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

Merge branch 'master' into faxstream

Browse Source
This commit is contained in:
sbarman 2011-06-23 09:48:57 -07:00
commit bf6f32baa8
4 changed files with 238 additions and 87 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

108
fonts.js
View File

@ -103,7 +103,7 @@ var Font = (function () {
// If the font is to be ignored, register it like an already loaded font
// to avoid the cost of waiting for it be be loaded by the platform.
if (properties.ignore || properties.type == "TrueType" || kDisableFonts) {
if (properties.ignore || kDisableFonts) {
Fonts[name] = {
data: file,
loading: false,
@ -242,7 +242,7 @@ var Font = (function () {
return ranges;
};
function createCMAPTable(glyphs) {
function createCMapTable(glyphs) {
var ranges = getRanges(glyphs);
var headerSize = (12 * 2 + (ranges.length * 4 * 2));
@ -274,7 +274,7 @@ var Font = (function () {
var bias = 0;
for (var i = 0; i < segCount - 1; i++) {
var range = ranges[i];
var start = range[0];
var start = range[0];
var end = range[1];
var delta = (((start - 1) - bias) ^ 0xffff) + 1;
bias += (end - start + 1);
@ -284,8 +284,8 @@ var Font = (function () {
idDeltas += string16(delta);
idRangeOffsets += string16(0);
for (var j = start; j <= end; j++)
glyphsIds += String.fromCharCode(j);
for (var j = 0; j < range.length; j++)
glyphsIds += String.fromCharCode(range[j]);
}
startCount += "\xFF\xFF";
@ -368,11 +368,11 @@ var Font = (function () {
var length = FontsUtils.bytesToInteger(file.getBytes(4));
// Read the table associated data
var currentPosition = file.pos;
file.pos = file.start + offset;
var previousPosition = file.pos;
file.pos = file.start ? file.start : 0;
file.skip(offset);
var data = file.getBytes(length);
file.pos = currentPosition;
file.pos = previousPosition;
return {
tag: tag,
@ -393,6 +393,76 @@ var Font = (function () {
}
};
function replaceCMapTable(cmap, font, properties) {
var version = FontsUtils.bytesToInteger(font.getBytes(2));
var numTables = FontsUtils.bytesToInteger(font.getBytes(2));
for (var i = 0; i < numTables; i++) {
var platformID = FontsUtils.bytesToInteger(font.getBytes(2));
var encodingID = FontsUtils.bytesToInteger(font.getBytes(2));
var offset = FontsUtils.bytesToInteger(font.getBytes(4));
var format = FontsUtils.bytesToInteger(font.getBytes(2));
var length = FontsUtils.bytesToInteger(font.getBytes(2));
var language = FontsUtils.bytesToInteger(font.getBytes(2));
if ((format == 0 && numTables == 1) ||
(format == 6 && numTables == 1 && !properties.encoding.empty)) {
// Format 0 alone is not allowed by the sanitizer so let's rewrite
// that to a 3-1-4 Unicode BMP table
var charset = properties.charset;
var glyphs = [];
for (var j = 0; j < charset.length; j++) {
glyphs.push({
unicode: GlyphsUnicode[charset[j]] || 0
});
}
cmap.data = createCMapTable(glyphs);
} else if (format == 6 && numTables == 1) {
// Format 6 is a 2-bytes dense mapping, which means the font data
// lives glue together even if they are pretty far in the unicode
// table. (This looks weird, so I can have missed something), this
// works on Linux but seems to fails on Mac so let's rewrite the
// cmap table to a 3-1-4 style
var firstCode = FontsUtils.bytesToInteger(font.getBytes(2));
var entryCount = FontsUtils.bytesToInteger(font.getBytes(2));
var glyphs = [];
var min = 0xffff, max = 0;
for (var j = 0; j < entryCount; j++) {
var charcode = FontsUtils.bytesToInteger(font.getBytes(2));
glyphs.push(charcode);
if (charcode < min)
min = charcode;
if (charcode > max)
max = charcode;
}
// Since Format 6 is a dense array, check for gaps
for (var j = min; j < max; j++) {
if (glyphs.indexOf(j) == -1)
glyphs.push(j);
}
for (var j = 0; j < glyphs.length; j++)
glyphs[j] = { unicode: glyphs[j] + firstCode };
var ranges= getRanges(glyphs);
assert(ranges.length == 1, "Got " + ranges.length + " ranges in a dense array");
var encoding = properties.encoding;
var denseRange = ranges[0];
var start = denseRange[0];
var end = denseRange[1];
var index = firstCode;
for (var j = start; j <= end; j++)
encoding[index++] = glyphs[j - firstCode - 1].unicode;
cmap.data = createCMapTable(glyphs);
}
}
};
// Check that required tables are present
var requiredTables = [ "OS/2", "cmap", "head", "hhea",
"hmtx", "maxp", "name", "post" ];
@ -425,7 +495,7 @@ var Font = (function () {
if (requiredTables.length && requiredTables[0] == "OS/2") {
// Create a new file to hold the new version of our truetype with a new
// header and new offsets
var ttf = Uint8Array(kMaxFontFileSize);
var ttf = new Uint8Array(kMaxFontFileSize);
// The offsets object holds at the same time a representation of where
// to write the table entry information about a table and another offset
@ -448,18 +518,8 @@ var Font = (function () {
data: OS2
});
// If the font is missing a OS/2 table it's could be an old mac font
// without a 3-1-4 Unicode BMP table, so let's rewrite it.
var charset = properties.charset;
var glyphs = [];
for (var i = 0; i < charset.length; i++) {
glyphs.push({
unicode: GlyphsUnicode[charset[i]]
});
}
// Replace the old CMAP table with a shiny new one
cmap.data = createCMAPTable(glyphs);
replaceCMapTable(cmap, font, properties);
// Rewrite the 'post' table if needed
if (!post) {
@ -521,7 +581,7 @@ var Font = (function () {
},
convert: function font_convert(name, font, properties) {
var otf = Uint8Array(kMaxFontFileSize);
var otf = new Uint8Array(kMaxFontFileSize);
function createNameTable(name) {
var names = [
@ -599,7 +659,7 @@ var Font = (function () {
var charstrings = font.getOrderedCharStrings(properties.glyphs);
/** CMAP */
cmap = createCMAPTable(charstrings);
cmap = createCMapTable(charstrings);
createTableEntry(otf, offsets, "cmap", cmap);
/** HEAD */
@ -1088,7 +1148,7 @@ var Type1Parser = function() {
* The CFF class takes a Type1 file and wrap it into a 'Compact Font Format',
* which itself embed Type2 charstrings.
*/
const CFFStrings = [
var CFFStrings = [
".notdef","space","exclam","quotedbl","numbersign","dollar","percent","ampersand",
"quoteright","parenleft","parenright","asterisk","plus","comma","hyphen","period",
"slash","zero","one","two","three","four","five","six","seven","eight","nine",

122
pdf.js
View File

@ -58,7 +58,7 @@ function bytesToString(bytes) {
var Stream = (function() {
function constructor(arrayBuffer, start, length, dict) {
this.bytes = Uint8Array(arrayBuffer);
this.bytes = new Uint8Array(arrayBuffer);
this.start = start || 0;
this.pos = this.start;
this.end = (start + length) || this.bytes.length;
@ -125,7 +125,7 @@ var Stream = (function() {
var StringStream = (function() {
function constructor(str) {
var length = str.length;
var bytes = Uint8Array(length);
var bytes = new Uint8Array(length);
for (var n = 0; n < length; ++n)
bytes[n] = str.charCodeAt(n);
Stream.call(this, bytes);
@ -154,7 +154,7 @@ var DecodeStream = (function() {
var size = 512;
while (size < requested)
size <<= 1;
var buffer2 = Uint8Array(size);
var buffer2 = new Uint8Array(size);
for (var i = 0; i < current; ++i)
buffer2[i] = buffer[i];
return this.buffer = buffer2;
@ -220,13 +220,50 @@ var DecodeStream = (function() {
})();
var FakeStream = (function() {
function constructor(stream) {
this.dict = stream.dict;
DecodeStream.call(this);
};
constructor.prototype = Object.create(DecodeStream.prototype);
constructor.prototype.readBlock = function() {
var bufferLength = this.bufferLength;
bufferLength += 1024;
var buffer = this.ensureBuffer(bufferLength);
this.bufferLength = bufferLength;
};
constructor.prototype.getBytes = function(length) {
var pos = this.pos;
if (length) {
this.ensureBuffer(pos + length);
var end = pos + length;
while (!this.eof && this.bufferLength < end)
this.readBlock();
var bufEnd = this.bufferLength;
if (end > bufEnd)
end = bufEnd;
} else {
this.eof = true;
var end = this.bufferLength;
}
this.pos = end;
return this.buffer.subarray(pos, end)
};
return constructor;
})();
var FlateStream = (function() {
const codeLenCodeMap = Uint32Array([
var codeLenCodeMap = new Uint32Array([
16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15
]);
const lengthDecode = Uint32Array([
var lengthDecode = new Uint32Array([
0x00003, 0x00004, 0x00005, 0x00006, 0x00007, 0x00008, 0x00009,
0x0000a, 0x1000b, 0x1000d, 0x1000f, 0x10011, 0x20013, 0x20017,
0x2001b, 0x2001f, 0x30023, 0x3002b, 0x30033, 0x3003b, 0x40043,
@ -234,7 +271,7 @@ var FlateStream = (function() {
0x00102, 0x00102, 0x00102
]);
const distDecode = Uint32Array([
var distDecode = new Uint32Array([
0x00001, 0x00002, 0x00003, 0x00004, 0x10005, 0x10007, 0x20009,
0x2000d, 0x30011, 0x30019, 0x40021, 0x40031, 0x50041, 0x50061,
0x60081, 0x600c1, 0x70101, 0x70181, 0x80201, 0x80301, 0x90401,
@ -242,7 +279,7 @@ var FlateStream = (function() {
0xd4001, 0xd6001
]);
const fixedLitCodeTab = [Uint32Array([
var fixedLitCodeTab = [new Uint32Array([
0x70100, 0x80050, 0x80010, 0x80118, 0x70110, 0x80070, 0x80030,
0x900c0, 0x70108, 0x80060, 0x80020, 0x900a0, 0x80000, 0x80080,
0x80040, 0x900e0, 0x70104, 0x80058, 0x80018, 0x90090, 0x70114,
@ -319,7 +356,7 @@ var FlateStream = (function() {
0x900ff
]), 9];
const fixedDistCodeTab = [Uint32Array([
var fixedDistCodeTab = [new Uint32Array([
0x50000, 0x50010, 0x50008, 0x50018, 0x50004, 0x50014, 0x5000c,
0x5001c, 0x50002, 0x50012, 0x5000a, 0x5001a, 0x50006, 0x50016,
0x5000e, 0x00000, 0x50001, 0x50011, 0x50009, 0x50019, 0x50005,
@ -410,7 +447,7 @@ var FlateStream = (function() {
// build the table
var size = 1 << maxLen;
var codes = Uint32Array(size);
var codes = new Uint32Array(size);
for (var len = 1, code = 0, skip = 2;
len <= maxLen;
++len, code <<= 1, skip <<= 1) {
@ -597,9 +634,6 @@ var PredictorStream = (function() {
constructor.prototype = Object.create(DecodeStream.prototype);
constructor.prototype.readBlockTiff = function() {
var buffer = this.buffer;
var pos = this.pos;
var rowBytes = this.rowBytes;
var pixBytes = this.pixBytes;
@ -660,9 +694,6 @@ var PredictorStream = (function() {
this.bufferLength += rowBytes;
};
constructor.prototype.readBlockPng = function() {
var buffer = this.buffer;
var pos = this.pos;
var rowBytes = this.rowBytes;
var pixBytes = this.pixBytes;
@ -782,8 +813,8 @@ var Ascii85Stream = (function() {
constructor.prototype = Object.create(DecodeStream.prototype);
constructor.prototype.readBlock = function() {
const tildaCode = "~".charCodeAt(0);
const zCode = "z".charCodeAt(0);
var tildaCode = "~".charCodeAt(0);
var zCode = "z".charCodeAt(0);
var str = this.str;
var c = str.getByte();
@ -1945,10 +1976,10 @@ var Lexer = (function() {
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 // fx
];
const MIN_INT = (1<<31) | 0;
const MAX_INT = (MIN_INT - 1) | 0;
const MIN_UINT = 0;
const MAX_UINT = ((1<<30) * 4) - 1;
var MIN_INT = (1<<31) | 0;
var MAX_INT = (MIN_INT - 1) | 0;
var MIN_UINT = 0;
var MAX_UINT = ((1<<30) * 4) - 1;
function ToHexDigit(ch) {
if (ch >= "0" && ch <= "9")
@ -2392,6 +2423,9 @@ var Parser = (function() {
return new JpegStream(bytes, stream.dict);
} else if (name == "ASCII85Decode") {
return new Ascii85Stream(stream);
} else if (name == "CCITTFaxDecode") {
TODO("implement fax stream");
return new FakeStream(stream);
} else {
error("filter '" + name + "' not supported yet");
}
@ -2975,7 +3009,7 @@ var PDFDoc = (function() {
return constructor;
})();
const IDENTITY_MATRIX = [ 1, 0, 0, 1, 0, 0 ];
var IDENTITY_MATRIX = [ 1, 0, 0, 1, 0, 0 ];
// <canvas> contexts store most of the state we need natively.
// However, PDF needs a bit more state, which we store here.
@ -2999,7 +3033,7 @@ var CanvasExtraState = (function() {
return constructor;
})();
const Encodings = {
var Encodings = {
get ExpertEncoding() {
return shadow(this, "ExpertEncoding", [ ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
"space","exclamsmall","Hungarumlautsmall",,"dollaroldstyle","dollarsuperior",
@ -3277,13 +3311,13 @@ var CanvasGraphics = (function() {
};
}
const LINE_CAP_STYLES = [ "butt", "round", "square" ];
const LINE_JOIN_STYLES = [ "miter", "round", "bevel" ];
const NORMAL_CLIP = {};
const EO_CLIP = {};
var LINE_CAP_STYLES = [ "butt", "round", "square" ];
var LINE_JOIN_STYLES = [ "miter", "round", "bevel" ];
var NORMAL_CLIP = {};
var EO_CLIP = {};
// Used for tiling patterns
const PAINT_TYPE_COLORED = 1, PAINT_TYPE_UNCOLORED = 2;
var PAINT_TYPE_COLORED = 1, PAINT_TYPE_UNCOLORED = 2;
constructor.prototype = {
translateFont: function(fontDict, xref, resources) {
@ -3306,7 +3340,7 @@ var CanvasGraphics = (function() {
// Fonts with an embedded cmap but without any assignment in
// it are not yet supported, so ask the fonts loader to ignore
// them to not pay a stupid one sec latence.
var ignoreFont = true;
var ignoreFont = false;
var encodingMap = {};
var charset = [];
@ -3350,20 +3384,24 @@ var CanvasGraphics = (function() {
}
}
} else if (fontDict.has("ToUnicode")) {
encodingMap = {empty: true};
var cmapObj = xref.fetchIfRef(fontDict.get("ToUnicode"));
if (IsName(cmapObj)) {
error("ToUnicode file cmap translation not implemented");
} else if (IsStream(cmapObj)) {
var encoding = Encodings["WinAnsiEncoding"];
var firstChar = xref.fetchIfRef(fontDict.get("FirstChar"));
for (var i = firstChar; i < encoding.length; i++)
encodingMap[i] = new Name(encoding[i]);
var tokens = [];
var token = "";
var buffer = cmapObj.ensureBuffer ? cmapObj.ensureBuffer() : cmapObj;
var cmap = cmapObj.getBytes(buffer.byteLength);
var length = cmapObj.length;
if (cmapObj instanceof FlateStream) {
cmapObj.readBlock();
length = cmapObj.bufferLength;
}
var cmap = cmapObj.getBytes(length);
for (var i =0; i < cmap.length; i++) {
var byte = cmap[i];
if (byte == 0x20 || byte == 0x0A || byte == 0x3C || byte == 0x3E) {
@ -3390,8 +3428,10 @@ var CanvasGraphics = (function() {
var code = parseInt("0x" + tokens[j+2]);
for (var k = startRange; k <= endRange; k++) {
encodingMap[k] = GlyphsUnicode[encoding[code]];
charset.push(encoding[code++]);
// The encoding mapping table will be filled
// later during the building phase
//encodingMap[k] = GlyphsUnicode[encoding[code]];
charset.push(encoding[code++] || ".notdef");
}
}
break;
@ -3696,7 +3736,7 @@ var CanvasGraphics = (function() {
}
this.current.fontSize = size;
this.ctx.font = this.current.fontSize +'px "' + fontName + '"';
this.ctx.font = this.current.fontSize +'px "' + fontName + '", Symbol';
},
setTextRenderingMode: function(mode) {
TODO("text rendering mode");
@ -3725,7 +3765,9 @@ var CanvasGraphics = (function() {
this.ctx.transform.apply(this.ctx, this.current.textMatrix);
this.ctx.scale(1, -1);
this.ctx.translate(0, -2 * this.current.y);
this.ctx.fillText(Fonts.charsToUnicode(text), this.current.x, this.current.y);
text = Fonts.charsToUnicode(text);
this.ctx.fillText(text, this.current.x, this.current.y);
this.current.x += this.ctx.measureText(text).width;
this.ctx.restore();
@ -3812,7 +3854,7 @@ var CanvasGraphics = (function() {
var pattern = xref.fetchIfRef(patternRes.get(patternName.name));
var patternDict = IsStream(pattern) ? pattern.dict : pattern;
const types = [null, this.tilingFill,
var types = [null, this.tilingFill,
function() { TODO("Shading Patterns"); }];
var typeNum = patternDict.get("PatternType");
var patternFn = types[typeNum];
@ -3961,7 +4003,7 @@ var CanvasGraphics = (function() {
if (background)
TODO("handle background colors");
const types = [null,
var types = [null,
this.fillFunctionShading,
this.fillAxialShading,
this.fillRadialShading];
@ -4398,7 +4440,7 @@ var PDFFunction = (function() {
if (!dict)
dict = fn;
const types = [this.constructSampled,
var types = [this.constructSampled,
null,
this.constructInterpolated,
this.constructStiched,

27
test.py
View File

@ -9,6 +9,7 @@ def prompt(question):
ANAL = True
DEFAULT_MANIFEST_FILE = 'test_manifest.json'
EQLOG_FILE = 'eq.log'
REFDIR = 'ref'
TMPDIR = 'tmp'
VERBOSE = False
@ -16,7 +17,7 @@ VERBOSE = False
MIMEs = {
'.css': 'text/css',
'.html': 'text/html',
'.js': 'application/json',
'.js': 'application/javascript',
'.json': 'application/json',
'.pdf': 'application/pdf',
'.xhtml': 'application/xhtml+xml',
@ -35,6 +36,7 @@ class State:
numEqNoSnapshot = 0
numFBFFailures = 0
numLoadFailures = 0
eqLog = None
class Result:
def __init__(self, snapshot, failure):
@ -115,8 +117,8 @@ def setUp(manifestFile, masterMode):
assert not os.path.isdir(TMPDIR)
testBrowsers = [ b for b in
( 'firefox5', )
#'chrome12', 'chrome13', 'firefox4', 'firefox6','opera11' ):
( 'firefox5', 'firefox6', )
#'chrome12', 'chrome13', 'firefox4', 'opera11' ):
if os.access(b, os.R_OK | os.X_OK) ]
mf = open(manifestFile)
@ -150,7 +152,7 @@ def setUp(manifestFile, masterMode):
taskResults.append([ ])
State.taskResults[b][id] = taskResults
State.remaining = len(manifestList)
State.remaining = len(testBrowsers) * len(manifestList)
for b in testBrowsers:
print 'Launching', b
@ -190,6 +192,7 @@ def check(task, results, browser):
def checkEq(task, results, browser):
pfx = os.path.join(REFDIR, sys.platform, browser, task['id'])
results = results[0]
taskId = task['id']
passed = True
for page in xrange(len(results)):
@ -208,7 +211,21 @@ def checkEq(task, results, browser):
eq = (ref == snapshot)
if not eq:
print 'TEST-UNEXPECTED-FAIL | eq', task['id'], '| in', browser, '| rendering of page', page + 1, '!= reference rendering'
print 'TEST-UNEXPECTED-FAIL | eq', taskId, '| in', browser, '| rendering of page', page + 1, '!= reference rendering'
# XXX need to dump this always, somehow, when we have
# the reference repository
if State.masterMode:
if not State.eqLog:
State.eqLog = open(EQLOG_FILE, 'w')
eqLog = State.eqLog
# NB: this follows the format of Mozilla reftest
# output so that we can reuse its reftest-analyzer
# script
print >>eqLog, 'REFTEST TEST-UNEXPECTED-FAIL |', browser +'-'+ taskId +'-page'+ str(page + 1), '| image comparison (==)'
print >>eqLog, 'REFTEST IMAGE 1 (TEST):', snapshot
print >>eqLog, 'REFTEST IMAGE 2 (REFERENCE):', ref
passed = False
State.numEqFailures += 1

68
test_slave.html
View File

@ -1,6 +1,7 @@
<html>
<head>
<title>pdf.js test slave</title>
<style type="text/css"></style>
<script type="text/javascript" src="pdf.js"></script>
<script type="text/javascript" src="fonts.js"></script>
<script type="text/javascript" src="glyphlist.js"></script>
@ -31,7 +32,7 @@ function load() {
stdout = document.getElementById("stdout");
log("Harness thinks this browser is '"+ browser +"'\n");
log("Fetching manifest ...");
log("Fetching manifest "+ manifestFile +"...");
var r = new XMLHttpRequest();
r.open("GET", manifestFile, false);
@ -81,38 +82,69 @@ function nextPage() {
}
failure = '';
log(" drawing page "+ currentTask.pageNum +"...");
log(" loading page "+ currentTask.pageNum +"... ");
var ctx = canvas.getContext("2d");
clear(ctx);
var fonts = [];
var fontsReady = true;
var gfx = new CanvasGraphics(ctx);
try {
currentPage = pdfDoc.getPage(currentTask.pageNum);
currentPage.compile(gfx, fonts);
// Inspect fonts and translate the missing ones
var count = fonts.length;
for (var i = 0; i < count; ++i) {
var font = fonts[i];
if (Fonts[font.name]) {
fontsReady = fontsReady && !Fonts[font.name].loading;
continue;
}
new Font(font.name, font.file, font.properties);
fontsReady = false;
}
} catch(e) {
failure = 'compile: '+ e.toString();
}
// TODO load fonts
setTimeout(function() {
if (!failure) {
try {
currentPage.display(gfx);
} catch(e) {
failure = 'render: '+ e.toString();
}
var checkFontsLoadedIntervalTimer = null;
function checkFontsLoaded() {
for (var i = 0; i < count; i++) {
if (Fonts[font.name].loading) {
return;
}
currentTask.taskDone = (currentTask.pageNum == pdfDoc.numPages
&& (1 + currentTask.round) == currentTask.rounds);
sendTaskResult(canvas.toDataURL("image/png"));
log("done"+ (failure ? " (failed!)" : "") +"\n");
}
window.clearInterval(checkFontsLoadedIntervalTimer);
++currentTask.pageNum, nextPage();
},
0
);
snapshotCurrentPage(gfx);
}
if (failure || fontsReady) {
snapshotCurrentPage(gfx);
} else {
checkFontsLoadedIntervalTimer = setInterval(checkFontsLoaded, 10);
}
}
function snapshotCurrentPage(gfx) {
log("done, snapshotting... ");
if (!failure) {
try {
currentPage.display(gfx);
} catch(e) {
failure = 'render: '+ e.toString();
}
}
currentTask.taskDone = (currentTask.pageNum == pdfDoc.numPages
&& (1 + currentTask.round) == currentTask.rounds);
sendTaskResult(canvas.toDataURL("image/png"));
log("done"+ (failure ? " (failed!)" : "") +"\n");
++currentTask.pageNum, nextPage();
}
function done() {