From 1006943d8eb503a21fb1c9e9d83d5612da91f685 Mon Sep 17 00:00:00 2001
From: Damien Elmes <git@ichi2.net>
Date: Sun, 26 Jul 2009 12:38:36 +0900
Subject: [PATCH] SuperMemo import fix from Petr

Fix supermemo import of Q&A part in escaped html. Supermemo 2004/2006 export
can contain unescaped chars > or < that confuse beautifulsoap library. Switch
autocreation of tags from all titles to True.
---
 anki/importing/supermemo_xml.py                    | 14 ++++++++++----
 .../supermemo_ENGLISHFORBEGGINERS_oem_1250.xml     |  4 ++--
 2 files changed, 12 insertions(+), 6 deletions(-)

diff --git a/anki/importing/supermemo_xml.py b/anki/importing/supermemo_xml.py
index b84fcc014..09086cc18 100644
--- a/anki/importing/supermemo_xml.py
+++ b/anki/importing/supermemo_xml.py
@@ -111,7 +111,7 @@ class SupermemoXmlImporter(Importer):
         self.META.resetLearningData  = False            # implemented
         self.META.onlyMemorizedItems = False            # implemented
         self.META.loggerLevel = 2                       # implemented 0no,1info,2error,3debug
-        self.META.tagAllTopics = False
+        self.META.tagAllTopics = True
         self.META.pathsToBeTagged = ['English for begginers', 'Advanced English 97', 'Phrasal Verbs']                # path patterns to be tagged - in gui entered like 'Advanced English 97|My Vocablary'
         self.META.tagMemorizedItems = True              # implemented
         self.META.logToStdOutput   = False              # implemented
@@ -132,10 +132,16 @@ class SupermemoXmlImporter(Importer):
 
     def _decode_htmlescapes(self,s):
         """Unescape HTML code."""
-        from BeautifulSoup import BeautifulStoneSoup
-        #my sm2004 also ecaped & chars in escaped sequences.
+        #In case of bad formated html you can import MinimalSoup etc.. see btflsoup source code
+        from BeautifulSoup import BeautifulStoneSoup as btflsoup    
+
+        #my sm2004 also ecaped & char in escaped sequences.
         s = re.sub(u'&amp;',u'&',s) 
-        return unicode(BeautifulStoneSoup(s,convertEntities=BeautifulStoneSoup.HTML_ENTITIES ))
+        #unescaped solitary chars < or > that were ok for minidom confuse btfl soup
+        s = re.sub(u'>',u'&gt;',s) 
+        s = re.sub(u'<',u'&lt;',s) 
+
+        return unicode(btflsoup(s,convertEntities=btflsoup.HTML_ENTITIES ))
 
 
     def _unescape(self,s,initilize):
diff --git a/tests/importing/supermemo_ENGLISHFORBEGGINERS_oem_1250.xml b/tests/importing/supermemo_ENGLISHFORBEGGINERS_oem_1250.xml
index 19f45a47a..1728c61a7 100644
--- a/tests/importing/supermemo_ENGLISHFORBEGGINERS_oem_1250.xml
+++ b/tests/importing/supermemo_ENGLISHFORBEGGINERS_oem_1250.xml
@@ -202,7 +202,7 @@
             <Type>Item</Type>
 
             <Content>
-              <Question>rozum&amp;#283;t</Question>
+              <Question>&lt;&gt;&lt;TEST&lt;TEST>TES>T&gt;TESTTEST rozum&amp;#283;t</Question>
 
               <Answer>understand [,and&amp;#273;r'st&amp;#281;nd]</Answer>
             </Content>
@@ -926,4 +926,4 @@
       </SuperMemoElement>
     </SuperMemoElement>
   </SuperMemoElement>
-</SuperMemoCollection>
\ No newline at end of file
+</SuperMemoCollection>