def write(self, stream, prev, startdata):
stream.write(pdf.b_("\r\n"))
positions = {2: 0}
for i in range(2, len(self._objects)):
idnum = i + 1
obj = self._objects[i]
if obj is None:
positions[idnum] = 0
continue
positions[idnum] = startdata + stream.tell()
stream.write(pdf.b_(str(idnum) + " 0 obj\n"))
key = None
if self._encrypt_key is not None:
pack1 = struct.pack("<i", i + 1)[:3]
pack2 = struct.pack("<i", 0)[:2]
key = self._encrypt_key + pack1 + pack2
assert len(key) == (len(self._encrypt_key) + 5)
md5_hash = hashlib.md5(key).digest()
key = md5_hash[:min(16, len(self._encrypt_key) + 5)]
obj.writeToStream(stream, key)
stream.write(pdf.b_("\nendobj\n"))
xref_location = startdata + stream.tell()
if not prev.xrefstream:
trailer = po.DictionaryObject()
else:
trailer = po.StreamObject()
self._addObject(trailer)
# xref table
trailer.update({
po.NameObject("/Size"):
po.NumberObject(len(self._objects) + 1),
po.NameObject("/Root"):
self.x_root,
po.NameObject("/Info"):
self.x_info,
po.NameObject("/Prev"):
po.NumberObject(prev.startxref),
po.NameObject("/ID"):
self._ID,
})
if prev.isEncrypted:
trailer[po.NameObject("/Encrypt")] = prev.trailer.raw_get(
"/Encrypt")
if not prev.xrefstream:
stream.write(pdf.b_("xref\n"))
stream.write(pdf.b_("0 1\n"))
stream.write(pdf.b_("0000000000 65535 f \n"))
keys = sorted(positions.keys())
i = 0
while i < len(keys):
off = positions[keys[i]]
if off == 0:
while i < len(keys) and positions[keys[i]] == 0:
i += 1
start = i
while i < len(keys) and positions[keys[i]] != 0:
i += 1
stream.write(pdf.b_("%d %d \n" % (keys[start], i - start)))
i = start
continue
else:
stream.write(pdf.b_("%010d %05d n \n" % (off, 0)))
i += 1
# trailer
stream.write(pdf.b_("trailer\n"))
trailer.writeToStream(stream, None)
else:
def pack(offset):
return struct.pack(">q", offset)
dataindex = ["0 1"]
dataxref = [b"\x00" + pack(0)]
keys = sorted(positions.keys())
i = 0
while i < len(keys):
off = positions[keys[i]]
if off != 0:
start = i
while i < len(keys) and positions[keys[i]] != 0:
dataxref.append(b"\x01" + pack(positions[keys[i]]))
i += 1
stop = i
dataindex.append("%d %d" % (keys[start], stop - start))
else:
i += 1
dataindex = " ".join(dataindex)
dataxref = b"".join(dataxref)
trailer[po.NameObject("/Type")] = po.NameObject("/XRef")
trailer[po.NameObject("/W")] = po.NameObject("[1 8 0]")
trailer[po.NameObject("/Index")] = po.NameObject("[%s]" %
dataindex)
trailer._data = dataxref
retval = trailer.flateEncode()
trailer.update(retval)
trailer._data = retval._data
stream.write(pdf.b_("%d 0 obj\n" % (len(self._objects))))
trailer.writeToStream(stream, None)
stream.write(pdf.b_("\nendobj"))
# eof
stream.write(pdf.b_("\nstartxref\n%s\n%%%%EOF\n" % (xref_location)))
def write(self, stream, prev, startxref):
externalReferenceMap = {}
# PDF objects sometimes have circular references to their /Page objects
# inside their object tree (for example, annotations). Those will be
# indirect references to objects that we've recreated in this PDF. To
# address this problem, PageObject's store their original object
# reference number, and we add it to the external reference map before
# we sweep for indirect references. This forces self-page-referencing
# trees to reference the correct new object location, rather than
# copying in a new copy of the page object.
for objIndex in range(len(self._objects)):
obj = self._objects[objIndex]
if isinstance(obj, pdf.PageObject) and obj.indirectRef != None:
data = obj.indirectRef
if data.pdf not in externalReferenceMap:
externalReferenceMap[data.pdf] = {}
if data.generation not in externalReferenceMap[data.pdf]:
externalReferenceMap[data.pdf][data.generation] = {}
externalReferenceMap[data.pdf][data.generation][
data.idnum] = po.IndirectObject(objIndex + 1, 0, self)
self.stack = []
self._sweepIndirectReferences(externalReferenceMap, self._root)
del self.stack
# Begin writing:
positions = {}
for i in range(2, len(self._objects)):
idnum = i + 1
obj = self._objects[i]
if obj is None:
continue
positions[idnum] = startxref + stream.tell()
stream.write(pdf.b_(str(idnum) + " 0 obj\n"))
key = None
if hasattr(self, "_encrypt") and idnum != self._encrypt.idnum:
pack1 = struct.pack("<i", i + 1)[:3]
pack2 = struct.pack("<i", 0)[:2]
key = self._encrypt_key + pack1 + pack2
assert len(key) == (len(self._encrypt_key) + 5)
md5_hash = hashlib.md5(key).digest()
key = md5_hash[:min(16, len(self._encrypt_key) + 5)]
obj.writeToStream(stream, key)
stream.write(pdf.b_("\nendobj\n"))
# xref table
xref_location = startxref + stream.tell()
stream.write(pdf.b_("xref\n"))
stream.write(pdf.b_("0 1\n"))
stream.write(pdf.b_("0000000000 65535 f\n"))
stream.write(pdf.b_("9 %d\n" % (len(positions))))
for key in sorted(positions.keys()):
stream.write(pdf.b_("%010d %05d n\n" % (positions[key], 0)))
# trailer
stream.write(pdf.b_("trailer\n"))
trailer = po.DictionaryObject()
trailer.update({
po.NameObject("/Size"):
po.NumberObject(len(self._objects) + 1),
po.NameObject("/Root"):
self.x_root,
po.NameObject("/Info"):
self.x_info,
po.NameObject("/Prev"):
po.NumberObject(prev.startxref),
})
if hasattr(self, "_ID"):
trailer[po.NameObject("/ID")] = self._ID
if hasattr(self, "_encrypt"):
trailer[po.NameObject("/Encrypt")] = self._encrypt
trailer.writeToStream(stream, None)
# eof
stream.write(pdf.b_("\nstartxref\n%s\n%%%%EOF\n" % (xref_location)))