Fix #5662 (< hexa entity problem)

src/calibre/ebooks/conversion/preprocess.py

@@ -14,8 +14,14 @@ XMLDECL_RE    = re.compile(r'^\s*<[?]xml.*?[?]>')
 SVG_NS       = 'http://www.w3.org/2000/svg'
 XLINK_NS     = 'http://www.w3.org/1999/xlink'
 
-convert_entities = functools.partial(entity_to_unicode, exceptions=['quot',
-    'apos', 'lt', 'gt', 'amp', '#60', '#62'])
+convert_entities = functools.partial(entity_to_unicode,
+        result_exceptions = {
+            u'<' : '&lt;',
+            u'>' : '&gt;',
+            u"'" : '&apos;',
+            u'"' : '&quot;',
+            u'&' : '&amp;',
+        })
 _span_pat = re.compile('<span.*?</span>', re.DOTALL|re.IGNORECASE)
 
 LIGATURES = {

src/calibre/ebooks/oeb/base.py

@@ -787,7 +787,6 @@ class Manifest(object):
             data = self.oeb.decode(data)
             data = self.oeb.html_preprocessor(data)
 
-
             # Remove DOCTYPE declaration as it messes up parsing
             # In particular, it causes tostring to insert xmlns
             # declarations, which messes up the coercing logic