Merge pull request #798 from Gluejar/dj111-errata

really fix pyepub

core/validation.py

@@ -4,6 +4,7 @@ methods to validate and clean identifiers
 '''
 import re
 import datetime
+import logging
 
 from dateutil.parser import parse
 from PyPDF2 import PdfFileReader
@@ -15,6 +16,8 @@ from regluit.pyepub import EPUB
 from regluit.mobi import Mobi
 from .isbn import ISBN
 
+logger = logging.getLogger(__name__)
+
 ID_VALIDATION = {
     'http': (re.compile(r"(https?|ftp)://(-\.)?([^\s/?\.#]+\.?)+(/[^\s]*)?$",
                         flags=re.IGNORECASE|re.S),
@@ -105,17 +108,21 @@ def test_file(the_file, fformat):
             try:
                 book = EPUB(the_file.file)
             except Exception as e:
+                logger.exception(e)
                 raise ValidationError(_('Are you sure this is an EPUB file?: %s' % e))
         elif fformat == 'mobi':
             try:
                 book = Mobi(the_file.file)
                 book.parse()
             except Exception as e:
-                raise ValidationError(_('Are you sure this is a MOBI file?: %s' % e))
+                logger.exception(e)
+                #raise ValidationError(_('Are you sure this is a MOBI file?: %s' % e))
+                raise e
         elif fformat == 'pdf':
             try:
                 PdfFileReader(the_file.file)
-            except Exception, e:
+            except Exception as e:
+                logger.exception(e)
                 raise ValidationError(_('%s is not a valid PDF file' % the_file.name))
     return True
 

pyepub/__init__.py

@@ -107,7 +107,7 @@ class EPUB(zipfile.ZipFile):
 
         # Iterate over <metadata> section, fill EPUB.info["metadata"] dictionary
         for i in self.opf.find("{0}metadata".format(NAMESPACE["opf"])):
-            if i.tag:
+            if i.tag and isinstance(i.tag, str):
                 tag = ns.sub('', i.tag)
                 if tag not in self.info["metadata"]:
                     self.info["metadata"][tag] = i.text or i.attrib