def readFromStream(stream, pdf):
idnum = b_("")
while True:
tok = stream.read(1)
if not tok:
# stream has truncated prematurely
raise PdfStreamError("Stream has ended unexpectedly")
if tok.isspace():
break
idnum += tok
generation = b_("")
while True:
tok = stream.read(1)
if not tok:
# stream has truncated prematurely
raise PdfStreamError("Stream has ended unexpectedly")
if tok.isspace():
break
generation += tok
r = stream.read(1)
if r != b_("R"):
raise utils.PdfReadError(
"Error reading indirect object reference at byte %s" %
utils.hexStr(stream.tell()))
return IndirectObject(int(idnum), int(generation), pdf)
def readStringFromStream(stream):
tok = stream.read(1)
parens = 1
txt = b_("")
while True:
tok = stream.read(1)
if not tok:
# stream has truncated prematurely
raise PdfStreamError("Stream has ended unexpectedly")
if tok == b_("("):
parens += 1
elif tok == b_(")"):
parens -= 1
if parens == 0:
break
elif tok == b_("\\"):
tok = stream.read(1)
if tok == b_("n"):
tok = b_("\n")
elif tok == b_("r"):
tok = b_("\r")
elif tok == b_("t"):
tok = b_("\t")
elif tok == b_("b"):
tok = b_("\b")
elif tok == b_("f"):
tok = b_("\f")
elif tok == b_("("):
tok = b_("(")
elif tok == b_(")"):
tok = b_(")")
elif tok == b_("\\"):
tok = b_("\\")
elif tok.isdigit():
# "The number ddd may consist of one, two, or three
# octal digits; high-order overflow shall be ignored.
# Three octal digits shall be used, with leading zeros
# as needed, if the next character of the string is also
# a digit." (PDF reference 7.3.4.2, p 16)
for i in range(2):
ntok = stream.read(1)
if ntok.isdigit():
tok += ntok
else:
break
tok = b_(chr(int(tok, base=8)))
elif tok in b_("\n\r"):
# This case is hit when a backslash followed by a line
# break occurs. If it's a multi-char EOL, consume the
# second character:
tok = stream.read(1)
if not tok in b_("\n\r"):
stream.seek(-1, 1)
# Then don't add anything to the actual string, since this
# line break was escaped:
tok = b_('')
else:
raise utils.PdfReadError("Unexpected escaped string")
txt += tok
return createStringObject(txt)
def readFromStream(stream):
word = stream.read(4)
if word == b_("true"):
return BooleanObject(True)
elif word == b_("fals"):
stream.read(1)
return BooleanObject(False)
else:
raise utils.PdfReadError('Could not read Boolean object')
def readFromStream(stream, pdf):
debug = False
if debug: print((stream.tell()))
name = stream.read(1)
if name != NameObject.surfix:
raise utils.PdfReadError("name read error")
name += utils.readUntilRegex(stream,
NameObject.delimiterPattern,
ignore_eof=True)
if debug: print(name)
try:
return NameObject(name.decode('utf-8'))
except (UnicodeEncodeError, UnicodeDecodeError) as e:
# Name objects should represent irregular characters
# with a '#' followed by the symbol's hex number
if not pdf.strict:
warnings.warn("Illegal character in Name Object",
utils.PdfReadWarning)
return NameObject(name)
else:
raise utils.PdfReadError("Illegal character in Name Object")
def readStringFromStream(stream):
tok = stream.read(1)
parens = 1
txt = ""
while True:
tok = stream.read(1)
if tok == "(":
parens += 1
elif tok == ")":
parens -= 1
if parens == 0:
break
elif tok == "\\":
tok = stream.read(1)
if tok == "n":
tok = "\n"
elif tok == "r":
tok = "\r"
elif tok == "t":
tok = "\t"
elif tok == "b":
tok == "\b"
elif tok == "f":
tok = "\f"
elif tok == "(":
tok = "("
elif tok == ")":
tok = ")"
elif tok == "\\":
tok = "\\"
elif tok.isdigit():
tok += stream.read(2)
tok = chr(int(tok, base=8))
elif tok in "\n\r":
# This case is hit when a backslash followed by a line
# break occurs. If it's a multi-char EOL, consume the
# second character:
tok = stream.read(1)
if not tok in "\n\r":
stream.seek(-1, 1)
# Then don't add anything to the actual string, since this
# line break was escaped:
tok = ''
else:
raise utils.PdfReadError("Unexpected escaped string")
txt += tok
return createStringObject(txt)
def readFromStream(stream, pdf):
idnum = ""
while True:
tok = stream.read(1)
if tok.isspace():
break
idnum += tok
generation = ""
while True:
tok = stream.read(1)
if tok.isspace():
break
generation += tok
r = stream.read(1)
if r != "R":
raise utils.PdfReadError("error reading indirect object reference")
return IndirectObject(int(idnum), int(generation), pdf)
def readFromStream(stream, pdf):
idnum = b_("")
while True:
tok = stream.read(1)
if tok.isspace():
break
idnum += tok
generation = b_("")
while True:
tok = stream.read(1)
if tok.isspace():
break
generation += tok
r = stream.read(1)
if r != b_("R"):
raise utils.PdfReadError(
"Error reading indirect object reference at byte %s" %
utils.hexStr(stream.tell()))
return IndirectObject(int(idnum), int(generation), pdf)
def readFromStream(stream, pdf):
arr = ArrayObject()
tmp = stream.read(1)
if tmp != b_("["):
raise utils.PdfReadError("Could not read array")
while True:
# skip leading whitespace
tok = stream.read(1)
while tok.isspace():
tok = stream.read(1)
stream.seek(-1, 1)
# check for array ending
peekahead = stream.read(1)
if peekahead == b_("]"):
break
stream.seek(-1, 1)
# read and append obj
arr.append(readObject(stream, pdf))
return arr
def __init__(self, title, page, typ, *args):
DictionaryObject.__init__(self)
self[NameObject("/Title")] = title
self[NameObject("/Page")] = page
self[NameObject("/Type")] = typ
# from table 8.2 of the PDF 1.6 reference.
if typ == "/XYZ":
(self[NameObject("/Left")], self[NameObject("/Top")],
self[NameObject("/Zoom")]) = args
elif typ == "/FitR":
(self[NameObject("/Left")], self[NameObject("/Bottom")],
self[NameObject("/Right")], self[NameObject("/Top")]) = args
elif typ in ["/FitH", "FitBH"]:
self[NameObject("/Top")], = args
elif typ in ["/FitV", "FitBV"]:
self[NameObject("/Left")], = args
elif typ in ["/Fit", "FitB"]:
pass
else:
raise utils.PdfReadError("Unknown Destination Type: %r" % typ)
def _buildOutline(self, node):
dest, title, outline = None, None, None
if "/A" in node and "/Title" in node:
# Action, section 8.5 (only type GoTo supported)
title = node["/Title"]
action = node["/A"]
if action["/S"] == "/GoTo":
dest = action["/D"]
elif "/Dest" in node and "/Title" in node:
# Destination, section 8.2.1
title = node["/Title"]
dest = node["/Dest"]
# if destination found, then create outline
if dest:
if isinstance(dest, ArrayObject):
outline = self._buildDestination(title, dest)
elif isinstance(dest, unicode) and dest in self._namedDests:
outline = self._namedDests[dest]
outline[NameObject("/Title")] = title
else:
raise utils.PdfReadError("Unexpected destination %r" % dest)
return outline
def setData(self, data):
raise utils.PdfReadError(
"Creating EncodedStreamObject is not currently supported")
def readFromStream(stream, pdf):
debug = False
tmp = stream.read(2)
if tmp != b_("<<"):
raise utils.PdfReadError(
"Dictionary read error at byte %s: stream must begin with '<<'"
% utils.hexStr(stream.tell()))
data = {}
while True:
tok = readNonWhitespace(stream)
if tok == b_('\x00'):
continue
elif tok == b_('%'):
stream.seek(-1, 1)
skipOverComment(stream)
continue
if not tok:
# stream has truncated prematurely
raise PdfStreamError("Stream has ended unexpectedly")
if debug: print(("Tok:", tok))
if tok == b_(">"):
stream.read(1)
break
stream.seek(-1, 1)
key = readObject(stream, pdf)
tok = readNonWhitespace(stream)
stream.seek(-1, 1)
value = readObject(stream, pdf)
if not data.get(key):
data[key] = value
elif pdf.strict:
# multiple definitions of key not permitted
raise utils.PdfReadError("Multiple definitions in dictionary at byte %s for key %s" \
% (utils.hexStr(stream.tell()), key))
else:
warnings.warn("Multiple definitions in dictionary at byte %s for key %s" \
% (utils.hexStr(stream.tell()), key), utils.PdfReadWarning)
pos = stream.tell()
s = readNonWhitespace(stream)
if s == b_('s') and stream.read(5) == b_('tream'):
eol = stream.read(1)
# odd PDF file output has spaces after 'stream' keyword but before EOL.
# patch provided by Danial Sandler
while eol == b_(' '):
eol = stream.read(1)
assert eol in (b_("\n"), b_("\r"))
if eol == b_("\r"):
# read \n after
if stream.read(1) != b_('\n'):
stream.seek(-1, 1)
# this is a stream object, not a dictionary
assert "/Length" in data
length = data["/Length"]
if debug: print(data)
if isinstance(length, IndirectObject):
t = stream.tell()
length = pdf.getObject(length)
stream.seek(t, 0)
data["__streamdata__"] = stream.read(length)
if debug: print("here")
#if debug: print(binascii.hexlify(data["__streamdata__"]))
e = readNonWhitespace(stream)
ndstream = stream.read(8)
if (e + ndstream) != b_("endstream"):
# (sigh) - the odd PDF file has a length that is too long, so
# we need to read backwards to find the "endstream" ending.
# ReportLab (unknown version) generates files with this bug,
# and Python users into PDF files tend to be our audience.
# we need to do this to correct the streamdata and chop off
# an extra character.
pos = stream.tell()
stream.seek(-10, 1)
end = stream.read(9)
if end == b_("endstream"):
# we found it by looking back one character further.
data["__streamdata__"] = data["__streamdata__"][:-1]
else:
if debug: print(("E", e, ndstream, debugging.toHex(end)))
stream.seek(pos, 0)
raise utils.PdfReadError(
"Unable to find 'endstream' marker after stream at byte %s."
% utils.hexStr(stream.tell()))
else:
stream.seek(pos, 0)
if "__streamdata__" in data:
return StreamObject.initializeFromDictionary(data)
else:
retval = DictionaryObject()
retval.update(data)
return retval
def readFromStream(stream):
nulltxt = stream.read(4)
if nulltxt != b_("null"):
raise utils.PdfReadError("Could not read Null object")
return NullObject()
def convertToInt(d, size):
if size > 8:
raise utils.PdfReadError("invalid size in convertToInt")
d = "\x00\x00\x00\x00\x00\x00\x00\x00" + d
d = d[-8:]
return struct.unpack(">q", d)[0]
def getObject(self, indirectReference):
retval = self.resolvedObjects.get(indirectReference.generation,
{}).get(indirectReference.idnum,
None)
if retval is not None:
return retval
if indirectReference.generation == 0 \
and indirectReference.idnum in self.xref_objStm:
# indirect reference to object in object stream
# read the entire object stream into memory
stmnum, idx = self.xref_objStm[indirectReference.idnum]
objStm = IndirectObject(stmnum, 0, self).getObject()
assert objStm['/Type'] == '/ObjStm'
assert idx < objStm['/N']
streamData = StringIO(objStm.getData())
for i in range(objStm['/N']):
objnum = NumberObject.readFromStream(streamData)
readNonWhitespace(streamData)
streamData.seek(-1, 1)
offset = NumberObject.readFromStream(streamData)
readNonWhitespace(streamData)
streamData.seek(-1, 1)
t = streamData.tell()
streamData.seek(objStm['/First'] + offset, 0)
obj = readObject(streamData, self)
self.resolvedObjects[0][objnum] = obj
streamData.seek(t, 0)
return self.resolvedObjects[0][indirectReference.idnum]
if indirectReference.idnum \
not in self.xref[indirectReference.generation]:
warnings.warn(
"Object %d %d not defined." %
(indirectReference.idnum, indirectReference.generation),
utils.PdfReadWarning)
return None
start = self.xref[indirectReference.generation][
indirectReference.idnum]
self.stream.seek(start, 0)
idnum, generation = self.readObjectHeader(self.stream)
try:
assert idnum == indirectReference.idnum
except AssertionError:
if self.xrefIndex:
# Xref table probably had bad indexes due to not
# being zero-indexed
if self.strict:
raise utils.PdfReadError(
"Expected object ID (%d %d) does "
"not match actual (%d %d); xref "
"table not zero-indexed." %
(indirectReference.idnum, indirectReference.generation,
idnum, generation))
else:
# should not happen since the xref table is corrected in
# non-strict mode
pass
else: # some other problem
raise utils.PdfReadError(
"Expected object ID (%d %d) does not "
" match actual (%d %d)." %
(indirectReference.idnum, indirectReference.generation,
idnum, generation))
assert generation == indirectReference.generation
retval = readObject(self.stream, self)
# override encryption is used for the /Encrypt dictionary
if not self._override_encryption and self.isEncrypted:
# if we don't have the encryption key:
if not hasattr(self, '_decryption_key'):
raise Exception("file has not been decrypted")
# otherwise, decrypt here...
pack1 = struct.pack("<i", indirectReference.idnum)[:3]
pack2 = struct.pack("<i", indirectReference.generation)[:2]
key = self._decryption_key + pack1 + pack2
assert len(key) == (len(self._decryption_key) + 5)
md5_hash = md5(key).digest()
key = md5_hash[:min(16, len(self._decryption_key) + 5)]
retval = self._decryptObject(retval, key)
self.cacheIndirectObject(generation, idnum, retval)
return retval
