Changeset 766

Timestamp:

12/12/06 22:01:32 (2 years ago)

Author:

mfenniak

Message:

update CHANGELOG. Begin work on adding encryption support for PDF files. It is incomplete and incorrect at this point.

Files:

• pypdf/trunk/CHANGELOG (modified) (1 diff)
• pypdf/trunk/pyPdf/pdf.py (modified) (3 diffs)
• pypdf/trunk/pyPdf/utils.py (modified) (1 diff)

pypdf/trunk/CHANGELOG

@@ +1 @@
+Version 1.8, YYYY-MM-DD
+-----------------------
+
+ - Add new pythondoc documentation.
+
+ - Fix bug in ASCII85 decode that occurs when whitespace exists inside the
+   two terminating characters of the stream.
+
-Version 1.7, 2006-12-10
------------------------

pypdf/trunk/pyPdf/pdf.py

@@ -44 +44 @@
-
-import filters
+import utils
-from generic import *
-from utils import readNonWhitespace, readUntilWhitespace, ConvertFunctionsToVirtualList
@@ -306 +307 @@
-            pageObj.update(pages)
-            self.flattenedPages.append(pageObj)
+
+    def safeGetObject(self, obj):
+        if isinstance(obj, IndirectObject):
+            return self.safeGetObject(self.getObject(obj))
+        return obj
-
-    def getObject(self, indirectReference):
@@ -504 +510 @@
-        return line
-
+    # 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'
+
+    def _alg32(self, password, rev, keylen):
+        import md5, struct
+        m = md5.new()
+        password = (password + self._encryption_padding)[:32]
+        m.update(password)
+        encrypt = self.safeGetObject(self.trailer['/Encrypt'])
+        owner_entry = self.safeGetObject(encrypt['/O'])
+        m.update(owner_entry)
+        p_entry = self.safeGetObject(encrypt['/P'])
+        p_entry = struct.pack('<i', p_entry)
+        m.update(p_entry)
+        id_entry = self.safeGetObject(self.trailer['/ID'])
+        id1_entry = self.safeGetObject(id_entry[0])
+        m.update(id1_entry)
+        #if rev >= 3:
+        #    if document metadata is not being encrypted, pass \xff\xff\xff\xff into hash
+        md5_hash = m.digest()
+        if rev >= 3:
+            for i in range(50):
+                m = md5.new()
+                m.update(md5_hash[:keylen])
+                md5_hash = m.digest()
+        return md5_hash[:keylen]
+
+    def _alg34(self, password):
+        rev = self.safeGetObject(self.safeGetObject(self.trailer['/Encrypt'])['/R'])
+        if rev == 2:
+            keylen = 5
+        key = self._alg32(password, rev, keylen)
+        U = utils.RC4_encrypt(key, self._encryption_padding)
+        return U
+
+    ##
+    # Decrypt file.
+    def decrypt(self, password):
+        U = self._alg34(password)
+        print "calc U: %r" % U
+        print "file U: %r" % self.safeGetObject(self.trailer['/Encrypt'])['/U']
-
-def getRectangle(self, name, defaults):

pypdf/trunk/pyPdf/utils.py

@@ -71 +71 @@
-        return self.getFunction(index)
-
+def RC4_encrypt(key, plaintext):
+    S = [i for i in range(256)]
+    j = 0
+    for i in range(256):
+        j = (j + S[i] + ord(key[i % len(key)])) % 256
+        S[i], S[j] = S[j], S[i]
+    i, j = 0, 0
+    retval = ""
+    for x in range(len(plaintext)):
+        i = (i + 1) % 256
+        j = (j + S[i]) % 256
+        S[i], S[j] = S[j], S[i]
+        t = S[(S[i] + S[j]) % 256]
+        retval += chr(ord(plaintext[x]) ^ t)
+    return retval
-
+if __name__ == "__main__":
+    # test RC4
+    out = RC4_encrypt("Key", "Plaintext")
+    print repr(out)
+    pt = RC4_encrypt("Key", out)
+    print repr(pt)