# vim: sw=4:expandtab:foldmethod=marker
#
# Copyright (c) 2006, Mathieu Fenniak
# Copyright (c) 2007, Ashish Kulkarni <kulkarni.ashish@gmail.com>
#
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
# * Redistributions of source code must retain the above copyright notice,
# this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright notice,
# this list of conditions and the following disclaimer in the documentation
# and/or other materials provided with the distribution.
# * The name of the author may not be used to endorse or promote products
# derived from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
# ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
# LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
# CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
# SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
# INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
# CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
# ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
# POSSIBILITY OF SUCH DAMAGE.


"""
A pure-Python PDF library with very minimal capabilities.  It was designed to
be able to split and merge PDF files by page, and that's about all it can do.
It may be a solid base for future PDF file work in Python.
"""
__author__ = "Mathieu Fenniak"
__author_email__ = "biziqe@mathieu.fenniak.net"

import struct
try:
    from cStringIO import StringIO
except ImportError:
    from StringIO import StringIO

import filters
import utils
import warnings
from generic import *
from utils import readNonWhitespace, readUntilWhitespace, ConvertFunctionsToVirtualList
from sets import ImmutableSet

##
# This class supports writing PDF files out, given pages produced by another
# class (typically {@link #PdfFileReader PdfFileReader}).
class PdfFileWriter(object):
    def __init__(self):
        self._header = "%PDF-1.3"
        self._objects = []  # array of indirect objects

        # The root of our page tree node.
        pages = DictionaryObject()
        pages.update({
                NameObject("/Type"): NameObject("/Pages"),
                NameObject("/Count"): NumberObject(0),
                NameObject("/Kids"): ArrayObject(),
                })
        self._pages = self._addObject(pages)

        # info object
        info = DictionaryObject()
        info.update({
                NameObject("/Producer"): createStringObject(u"Python PDF Library - http://pybrary.net/pyPdf/")
                })
        self._info = self._addObject(info)

        # root object
        root = DictionaryObject()
        root.update({
            NameObject("/Type"): NameObject("/Catalog"),
            NameObject("/Pages"): self._pages,
            })
        self._root = self._addObject(root)

    def _addObject(self, obj):
        self._objects.append(obj)
        return IndirectObject(len(self._objects), 0, self)

    def getObject(self, ido):
        if ido.pdf != self:
            raise ValueError("pdf must be self")
        return self._objects[ido.idnum - 1]

    ##
    # Adds a page to this PDF file.  The page is usually acquired from a
    # {@link #PdfFileReader PdfFileReader} instance.
    # <p>
    # Stability: Added in v1.0, will exist for all v1.x releases.
    #
    # @param page The page to add to the document.  This argument should be
    #             an instance of {@link #PageObject PageObject}.
    def addPage(self, page):
        assert page["/Type"] == "/Page"
        page[NameObject("/Parent")] = self._pages
        page = self._addObject(page)
        pages = self.getObject(self._pages)
        pages["/Kids"].append(page)
        pages[NameObject("/Count")] = NumberObject(pages["/Count"] + 1)

    ##
    # Encrypt this PDF file with the PDF Standard encryption handler.
    # @param user_pwd The "user password", which allows for opening and reading
    # the PDF file with the restrictions provided.
    # @param owner_pwd The "owner password", which allows for opening the PDF
    # files without any restrictions.  By default, the owner password is the
    # same as the user password.
    # @param use_128bit Boolean argument as to whether to use 128bit
    # encryption.  When false, 40bit encryption will be used.  By default, this
    # flag is on.
    def encrypt(self, user_pwd, owner_pwd = None, use_128bit = True):
        import md5, time, random
        if owner_pwd == None:
            owner_pwd = user_pwd
        if use_128bit:
            V = 2
            rev = 3
            keylen = 128 / 8
        else:
            V = 1
            rev = 2
            keylen = 40 / 8
        # permit everything:
        P = -1
        O = ByteStringObject(_alg33(owner_pwd, user_pwd, rev, keylen))
        ID_1 = md5.new(repr(time.time())).digest()
        ID_2 = md5.new(repr(random.random())).digest()
        self._ID = ArrayObject((ByteStringObject(ID_1), ByteStringObject(ID_2)))
        if rev == 2:
            U, key = _alg34(user_pwd, O, P, ID_1)
        else:
            assert rev == 3
            U, key = _alg35(user_pwd, rev, keylen, O, P, ID_1, False)
        encrypt = DictionaryObject()
        encrypt[NameObject("/Filter")] = NameObject("/Standard")
        encrypt[NameObject("/V")] = NumberObject(V)
        if V == 2:
            encrypt[NameObject("/Length")] = NumberObject(keylen * 8)
        encrypt[NameObject("/R")] = NumberObject(rev)
        encrypt[NameObject("/O")] = ByteStringObject(O)
        encrypt[NameObject("/U")] = ByteStringObject(U)
        encrypt[NameObject("/P")] = NumberObject(P)
        self._encrypt = self._addObject(encrypt)
        self._encrypt_key = key

    ##
    # Writes the collection of pages added to this object out as a PDF file.
    # <p>
    # Stability: Added in v1.0, will exist for all v1.x releases.
    # @param stream An object to write the file to.  The object must support
    # the write method, and the tell method, similar to a file object.
    def write(self, stream):
        import struct, md5

        externalReferenceMap = {}
        self.stack = []
        self._sweepIndirectReferences(externalReferenceMap, self._root)
        del self.stack

        # Begin writing:
        object_positions = []
        stream.write(self._header + "\n")
        for i in range(len(self._objects)):
            idnum = (i + 1)
            obj = self._objects[i]
            object_positions.append(stream.tell())
            stream.write(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 = md5.new(key).digest()
                key = md5_hash[:min(16, len(self._encrypt_key) + 5)]
            obj.writeToStream(stream, key)
            stream.write("\nendobj\n")

        # xref table
        xref_location = stream.tell()
        stream.write("xref\n")
        stream.write("0 %s\n" % (len(self._objects) + 1))
        stream.write("%010d %05d f \n" % (0, 65535))
        for offset in object_positions:
            stream.write("%010d %05d n \n" % (offset, 0))

        # trailer
        stream.write("trailer\n")
        trailer = DictionaryObject()
        trailer.update({
                NameObject("/Size"): NumberObject(len(self._objects) + 1),
                NameObject("/Root"): self._root,
                NameObject("/Info"): self._info,
                })
        if hasattr(self, "_ID"):
            trailer[NameObject("/ID")] = self._ID
        if hasattr(self, "_encrypt"):
            trailer[NameObject("/Encrypt")] = self._encrypt
        trailer.writeToStream(stream, None)
        
        # eof
        stream.write("\nstartxref\n%s\n%%%%EOF\n" % (xref_location))

    def _sweepIndirectReferences(self, externMap, data):
        if isinstance(data, DictionaryObject):
            for key, value in data.items():
                origvalue = value
                value = self._sweepIndirectReferences(externMap, value)
                if isinstance(value, StreamObject):
                    # a dictionary value is a stream.  streams must be indirect
                    # objects, so we need to change this value.
                    value = self._addObject(value)
                data[key] = value
            return data
        elif isinstance(data, ArrayObject):
            for i in range(len(data)):
                value = self._sweepIndirectReferences(externMap, data[i])
                if isinstance(value, StreamObject):
                    # an array value is a stream.  streams must be indirect
                    # objects, so we need to change this value
                    value = self._addObject(value)
                data[i] = value
            return data
        elif isinstance(data, IndirectObject):
            # internal indirect references are fine
            if data.pdf == self:
                if data.idnum in self.stack:
                    return data
                else:
                    self.stack.append(data.idnum)
                    realdata = self.getObject(data)
                    self._sweepIndirectReferences(externMap, realdata)
                    self.stack.pop()
                    return data
            else:
                newobj = externMap.get(data.pdf, {}).get(data.generation, {}).get(data.idnum, None)
                if newobj == None:
                    newobj = data.pdf.getObject(data)
                    self._objects.append(None) # placeholder
                    idnum = len(self._objects)
                    newobj_ido = IndirectObject(idnum, 0, self)
                    if not externMap.has_key(data.pdf):
                        externMap[data.pdf] = {}
                    if not externMap[data.pdf].has_key(data.generation):
                        externMap[data.pdf][data.generation] = {}
                    externMap[data.pdf][data.generation][data.idnum] = newobj_ido
                    newobj = self._sweepIndirectReferences(externMap, newobj)
                    self._objects[idnum-1] = newobj
                    return newobj_ido
                return newobj
        else:
            return data


##
# Initializes a PdfFileReader object.  This operation can take some time, as
# the PDF stream's cross-reference tables are read into memory.
# <p>
# Stability: Added in v1.0, will exist for all v1.x releases.
#
# @param stream An object that supports the standard read and seek methods
#               similar to a file object.
class PdfFileReader(object):
    def __init__(self, stream):
        self.flattenedPages = None
        self.resolvedObjects = {}
        self.read(stream)
        self.stream = stream
        self._override_encryption = False

    ##
    # Retrieves the PDF file's document information dictionary, if it exists.
    # Note that some PDF files use metadata streams instead of docinfo
    # dictionaries, and these metadata streams will not be accessed by this
    # function.
    # <p>
    # Stability: Added in v1.6, will exist for all future v1.x releases.
    # @return Returns a {@link #DocumentInformation DocumentInformation}
    #         instance, or None if none exists.
    def getDocumentInfo(self):
        if not self.trailer.has_key("/Info"):
            return None
        obj = self.trailer['/Info']
        retval = DocumentInformation()
        retval.update(obj)
        return retval

    ##
    # Read-only property that accesses the {@link
    # #PdfFileReader.getDocumentInfo getDocumentInfo} function.
    # <p>
    # Stability: Added in v1.7, will exist for all future v1.x releases.
    documentInfo = property(lambda self: self.getDocumentInfo(), None, None)

    ##
    # Retrieves XMP (Extensible Metadata Platform) data from the PDF document
    # root.
    # <p>
    # Stability: Added in v1.12, will exist for all future v1.x releases.
    # @return Returns a {@link #generic.XmpInformation XmlInformation}
    # instance that can be used to access XMP metadata from the document.
    # Can also return None if no metadata was found on the document root.
    def getXmpMetadata(self):
        try:
            self._override_encryption = True
            return self.trailer["/Root"].getXmpMetadata()
        finally:
            self._override_encryption = False

    ##
    # Read-only property that accesses the {@link #PdfFileReader.getXmpData
    # getXmpData} function.
    # <p>
    # Stability: Added in v1.12, will exist for all future v1.x releases.
    xmpMetadata = property(lambda self: self.getXmpMetadata(), None, None)

    ##
    # Calculates the number of pages in this PDF file.
    # <p>
    # Stability: Added in v1.0, will exist for all v1.x releases.
    # @return Returns an integer.
    def getNumPages(self):
        if self.flattenedPages == None:
            self._flatten()
        return len(self.flattenedPages)

    ##
    # Read-only property that accesses the {@link #PdfFileReader.getNumPages
    # getNumPages} function.
    # <p>
    # Stability: Added in v1.7, will exist for all future v1.x releases.
    numPages = property(lambda self: self.getNumPages(), None, None)

    ##
    # Retrieves a page by number from this PDF file.
    # <p>
    # Stability: Added in v1.0, will exist for all v1.x releases.
    # @return Returns a {@link #PageObject PageObject} instance.
    def getPage(self, pageNumber):
        ## ensure that we're not trying to access an encrypted PDF
        #assert not self.trailer.has_key("/Encrypt")
        if self.flattenedPages == None:
            self._flatten()
        return self.flattenedPages[pageNumber]

    ##
    # Read-only property that accesses the 
    # {@link #PdfFileReader.getNamedDestinations 
    # getNamedDestinations} function.
    # <p>
    # Stability: Added in v1.10, will exist for all future v1.x releases.
    namedDestinations = property(lambda self:
                                  self.getNamedDestinations(), None, None)

    ##
    # Retrieves the named destinations present in the document.
    # <p>
    # Stability: Added in v1.10, will exist for all future v1.x releases.
    # @return Returns a dict which maps names to {@link #Destination
    # destinations}.
    def getNamedDestinations(self, tree=None, retval=None):
        if retval == None:
            retval = {}
            catalog = self.trailer["/Root"]
            
            # get the name tree
            if catalog.has_key("/Dests"):
                tree = catalog["/Dests"]
            elif catalog.has_key("/Names"):
                names = catalog['/Names']
                if names.has_key("/Dests"):
                    tree = names['/Dests']
        
        if tree == None:
            return retval

        if tree.has_key("/Kids"):
            # recurse down the tree
            for kid in tree["/Kids"]:
                self.getNamedDestinations(kid.getObject(), retval)

        if tree.has_key("/Names"):
            names = tree["/Names"]
            for i in range(0, len(names), 2):
                key = names[i].getObject()
                val = names[i+1].getObject()
                if isinstance(val, DictionaryObject) and val.has_key('/D'):
                    val = val['/D']
                dest = self._buildDestination(key, val)
                if dest != None:
                    retval[key] = dest

        return retval

    ##
    # Read-only property that accesses the {@link #PdfFileReader.getOutlines
    # getOutlines} function.
    # <p>
    # Stability: Added in v1.10, will exist for all future v1.x releases.
    outlines = property(lambda self: self.getOutlines(), None, None)

    ##
    # Retrieves the document outline present in the document.
    # <p>
    # Stability: Added in v1.10, will exist for all future v1.x releases.
    # @return Returns a nested list of {@link #Destination destinations}.
    def getOutlines(self, node=None, outlines=None):
        if outlines == None:
            outlines = []
            catalog = self.trailer["/Root"]
            
            # get the outline dictionary and named destinations
            if catalog.has_key("/Outlines"):
                lines = catalog["/Outlines"]
                if lines.has_key("/First"):
                    node = lines["/First"]
            self._namedDests = self.getNamedDestinations()
            
        if node == None:
          return outlines
          
        # see if there are any more outlines
        while 1:
            outline = self._buildOutline(node)
            if outline:
                outlines.append(outline)

            # check for sub-outlines
            if node.has_key("/First"):
                subOutlines = []
                self.getOutlines(node["/First"], subOutlines)
                if subOutlines:
                    outlines.append(subOutlines)

            if not node.has_key("/Next"):
                break
            node = node["/Next"]

        return outlines

    def _buildDestination(self, title, array):
        page, typ = array[0:2]
        array = array[2:]
        return Destination(title, page, typ, *array)
          
    def _buildOutline(self, node):
        dest, title, outline = None, None, None
        
        if node.has_key("/A") and node.has_key("/Title"):
            # Action, section 8.5 (only type GoTo supported)
            title  = node["/Title"]
            action = node["/A"]
            if action["/S"] == "/GoTo":
                dest = action["/D"]
        elif node.has_key("/Dest") and node.has_key("/Title"):
            # 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 self._namedDests.has_key(dest):
                outline = self._namedDests[dest]
                outline[NameObject("/Title")] = title
            else:
                raise utils.PdfReadError("Unexpected destination %r" % dest)
        return outline

    ##
    # Read-only property that emulates a list based upon the {@link
    # #PdfFileReader.getNumPages getNumPages} and {@link #PdfFileReader.getPage
    # getPage} functions.
    # <p>
    # Stability: Added in v1.7, and will exist for all future v1.x releases.
    pages = property(lambda self: ConvertFunctionsToVirtualList(self.getNumPages, self.getPage),
            None, None)

    def _flatten(self, pages=None, inherit=None):
        inheritablePageAttributes = (
            NameObject("/Resources"), NameObject("/MediaBox"),
            NameObject("/CropBox"), NameObject("/Rotate")
            )
        if inherit == None:
            inherit = dict()
        if pages == None:
            self.flattenedPages = []
            catalog = self.trailer["/Root"].getObject()
            pages = catalog["/Pages"].getObject()
        t = pages["/Type"]
        if t == "/Pages":
            for attr in inheritablePageAttributes:
                if pages.has_key(attr):
                    inherit[attr] = pages[attr]
            for page in pages["/Kids"]:
                self._flatten(page.getObject(), inherit)
        elif t == "/Page":
            for attr,value in inherit.items():
                # if the page has it's own value, it does not inherit the
                # parent's value:
                if not pages.has_key(attr):
                    pages[attr] = value
            pageObj = PageObject(self)
            pageObj.update(pages)
            self.flattenedPages.append(pageObj)

    def getObject(self, indirectReference):
        retval = self.resolvedObjects.get(indirectReference.generation, {}).get(indirectReference.idnum, None)
        if retval != None:
            return retval
        if indirectReference.generation == 0 and \
           self.xref_objStm.has_key(indirectReference.idnum):
            # 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]
        start = self.xref[indirectReference.generation][indirectReference.idnum]
        self.stream.seek(start, 0)
        idnum, generation = self.readObjectHeader(self.stream)
        assert idnum == indirectReference.idnum
        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...
            import struct, md5
            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.new(key).digest()
            key = md5_hash[:min(16, len(self._decryption_key) + 5)]
            retval = self._decryptObject(retval, key)

        self.cacheIndirectObject(generation, idnum, retval)
        return retval

    def _decryptObject(self, obj, key):
        if isinstance(obj, ByteStringObject) or isinstance(obj, TextStringObject):
            obj = createStringObject(utils.RC4_encrypt(key, obj.original_bytes))
        elif isinstance(obj, StreamObject):
            obj._data = utils.RC4_encrypt(key, obj._data)
        elif isinstance(obj, DictionaryObject):
            for dictkey, value in obj.items():
                obj[dictkey] = self._decryptObject(value, key)
        elif isinstance(obj, ArrayObject):
            for i in range(len(obj)):
                obj[i] = self._decryptObject(obj[i], key)
        return obj

    def readObjectHeader(self, stream):
        # Should never be necessary to read out whitespace, since the
        # cross-reference table should put us in the right spot to read the
        # object header.  In reality... some files have stupid cross reference
        # tables that are off by whitespace bytes.
        readNonWhitespace(stream); stream.seek(-1, 1)
        idnum = readUntilWhitespace(stream)
        generation = readUntilWhitespace(stream)
        obj = stream.read(3)
        readNonWhitespace(stream)
        stream.seek(-1, 1)
        return int(idnum), int(generation)

    def cacheIndirectObject(self, generation, idnum, obj):
        if not self.resolvedObjects.has_key(generation):
            self.resolvedObjects[generation] = {}
        self.resolvedObjects[generation][idnum] = obj

    def read(self, stream):
        # start at the end:
        stream.seek(-1, 2)
        line = ''
        while not line:
            line = self.readNextEndLine(stream)
        if line[:5] != "%%EOF":
            raise utils.PdfReadError, "EOF marker not found"

        # find startxref entry - the location of the xref table
        line = self.readNextEndLine(stream)
        startxref = int(line)
        line = self.readNextEndLine(stream)
        if line[:9] != "startxref":
            raise utils.PdfReadError, "startxref not found"

        # read all cross reference tables and their trailers
        self.xref = {}
        self.xref_objStm = {}
        self.trailer = DictionaryObject()
        while 1:
            # load the xref table
            stream.seek(startxref, 0)
            x = stream.read(1)
            if x == "x":
                # standard cross-reference table
                ref = stream.read(4)
                if ref[:3] != "ref":
                    raise utils.PdfReadError, "xref table read error"
                readNonWhitespace(stream)
                stream.seek(-1, 1)
                while 1:
                    num = readObject(stream, self)
                    readNonWhitespace(stream)
                    stream.seek(-1, 1)
                    size = readObject(stream, self)
                    readNonWhitespace(stream)
                    stream.seek(-1, 1)
                    cnt = 0
                    while cnt < size:
                        line = stream.read(20)
                        # It's very clear in section 3.4.3 of the PDF spec
                        # that all cross-reference table lines are a fixed
                        # 20 bytes.  However... some malformed PDF files
                        # use a single character EOL without a preceeding
                        # space.  Detect that case, and seek the stream
                        # back one character.  (0-9 means we've bled into
                        # the next xref entry, t means we've bled into the
                        # text "trailer"):
                        if line[-1] in "0123456789t":
                            stream.seek(-1, 1)
                        offset, generation = line[:16].split(" ")
                        offset, generation = int(offset), int(generation)
                        if not self.xref.has_key(generation):
                            self.xref[generation] = {}
                        if self.xref[generation].has_key(num):
                            # It really seems like we should allow the last
                            # xref table in the file to override previous
                            # ones. Since we read the file backwards, assume
                            # any existing key is already set correctly.
                            pass
                        else:
                            self.xref[generation][num] = offset
                        cnt += 1
                        num += 1
                    readNonWhitespace(stream)
                    stream.seek(-1, 1)
                    trailertag = stream.read(7)
                    if trailertag != "trailer":
                        # more xrefs!
                        stream.seek(-7, 1)
                    else:
                        break
                readNonWhitespace(stream)
                stream.seek(-1, 1)
                newTrailer = readObject(stream, self)
                for key, value in newTrailer.items():
                    if not self.trailer.has_key(key):
                        self.trailer[key] = value
                if newTrailer.has_key("/Prev"):
                    startxref = newTrailer["/Prev"]
                else:
                    break
            elif x.isdigit():
                # PDF 1.5+ Cross-Reference Stream
                stream.seek(-1, 1)
                idnum, generation = self.readObjectHeader(stream)
                xrefstream = readObject(stream, self)
                assert xrefstream["/Type"] == "/XRef"
                self.cacheIndirectObject(generation, idnum, xrefstream)
                streamData = StringIO(xrefstream.getData())
                idx_pairs = xrefstream.get("/Index", [0, xrefstream.get("/Size")])
                entrySizes = xrefstream.get("/W")
                for num, size in self._pairs(idx_pairs):
                    cnt = 0
                    while cnt < size:
                        for i in range(len(entrySizes)):
                            d = streamData.read(entrySizes[i])
                            di = convertToInt(d, entrySizes[i])
                            if i == 0:
                                xref_type = di
                            elif i == 1:
                                if xref_type == 0:
                                    next_free_object = di
                                elif xref_type == 1:
                                    byte_offset = di
                                elif xref_type == 2:
                                    objstr_num = di
                            elif i == 2:
                                if xref_type == 0:
                                    next_generation = di
                                elif xref_type == 1:
                                    generation = di
                                elif xref_type == 2:
                                    obstr_idx = di
                        if xref_type == 0:
                            pass
                        elif xref_type == 1:
                            if not self.xref.has_key(generation):
                                self.xref[generation] = {}
                            if not num in self.xref[generation]:
                                self.xref[generation][num] = byte_offset
                        elif xref_type == 2:
                            if not num in self.xref_objStm:
                                self.xref_objStm[num] = [objstr_num, obstr_idx]
                        cnt += 1
                        num += 1
                trailerKeys = "/Root", "/Encrypt", "/Info", "/ID"
                for key in trailerKeys:
                    if xrefstream.has_key(key) and not self.trailer.has_key(key):
                        self.trailer[NameObject(key)] = xrefstream.raw_get(key)
                if xrefstream.has_key("/Prev"):
                    startxref = xrefstream["/Prev"]
                else:
                    break
            else:
                # bad xref character at startxref.  Let's see if we can find
                # the xref table nearby, as we've observed this error with an
                # off-by-one before.
                stream.seek(-11, 1)
                tmp = stream.read(20)
                xref_loc = tmp.find("xref")
                if xref_loc != -1:
                    startxref -= (10 - xref_loc)
                    continue
                else:
                    # no xref table found at specified location
                    assert False
                    break

    def _pairs(self, array):
        i = 0
        while True:
            yield array[i], array[i+1]
            i += 2
            if (i+1) >= len(array):
                break

    def readNextEndLine(self, stream):
        line = ""
        while True:
            x = stream.read(1)
            stream.seek(-2, 1)
            if x == '\n' or x == '\r':
                while x == '\n' or x == '\r':
                    x = stream.read(1)
                    stream.seek(-2, 1)
                stream.seek(1, 1)
                break
            else:
                line = x + line
        return line

    ##
    # When using an encrypted / secured PDF file with the PDF Standard
    # encryption handler, this function will allow the file to be decrypted.
    # It checks the given password against the document's user password and
    # owner password, and then stores the resulting decryption key if either
    # password is correct.
    # <p>
    # It does not matter which password was matched.  Both passwords provide
    # the correct decryption key that will allow the document to be used with
    # this library.
    # <p>
    # Stability: Added in v1.8, will exist for all future v1.x releases.
    #
    # @return 0 if the password failed, 1 if the password matched the user
    # password, and 2 if the password matched the owner password.
    #
    # @exception NotImplementedError Document uses an unsupported encryption
    # method.
    def decrypt(self, password):
        self._override_encryption = True
        try:
            return self._decrypt(password)
        finally:
            self._override_encryption = False

    def _decrypt(self, password):
        encrypt = self.trailer['/Encrypt'].getObject()
        if encrypt['/Filter'] != '/Standard':
            raise NotImplementedError, "only Standard PDF encryption handler is available"
        if not (encrypt['/V'] in (1, 2)):
            raise NotImplementedError, "only algorithm code 1 and 2 are supported"
        user_password, key = self._authenticateUserPassword(password)
        if user_password:
            self._decryption_key = key
            return 1
        else:
            rev = encrypt['/R'].getObject()
            if rev == 2:
                keylen = 5
            else:
                keylen = encrypt['/Length'].getObject() / 8
            key = _alg33_1(password, rev, keylen)
            real_O = encrypt["/O"].getObject()
            if rev == 2:
                userpass = utils.RC4_encrypt(key, real_O)
            else:
                val = real_O
                for i in range(19, -1, -1):
                    new_key = ''
                    for l in range(len(key)):
                        new_key += chr(ord(key[l]) ^ i)
                    val = utils.RC4_encrypt(new_key, val)
                userpass = val
            owner_password, key = self._authenticateUserPassword(userpass)
            if owner_password:
                self._decryption_key = key
                return 2
        return 0

    def _authenticateUserPassword(self, password):
        encrypt = self.trailer['/Encrypt'].getObject()
        rev = encrypt['/R'].getObject()
        owner_entry = encrypt['/O'].getObject().original_bytes
        p_entry = encrypt['/P'].getObject()
        id_entry = self.trailer['/ID'].getObject()
        id1_entry = id_entry[0].getObject()
        if rev == 2:
            U, key = _alg34(password, owner_entry, p_entry, id1_entry)
        elif rev >= 3:
            U, key = _alg35(password, rev,
                    encrypt["/Length"].getObject() / 8, owner_entry,
                    p_entry, id1_entry,
                    encrypt.get("/EncryptMetadata", BooleanObject(False)).getObject())
        real_U = encrypt['/U'].getObject().original_bytes
        return U == real_U, key

    def getIsEncrypted(self):
        return self.trailer.has_key("/Encrypt")

    ##
    # Read-only boolean property showing whether this PDF file is encrypted.
    # Note that this property, if true, will remain true even after the {@link
    # #PdfFileReader.decrypt decrypt} function is called.
    isEncrypted = property(lambda self: self.getIsEncrypted(), None, None)


def getRectangle(self, name, defaults):
    retval = self.get(name)
    if isinstance(retval, RectangleObject):
        return retval
    if retval == None:
        for d in defaults:
            retval = self.get(d)
            if retval != None:
                break
    if isinstance(retval, IndirectObject):
        retval = self.pdf.getObject(retval)
    retval = RectangleObject(retval)
    setRectangle(self, name, retval)
    return retval

def setRectangle(self, name, value):
    if not isinstance(name, NameObject):
        name = NameObject(name)
    self[name] = value

def deleteRectangle(self, name):
    del self[name]

def createRectangleAccessor(name, fallback):
    return \
        property(
            lambda self: getRectangle(self, name, fallback),
            lambda self, value: setRectangle(self, name, value),
            lambda self: deleteRectangle(self, name)
            )

##
# This class represents a single page within a PDF file.  Typically this object
# will be created by accessing the {@link #PdfFileReader.getPage getPage}
# function of the {@link #PdfFileReader PdfFileReader} class.
class PageObject(DictionaryObject):
    def __init__(self, pdf):
        DictionaryObject.__init__(self)
        self.pdf = pdf

    ##
    # Rotates a page clockwise by increments of 90 degrees.
    # <p>
    # Stability: Added in v1.1, will exist for all future v1.x releases.
    # @param angle Angle to rotate the page.  Must be an increment of 90 deg.
    def rotateClockwise(self, angle):
        assert angle % 90 == 0
        self._rotate(angle)
        return self

    ##
    # Rotates a page counter-clockwise by increments of 90 degrees.
    # <p>
    # Stability: Added in v1.1, will exist for all future v1.x releases.
    # @param angle Angle to rotate the page.  Must be an increment of 90 deg.
    def rotateCounterClockwise(self, angle):
        assert angle % 90 == 0
        self._rotate(-angle)
        return self

    def _rotate(self, angle):
        currentAngle = self.get("/Rotate", 0)
        self[NameObject("/Rotate")] = NumberObject(currentAngle + angle)

    def _mergeResources(res1, res2, resource):
        newRes = DictionaryObject()
        newRes.update(res1.get(resource, DictionaryObject()).getObject())
        page2Res = res2.get(resource, DictionaryObject()).getObject()
        renameRes = {}
        for key in page2Res.keys():
            if newRes.has_key(key) and newRes[key] != page2Res[key]:
                newname = NameObject(key + "renamed")
                renameRes[key] = newname
                newRes[newname] = page2Res[key]
            elif not newRes.has_key(key):
                newRes[key] = page2Res[key]
        return newRes, renameRes
    _mergeResources = staticmethod(_mergeResources)

    def _contentStreamRename(stream, rename, pdf):
        if not rename:
            return stream
        stream = ContentStream(stream, pdf)
        for operands,operator in stream.operations:
            for i in range(len(operands)):
                op = operands[i]
                if isinstance(op, NameObject):
                    operands[i] = rename.get(op, op)
        return stream
    _contentStreamRename = staticmethod(_contentStreamRename)

    def _pushPopGS(contents, pdf):
        # adds a graphics state "push" and "pop" to the beginning and end
        # of a content stream.  This isolates it from changes such as 
        # transformation matricies.
        stream = ContentStream(contents, pdf)
        stream.operations.insert(0, [[], "q"])
        stream.operations.append([[], "Q"])
        return stream
    _pushPopGS = staticmethod(_pushPopGS)

    ##
    # Merges the content streams of two pages into one.  Resource references
    # (i.e. fonts) are maintained from both pages.  The mediabox/cropbox/etc
    # of this page are not altered.  The parameter page's content stream will
    # be added to the end of this page's content stream, meaning that it will
    # be drawn after, or "on top" of this page.
    # <p>
    # Stability: Added in v1.4, will exist for all future 1.x releases.
    # @param page2 An instance of {@link #PageObject PageObject} to be merged
    #              into this one.
    def mergePage(self, page2):

        # First we work on merging the resource dictionaries.  This allows us
        # to find out what symbols in the content streams we might need to
        # rename.

        newResources = DictionaryObject()
        rename = {}
        originalResources = self["/Resources"].getObject()
        page2Resources = page2["/Resources"].getObject()

        for res in "/ExtGState", "/Font", "/XObject", "/ColorSpace", "/Pattern", "/Shading":
            new, newrename = PageObject._mergeResources(originalResources, page2Resources, res)
            if new:
                newResources[NameObject(res)] = new
                rename.update(newrename)

        # Combine /ProcSet sets.
        newResources[NameObject("/ProcSet")] = ArrayObject(
            ImmutableSet(originalResources.get("/ProcSet", ArrayObject()).getObject()).union(
                ImmutableSet(page2Resources.get("/ProcSet", ArrayObject()).getObject())
            )
        )

        newContentArray = ArrayObject()

        originalContent = self["/Contents"].getObject()
        newContentArray.append(PageObject._pushPopGS(originalContent, self.pdf))

        page2Content = page2['/Contents'].getObject()
        page2Content = PageObject._contentStreamRename(page2Content, rename, self.pdf)
        page2Content = PageObject._pushPopGS(page2Content, self.pdf)
        newContentArray.append(page2Content)

        self[NameObject('/Contents')] = ContentStream(newContentArray, self.pdf)
        self[NameObject('/Resources')] = newResources

    ##
    # Compresses the size of this page by joining all content streams and
    # applying a FlateDecode filter.
    # <p>
    # Stability: Added in v1.6, will exist for all future v1.x releases.
    # However, it is possible that this function will perform no action if
    # content stream compression becomes "automatic" for some reason.
    def compressContentStreams(self):
        content = self["/Contents"].getObject()
        if not isinstance(content, ContentStream):
            content = ContentStream(content, self.pdf)
        self[NameObject("/Contents")] = content.flateEncode()

    ##
    # Locate all text drawing commands, in the order they are provided in the
    # content stream, and extract the text.  This works well for some PDF
    # files, but poorly for others, depending on the generator used.  This will
    # be refined in the future.  Do not rely on the order of text coming out of
    # this function, as it will change if this function is made more
    # sophisticated.
    # <p>
    # Stability: Added in v1.7, will exist for all future v1.x releases.  May
    # be overhauled to provide more ordered text in the future.
    # @return a unicode string object
    def extractText(self):
        text = u""
        content = self["/Contents"].getObject()
        if not isinstance(content, ContentStream):
            content = ContentStream(content, self.pdf)
        # Note: we check all strings are TextStringObjects.  ByteStringObjects
        # are strings where the byte->string encoding was unknown, so adding
        # them to the text here would be gibberish.
        for operands,operator in content.operations:
            if operator == "Tj":
                _text = operands[0]
                if isinstance(_text, TextStringObject):
                    text += _text
            elif operator == "T*":
                text += "\n"
            elif operator == "'":
                text += "\n"
                _text = operands[0]
                if isinstance(_text, TextStringObject):
                    text += operands[0]
            elif operator == '"':
                _text = operands[2]
                if isinstance(_text, TextStringObject):
                    text += "\n"
                    text += _text
            elif operator == "TJ":
                for i in operands[0]:
                    if isinstance(i, TextStringObject):
                        text += i
        return text

    ##
    # A rectangle (RectangleObject), expressed in default user space units,
    # defining the boundaries of the physical medium on which the page is
    # intended to be displayed or printed.
    # <p>
    # Stability: Added in v1.4, will exist for all future v1.x releases.
    mediaBox = createRectangleAccessor("/MediaBox", ())

    ##
    # A rectangle (RectangleObject), expressed in default user space units,
    # defining the visible region of default user space.  When the page is
    # displayed or printed, its contents are to be clipped (cropped) to this
    # rectangle and then imposed on the output medium in some
    # implementation-defined manner.  Default value: same as MediaBox.
    # <p>
    # Stability: Added in v1.4, will exist for all future v1.x releases.
    cropBox = createRectangleAccessor("/CropBox", ("/MediaBox",))

    ##
    # A rectangle (RectangleObject), expressed in default user space units,
    # defining the region to which the contents of the page should be clipped
    # when output in a production enviroment.
    # <p>
    # Stability: Added in v1.4, will exist for all future v1.x releases.
    bleedBox = createRectangleAccessor("/BleedBox", ("/CropBox", "/MediaBox"))

    ##
    # A rectangle (RectangleObject), expressed in default user space units,
    # defining the intended dimensions of the finished page after trimming.
    # <p>
    # Stability: Added in v1.4, will exist for all future v1.x releases.
    trimBox = createRectangleAccessor("/TrimBox", ("/CropBox", "/MediaBox"))

    ##
    # A rectangle (RectangleObject), expressed in default user space units,
    # defining the extent of the page's meaningful content as intended by the
    # page's creator.
    # <p>
    # Stability: Added in v1.4, will exist for all future v1.x releases.
    artBox = createRectangleAccessor("/ArtBox", ("/CropBox", "/MediaBox"))


class ContentStream(DecodedStreamObject):
    def __init__(self, stream, pdf):
        self.pdf = pdf
        self.operations = []
        # stream may be a StreamObject or an ArrayObject containing
        # multiple StreamObjects to be cat'd together.
        stream = stream.getObject()
        if isinstance(stream, ArrayObject):
            data = ""
            for s in stream:
                data += s.getObject().getData()
            stream = StringIO(data)
        else:
            stream = StringIO(stream.getData())
        self.__parseContentStream(stream)

    def __parseContentStream(self, stream):
        # file("f:\\tmp.txt", "w").write(stream.read())
        stream.seek(0, 0)
        operands = []
        while True:
            peek = readNonWhitespace(stream)
            if peek == '':
                break
            stream.seek(-1, 1)
            if peek.isalpha() or peek == "'" or peek == '"':
                operator = ""
                while True:
                    tok = stream.read(1)
                    if tok.isspace() or tok in NameObject.delimiterCharacters:
                        stream.seek(-1, 1)
                        break
                    elif tok == '':
                        break
                    operator += tok
                if operator == "BI":
                    # begin inline image - a completely different parsing
                    # mechanism is required, of course... thanks buddy...
                    assert operands == []
                    ii = self._readInlineImage(stream)
                    self.operations.append((ii, "INLINE IMAGE"))
                else:
                    self.operations.append((operands, operator))
                    operands = []
            elif peek == '%':
                # If we encounter a comment in the content stream, we have to
                # handle it here.  Typically, readObject will handle
                # encountering a comment -- but readObject assumes that
                # following the comment must be the object we're trying to
                # read.  In this case, it could be an operator instead.
                while peek not in ('\r', '\n'):
                    peek = stream.read(1)
            else:
                operands.append(readObject(stream, None))

    def _readInlineImage(self, stream):
        # begin reading just after the "BI" - begin image
        # first read the dictionary of settings.
        settings = DictionaryObject()
        while True:
            tok = readNonWhitespace(stream)
            stream.seek(-1, 1)
            if tok == "I":
                # "ID" - begin of image data
                break
            key = readObject(stream, self.pdf)
            tok = readNonWhitespace(stream)
            stream.seek(-1, 1)
            value = readObject(stream, self.pdf)
            settings[key] = value
        # left at beginning of ID
        tmp = stream.read(3)
        assert tmp[:2] == "ID"
        data = ""
        while True:
            tok = stream.read(1)
            if tok == "E":
                next = stream.read(1)
                if next == "I":
                    break
                else:
                    stream.seek(-1, 1)
                    data += tok
            else:
                data += tok
        x = readNonWhitespace(stream)
        stream.seek(-1, 1)
        return {"settings": settings, "data": data}

    def _getData(self):
        newdata = StringIO()
        for operands,operator in self.operations:
            if operator == "INLINE IMAGE":
                newdata.write("BI")
                dicttext = StringIO()
                operands["settings"].writeToStream(dicttext, None)
                newdata.write(dicttext.getvalue()[2:-2])
                newdata.write("ID ")
                newdata.write(operands["data"])
                newdata.write("EI")
            else:
                for op in operands:
                    op.writeToStream(newdata, None)
                    newdata.write(" ")
                newdata.write(operator)
            newdata.write("\n")
        return newdata.getvalue()

    def _setData(self, value):
        self.__parseContentStream(StringIO(value))

    _data = property(_getData, _setData)


##
# A class representing the basic document metadata provided in a PDF File.
# <p>
# As of pyPdf v1.10, all text properties of the document metadata have two
# properties, eg. author and author_raw.  The non-raw property will always
# return a TextStringObject, making it ideal for a case where the metadata is
# being displayed.  The raw property can sometimes return a ByteStringObject,
# if pyPdf was unable to decode the string's text encoding; this requires
# additional safety in the caller and therefore is not as commonly accessed.
class DocumentInformation(DictionaryObject):
    def __init__(self):
        DictionaryObject.__init__(self)

    def getText(self, key):
        retval = self.get(key, None)
        if isinstance(retval, TextStringObject):
            return retval
        return None

    ##
    # Read-only property accessing the document's title.  Added in v1.6, will
    # exist for all future v1.x releases.  Modified in v1.10 to always return a
    # unicode string (TextStringObject).
    # @return A unicode string, or None if the title is not provided.
    title = property(lambda self: self.getText("/Title"))
    title_raw = property(lambda self: self.get("/Title"))

    ##
    # Read-only property accessing the document's author.  Added in v1.6, will
    # exist for all future v1.x releases.  Modified in v1.10 to always return a
    # unicode string (TextStringObject).
    # @return A unicode string, or None if the author is not provided.
    author = property(lambda self: self.getText("/Author"))
    author_raw = property(lambda self: self.get("/Author"))

    ##
    # Read-only property accessing the subject of the document.  Added in v1.6,
    # will exist for all future v1.x releases.  Modified in v1.10 to always
    # return a unicode string (TextStringObject).
    # @return A unicode string, or None if the subject is not provided.
    subject = property(lambda self: self.getText("/Subject"))
    subject_raw = property(lambda self: self.get("/Subject"))

    ##
    # Read-only property accessing the document's creator.  If the document was
    # converted to PDF from another format, the name of the application (for
    # example, OpenOffice) that created the original document from which it was
    # converted.  Added in v1.6, will exist for all future v1.x releases.
    # Modified in v1.10 to always return a unicode string (TextStringObject).
    # @return A unicode string, or None if the creator is not provided.
    creator = property(lambda self: self.getText("/Creator"))
    creator_raw = property(lambda self: self.get("/Creator"))

    ##
    # Read-only property accessing the document's producer.  If the document
    # was converted to PDF from another format, the name of the application
    # (for example, OSX Quartz) that converted it to PDF.  Added in v1.6, will
    # exist for all future v1.x releases.  Modified in v1.10 to always return a
    # unicode string (TextStringObject).
    # @return A unicode string, or None if the producer is not provided.
    producer = property(lambda self: self.getText("/Producer"))
    producer_raw = property(lambda self: self.get("/Producer"))


##
# A class representing a destination within a PDF file.
# See section 8.2.1 of the PDF 1.6 reference.
# Stability: Added in v1.10, will exist for all v1.x releases.
class Destination(DictionaryObject):
    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)
          
    ##
    # Read-only property accessing the destination title.
    # @return A string.
    title = property(lambda self: self.get("/Title"))

    ##
    # Read-only property accessing the destination page.
    # @return An integer.
    page = property(lambda self: self.get("/Page"))

    ##
    # Read-only property accessing the destination type.
    # @return A string.
    typ = property(lambda self: self.get("/Type"))

    ##
    # Read-only property accessing the zoom factor.
    # @return A number, or None if not available.
    zoom = property(lambda self: self.get("/Zoom", None))

    ##
    # Read-only property accessing the left horizontal coordinate.
    # @return A number, or None if not available.
    left = property(lambda self: self.get("/Left", None))

    ##
    # Read-only property accessing the right horizontal coordinate.
    # @return A number, or None if not available.
    right = property(lambda self: self.get("/Right", None))

    ##
    # Read-only property accessing the top vertical coordinate.
    # @return A number, or None if not available.
    top = property(lambda self: self.get("/Top", None))

    ##
    # Read-only property accessing the bottom vertical coordinate.
    # @return A number, or None if not available.
    bottom = property(lambda self: self.get("/Bottom", None))

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]

# ref: pdf1.8 spec section 3.5.2 algorithm 3.2
_encryption_padding = '\x28\xbf\x4e\x5e\x4e\x75\x8a\x41\x64\x00\x4e\x56' + \
        '\xff\xfa\x01\x08\x2e\x2e\x00\xb6\xd0\x68\x3e\x80\x2f\x0c' + \
        '\xa9\xfe\x64\x53\x69\x7a'

# Implementation of algorithm 3.2 of the PDF standard security handler,
# section 3.5.2 of the PDF 1.6 reference.
def _alg32(password, rev, keylen, owner_entry, p_entry, id1_entry, metadata_encrypt=True):
    # 1. Pad or truncate the password string to exactly 32 bytes.  If the
    # password string is more than 32 bytes long, use only its first 32 bytes;
    # if it is less than 32 bytes long, pad it by appending the required number
    # of additional bytes from the beginning of the padding string
    # (_encryption_padding).
    password = (password + _encryption_padding)[:32]
    # 2. Initialize the MD5 hash function and pass the result of step 1 as
    # input to this function.
    import md5, struct
    m = md5.new(password)
    # 3. Pass the value of the encryption dictionary's /O entry to the MD5 hash
    # function.
    m.update(owner_entry)
    # 4. Treat the value of the /P entry as an unsigned 4-byte integer and pass
    # these bytes to the MD5 hash function, low-order byte first.
    p_entry = struct.pack('<i', p_entry)
    m.update(p_entry)
    # 5. Pass the first element of the file's file identifier array to the MD5
    # hash function.
    m.update(id1_entry)
    # 6. (Revision 3 or greater) If document metadata is not being encrypted,
    # pass 4 bytes with the value 0xFFFFFFFF to the MD5 hash function.
    if rev >= 3 and not metadata_encrypt:
        m.update("\xff\xff\xff\xff")
    # 7. Finish the hash.
    md5_hash = m.digest()
    # 8. (Revision 3 or greater) Do the following 50 times: Take the output
    # from the previous MD5 hash and pass the first n bytes of the output as
    # input into a new MD5 hash, where n is the number of bytes of the
    # encryption key as defined by the value of the encryption dictionary's
    # /Length entry.
    if rev >= 3:
        for i in range(50):
            md5_hash = md5.new(md5_hash[:keylen]).digest()
    # 9. Set the encryption key to the first n bytes of the output from the
    # final MD5 hash, where n is always 5 for revision 2 but, for revision 3 or
    # greater, depends on the value of the encryption dictionary's /Length
    # entry.
    return md5_hash[:keylen]

# Implementation of algorithm 3.3 of the PDF standard security handler,
# section 3.5.2 of the PDF 1.6 reference.
def _alg33(owner_pwd, user_pwd, rev, keylen):
    # steps 1 - 4
    key = _alg33_1(owner_pwd, rev, keylen)
    # 5. Pad or truncate the user password string as described in step 1 of
    # algorithm 3.2.
    user_pwd = (user_pwd + _encryption_padding)[:32]
    # 6. Encrypt the result of step 5, using an RC4 encryption function with
    # the encryption key obtained in step 4.
    val = utils.RC4_encrypt(key, user_pwd)
    # 7. (Revision 3 or greater) Do the following 19 times: Take the output
    # from the previous invocation of the RC4 function and pass it as input to
    # a new invocation of the function; use an encryption key generated by
    # taking each byte of the encryption key obtained in step 4 and performing
    # an XOR operation between that byte and the single-byte value of the
    # iteration counter (from 1 to 19).
    if rev >= 3:
        for i in range(1, 20):
            new_key = ''
            for l in range(len(key)):
                new_key += chr(ord(key[l]) ^ i)
            val = utils.RC4_encrypt(new_key, val)
    # 8. Store the output from the final invocation of the RC4 as the value of
    # the /O entry in the encryption dictionary.
    return val

# Steps 1-4 of algorithm 3.3
def _alg33_1(password, rev, keylen):
    # 1. Pad or truncate the owner password string as described in step 1 of
    # algorithm 3.2.  If there is no owner password, use the user password
    # instead.
    password = (password + _encryption_padding)[:32]
    # 2. Initialize the MD5 hash function and pass the result of step 1 as
    # input to this function.
    import md5
    m = md5.new(password)
    # 3. (Revision 3 or greater) Do the following 50 times: Take the output
    # from the previous MD5 hash and pass it as input into a new MD5 hash.
    md5_hash = m.digest()
    if rev >= 3:
        for i in range(50):
            md5_hash = md5.new(md5_hash).digest()
    # 4. Create an RC4 encryption key using the first n bytes of the output
    # from the final MD5 hash, where n is always 5 for revision 2 but, for
    # revision 3 or greater, depends on the value of the encryption
    # dictionary's /Length entry.
    key = md5_hash[:keylen]
    return key

# Implementation of algorithm 3.4 of the PDF standard security handler,
# section 3.5.2 of the PDF 1.6 reference.
def _alg34(password, owner_entry, p_entry, id1_entry):
    # 1. Create an encryption key based on the user password string, as
    # described in algorithm 3.2.
    key = _alg32(password, 2, 5, owner_entry, p_entry, id1_entry)
    # 2. Encrypt the 32-byte padding string shown in step 1 of algorithm 3.2,
    # using an RC4 encryption function with the encryption key from the
    # preceding step.
    U = utils.RC4_encrypt(key, _encryption_padding)
    # 3. Store the result of step 2 as the value of the /U entry in the
    # encryption dictionary.
    return U, key

# Implementation of algorithm 3.4 of the PDF standard security handler,
# section 3.5.2 of the PDF 1.6 reference.
def _alg35(password, rev, keylen, owner_entry, p_entry, id1_entry, metadata_encrypt):
    # 1. Create an encryption key based on the user password string, as
    # described in Algorithm 3.2.
    key = _alg32(password, rev, keylen, owner_entry, p_entry, id1_entry)
    # 2. Initialize the MD5 hash function and pass the 32-byte padding string
    # shown in step 1 of Algorithm 3.2 as input to this function. 
    import md5
    m = md5.new()
    m.update(_encryption_padding)
    # 3. Pass the first element of the file's file identifier array (the value
    # of the ID entry in the document's trailer dictionary; see Table 3.13 on
    # page 73) to the hash function and finish the hash.  (See implementation
    # note 25 in Appendix H.) 
    m.update(id1_entry)
    md5_hash = m.digest()
    # 4. Encrypt the 16-byte result of the hash, using an RC4 encryption
    # function with the encryption key from step 1. 
    val = utils.RC4_encrypt(key, md5_hash)
    # 5. Do the following 19 times: Take the output from the previous
    # invocation of the RC4 function and pass it as input to a new invocation
    # of the function; use an encryption key generated by taking each byte of
    # the original encryption key (obtained in step 2) and performing an XOR
    # operation between that byte and the single-byte value of the iteration
    # counter (from 1 to 19). 
    for i in range(1, 20):
        new_key = ''
        for l in range(len(key)):
            new_key += chr(ord(key[l]) ^ i)
        val = utils.RC4_encrypt(new_key, val)
    # 6. Append 16 bytes of arbitrary padding to the output from the final
    # invocation of the RC4 function and store the 32-byte result as the value
    # of the U entry in the encryption dictionary. 
    # (implementator note: I don't know what "arbitrary padding" is supposed to
    # mean, so I have used null bytes.  This seems to match a few other
    # people's implementations)
    return val + ('\x00' * 16), key

#if __name__ == "__main__":
#    output = PdfFileWriter()
#
#    input1 = PdfFileReader(file("test\\5000-s1-05e.pdf", "rb"))
#    page1 = input1.getPage(0)
#
#    input2 = PdfFileReader(file("test\\PDFReference16.pdf", "rb"))
#    page2 = input2.getPage(0)
#    page3 = input2.getPage(1)
#    page1.mergePage(page2)
#    page1.mergePage(page3)
#
#    input3 = PdfFileReader(file("test\\cc-cc.pdf", "rb"))
#    page1.mergePage(input3.getPage(0))
#
#    page1.compressContentStreams()
#
#    output.addPage(page1)
#    output.write(file("test\\merge-test.pdf", "wb"))