pdf.js/src/obj.js
2011-11-02 15:08:19 -04:00

743 lines
23 KiB
JavaScript

/* -*- Mode: Java; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* vim: set shiftwidth=2 tabstop=2 autoindent cindent expandtab: */
'use strict';
var Name = (function nameName() {
function constructor(name) {
this.name = name;
}
constructor.prototype = {
};
return constructor;
})();
var Cmd = (function cmdCmd() {
function constructor(cmd) {
this.cmd = cmd;
}
constructor.prototype = {
};
return constructor;
})();
var Dict = (function dictDict() {
function constructor() {
this.map = Object.create(null);
}
constructor.prototype = {
get: function dictGet(key1, key2, key3) {
var value;
if (typeof (value = this.map[key1]) != 'undefined' || key1 in this.map ||
typeof key2 == 'undefined') {
return value;
}
if (typeof (value = this.map[key2]) != 'undefined' || key2 in this.map ||
typeof key3 == 'undefined') {
return value;
}
return this.map[key3] || null;
},
set: function dictSet(key, value) {
this.map[key] = value;
},
has: function dictHas(key) {
return key in this.map;
},
forEach: function dictForEach(callback) {
for (var key in this.map) {
callback(key, this.map[key]);
}
}
};
return constructor;
})();
var Ref = (function refRef() {
function constructor(num, gen) {
this.num = num;
this.gen = gen;
}
constructor.prototype = {
};
return constructor;
})();
// The reference is identified by number and generation,
// this structure stores only one instance of the reference.
var RefSet = (function refSet() {
function constructor() {
this.dict = {};
}
constructor.prototype = {
has: function refSetHas(ref) {
return !!this.dict['R' + ref.num + '.' + ref.gen];
},
put: function refSetPut(ref) {
this.dict['R' + ref.num + '.' + ref.gen] = ref;
}
};
return constructor;
})();
var Catalog = (function catalogCatalog() {
function constructor(xref) {
this.xref = xref;
var obj = xref.getCatalogObj();
assertWellFormed(isDict(obj), 'catalog object is not a dictionary');
this.catDict = obj;
}
constructor.prototype = {
get toplevelPagesDict() {
var pagesObj = this.catDict.get('Pages');
assertWellFormed(isRef(pagesObj), 'invalid top-level pages reference');
var xrefObj = this.xref.fetch(pagesObj);
assertWellFormed(isDict(xrefObj), 'invalid top-level pages dictionary');
// shadow the prototype getter
return shadow(this, 'toplevelPagesDict', xrefObj);
},
get documentOutline() {
var obj = this.catDict.get('Outlines');
var xref = this.xref;
var root = { items: [] };
if (isRef(obj)) {
obj = xref.fetch(obj).get('First');
var processed = new RefSet();
if (isRef(obj)) {
var queue = [{obj: obj, parent: root}];
// to avoid recursion keeping track of the items
// in the processed dictionary
processed.put(obj);
while (queue.length > 0) {
var i = queue.shift();
var outlineDict = xref.fetch(i.obj);
if (!outlineDict.has('Title'))
error('Invalid outline item');
var dest = outlineDict.get('A');
if (dest)
dest = xref.fetchIfRef(dest).get('D');
else if (outlineDict.has('Dest')) {
dest = outlineDict.get('Dest');
if (isName(dest))
dest = dest.name;
}
var title = xref.fetchIfRef(outlineDict.get('Title'));
var outlineItem = {
dest: dest,
title: stringToPDFString(title),
color: outlineDict.get('C') || [0, 0, 0],
count: outlineDict.get('Count'),
bold: !!(outlineDict.get('F') & 2),
italic: !!(outlineDict.get('F') & 1),
items: []
};
i.parent.items.push(outlineItem);
obj = outlineDict.get('First');
if (isRef(obj) && !processed.has(obj)) {
queue.push({obj: obj, parent: outlineItem});
processed.put(obj);
}
obj = outlineDict.get('Next');
if (isRef(obj) && !processed.has(obj)) {
queue.push({obj: obj, parent: i.parent});
processed.put(obj);
}
}
}
}
obj = root.items.length > 0 ? root.items : null;
return shadow(this, 'documentOutline', obj);
},
get numPages() {
var obj = this.toplevelPagesDict.get('Count');
assertWellFormed(
isInt(obj),
'page count in top level pages object is not an integer'
);
// shadow the prototype getter
return shadow(this, 'num', obj);
},
traverseKids: function catalogTraverseKids(pagesDict) {
var pageCache = this.pageCache;
var kids = pagesDict.get('Kids');
assertWellFormed(isArray(kids),
'page dictionary kids object is not an array');
for (var i = 0, ii = kids.length; i < ii; ++i) {
var kid = kids[i];
assertWellFormed(isRef(kid),
'page dictionary kid is not a reference');
var obj = this.xref.fetch(kid);
if (isDict(obj, 'Page') || (isDict(obj) && !obj.has('Kids'))) {
pageCache.push(new Page(this.xref, pageCache.length, obj, kid));
} else { // must be a child page dictionary
assertWellFormed(
isDict(obj),
'page dictionary kid reference points to wrong type of object'
);
this.traverseKids(obj);
}
}
},
get destinations() {
function fetchDestination(xref, ref) {
var dest = xref.fetchIfRef(ref);
return isDict(dest) ? dest.get('D') : dest;
}
var xref = this.xref;
var dests = {}, nameTreeRef, nameDictionaryRef;
var obj = this.catDict.get('Names');
if (obj)
nameTreeRef = xref.fetchIfRef(obj).get('Dests');
else if (this.catDict.has('Dests'))
nameDictionaryRef = this.catDict.get('Dests');
if (nameDictionaryRef) {
// reading simple destination dictionary
obj = xref.fetchIfRef(nameDictionaryRef);
obj.forEach(function catalogForEach(key, value) {
if (!value) return;
dests[key] = fetchDestination(xref, value);
});
}
if (nameTreeRef) {
// reading name tree
var processed = new RefSet();
processed.put(nameTreeRef);
var queue = [nameTreeRef];
while (queue.length > 0) {
var i, n;
obj = xref.fetch(queue.shift());
if (obj.has('Kids')) {
var kids = obj.get('Kids');
for (i = 0, n = kids.length; i < n; i++) {
var kid = kids[i];
if (processed.has(kid))
error('invalid destinations');
queue.push(kid);
processed.put(kid);
}
continue;
}
var names = obj.get('Names');
for (i = 0, n = names.length; i < n; i += 2) {
dests[names[i]] = fetchDestination(xref, names[i + 1]);
}
}
}
return shadow(this, 'destinations', dests);
},
getPage: function catalogGetPage(n) {
var pageCache = this.pageCache;
if (!pageCache) {
pageCache = this.pageCache = [];
this.traverseKids(this.toplevelPagesDict);
}
return this.pageCache[n - 1];
}
};
return constructor;
})();
var XRef = (function xRefXRef() {
function constructor(stream, startXRef, mainXRefEntriesOffset) {
this.stream = stream;
this.entries = [];
this.xrefstms = {};
var trailerDict = this.readXRef(startXRef);
// prepare the XRef cache
this.cache = [];
var encrypt = trailerDict.get('Encrypt');
if (encrypt) {
var fileId = trailerDict.get('ID');
this.encrypt = new CipherTransformFactory(this.fetch(encrypt),
fileId[0] /*, password */);
}
// get the root dictionary (catalog) object
if (!isRef(this.root = trailerDict.get('Root')))
error('Invalid root reference');
}
constructor.prototype = {
readXRefTable: function readXRefTable(parser) {
var obj;
while (true) {
if (isCmd(obj = parser.getObj(), 'trailer'))
break;
if (!isInt(obj))
error('Invalid XRef table');
var first = obj;
if (!isInt(obj = parser.getObj()))
error('Invalid XRef table');
var n = obj;
if (first < 0 || n < 0 || (first + n) != ((first + n) | 0))
error('Invalid XRef table: ' + first + ', ' + n);
for (var i = first; i < first + n; ++i) {
var entry = {};
if (!isInt(obj = parser.getObj()))
error('Invalid XRef table: ' + first + ', ' + n);
entry.offset = obj;
if (!isInt(obj = parser.getObj()))
error('Invalid XRef table: ' + first + ', ' + n);
entry.gen = obj;
obj = parser.getObj();
if (isCmd(obj, 'n')) {
entry.uncompressed = true;
} else if (isCmd(obj, 'f')) {
entry.free = true;
} else {
error('Invalid XRef table: ' + first + ', ' + n);
}
if (!this.entries[i]) {
// In some buggy PDF files the xref table claims to start at 1
// instead of 0.
if (i == 1 && first == 1 &&
entry.offset == 0 && entry.gen == 65535 && entry.free) {
i = first = 0;
}
this.entries[i] = entry;
}
}
}
// read the trailer dictionary
var dict;
if (!isDict(dict = parser.getObj()))
error('Invalid XRef table');
// get the 'Prev' pointer
var prev;
obj = dict.get('Prev');
if (isInt(obj)) {
prev = obj;
} else if (isRef(obj)) {
// certain buggy PDF generators generate "/Prev NNN 0 R" instead
// of "/Prev NNN"
prev = obj.num;
}
if (prev) {
this.readXRef(prev);
}
// check for 'XRefStm' key
if (isInt(obj = dict.get('XRefStm'))) {
var pos = obj;
// ignore previously loaded xref streams (possible infinite recursion)
if (!(pos in this.xrefstms)) {
this.xrefstms[pos] = 1;
this.readXRef(pos);
}
}
return dict;
},
readXRefStream: function readXRefStream(stream) {
var streamParameters = stream.parameters;
var byteWidths = streamParameters.get('W');
var range = streamParameters.get('Index');
if (!range)
range = [0, streamParameters.get('Size')];
var i, j;
while (range.length > 0) {
var first = range[0], n = range[1];
if (!isInt(first) || !isInt(n))
error('Invalid XRef range fields: ' + first + ', ' + n);
var typeFieldWidth = byteWidths[0];
var offsetFieldWidth = byteWidths[1];
var generationFieldWidth = byteWidths[2];
if (!isInt(typeFieldWidth) || !isInt(offsetFieldWidth) ||
!isInt(generationFieldWidth)) {
error('Invalid XRef entry fields length: ' + first + ', ' + n);
}
for (i = 0; i < n; ++i) {
var type = 0, offset = 0, generation = 0;
for (j = 0; j < typeFieldWidth; ++j)
type = (type << 8) | stream.getByte();
// if type field is absent, its default value = 1
if (typeFieldWidth == 0)
type = 1;
for (j = 0; j < offsetFieldWidth; ++j)
offset = (offset << 8) | stream.getByte();
for (j = 0; j < generationFieldWidth; ++j)
generation = (generation << 8) | stream.getByte();
var entry = {};
entry.offset = offset;
entry.gen = generation;
switch (type) {
case 0:
entry.free = true;
break;
case 1:
entry.uncompressed = true;
break;
case 2:
break;
default:
error('Invalid XRef entry type: ' + type);
}
if (!this.entries[first + i])
this.entries[first + i] = entry;
}
range.splice(0, 2);
}
var prev = streamParameters.get('Prev');
if (isInt(prev))
this.readXRef(prev);
return streamParameters;
},
indexObjects: function indexObjects() {
// Simple scan through the PDF content to find objects,
// trailers and XRef streams.
function readToken(data, offset) {
var token = '', ch = data[offset];
while (ch !== 13 && ch !== 10) {
if (++offset >= data.length)
break;
token += String.fromCharCode(ch);
ch = data[offset];
}
return token;
}
function skipUntil(data, offset, what) {
var length = what.length, dataLength = data.length;
var skipped = 0;
// finding byte sequence
while (offset < dataLength) {
var i = 0;
while (i < length && data[offset + i] == what[i])
++i;
if (i >= length)
break; // sequence found
offset++;
skipped++;
}
return skipped;
}
var trailerBytes = new Uint8Array([116, 114, 97, 105, 108, 101, 114]);
var startxrefBytes = new Uint8Array([115, 116, 97, 114, 116, 120, 114,
101, 102]);
var endobjBytes = new Uint8Array([101, 110, 100, 111, 98, 106]);
var xrefBytes = new Uint8Array([47, 88, 82, 101, 102]);
var stream = this.stream;
stream.pos = 0;
var buffer = stream.getBytes();
var position = stream.start, length = buffer.length;
var trailers = [], xrefStms = [];
var state = 0;
var currentToken;
while (position < length) {
var ch = buffer[position];
if (ch === 32 || ch === 9 || ch === 13 || ch === 10) {
++position;
continue;
}
if (ch === 37) { // %-comment
do {
++position;
ch = buffer[position];
} while (ch !== 13 && ch !== 10);
continue;
}
var token = readToken(buffer, position);
var m;
if (token === 'xref') {
position += skipUntil(buffer, position, trailerBytes);
trailers.push(position);
position += skipUntil(buffer, position, startxrefBytes);
} else if ((m = /^(\d+)\s+(\d+)\s+obj\b/.exec(token))) {
this.entries[m[1]] = {
offset: position,
gen: m[2] | 0,
uncompressed: true
};
var contentLength = skipUntil(buffer, position, endobjBytes) + 7;
var content = buffer.subarray(position, position + contentLength);
// checking XRef stream suspect
// (it shall have '/XRef' and next char is not a letter)
var xrefTagOffset = skipUntil(content, 0, xrefBytes);
if (xrefTagOffset < contentLength &&
content[xrefTagOffset + 5] < 64) {
xrefStms.push(position);
this.xrefstms[position] = 1; // don't read it recursively
}
position += contentLength;
} else
position += token.length + 1;
}
// reading XRef streams
for (var i = 0, ii = xrefStms.length; i < ii; ++i) {
this.readXRef(xrefStms[i]);
}
// finding main trailer
var dict;
for (var i = 0, ii = trailers.length; i < ii; ++i) {
stream.pos = trailers[i];
var parser = new Parser(new Lexer(stream), true);
var obj = parser.getObj();
if (!isCmd(obj, 'trailer'))
continue;
// read the trailer dictionary
if (!isDict(dict = parser.getObj()))
continue;
// taking the first one with 'ID'
if (dict.has('ID'))
return dict;
}
// no tailer with 'ID', taking last one (if exists)
if (dict)
return dict;
// nothing helps
error('Invalid PDF structure');
return null;
},
readXRef: function readXref(startXRef) {
var stream = this.stream;
stream.pos = startXRef;
var parser = new Parser(new Lexer(stream), true);
var obj = parser.getObj();
// parse an old-style xref table
if (isCmd(obj, 'xref'))
return this.readXRefTable(parser);
// parse an xref stream
if (isInt(obj)) {
if (!isInt(parser.getObj()) ||
!isCmd(parser.getObj(), 'obj') ||
!isStream(obj = parser.getObj())) {
error('Invalid XRef stream');
}
return this.readXRefStream(obj);
}
return this.indexObjects();
},
getEntry: function xRefGetEntry(i) {
var e = this.entries[i];
if (e.free)
error('reading an XRef stream not implemented yet');
return e;
},
fetchIfRef: function xRefFetchIfRef(obj) {
if (!isRef(obj))
return obj;
return this.fetch(obj);
},
fetch: function xRefFetch(ref, suppressEncryption) {
var num = ref.num;
var e = this.cache[num];
if (e)
return e;
e = this.getEntry(num);
var gen = ref.gen;
var stream, parser;
if (e.uncompressed) {
if (e.gen != gen)
throw ('inconsistent generation in XRef');
stream = this.stream.makeSubStream(e.offset);
parser = new Parser(new Lexer(stream), true, this);
var obj1 = parser.getObj();
var obj2 = parser.getObj();
var obj3 = parser.getObj();
if (!isInt(obj1) || obj1 != num ||
!isInt(obj2) || obj2 != gen ||
!isCmd(obj3)) {
error('bad XRef entry');
}
if (!isCmd(obj3, 'obj')) {
// some bad pdfs use "obj1234" and really mean 1234
if (obj3.cmd.indexOf('obj') == 0) {
num = parseInt(obj3.cmd.substring(3), 10);
if (!isNaN(num))
return num;
}
error('bad XRef entry');
}
if (this.encrypt && !suppressEncryption) {
try {
e = parser.getObj(this.encrypt.createCipherTransform(num, gen));
} catch (ex) {
// almost all streams must be encrypted, but sometimes
// they are not probably due to some broken generators
// re-trying without encryption
return this.fetch(ref, true);
}
} else {
e = parser.getObj();
}
// Don't cache streams since they are mutable (except images).
if (!isStream(e) || e.getImage)
this.cache[num] = e;
return e;
}
// compressed entry
stream = this.fetch(new Ref(e.offset, 0));
if (!isStream(stream))
error('bad ObjStm stream');
var first = stream.parameters.get('First');
var n = stream.parameters.get('N');
if (!isInt(first) || !isInt(n)) {
error('invalid first and n parameters for ObjStm stream');
}
parser = new Parser(new Lexer(stream), false);
var i, entries = [], nums = [];
// read the object numbers to populate cache
for (i = 0; i < n; ++i) {
num = parser.getObj();
if (!isInt(num)) {
error('invalid object number in the ObjStm stream: ' + num);
}
nums.push(num);
var offset = parser.getObj();
if (!isInt(offset)) {
error('invalid object offset in the ObjStm stream: ' + offset);
}
}
// read stream objects for cache
for (i = 0; i < n; ++i) {
entries.push(parser.getObj());
this.cache[nums[i]] = entries[i];
}
e = entries[e.gen];
if (!e) {
error('bad XRef entry for compressed object');
}
return e;
},
getCatalogObj: function xRefGetCatalogObj() {
return this.fetch(this.root);
}
};
return constructor;
})();
/**
* A PDF document and page is built of many objects. E.g. there are objects
* for fonts, images, rendering code and such. These objects might get processed
* inside of a worker. The `PDFObjects` implements some basic functions to
* manage these objects.
*/
var PDFObjects = (function pdfObjects() {
function PDFObjects() {
this.objs = {};
}
PDFObjects.prototype = {
objs: null,
/**
* Internal function.
* Ensures there is an object defined for `objId`. Stores `data` on the
* object *if* it is created.
*/
ensureObj: function pdfObjectsEnsureObj(objId, data) {
if (this.objs[objId])
return this.objs[objId];
return this.objs[objId] = new Promise(objId, data);
},
/**
* If called *without* callback, this returns the data of `objId` but the
* object needs to be resolved. If it isn't, this function throws.
*
* If called *with* a callback, the callback is called with the data of the
* object once the object is resolved. That means, if you call this
* function and the object is already resolved, the callback gets called
* right away.
*/
get: function pdfObjectsGet(objId, callback) {
// If there is a callback, then the get can be async and the object is
// not required to be resolved right now
if (callback) {
this.ensureObj(objId).then(callback);
return null;
}
// If there isn't a callback, the user expects to get the resolved data
// directly.
var obj = this.objs[objId];
// If there isn't an object yet or the object isn't resolved, then the
// data isn't ready yet!
if (!obj || !obj.isResolved) {
throw 'Requesting object that isn\'t resolved yet ' + objId;
return null;
} else {
return obj.data;
}
},
/**
* Resolves the object `objId` with optional `data`.
*/
resolve: function pdfObjectsResolve(objId, data) {
var objs = this.objs;
// In case there is a promise already on this object, just resolve it.
if (objs[objId]) {
objs[objId].resolve(data);
} else {
this.ensureObj(objId, data);
}
},
onData: function pdfObjectsOnData(objId, callback) {
this.ensureObj(objId).onData(callback);
},
isResolved: function pdfObjectsIsResolved(objId) {
var objs = this.objs;
if (!objs[objId]) {
return false;
} else {
return objs[objId].isResolved;
}
},
hasData: function pdfObjectsHasData(objId) {
var objs = this.objs;
if (!objs[objId]) {
return false;
} else {
return objs[objId].hasData;
}
},
/**
* Sets the data of an object but *doesn't* resolve it.
*/
setData: function pdfObjectsSetData(objId, data) {
// Watchout! If you call `this.ensureObj(objId, data)` you're going to
// create a *resolved* promise which shouldn't be the case!
this.ensureObj(objId).data = data;
}
};
return PDFObjects;
})();