Support Adobe's idiotic font encryption scheme for EPUB books created by InDesign

src/calibre/ebooks/epub/__init__.py

@@ -1,4 +1,4 @@
-#!/usr/bin/env  python
+from __future__ import with_statement
 __license__   = 'GPL v3'
 __copyright__ = '2008, Kovid Goyal kovid@kovidgoyal.net'
 __docformat__ = 'restructuredtext en'
@@ -6,10 +6,12 @@ __docformat__ = 'restructuredtext en'
 '''
 Conversion to EPUB.
 '''
-import sys, textwrap, re
+import sys, textwrap, re, os, uuid
+from itertools import cycle
 from calibre.utils.config import Config, StringConfig
 from calibre.utils.zipfile import ZipFile, ZIP_STORED
 from calibre.ebooks.html import config as common_config, tostring
+from lxml import etree
 
 class DefaultProfile(object):
     
@@ -36,6 +38,38 @@ def rules(stylesheets):
                 if r.type == r.STYLE_RULE:
                     yield r
 
+def decrypt_font(key, path):
+    raw = open(path, 'rb').read()
+    crypt = raw[:1024]
+    key = cycle(iter(key))
+    decrypt = ''.join([chr(ord(x)^key.next()) for x in crypt])
+    with open(path, 'wb') as f:
+        f.write(decrypt)
+        f.write(raw[1024:])
+
+def process_encryption(encfile, opf):
+    key = None
+    m = re.search(r'(?i)(urn:uuid:[0-9a-f-]+)', open(opf, 'rb').read())
+    if m:
+        key = m.group(1)
+        key = list(map(ord, uuid.UUID(key).bytes))
+    try:
+        root = etree.parse(encfile)
+        for em in root.xpath('descendant::*[contains(name(), "EncryptionMethod")]'):
+            algorithm = em.get('Algorithm', '')
+            if algorithm != 'http://ns.adobe.com/pdf/enc#RC':
+                return False
+            cr = em.getparent().xpath('descendant::*[contains(name(), "CipherReference")]')[0]
+            uri = cr.get('URI')
+            path = os.path.abspath(os.path.join(os.path.dirname(encfile), '..', *uri.split('/')))
+            if os.path.exists(path):
+                decrypt_font(key, path)
+        return True
+    except:
+        import traceback
+        traceback.print_exc()
+    return False
+
 def initialize_container(path_to_container, opf_name='metadata.opf'):
     '''
     Create an empty EPUB document, with a default skeleton.

src/calibre/ebooks/epub/from_any.py

@@ -12,7 +12,7 @@ from contextlib import nested
 
 from calibre import extract, walk
 from calibre.ebooks import DRMError
-from calibre.ebooks.epub import config as common_config
+from calibre.ebooks.epub import config as common_config, process_encryption
 from calibre.ebooks.epub.from_html import convert as html2epub, find_html_index
 from calibre.ptempfile import TemporaryDirectory
 from calibre.ebooks.metadata import MetaInformation
@@ -72,12 +72,19 @@ def epub2opf(path, tdir, opts):
     zf = ZipFile(path)
     zf.extractall(tdir)
     opts.chapter_mark = 'none'
-    if os.path.exists(os.path.join(tdir, 'META-INF', 'encryption.xml')):
-        raise DRMError(os.path.basename(path))
+    encfile = os.path.join(tdir, 'META-INF', 'encryption.xml')
+    opf = None
     for f in walk(tdir):
         if f.lower().endswith('.opf'):
-            return f
+            opf = f
+            break
+    if opf and os.path.exists(encfile):
+        if not process_encryption(encfile, opf):
+            raise DRMError(os.path.basename(path))
+        
+    if opf is None:
         raise ValueError('%s is not a valid EPUB file'%path)
+    return opf
     
 def odt2epub(path, tdir, opts):
     from calibre.ebooks.odt.to_oeb import Extract

src/calibre/ebooks/epub/iterator.py

@@ -37,14 +37,16 @@ class UnsupportedFormatError(Exception):
 
 class SpineItem(unicode):
     
-    def __init__(self, path):
-        unicode.__init__(self, path)
+    def __new__(cls, *args):
+        obj = super(SpineItem, cls).__new__(cls, *args)
+        path = args[0]
         raw = open(path, 'rb').read()
-        raw, self.encoding = xml_to_unicode(raw)
-        self.character_count = character_count(raw)
-        self.start_page = -1
-        self.pages      = -1
-        self.max_page   = -1
+        raw, obj.encoding = xml_to_unicode(raw)
+        obj.character_count = character_count(raw)
+        obj.start_page = -1
+        obj.pages      = -1
+        obj.max_page   = -1
+        return obj
 
 def html2opf(path, tdir, opts):
     opts = copy.copy(opts)