def _addPage(self, page, action):
assert page["/Type"] == "/Page"
page[NameObject("/Parent")] = self._pages
page = self._addObject(page)
pages = self.getObject(self._pages)
action(pages["/Kids"], page)
pages[NameObject("/Count")] = NumberObject(pages["/Count"] + 1)
def _flatten(self, pages=None, inherit=None, indirectRef=None):
inheritablePageAttributes = (NameObject("/Resources"),
NameObject("/MediaBox"),
NameObject("/CropBox"),
NameObject("/Rotate"))
if inherit is None:
inherit = dict()
if pages is None:
self.flattenedPages = []
catalog = self.trailer["/Root"].getObject()
pages = catalog["/Pages"].getObject()
t = pages["/Type"]
if t == "/Pages":
for attr in inheritablePageAttributes:
if attr in pages:
inherit[attr] = pages[attr]
for page in pages["/Kids"]:
addt = {}
if isinstance(page, IndirectObject):
addt["indirectRef"] = page
self._flatten(page.getObject(), inherit, **addt)
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 attr not in pages:
pages[attr] = value
pageObj = PageObject(self, indirectRef)
pageObj.update(pages)
self.flattenedPages.append(pageObj)
def __init__(self):
self._header = b_("%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 addBookmark(self, title, pagenum, parent=None):
"""
Add a bookmark to the pdf, using the specified title and pointing at
the specified page number. A parent can be specified to make this a
nested bookmark below the parent.
"""
if parent is None:
iloc = [len(self.bookmarks)-1]
elif type(parent) == list:
iloc = parent
else:
iloc = self.findBookmark(parent)
dest = Bookmark(TextStringObject(title), NumberObject(pagenum),
NameObject('/FitH'), NumberObject(826))
if parent is None:
self.bookmarks.append(dest)
else:
bmparent = self.bookmarks
for i in iloc[:-1]:
bmparent = bmparent[i]
npos = iloc[-1]+1
if npos < len(bmparent) and type(bmparent[npos]) == list:
bmparent[npos].append(dest)
else:
bmparent.insert(npos, [dest])
def addTransformation(self, ctm):
originalContent = self.getContents()
if originalContent is not None:
newContent = PageObject._addTransformationMatrix(
originalContent, self.pdf, ctm)
newContent = PageObject._pushPopGS(newContent, self.pdf)
self[NameObject('/Contents')] = newContent
def addNamedDestination(self, title, pagenum):
pageRef = self.getObject(self._pages)['/Kids'][pagenum]
dest = DictionaryObject()
dest.update({
NameObject('/D'):
ArrayObject([pageRef,
NameObject('/FitH'),
NumberObject(826)]),
NameObject('/S'):
NameObject('/GoTo')
})
destRef = self._addObject(dest)
nd = self.getNamedDestRoot()
nd.extend([title, destRef])
return destRef
def addNamedDestination(self, title, pagenum):
"""
Add a destination to the pdf, using the specified title and pointing
at the specified page number.
"""
dest = Destination(TextStringObject(title), NumberObject(pagenum),
NameObject('/FitH'), NumberObject(826))
self.named_dests.append(dest)
def addBookmarkDict(self, bookmark, parent=None):
bookmarkObj = TreeObject()
for k, v in bookmark.items():
bookmarkObj[NameObject(str(k))] = v
bookmarkObj.update(bookmark)
if '/A' in bookmark:
action = DictionaryObject()
for k, v in bookmark['/A'].items():
action[NameObject(str(k))] = v
actionRef = self._addObject(action)
bookmarkObj['/A'] = actionRef
bookmarkRef = self._addObject(bookmarkObj)
outlineRef = self.getOutlineRoot()
if parent is None:
parent = outlineRef
parent = parent.getObject()
parent.addChild(bookmarkRef, self)
return bookmarkRef
def createBlankPage(pdf=None, width=None, height=None):
page = PageObject(pdf)
# Creates a new page (cf PDF Reference 7.7.3.3)
page.__setitem__(NameObject('/Type'), NameObject('/Page'))
page.__setitem__(NameObject('/Parent'), NullObject())
page.__setitem__(NameObject('/Resources'), DictionaryObject())
if width is None or height is None:
if pdf is not None and pdf.getNumPages() > 0:
lastpage = pdf.getPage(pdf.getNumPages() - 1)
width = lastpage.mediaBox.getWidth()
height = lastpage.mediaBox.getHeight()
else:
raise utils.PageSizeNotDefinedError()
page.__setitem__(NameObject('/MediaBox'),
RectangleObject([0, 0, width, height]))
return page
def _write_dests(self):
dests = self.named_dests
for v in dests:
pageno = None
if '/Page' in v:
for i, p in enumerate(self.pages):
if p.id == v['/Page']:
v[NameObject('/Page')] = p.out_pagedata
pageno = i
if pageno is not None:
self.output.addNamedDestinationObject(v)
def getNamedDestRoot(self):
root = self.getObject(self._root)
if '/Names' in root and isinstance(root['/Names'], DictionaryObject):
names = root['/Names']
idnum = self._objects.index(names) + 1
namesRef = IndirectObject(idnum, 0, self)
assert namesRef.getObject() == names
if '/Dests' in names and isinstance(names['/Dests'],
DictionaryObject):
dests = names['/Dests']
idnum = self._objects.index(dests) + 1
destsRef = IndirectObject(idnum, 0, self)
assert destsRef.getObject() == dests
if '/Names' in dests:
nd = dests['/Names']
else:
nd = ArrayObject()
dests[NameObject('/Names')] = nd
else:
dests = DictionaryObject()
destsRef = self._addObject(dests)
names[NameObject('/Dests')] = destsRef
nd = ArrayObject()
dests[NameObject('/Names')] = nd
else:
names = DictionaryObject()
namesRef = self._addObject(names)
root[NameObject('/Names')] = namesRef
dests = DictionaryObject()
destsRef = self._addObject(dests)
names[NameObject('/Dests')] = destsRef
nd = ArrayObject()
dests[NameObject('/Names')] = nd
return nd
def addBookmark(self, title, pagenum, parent=None):
"""
Add a bookmark to the pdf, using the specified title and pointing at
the specified page number. A parent can be specified to make this a
nested bookmark below the parent.
"""
pageRef = self.getObject(self._pages)['/Kids'][pagenum]
action = DictionaryObject()
action.update({
NameObject('/D'):
ArrayObject([pageRef,
NameObject('/FitH'),
NumberObject(826)]),
NameObject('/S'):
NameObject('/GoTo')
})
actionRef = self._addObject(action)
outlineRef = self.getOutlineRoot()
if parent is None:
parent = outlineRef
bookmark = TreeObject()
bookmark.update({
NameObject('/A'): actionRef,
NameObject('/Title'): createStringObject(title)
})
bookmarkRef = self._addObject(bookmark)
parent = parent.getObject()
parent.addChild(bookmarkRef, self)
return bookmarkRef
def getOutlineRoot(self):
root = self.getObject(self._root)
if '/Outlines' in root:
outline = root['/Outlines']
idnum = self._objects.index(outline) + 1
outlineRef = IndirectObject(idnum, 0, self)
assert outlineRef.getObject() == outline
else:
outline = TreeObject()
outline.update({})
outlineRef = self._addObject(outline)
root[NameObject('/Outlines')] = outlineRef
return outline
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 key in newRes and newRes[key] != page2Res[key]:
newname = NameObject(key + "renamed")
renameRes[key] = newname
newRes[newname] = page2Res[key]
elif key not in newRes:
newRes[key] = page2Res.raw_get(key)
return newRes, renameRes
def _trim_dests(self, pdf, dests, pages):
"""
Removes any named destinations that are not a part of the specified
page set
"""
new_dests = []
for k, o in dests.items():
for j in range(*pages):
if pdf.getPage(j).getObject() == o['/Page'].getObject():
o[NameObject('/Page')] = o['/Page'].getObject()
assert str(k) == str(o['/Title'])
new_dests.append(o)
break
return new_dests
def _associate_dests_to_pages(self, pages):
for nd in self.named_dests:
pageno = None
np = nd['/Page']
if type(np) == NumberObject:
continue
for p in pages:
if np.getObject() == p.pagedata.getObject():
pageno = p.id
if pageno is not None:
nd[NameObject('/Page')] = NumberObject(pageno)
else:
raise ValueError("Unresolved named destination '%s'" %
nd['/Title'])
def _write_bookmarks(self, bookmarks=None, parent=None):
if bookmarks is None:
bookmarks = self.bookmarks
last_added = None
for b in bookmarks:
if type(b) == list:
self._write_bookmarks(b, last_added)
continue
pageno = None
if '/Page' in b:
for i, p in enumerate(self.pages):
if p.id == b['/Page']:
b[NameObject('/Page')] = p.out_pagedata
pageno = i
if pageno is not None:
last_added = self.output.addBookmarkDestination(b, parent)
def add(self, title, pagenum):
pageRef = self.pdf.getObject(self.pdf._pages)['/Kids'][pagenum]
action = DictionaryObject()
action.update({NameObject('/D'): ArrayObject([pageRef,
NameObject('/FitH'),
NumberObject(826)]),
NameObject('/S'): NameObject('/GoTo')})
actionRef = self.pdf._addObject(action)
bookmark = TreeObject()
bookmark.update({NameObject('/A'): actionRef,
NameObject('/Title'): createStringObject(title)})
self.pdf._addObject(bookmark)
self.tree.addChild(bookmark)
def _associate_bookmarks_to_pages(self, pages, bookmarks=None):
if bookmarks is None:
bookmarks = self.bookmarks
for b in bookmarks:
if type(b) == list:
self._associate_bookmarks_to_pages(pages, b)
continue
pageno = None
bp = b['/Page']
if type(bp) == NumberObject:
continue
for p in pages:
if bp.getObject() == p.pagedata.getObject():
pageno = p.id
if pageno is not None:
b[NameObject('/Page')] = NumberObject(pageno)
else:
raise ValueError("Unresolved bookmark '%s'" % b['/Title'])
def encrypt(self, user_pwd, owner_pwd=None, use_128bit=True):
import time
import random
if owner_pwd is 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(repr(time.time())).digest()
ID_2 = md5(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
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 _trim_outline(self, pdf, outline, pages):
"""
Removes any outline/bookmark entries that are not a part of the
specified page set
"""
new_outline = []
prev_header_added = True
for i, o in enumerate(outline):
if type(o) == list:
sub = self._trim_outline(pdf, o, pages)
if sub:
if not prev_header_added:
new_outline.append(outline[i-1])
new_outline.append(sub)
else:
prev_header_added = False
for j in range(*pages):
if pdf.getPage(j).getObject() == o['/Page'].getObject():
o[NameObject('/Page')] = o['/Page'].getObject()
new_outline.append(o)
prev_header_added = True
break
return new_outline
def _mergePage(self,
page2,
page2transformation=None,
ctm=None,
expand=False):
# 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", "/Properties"
]:
new, newrename = PageObject._mergeResources(
originalResources, page2Resources, res)
if new:
newResources[NameObject(res)] = new
rename.update(newrename)
# Combine /ProcSet sets.
newResources[NameObject("/ProcSet")] = ArrayObject(
frozenset(
originalResources.get("/ProcSet",
ArrayObject()).getObject()).union(
frozenset(
page2Resources.get(
"/ProcSet",
ArrayObject()).getObject())))
newContentArray = ArrayObject()
originalContent = self.getContents()
if originalContent is not None:
newContentArray.append(
PageObject._pushPopGS(originalContent, self.pdf))
page2Content = page2.getContents()
if page2Content is not None:
if page2transformation is not None:
page2Content = page2transformation(page2Content)
page2Content = PageObject._contentStreamRename(
page2Content, rename, self.pdf)
page2Content = PageObject._pushPopGS(page2Content, self.pdf)
newContentArray.append(page2Content)
# if expanding the page to fit a new page,
# calculate the new media box size
if expand:
corners1 = [
self.mediaBox.getLowerLeft_x().as_numeric(),
self.mediaBox.getLowerLeft_y().as_numeric(),
self.mediaBox.getUpperRight_x().as_numeric(),
self.mediaBox.getUpperRight_y().as_numeric()
]
corners2 = [
page2.mediaBox.getLowerLeft_x().as_numeric(),
page2.mediaBox.getLowerLeft_y().as_numeric(),
page2.mediaBox.getUpperLeft_x().as_numeric(),
page2.mediaBox.getUpperLeft_y().as_numeric(),
page2.mediaBox.getUpperRight_x().as_numeric(),
page2.mediaBox.getUpperRight_y().as_numeric(),
page2.mediaBox.getLowerRight_x().as_numeric(),
page2.mediaBox.getLowerRight_y().as_numeric()
]
if ctm is not None:
new_x = map(
lambda i: ctm[0] * corners2[i] + ctm[2] * corners2[i + 1] +
ctm[4], range(0, 8, 2))
new_y = map(
lambda i: ctm[1] * corners2[i] + ctm[3] * corners2[i + 1] +
ctm[5], range(0, 8, 2))
else:
new_x = corners2[0:8:2]
new_y = corners2[1:8:2]
lowerleft = [min(new_x), min(new_y)]
upperright = [max(new_x), max(new_y)]
lowerleft = [
min(corners1[0], lowerleft[0]),
min(corners1[1], lowerleft[1])
]
upperright = [
max(corners1[2], upperright[0]),
max(corners1[3], upperright[1])
]
self.mediaBox.setLowerLeft(lowerleft)
self.mediaBox.setUpperRight(upperright)
self[NameObject('/Contents')] = ContentStream(newContentArray,
self.pdf)
self[NameObject('/Resources')] = newResources
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 _rotate(self, angle):
currentAngle = self.get("/Rotate", 0)
self[NameObject("/Rotate")] = NumberObject(currentAngle + angle)
def merge(self, position, fileobj, bookmark=None,
pages=None, import_bookmarks=True):
"""
>>> merge(position, file, bookmark=None,
pages=None, import_bookmarks=True)
Merges the pages from the source document specified by "file"
into the output file at the page number specified by "position".
Optionally, you may specify a bookmark to be applied at the beginning
of the included file by supplying the text of the bookmark in the
"bookmark" parameter.
You may prevent the source document's bookmarks from being imported by
specifying "import_bookmarks" as False.
You may also use the "pages" parameter to merge only the specified
range of pages from the source document into the output document.
"""
# This parameter is passed to self.inputs.append and means
# that the stream used was created in this method.
my_file = False
# If the fileobj parameter is a string, assume it is a path
# and create a file object at that location. If it is a file,
# copy the file's contents into a StringIO stream object; if
# it is a PdfFileReader, copy that reader's stream into a
# StringIO stream.
# If fileobj is none of the above types, it is not modified
if type(fileobj) in (str, unicode):
fileobj = file(fileobj, 'rb')
my_file = True
elif type(fileobj) == file:
fileobj.seek(0)
filecontent = fileobj.read()
fileobj = StringIO(filecontent)
my_file = True
elif type(fileobj) == PdfFileReader:
orig_tell = fileobj.stream.tell()
fileobj.stream.seek(0)
filecontent = StringIO(fileobj.stream.read())
# reset the stream to its original location
fileobj.stream.seek(orig_tell)
fileobj = filecontent
my_file = True
# Create a new PdfFileReader instance using the stream
# (either file or StringIO) created above
pdfr = PdfFileReader(fileobj, strict=self.strict)
# Find the range of pages to merge
if pages is None:
pages = (0, pdfr.getNumPages())
elif type(pages) in (int, float, str, unicode):
raise TypeError('"pages" must be a tuple of (start, end)')
srcpages = []
if bookmark:
bookmark = Bookmark(TextStringObject(bookmark),
NumberObject(self.id_count),
NameObject('/Fit'))
outline = []
if import_bookmarks:
outline = pdfr.getOutlines()
outline = self._trim_outline(pdfr, outline, pages)
if bookmark:
self.bookmarks += [bookmark, outline]
else:
self.bookmarks += outline
dests = pdfr.namedDestinations
dests = self._trim_dests(pdfr, dests, pages)
self.named_dests += dests
# Gather all the pages that are going to be merged
for i in range(*pages):
pg = pdfr.getPage(i)
id = self.id_count
self.id_count += 1
mp = _MergedPage(pg, pdfr, id)
srcpages.append(mp)
self._associate_dests_to_pages(srcpages)
self._associate_bookmarks_to_pages(srcpages)
# Slice to insert the pages at the specified position
self.pages[position:position] = srcpages
# Keep track of our input files so we can close them later
self.inputs.append((fileobj, pdfr, my_file))
def getNumPages(self):
pages = self.getObject(self._pages)
return int(pages[NameObject("/Count")])
def compressContentStreams(self):
content = self.getContents()
if content is not None:
if not isinstance(content, ContentStream):
content = ContentStream(content, self.pdf)
self[NameObject("/Contents")] = content.flateEncode()
def setRectangle(self, name, value):
if not isinstance(name, NameObject):
name = NameObject(name)
self[name] = value
def write(self, stream):
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 xrange(len(self._objects)):
obj = self._objects[objIndex]
if isinstance(obj, PageObject) and obj.indirectRef is not None:
data = obj.indirectRef
externalReferenceMap.setdefault(data.pdf, {})
externalReferenceMap[data.pdf].setdefault(data.generation, {})
externalReferenceMap[data.pdf][data.generation][data.idnum] = \
IndirectObject(objIndex + 1, 0, self)
self.stack = []
self._sweepIndirectReferences(externalReferenceMap, self._root)
del self.stack
# Begin writing:
object_positions = []
stream.write(self._header + b_("\n"))
for i in range(len(self._objects)):
idnum = (i + 1)
obj = self._objects[i]
object_positions.append(stream.tell())
stream.write(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 = md5(key).digest()
key = md5_hash[:min(16, len(self._encrypt_key) + 5)]
if obj is not None:
obj.writeToStream(stream, key)
stream.write(b_("\nendobj\n"))
# xref table
xref_location = stream.tell()
stream.write(b_("xref\n"))
stream.write(b_("0 %s\n" % (len(self._objects) + 1)))
stream.write(b_("%010d %05d f \n" % (0, 65535)))
for offset in object_positions:
stream.write(b_("%010d %05d n \n" % (offset, 0)))
# trailer
stream.write(b_("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(b_("\nstartxref\n%s\n%%%%EOF\n" % (xref_location)))
