Add basic HTML -> plain text parser

[rss2maildir.git] / rss2maildir.py

diff --git a/rss2maildir.py b/rss2maildir.py
index b48498efd49801f4cee0554e7c2b3b06e7eed986..ff7be81b84c4f0cdb97c1b95d7b1742a1f9d96ee 100755 (executable)
--- a/rss2maildir.py
+++ b/rss2maildir.py
@@ -1,4 +1,5 @@
 #!/usr/bin/python
 #!/usr/bin/python

+# coding=utf8

 
 import mailbox
 import sys
 
 import mailbox
 import sys


@@ -8,33 +9,144 @@ import urllib
 
 import feedparser
 
 
 import feedparser
 

-import email
+from email.MIMEMultipart import MIMEMultipart
+from email.MIMEText import MIMEText

 
 import datetime
 import random
 import string
 
 
 import datetime
 import random
 import string
 

+import socket
+

 from optparse import OptionParser
 from ConfigParser import SafeConfigParser
 
 from optparse import OptionParser
 from ConfigParser import SafeConfigParser
 

+from base64 import b64encode
+import md5
+
+import cgi
+import dbm
+
+from HTMLParser import HTMLParser
+
+entities = {
+    "amp": "&",
+    "lt": "<",
+    "gt": ">",
+    "pound": "£",
+    "copy": "©",
+    "apos": "'",
+    "quote": "\"",
+    }
+
+class HTML2Text(HTMLParser):
+    
+    def __init__(self):
+        self.inheadingone = False
+        self.inheadingtwo = False
+        self.inotherheading = False
+        self.inlink = False
+        self.text = ""
+        self.headingtext = ""
+        HTMLParser.__init__(self)
+
+    def handle_starttag(self, tag, attrs):
+        if tag.lower() == "h1":
+            self.inheadingone = True
+        elif tag.lower() == "h2":
+            self.inheadingtwo = True
+        elif tag.lower() in ["h3", "h4", "h5", "h6"]:
+            self.inotherheading = True
+        elif tag.lower() == "a":
+            self.inlink = True
+        elif tag.lower() == "br":
+            self.text = self.text + "\n"
+        elif tag.lower() == "p":
+            if self.text != "":
+                self.text = self.text + "\n\n"
+
+    def handle_startendtag(self, tag, attrs):
+        if tag.lower() == "br":
+            self.text = self.text + "\n"
+
+    def handle_endtag(self, tag):
+        if tag.lower() == "h1":
+            self.inheadingone = False
+            self.text = self.text + self.headingtext + "\n" + "=" * len(self.headingtext)
+            self.headingtext = ""
+        elif tag.lower() == "h2":
+            self.inheadingtwo = False
+            self.text = self.text + self.headingtext + "\n" + "-" * len(self.headingtext)
+            self.headingtext = ""
+        elif tag.lower() in ["h3", "h4", "h5", "h6"]:
+            self.inotherheading = False
+            self.text = self.text + self.headingtext + "\n" + "~" * len(self.headingtext)
+            self.headingtext = ""
+
+    def handle_data(self, data):
+        if not self.inheadingone and not self.inheadingtwo and not self.inotherheading:
+            self.text = self.text + data.strip() + " "
+        else:
+            self.headingtext = self.headingtext + data.strip() + " "
+
+    def handle_entityref(self, name):
+        if entities.has_key(name.lower()):
+            self.text = self.text + entities[name.lower()]
+        else:
+            self.text = self.text + "&" + name + ";"
+
+    def gettext(self):
+        return self.text

 
 def parse_and_deliver(maildir, url, statedir):
     md = mailbox.Maildir(maildir)
     fp = feedparser.parse(url)
 
 def parse_and_deliver(maildir, url, statedir):
     md = mailbox.Maildir(maildir)
     fp = feedparser.parse(url)

+    db = dbm.open(os.path.join(statedir, "seen"), "c")

     for item in fp["items"]:
     for item in fp["items"]:

-        # things that we need in the message
-        msg = email.message_from_string("")
+        # have we seen it before?
+        # need to work out what the content is first...
+
+        if item.has_key("content"):
+            content = item["content"][0]["value"]
+        else:
+            content = item["summary"]
+
+        md5sum = md5.md5(content.encode("utf8")).hexdigest()
+
+        if db.has_key(item["link"]):
+            data = db[item["link"]]
+            data = cgi.parse_qs(data)
+            if data["contentmd5"][0] == md5sum:
+                continue
+
+        try:
+            author = item["author"]
+        except:
+            author = url
+
+        # create a basic email message
+        msg = MIMEMultipart("alternative")
+        messageid = "<" + datetime.datetime.now().strftime("%Y%m%d%H%M") + "." + "".join([random.choice(string.ascii_letters + string.digits) for a in range(0,6)]) + "@" + socket.gethostname() + ">"
+        msg.add_header("Message-ID", messageid)
+        msg.set_unixfrom("\"%s\" <rss2maildir@localhost>" %(url))
+        msg.add_header("From", "\"%s\" <rss2maildir@localhost>" %(author))
+        msg.add_header("To", "\"%s\" <rss2maildir@localhost>" %(url))
+        createddate = datetime.datetime(*item["updated_parsed"][0:6]).strftime("%a, %e %b %Y %T -0000")
+        msg.add_header("Date", createddate)

         msg.add_header("Subject", item["title"])
         msg.add_header("Subject", item["title"])

-        msg.set_unixfrom("Brett Parker <iDunno@sommitrealweird.co.uk>")
-        msg.add_header("Date", datetime.datetime(*item["created_parsed"][0:6]).strftime("%a, %e %b %Y %T -0000"))
-        msg.add_header("To", url)
-        msg.set_payload(item["content"][0]["value"])
-        msg.set_charset("utf8")

         msg.set_default_type("text/plain")
 
         msg.set_default_type("text/plain")
 

+        htmlpart = MIMEText(content.encode("utf8"), "html", "utf8")
+        textparser = HTML2Text()
+        textparser.feed(content.encode("utf8"))
+        textcontent = textparser.gettext()
+        textpart = MIMEText(textcontent, "plain", "utf8")
+        msg.attach(textpart)
+        msg.attach(htmlpart)
+

         # start by working out the filename we should be writting to, we do
         # this following the normal maildir style rules
         # start by working out the filename we should be writting to, we do
         # this following the normal maildir style rules

-        fname = str(os.getpid()) + "".join([random.choice(string.ascii_letters + string.digits) for a in range(0,10)]) + datetime.datetime.now().strftime('%s')
+        fname = str(os.getpid()) + "." + socket.gethostname() + "." + "".join([random.choice(string.ascii_letters + string.digits) for a in range(0,10)]) + "." + datetime.datetime.now().strftime('%s')

         fn = os.path.join(maildir, "tmp", fname)
         fh = open(fn, "w")
         fh.write(msg.as_string())
         fn = os.path.join(maildir, "tmp", fname)
         fh = open(fn, "w")
         fh.write(msg.as_string())


@@ -44,6 +156,12 @@ def parse_and_deliver(maildir, url, statedir):
         os.link(fn, newfn)
         os.unlink(fn)
 
         os.link(fn, newfn)
         os.unlink(fn)
 

+        # now add to the database about the item
+        data = urllib.urlencode((("message-id", messageid), ("created", createddate), ("contentmd5", md5sum)))
+        db[item["link"]] = data
+
+    db.close()
+

 # first off, parse the command line arguments
 
 oparser = OptionParser()
 # first off, parse the command line arguments
 
 oparser = OptionParser()


@@ -122,6 +240,18 @@ elif scp.has_option("general", "state_dir"):
         except:
             sys.stderr.write("Couldn't create state directory %s\n" %(new_state_dir))
             sys.exit(1)
         except:
             sys.stderr.write("Couldn't create state directory %s\n" %(new_state_dir))
             sys.exit(1)

+else:
+    try:
+        mode = os.stat(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)
+    except:
+        try:
+            os.mkdir(state_dir)
+        except:
+            sys.stderr.write("State directory %s could not be created\n" %(state_dir))
+            sys.exit(1)

 
 if scp.has_option("general", "maildir_root"):
     maildir_root = scp.get("general", "maildir_root")
 
 if scp.has_option("general", "maildir_root"):
     maildir_root = scp.get("general", "maildir_root")