Only download feeds that have changed (or that don't give us enough data to

[rss2maildir.git] / rss2maildir.py

diff --git a/rss2maildir.py b/rss2maildir.py
index 268d1922561cb8aacde58ba3d0fa7bbec812b8f8..4402c43884ae5271a201de8529ebc46271a5fcdc 100755 (executable)
--- a/rss2maildir.py
+++ b/rss2maildir.py
@@ -20,6 +20,7 @@
 import sys
 import os
 import stat
 import sys
 import os
 import stat

+import httplib

 import urllib
 
 import feedparser
 import urllib
 
 import feedparser


@@ -57,7 +58,7 @@ entities = {
     }
 
 class HTML2Text(HTMLParser):
     }
 
 class HTML2Text(HTMLParser):

-    
+

     def __init__(self):
         self.inheadingone = False
         self.inheadingtwo = False
     def __init__(self):
         self.inheadingone = False
         self.inheadingtwo = False


@@ -88,23 +89,7 @@ class HTML2Text(HTMLParser):
         elif tag.lower() == "a":
             self.inlink = True
         elif tag.lower() == "br":
         elif tag.lower() == "a":
             self.inlink = True
         elif tag.lower() == "br":

-            if self.inparagraph:
-                self.text = self.text \
-                    + u'\n'.join( \
-                        textwrap.wrap(self.currentparagraph, 70)) \
-                    + u'\n'
-                self.currentparagraph = ""
-            elif self.inblockquote:
-                self.text = self.text \
-                    + u'\n> ' \
-                    + u'\n> '.join( \
-                        [a.strip() \
-                            for a in textwrap.wrap(self.blockquote, 68) \
-                        ]) \
-                    + u'\n'
-                self.blockquote = u''
-            else:
-                self.text = self.text + u'\n'
+            self.handle_br()

         elif tag.lower() == "blockquote":
             self.inblockquote = True
             self.text = self.text + u'\n'
         elif tag.lower() == "blockquote":
             self.inblockquote = True
             self.text = self.text + u'\n'


@@ -139,6 +124,9 @@ class HTML2Text(HTMLParser):
 
     def handle_startendtag(self, tag, attrs):
         if tag.lower() == "br":
 
     def handle_startendtag(self, tag, attrs):
         if tag.lower() == "br":

+            self.handle_br()
+
+    def handle_br(self):

             if self.inparagraph:
                 self.text = self.text \
                 + u'\n'.join( \
             if self.inparagraph:
                 self.text = self.text \
                 + u'\n'.join( \


@@ -265,7 +253,57 @@ class HTML2Text(HTMLParser):
         return data
 
 def parse_and_deliver(maildir, url, statedir):
         return data
 
 def parse_and_deliver(maildir, url, statedir):

-    fp = feedparser.parse(url)
+    feedhandle = None
+    headers = None
+    # first check if we know about this feed already
+    feeddb = dbm.open(os.path.join(statedir, "feeds"), "c")
+    # we need all the parts of the url 
+    (type, rest) = urllib.splittype(url)
+    (host, path) = urllib.splithost(rest)
+    (host, port) = urllib.splitport(host)
+    if port == None:
+        port = 80
+    if feeddb.has_key(url):
+        data = feeddb[url]
+        data = cgi.parse_qs(data)
+        # now do a head on the feed to see if it's been updated
+        conn = httplib.HTTPConnection("%s:%s" %(host, port))
+        conn.request("HEAD", path)
+        response = conn.getresponse()
+        headers = response.getheaders()
+        ischanged = False
+        try:
+            for header in headers:
+                if header[0] == "content-length":
+                    if header[1] != data["content-length"][0]:
+                        ischanged = True
+                elif header[0] == "etag":
+                    if header[1] != data["etag"][0]:
+                        ischanged = True
+                elif header[0] == "last-modified":
+                    if header[1] != data["last-modified"][0]:
+                        ischanged = True
+                elif header[0] == "content-md5":
+                    if header[1] != data["content-md5"][0]:
+                        ischanged = True
+        except:
+            ischanged = True
+        if ischanged:
+            conn = httplib.HTTPConnection("%s:%s" %(host, port))
+            conn.request("GET", path)
+            response = conn.getresponse()
+            headers = response.getheaders()
+            feedhandle = response
+        else:
+            return # don't need to do anything, nothings changed.
+    else:
+        conn = httplib.HTTPConnection("%s:%s" %(host, port))
+        conn.request("GET", path)
+        response = conn.getresponse()
+        headers = response.getheaders()
+        feedhandle = response
+
+    fp = feedparser.parse(feedhandle)

     db = dbm.open(os.path.join(statedir, "seen"), "c")
     for item in fp["items"]:
         # have we seen it before?
     db = dbm.open(os.path.join(statedir, "seen"), "c")
     for item in fp["items"]:
         # have we seen it before?


@@ -278,9 +316,13 @@ def parse_and_deliver(maildir, url, statedir):
 
         md5sum = md5.md5(content.encode("utf-8")).hexdigest()
 
 
         md5sum = md5.md5(content.encode("utf-8")).hexdigest()
 

+        prevmessageid = None
+

         if db.has_key(url + "|" + item["link"]):
             data = db[url + "|" + item["link"]]
             data = cgi.parse_qs(data)
         if db.has_key(url + "|" + item["link"]):
             data = db[url + "|" + item["link"]]
             data = cgi.parse_qs(data)

+            if data.has_key("message-id"):
+                prevmessageid = data["message-id"][0]

             if data["contentmd5"][0] == md5sum:
                 continue
 
             if data["contentmd5"][0] == md5sum:
                 continue
 


@@ -303,6 +345,8 @@ def parse_and_deliver(maildir, url, statedir):
         msg.set_unixfrom("\"%s\" <rss2maildir@localhost>" %(url))
         msg.add_header("From", "\"%s\" <rss2maildir@localhost>" %(author))
         msg.add_header("To", "\"%s\" <rss2maildir@localhost>" %(url))
         msg.set_unixfrom("\"%s\" <rss2maildir@localhost>" %(url))
         msg.add_header("From", "\"%s\" <rss2maildir@localhost>" %(author))
         msg.add_header("To", "\"%s\" <rss2maildir@localhost>" %(url))

+        if prevmessageid:
+            msg.add_header("References", prevmessageid)

         createddate = datetime.datetime(*item["updated_parsed"][0:6]) \
             .strftime("%a, %e %b %Y %T -0000")
         msg.add_header("Date", createddate)
         createddate = datetime.datetime(*item["updated_parsed"][0:6]) \
             .strftime("%a, %e %b %Y %T -0000")
         msg.add_header("Date", createddate)


@@ -337,6 +381,8 @@ def parse_and_deliver(maildir, url, statedir):
         os.unlink(fn)
 
         # now add to the database about the item
         os.unlink(fn)
 
         # now add to the database about the item

+        if prevmessageid:
+            messageid = prevmessageid + " " + messageid

         data = urllib.urlencode((
             ("message-id", messageid), \
             ("created", createddate), \
         data = urllib.urlencode((
             ("message-id", messageid), \
             ("created", createddate), \


@@ -344,7 +390,17 @@ def parse_and_deliver(maildir, url, statedir):
             ))
         db[url + "|" + item["link"]] = data
 
             ))
         db[url + "|" + item["link"]] = data
 

+    if headers:
+        data = []
+        for header in headers:
+            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)
+            feeddb[url] = data
+

     db.close()
     db.close()

+    feeddb.close()

 
 # first off, parse the command line arguments
 
 
 # first off, parse the command line arguments