From 7dd371e5e37465475164a5475ca177abde84212d Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Sun, 22 Jan 2012 16:32:20 +0530
Subject: [PATCH] =?UTF-8?q?HTML=20Input:=20Fix=20handling=20of=20files=20w?=
 =?UTF-8?q?ith=20=C3=A4=20characters=20in=20their=20filenames.=20Fixes=20#?=
 =?UTF-8?q?919931=20(ebook-convert=20crash=20on=20converting=20from=20html?=
 =?UTF-8?q?-source)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 src/calibre/ebooks/html/input.py |  4 +++-
 src/calibre/ebooks/oeb/base.py   | 17 +++++++++++++----
 2 files changed, 16 insertions(+), 5 deletions(-)

diff --git a/src/calibre/ebooks/html/input.py b/src/calibre/ebooks/html/input.py
index d60baf8bce..d303dd66a5 100644
--- a/src/calibre/ebooks/html/input.py
+++ b/src/calibre/ebooks/html/input.py
@@ -475,7 +475,9 @@ class HTMLInput(InputFormatPlugin):
             # bhref refers to an already existing file. The read() method of
             # DirContainer will call unquote on it before trying to read the
             # file, therefore we quote it here.
-            item.html_input_href = quote(bhref)
+            if isinstance(bhref, unicode):
+                bhref = bhref.encode('utf-8')
+            item.html_input_href = quote(bhref).decode('utf-8')
             if guessed in self.OEB_STYLES:
                 item.override_css_fetch = partial(
                         self.css_import_handler, os.path.dirname(link))
diff --git a/src/calibre/ebooks/oeb/base.py b/src/calibre/ebooks/oeb/base.py
index bc01cc13cd..8924f1e913 100644
--- a/src/calibre/ebooks/oeb/base.py
+++ b/src/calibre/ebooks/oeb/base.py
@@ -425,15 +425,24 @@ class DirContainer(object):
                     self.opfname = path
                     return
 
+    def _unquote(self, path):
+        # urlunquote must run on a bytestring and will return a bytestring
+        # if it runs on a unicode object, it returns a double encoded unicode
+        # string: unquote(u'%C3%A4') != unquote(b'%C3%A4').decode('utf-8')
+        # and the latter is correct
+        if isinstance(path, unicode):
+            path = path.encode('utf-8')
+        return urlunquote(path).decode('utf-8')
+
     def read(self, path):
         if path is None:
             path = self.opfname
-        path = os.path.join(self.rootdir, path)
-        with open(urlunquote(path), 'rb') as f:
+        path = os.path.join(self.rootdir, self._unquote(path))
+        with open(path, 'rb') as f:
             return f.read()
 
     def write(self, path, data):
-        path = os.path.join(self.rootdir, urlunquote(path))
+        path = os.path.join(self.rootdir, self._unquote(path))
         dir = os.path.dirname(path)
         if not os.path.isdir(dir):
             os.makedirs(dir)
@@ -442,7 +451,7 @@ class DirContainer(object):
 
     def exists(self, path):
         try:
-            path = os.path.join(self.rootdir, urlunquote(path))
+            path = os.path.join(self.rootdir, self._unquote(path))
         except ValueError: #Happens if path contains quoted special chars
             return False
         return os.path.isfile(path)