X-Git-Url: https://git.sommitrealweird.co.uk/rss2maildir.git/blobdiff_plain/82cab2ee7efbfde9b74a65ff10d0c4d1c05a0d7b..HEAD:/rss2maildir.py

diff --git a/rss2maildir.py b/rss2maildir.py
index 8a59c85..314a9c1 100755
--- a/rss2maildir.py
+++ b/rss2maildir.py
@@ -40,6 +40,8 @@ from ConfigParser import SafeConfigParser
 
 from base64 import b64encode
 
+import chardet
+
 if sys.version_info[0] == 2 and sys.version_info[1] >= 6:
     import hashlib as md5
 else:
@@ -275,7 +277,7 @@ class HTML2Text(HTMLParser):
             elif tag_name == u'a':
                 for attr in attrs:
                     if attr[0].lower() == u'href':
-                        self.urls.append(attr[1].decode('utf-8'))
+                        self.urls.append(attr[1])
                 self.curdata = self.curdata + u'`'
                 self.opentags.append(tag_name)
                 return
@@ -307,9 +309,15 @@ class HTML2Text(HTMLParser):
         url = u''
         for attr in attrs:
             if attr[0] == 'alt':
-                alt = attr[1].decode('utf-8')
+                if isinstance(attr[1], str):
+                    alt = u'%s' %(attr[1])
+                else:
+                    alt = attr[1]
             elif attr[0] == 'src':
-                url = attr[1].decode('utf-8')
+                if isinstance(attr[1], str):
+                    url = u'%s' %(attr[1])
+                else:
+                    url = attr[1]
         if url:
             if alt:
                 if self.images.has_key(alt):
@@ -560,7 +568,7 @@ class HTML2Text(HTMLParser):
     def handle_data(self, data):
         if len(self.opentags) == 0:
             self.opentags.append(u'p')
-        self.curdata = self.curdata + data.decode("utf-8")
+        self.curdata = "%s%s" %(self.curdata, data)
 
     def handle_charref(self, name):
         try:
@@ -701,9 +709,18 @@ def parse_and_deliver(maildir, url, statedir):
 
         md5sum = md5.md5(content.encode("utf-8")).hexdigest()
 
+        # make sure content is unicode encoded
+        if not isinstance(content, unicode):
+            cd_res = chardet.detect(content)
+            chrset = cd_res['encoding']
+            print "detected charset %s for item %s" %(chrset, item["link"])
+            content = content.decode(chrset)
+
         prevmessageid = None
 
         db_guid_key = None
+        if not item.has_key("link"):
+            item["link"] = u'#' + md5sum
         db_link_key = (url + u'|' + item["link"]).encode("utf-8")
 
         # check if there's a guid too - if that exists and we match the md5,
@@ -759,7 +776,7 @@ def parse_and_deliver(maildir, url, statedir):
         title = item["title"]
         title = re.sub(u'<', u'&lt;', title)
         title = re.sub(u'>', u'&gt;', title)
-        subj_gen.feed(title.encode("utf-8"))
+        subj_gen.feed(title)
         msg.add_header("Subject", subj_gen.gettext())
         msg.set_default_type("text/plain")
 
@@ -770,7 +787,7 @@ def parse_and_deliver(maildir, url, statedir):
             item["link"] )
         htmlpart = MIMEText(htmlcontent.encode("utf-8"), "html", "utf-8")
         textparser = HTML2Text()
-        textparser.feed(content.encode("utf-8"))
+        textparser.feed(content)
         textcontent = textparser.gettext()
         textcontent = "%s\n\nItem URL: %s" %( \
             textcontent, \