqual(String.fromCharCode(0x05)); expect(cmap.lookup(0x1B)).toEqual(0x05);
expect(cmap.lookup(0x1C)).toBeUndefined(); expect(cmap.lookup(0x1C)).toBeUndefined();
}); });
it('decodes codespace ranges', function() { it('decodes codespace ranges', function() {