def readindirect(self, objnum, gennum):
''' Read an indirect object. If it has already
been read, return it from the cache.
'''
def setobj(obj):
# Store the new object in the dictionary
# once we have its value
record[1] = obj
def ordinary(source, setobj, obj):
# Deal with an ordinary (non-array, non-dict) object
setobj(obj)
return obj
fdata, objnum, gennum = self.fdata, int(objnum), int(gennum)
record = self.indirect_objects[fdata, objnum, gennum]
if record[1] is not self.unresolved:
return record[1]
# Read the object header and validate it
source = PdfTokens(fdata, record[0])
objid = source.multiple(3)
assert int(objid[0]) == objnum, objid
assert int(objid[1]) == gennum, objid
assert objid[2] == 'obj', objid
# Read the object, and call special code if it starts
# an array or dictionary
obj = source.next()
obj = self.special.get(obj, ordinary)(source, setobj, obj)
self.readstream(obj, source)
obj.indirect = True
return obj
def readindirect(self, objnum, gennum):
''' Read an indirect object. If it has already
been read, return it from the cache.
'''
def setobj(obj):
# Store the new object in the dictionary
# once we have its value
record[1] = obj
def ordinary(source, setobj, obj):
# Deal with an ordinary (non-array, non-dict) object
setobj(obj)
return obj
fdata, objnum, gennum = self.fdata, int(objnum), int(gennum)
record = self.indirect_objects[fdata, objnum, gennum]
if record[1] is not self.unresolved:
return record[1]
# Read the object header and validate it
source = PdfTokens(fdata, record[0])
objid = source.multiple(3)
assert int(objid[0]) == objnum, objid
assert int(objid[1]) == gennum, objid
assert objid[2] == 'obj', objid
# Read the object, and call special code if it starts
# an array or dictionary
obj = source.next()
obj = self.special.get(obj, ordinary)(source, setobj, obj)
self.readstream(obj, source)
obj.indirect = True
return obj
def readxref(fdata):
startloc = fdata.rindex('startxref')
xrefinfo = list(PdfTokens(fdata, startloc, False))
assert len(xrefinfo) == 3, xrefinfo
assert xrefinfo[0] == 'startxref', xrefinfo[0]
assert xrefinfo[1].isdigit(), xrefinfo[1]
assert xrefinfo[2].rstrip() == '%%EOF', repr(xrefinfo[2])
return startloc, PdfTokens(fdata, int(xrefinfo[1]))
def readstream(obj, source):
''' Read optional stream following a dictionary
object.
'''
tok = source.next()
if tok == 'endobj':
return # No stream
assert isinstance(obj, PdfDict)
assert tok == 'stream', tok
fdata = source.fdata
floc = fdata.rindex(tok, 0, source.floc) + len(tok)
ch = fdata[floc]
if ch == '\r':
floc += 1
ch = fdata[floc]
assert ch == '\n'
startstream = floc + 1
endstream = startstream + int(obj.Length)
obj._stream = fdata[startstream:endstream]
source = PdfTokens(fdata, endstream)
endit = source.multiple(2)
if endit != 'endstream endobj'.split():
# /Length attribute is broken, try to read stream
# anyway disregarding the specified value
# TODO: issue warning here once we have some kind of
# logging
endstream = fdata.index('endstream', startstream)
if fdata[endstream-2:endstream] == '\r\n':
endstream -= 2
elif fdata[endstream-1] in ['\n', '\r']:
endstream -= 1
source = PdfTokens(fdata, endstream)
endit = source.multiple(2)
assert endit == 'endstream endobj'.split()
obj.Length = str(endstream-startstream)
obj._stream = fdata[startstream:endstream]
def readstream(obj, source):
''' Read optional stream following a dictionary
object.
'''
tok = source.next()
if tok == 'endobj':
return # No stream
assert isinstance(obj, PdfDict)
assert tok == 'stream', tok
fdata = source.fdata
floc = fdata.rindex(tok, 0, source.floc) + len(tok)
ch = fdata[floc]
if ch == '\r':
floc += 1
ch = fdata[floc]
assert ch == '\n'
startstream = floc + 1
endstream = startstream + int(obj.Length)
obj._stream = fdata[startstream:endstream]
source = PdfTokens(fdata, endstream)
endit = source.multiple(2)
if endit != 'endstream endobj'.split():
# /Length attribute is broken, try to read stream
# anyway disregarding the specified value
# TODO: issue warning here once we have some kind of
# logging
endstream = fdata.index('endstream', startstream)
if fdata[endstream - 2:endstream] == '\r\n':
endstream -= 2
elif fdata[endstream - 1] in ['\n', '\r']:
endstream -= 1
source = PdfTokens(fdata, endstream)
endit = source.multiple(2)
assert endit == 'endstream endobj'.split()
obj.Length = str(endstream - startstream)
obj._stream = fdata[startstream:endstream]