akira
/
pdfjs-dist
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
Browse Source

Merge pull request #1148 from arturadib/readXRefTable 

Rewrite of readXRefTable [obj.js]
notmasteryet 13 years ago
parent
13f207424c 775290d698
commit
ff01faa86f
1 changed files with 90 additions and 73 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 163
      src/obj.js

163
src/obj.js
Unescape View File

@ -287,75 +287,70 @@ var XRef = (function XRefClosure() { @@ -287,75 +287,70 @@ var XRef = (function XRefClosure() {
XRef.prototype = {
readXRefTable: function readXRefTable(parser) {
// Example of cross-reference table:
// xref
// 0 1 <-- subsection header (first obj #, obj count)
// 0000000000 65535 f <-- actual object (offset, generation #, f/n)
// 23 2 <-- subsection header ... and so on ...
// 0000025518 00002 n
// 0000025635 00000 n
// trailer
// ...
// Outer loop is over subsection headers
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) {
while (!isCmd(obj = parser.getObj(), 'trailer')) {
var first = obj,
count = parser.getObj();
if (!isInt(first) || !isInt(count))
error('Invalid XRef table: wrong types in subsection header');
// Inner loop is over objects themselves
for (var i = 0; i < count; 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.offset = parser.getObj();
entry.gen = parser.getObj();
var type = parser.getObj();
if (isCmd(type, '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;
}
}
}
else if (isCmd(type, 'n'))
entry.uncompressed = true;
// 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);
}
// Validate entry obj
if (!isInt(entry.offset) || !isInt(entry.gen) ||
!(entry.free || entry.uncompressed)) {
error('Invalid entry in XRef subsection: ' + first + ', ' + count);
}
// 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);
if (!this.entries[i + first])
this.entries[i + first] = entry;
}
}
// Sanity check: as per spec, first object must have these properties
if (this.entries[0] &&
!(this.entries[0].gen === 65535 && this.entries[0].free))
error('Invalid XRef table: unexpected first object');
// Sanity check
if (!isCmd(obj, 'trailer'))
error('Invalid XRef table: could not find trailer dictionary');
// Read trailer dictionary, e.g.
// trailer
// << /Size 22
// /Root 20R
// /Info 10R
// /ID [ <81b14aafa313db63dbd6f981e49f94f4> ]
// >>
// The parser goes through the entire stream << ... >> and provides
// a getter interface for the key-value table
var dict = parser.getObj();
if (!isDict(dict))
error('Invalid XRef table: could not parse trailer dictionary');
return dict;
},
readXRefStream: function readXRefStream(stream) {
@ -407,9 +402,6 @@ var XRef = (function XRefClosure() { @@ -407,9 +402,6 @@ var XRef = (function XRefClosure() {
}
range.splice(0, 2);
}
var prev = streamParameters.get('Prev');
if (isInt(prev))
this.readXRef(prev);
return streamParameters;
},
indexObjects: function indexObjects() {
@ -529,22 +521,47 @@ var XRef = (function XRefClosure() { @@ -529,22 +521,47 @@ var XRef = (function XRefClosure() {
try {
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)) {
var dict;
// Get dictionary
if (isCmd(obj, 'xref')) {
// Parse end-of-file XRef
dict = this.readXRefTable(parser);
// Recursively get other XRefs 'XRefStm', if any
obj = dict.get('XRefStm');
if (isInt(obj)) {
var pos = obj;
// ignore previously loaded xref streams
// (possible infinite recursion)
if (!(pos in this.xrefstms)) {
this.xrefstms[pos] = 1;
this.readXRef(pos);
}
}
} else if (isInt(obj)) {
// Parse in-stream XRef
if (!isInt(parser.getObj()) ||
!isCmd(parser.getObj(), 'obj') ||
!isStream(obj = parser.getObj())) {
error('Invalid XRef stream');
}
return this.readXRefStream(obj);
dict = this.readXRefStream(obj);
}
// Recursively get previous dictionary, if any
obj = dict.get('Prev');
if (isInt(obj))
this.readXRef(obj);
else if (isRef(obj)) {
// The spec says Prev must not be a reference, i.e. "/Prev NNN"
// This is a fallback for non-compliant PDFs, i.e. "/Prev NNN 0 R"
this.readXRef(obj.num);
}
return dict;
} catch (e) {
log('Reading of the xref table/stream failed: ' + e);
log('(while reading XRef): ' + e);
}
warn('Indexing all PDF objects');

Write Preview
Loading…
Cancel
Save