Unicode handling of URLs fix

[rss2maildir.git] / rss2maildir.py

diff --git a/rss2maildir.py b/rss2maildir.py
index a51209ce0749e70fd5e9d40ff696d8cf6a2243e1..2af32bc9105a49201e2bbba49d87b5e2cbcda1da 100755 (executable)
--- a/rss2maildir.py
+++ b/rss2maildir.py
@@ -48,42 +48,46 @@ from HTMLParser import HTMLParser
 
 class HTML2Text(HTMLParser):
     entities = {
 
 class HTML2Text(HTMLParser):
     entities = {

-        "amp": "&",
-        "lt": "<",
-        "gt": ">",
-        "pound": "£",
-        "copy": "©",
-        "apos": "'",
-        "quot": "\"",
-        "nbsp": " ",
+        u'amp': "&",
+        u'lt': "<",
+        u'gt': ">",
+        u'pound': "£",
+        u'copy': "©",
+        u'apos': "'",
+        u'quot': "\"",
+        u'nbsp': " ",

         }
 
     blockleveltags = [
         }
 
     blockleveltags = [

-        "h1",
-        "h2",
-        "h3",
-        "h4",
-        "h5",
-        "h6",
-        "pre",
-        "p",
-        "ul",
-        "ol",
-        "dl",
-        "br",
+        u'h1',
+        u'h2',
+        u'h3',
+        u'h4',
+        u'h5',
+        u'h6',
+        u'pre',
+        u'p',
+        u'ul',
+        u'ol',
+        u'dl',
+        u'li',
+        u'dt',
+        u'dd',
+        u'div',
+        #u'blockquote',

         ]
 
     liststarttags = [
         ]
 
     liststarttags = [

-        "ul",
-        "ol",
-        "dl",
+        u'ul',
+        u'ol',
+        u'dl',

         ]
 
     cancontainflow = [
         ]
 
     cancontainflow = [

-        "div",
-        "li",
-        "dd",
-        "blockquote",
+        u'div',
+        u'li',
+        u'dd',
+        u'blockquote',

     ]
 
     def __init__(self,textwidth=70):
     ]
 
     def __init__(self,textwidth=70):


@@ -92,6 +96,9 @@ class HTML2Text(HTMLParser):
         self.textwidth = textwidth
         self.opentags = []
         self.indentlevel = 0
         self.textwidth = textwidth
         self.opentags = []
         self.indentlevel = 0

+        self.ignorenodata = False
+        self.listcount = []
+        self.urls = []

         HTMLParser.__init__(self)
 
     def handle_starttag(self, tag, attrs):
         HTMLParser.__init__(self)
 
     def handle_starttag(self, tag, attrs):


@@ -100,45 +107,142 @@ class HTML2Text(HTMLParser):
             # handle starting a new block - unless we're in a block element
             # that can contain other blocks, we'll assume that we want to close
             # the container
             # handle starting a new block - unless we're in a block element
             # that can contain other blocks, we'll assume that we want to close
             # the container

-            if tag_name == u'br':
+            if len(self.opentags) > 1 and self.opentags[-1] == u'li':

                 self.handle_curdata()
                 self.handle_curdata()

-                self.opentags.append(tag_name)
-                self.opentags.pop()
+
+            if tag_name == u'ol':
+                self.handle_curdata()
+                self.listcount.append(1)
+                self.listlevel = len(self.listcount) - 1
+
+            if tag_name in self.liststarttags:
+                smallist = self.opentags[-3:-1]
+                smallist.reverse()
+                for prev_listtag in smallist:
+                    if prev_listtag in [u'dl', u'ol']:
+                        self.indentlevel = self.indentlevel + 4
+                        break
+                    elif prev_listtag == u'ul':
+                        self.indentlevel = self.indentlevel + 3
+                        break

 
             if len(self.opentags) > 0:
                 self.handle_curdata()
 
             if len(self.opentags) > 0:
                 self.handle_curdata()

-                self.opentags.pop()
+                if tag_name not in self.cancontainflow:
+                    self.opentags.pop()

             self.opentags.append(tag_name)
         else:
             self.opentags.append(tag_name)
         else:

-            self.handle_curdata()
-            self.opentags.append(tag_name)
+            if tag_name == "span":
+                return
+            listcount = 0
+            try:
+                listcount = self.listcount[-1]
+            except:
+                pass
+
+            if tag_name == u'dd' and len(self.opentags) > 1 \
+                and self.opentags[-1] == u'dt':
+                self.handle_curdata()
+                self.opentags.pop()
+            elif tag_name == u'dt' and len(self.opentags) > 1 \
+                and self.opentags[-1] == u'dd':
+                self.handle_curdata()
+                self.opentags.pop()
+            elif tag_name == u'a':
+                for attr in attrs:
+                    if attr[0].lower() == u'href':
+                        self.urls.append(attr[1].decode('utf-8'))
+                self.curdata = self.curdata + u'`'
+                self.opentags.append(tag_name)
+                return
+            elif tag_name == u'img':
+                self.handle_image(attrs)
+                return
+            elif tag_name == u'br':
+                self.handle_br()
+                return
+            else:
+                # we don't know the tag, so lets avoid handling it!
+                return 

 
     def handle_startendtag(self, tag, attrs):
         if tag.lower() == u'br':
 
     def handle_startendtag(self, tag, attrs):
         if tag.lower() == u'br':

-            self.tags.append(u'br')
-            self.handle_curdata() # just handle the data, don't do anything else
-            self.tags.pop()
+            self.handle_br()
+        elif tag.lower() == u'img':
+            self.handle_image(attrs)
+            return
+
+    def handle_br(self):
+            self.handle_curdata()
+            self.opentags.append(u'br')
+            self.handle_curdata()
+            self.opentags.pop()
+
+    def handle_image(self, attrs):
+        alt = u''
+        url = u''
+        for attr in attrs:
+            if attr[0] == 'alt':
+                alt = attr[1].decode('utf-8')
+            elif attr[0] == 'src':
+                url = attr[1].decode('utf-8')
+        if url:
+            self.curdata = self.curdata \
+                + u' [img:' \
+                + url
+            if alt:
+                self.curdata = self.curdata \
+                    + u'(' \
+                    + alt \
+                    + u')'
+            self.curdata = self.curdata \
+                + u']'

 
     def handle_curdata(self):
 
     def handle_curdata(self):

+

         if len(self.opentags) == 0:
             return
 
         if len(self.opentags) == 0:
             return
 

+        tag_thats_done = self.opentags[-1]
+

         if len(self.curdata) == 0:
             return
 
         if len(self.curdata) == 0:
             return
 

-        tag_thats_done = self.opentags[-1]
+        if tag_thats_done == u'br':
+            if len(self.text) == 0 or self.text[-1] != '\n':
+                self.text = self.text + '\n'
+                self.ignorenodata = True
+            return
+
+        if len(self.curdata.strip()) == 0:
+            return

 
         if tag_thats_done in self.blockleveltags:
             newlinerequired = self.text != u''
 
         if tag_thats_done in self.blockleveltags:
             newlinerequired = self.text != u''

+            if self.ignorenodata:
+                newlinerequired = False
+            self.ignorenodata = False

             if newlinerequired:
             if newlinerequired:

-                self.text = self.text + u'\n\n'
+                if tag_thats_done in [u'dt', u'dd', u'li'] \
+                    and len(self.text) > 1 \
+                    and self.text[-1] != u'\n':
+                        self.text = self.text + u'\n'
+                elif len(self.text) > 2 \
+                    and self.text[-1] != u'\n' \
+                    and self.text[-2] != u'\n':
+                    self.text = self.text + u'\n\n'

 
         if tag_thats_done in ["h1", "h2", "h3", "h4", "h5", "h6"]:
             underline = u''
             underlinechar = u'='
 
         if tag_thats_done in ["h1", "h2", "h3", "h4", "h5", "h6"]:
             underline = u''
             underlinechar = u'='

-            headingtext = self.curdata.encode("utf-8").strip()
-            headingtext = u'\n'.join( \
-                textwrap.wrap(headingtext, self.textwidth))
+            headingtext = self.curdata
+            seperator = u'\n' + u' '*self.indentlevel
+            headingtext = seperator.join( \
+                textwrap.wrap( \
+                    headingtext, \
+                    self.textwidth - self.indentlevel \
+                    ) \
+                )

 
             if tag_thats_done == u'h2':
                 underlinechar = u'-'
 
             if tag_thats_done == u'h2':
                 underlinechar = u'-'


@@ -146,74 +250,148 @@ class HTML2Text(HTMLParser):
                 underlinechar = u'~'
 
             if u'\n' in headingtext:
                 underlinechar = u'~'
 
             if u'\n' in headingtext:

-                underline = underlinechar * self.textwidth
+                underline = u' ' * self.indentlevel \
+                    + underlinechar * (self.textwidth - self.indentlevel)

             else:
             else:

-                underline = underlinechar * len(headingtext)
+                underline = u' ' * self.indentlevel \
+                    + underlinechar * len(headingtext)

             self.text = self.text \
             self.text = self.text \

-                + headingtext.encode("utf-8") + u'\n' \
+                + headingtext + u'\n' \

                 + underline
                 + underline

-        elif tag_thats_done == "p":
-            paragraph = self.curdata.encode("utf-8").strip()
+        elif tag_thats_done in [u'p', u'div']:
+            paragraph = unicode( \
+                self.curdata.strip().encode("utf-8"), "utf-8")
+            seperator = u'\n' + u' ' * self.indentlevel

             self.text = self.text \
             self.text = self.text \

-                + u'\n'.join(textwrap.wrap(paragraph, self.textwidth))
+                + u' ' * self.indentlevel \
+                + seperator.join( \
+                    textwrap.wrap( \
+                        paragraph, self.textwidth - self.indentlevel))

         elif tag_thats_done == "pre":
         elif tag_thats_done == "pre":

-            self.text = self.text + self.curdata
-        elif tag_thats_done == "blockquote":
-            quote = self.curdata.encode("utf-8").strip()
-            self.text = self.text \
-                + u'> ' \
-                + u'> '.join(textwrap.wrap(quote, self.textwidth - 2))
-        elif tag_thats_done == "li":
-            item = self.curdata.encode("utf-8").strip()
+            self.text = self.text + unicode( \
+                self.curdata.encode("utf-8"), "utf-8")
+        elif tag_thats_done == u'blockquote':
+            quote = unicode( \
+                self.curdata.encode("utf-8").strip(), "utf-8")
+            seperator = u'\n' + u' ' * self.indentlevel + u'> '

             if len(self.text) > 0 and self.text[-1] != u'\n':
                 self.text = self.text + u'\n'
             self.text = self.text \
             if len(self.text) > 0 and self.text[-1] != u'\n':
                 self.text = self.text + u'\n'
             self.text = self.text \

-                + u' * ' \
-                + u'\n   '.join( \
-                    textwrap.wrap(item, self.textwidth - 3))
+                + u'> ' \
+                + seperator.join( \
+                    textwrap.wrap( \
+                        quote, \
+                        self.textwidth - self.indentlevel - 2 \
+                    )
+                )

             self.curdata = u''
             self.curdata = u''

-        elif tag_thats_done == "dt":
-            definition = self.curdata.encode("utf-8").strip()
+        elif tag_thats_done == "li":
+            item = unicode(self.curdata.encode("utf-8").strip(), "utf-8")

             if len(self.text) > 0 and self.text[-1] != u'\n':
             if len(self.text) > 0 and self.text[-1] != u'\n':

-                self.text = self.text + u'\n\n'
-            elif len(self.text) > 0 and self.text[-2] != u'\n':

                 self.text = self.text + u'\n'
                 self.text = self.text + u'\n'

-            definition = definition + "::"
+            # work out if we're in an ol rather than a ul
+            latesttags = self.opentags[-4:]
+            latesttags.reverse()
+            isul = None
+            for thing in latesttags:
+                if thing == 'ul':
+                    isul = True
+                    break
+                elif thing == 'ol':
+                    isul = False
+                    break
+
+            listindent = 3
+            if not isul:
+                listindent = 4
+
+            listmarker = u' * '
+            if isul == False:
+                listmarker = u' %2d. ' %(self.listcount[-1])
+                self.listcount[-1] = self.listcount[-1] + 1
+
+            seperator = u'\n' \
+                + u' ' * self.indentlevel \
+                + u' ' * listindent

             self.text = self.text \
             self.text = self.text \

-                + '\n '.join(
-                    textwrap.wrap(definition, self.textwidth - 1))
+                + u' ' * self.indentlevel \
+                + listmarker \
+                + seperator.join( \
+                    textwrap.wrap( \
+                        item, \
+                        self.textwidth - self.indentlevel - listindent \
+                    ) \
+                )

             self.curdata = u''
             self.curdata = u''

-        elif tag_thats_done == "dd":
-            definition = self.curdata.encode("utf-8").strip()
+        elif tag_thats_done == u'dt':
+            definition = unicode(self.curdata.encode("utf-8").strip(), "utf-8")

             if len(self.text) > 0 and self.text[-1] != u'\n':
             if len(self.text) > 0 and self.text[-1] != u'\n':

+                self.text = self.text + u'\n\n'
+            elif len(self.text) > 1 and self.text[-2] != u'\n':

                 self.text = self.text + u'\n'
                 self.text = self.text + u'\n'

+            definition = u' ' * self.indentlevel + definition + "::"
+            indentstring = u'\n' + u' ' * (self.indentlevel + 1)

             self.text = self.text \
             self.text = self.text \

-                + '    ' \
-                + '\n    '.join( \
-                    textwrap.wrap(definition, self.textwidth - 4))
+                + indentstring.join(
+                    textwrap.wrap(definition, \
+                        self.textwidth - self.indentlevel - 1))

             self.curdata = u''
             self.curdata = u''

+        elif tag_thats_done == u'dd':
+            definition = unicode(self.curdata.encode("utf-8").strip(), "utf-8")
+            if len(definition) > 0:
+                if len(self.text) > 0 and self.text[-1] != u'\n':
+                    self.text = self.text + u'\n'
+                indentstring = u'\n' + u' ' * (self.indentlevel + 4)
+                self.text = self.text \
+                    + u' ' * (self.indentlevel + 4) \
+                    + indentstring.join( \
+                        textwrap.wrap( \
+                            definition, \
+                            self.textwidth - self.indentlevel - 4 \
+                            ) \
+                        )
+                self.curdata = u''
+        elif tag_thats_done == u'a':
+            self.curdata = self.curdata + u'`__'
+            pass

         elif tag_thats_done in self.liststarttags:
             pass
         elif tag_thats_done in self.liststarttags:
             pass

-        else:
-            # we've got no idea what this tag does, so we'll
-            # make an assumption that we're not going to know later
-            if len(self.curdata) > 0:
-                self.text = self.text \
-                    + u' ... ' \
-                    + u'\n ... '.join( \
-                        textwrap.wrap(self.curdata, self.textwidth - 5))
-            self.curdata = u''

 
         if tag_thats_done in self.blockleveltags:
             self.curdata = u''
 
 
         if tag_thats_done in self.blockleveltags:
             self.curdata = u''
 

+        self.ignorenodata = False
+

     def handle_endtag(self, tag):
     def handle_endtag(self, tag):

+        self.ignorenodata = False
+        if tag == "span":
+            return
+

         try:
             tagindex = self.opentags.index(tag)
         except:
         try:
             tagindex = self.opentags.index(tag)
         except:

-            # closing tag we know nothing about.
-            # err. weird.
-            tagindex = 0
+            return
+        tag = tag.lower()
+
+        if tag in [u'br', u'img']:
+            return
+
+        if tag in self.liststarttags:
+            if tag in [u'ol', u'dl', u'ul']:
+                self.handle_curdata()
+                # find if there was a previous list level
+                smalllist = self.opentags[:-1]
+                smalllist.reverse()
+                for prev_listtag in smalllist:
+                    if prev_listtag in [u'ol', u'dl']:
+                        self.indentlevel = self.indentlevel - 4
+                        break
+                    elif prev_listtag == u'ul':
+                        self.indentlevel = self.indentlevel - 3
+                        break
+
+        if tag == u'ol':
+            self.listcount = self.listcount[:-1]

 
         while tagindex < len(self.opentags) \
             and tag in self.opentags[tagindex+1:]:
 
         while tagindex < len(self.opentags) \
             and tag in self.opentags[tagindex+1:]:


@@ -226,10 +404,16 @@ class HTML2Text(HTMLParser):
             # Assuming the data was for the last opened tag first
             self.handle_curdata()
             # Now kill the list to be a slice before this tag was opened
             # Assuming the data was for the last opened tag first
             self.handle_curdata()
             # Now kill the list to be a slice before this tag was opened

-            self.opentags = self.opentags[:tagindex]
+            self.opentags = self.opentags[:tagindex + 1]
+        else:
+            self.handle_curdata()
+            if self.opentags[-1] == tag:
+                self.opentags.pop()

 
     def handle_data(self, data):
 
     def handle_data(self, data):

-        self.curdata = self.curdata + unicode(data, "utf-8")
+        if len(self.opentags) == 0:
+            self.opentags.append(u'p')
+        self.curdata = self.curdata + data.decode("utf-8")

 
     def handle_entityref(self, name):
         entity = name
 
     def handle_entityref(self, name):
         entity = name


@@ -251,6 +435,9 @@ class HTML2Text(HTMLParser):
             while len(self.text) > 1 and self.text[-1] == u'\n':
                 self.text = self.text[:-1]
             self.text = self.text + u'\n'
             while len(self.text) > 1 and self.text[-1] == u'\n':
                 self.text = self.text[:-1]
             self.text = self.text + u'\n'

+        if len(self.urls) > 0:
+            self.text = self.text + u'\n__ ' + u'\n__ '.join(self.urls) + u'\n'
+            self.urls = []

         return self.text
 
 def open_url(method, url):
         return self.text
 
 def open_url(method, url):


@@ -456,7 +643,8 @@ def parse_and_deliver(maildir, url, statedir):
     if headers:
         data = []
         for header in headers:
     if headers:
         data = []
         for header in headers:

-            if header[0] in ["content-md5", "etag", "last-modified", "content-length"]:
+            if header[0] in \
+                ["content-md5", "etag", "last-modified", "content-length"]:

                 data.append((header[0], header[1]))
         if len(data) > 0:
             data = urllib.urlencode(data)
                 data.append((header[0], header[1]))
         if len(data) > 0:
             data = urllib.urlencode(data)


@@ -535,11 +723,13 @@ if __name__ == "__main__":
     elif scp.has_option("general", "state_dir"):
         new_state_dir = scp.get("general", "state_dir")
         try:
     elif scp.has_option("general", "state_dir"):
         new_state_dir = scp.get("general", "state_dir")
         try:

-            mode = os.stat(state_dir)[stat.ST_MODE]
+            mode = os.stat(new_state_dir)[stat.ST_MODE]

             if not stat.S_ISDIR(mode):
                 sys.stderr.write( \
                     "State directory (%s) is not a directory\n" %(state_dir))
                 sys.exit(1)
             if not stat.S_ISDIR(mode):
                 sys.stderr.write( \
                     "State directory (%s) is not a directory\n" %(state_dir))
                 sys.exit(1)

+            else:
+                state_dir = new_state_dir

         except:
             # try to create it
             try:
         except:
             # try to create it
             try: