#! /usr/local/bin/python #-*- coding: utf-8 -*- __author__ = "Cedric Bonhomme" __version__ = "$Revision: 0.5 $" __date__ = "$Date: 2010/02/02 $" __copyright__ = "Copyright (c) 2010 Cedric Bonhomme" __license__ = "GPLv3" import re import sqlite3 import hashlib import threading import feedparser from datetime import datetime url_finders = [ \ re.compile("([0-9]{1,3}\\.[0-9]{1,3}\\.[0-9]{1,3}\\.[0-9]{1,3}|(((news|telnet|nttp|file|http|ftp|https)://)|(www|ftp)[-A-Za-z0-9]*\\.)[-A-Za-z0-9\\.]+)(:[0-9]*)?/[-A-Za-z0-9_\\$\\.\\+\\!\\*\\(\\),;:@&=\\?/~\\#\\%]*[^]'\\.}>\\),\\\"]"), \ re.compile("([0-9]{1,3}\\.[0-9]{1,3}\\.[0-9]{1,3}\\.[0-9]{1,3}|(((news|telnet|nttp|file|http|ftp|https)://)|(www|ftp)[-A-Za-z0-9]*\\.)[-A-Za-z0-9\\.]+)(:[0-9]*)?"), \ re.compile("(~/|/|\\./)([-A-Za-z0-9_\\$\\.\\+\\!\\*\\(\\),;:@&=\\?/~\\#\\%]|\\\\)+"), \ re.compile("'\\<((mailto:)|)[-A-Za-z0-9\\.]+@[-A-Za-z0-9\\.]+"), \ ] feeds_list = [] list_of_threads = [] class FeedGetter(object): """ """ def __init__(self): """ Initializes the base and variables. """ try: self.feeds_file = open("./var/feed.lst") except: print "./feed.lst not found" exit(0) # Create the base if not exists. self.conn = sqlite3.connect("./var/feed.db", isolation_level = None) self.c = self.conn.cursor() self.c.execute('''create table if not exists rss_feed (article_date text, article_title text, \ article_link text PRIMARY KEY, article_description text, \ feed_title text, feed_site_link text)''') self.conn.commit() self.c.close() # mutex to protect the SQLite base self.locker = threading.Lock() def retrieve_feed(self): """ Parse the file 'feeds.lst' and launch a thread for each RSS feed. """ for a_feed in self.feeds_file.readlines(): # test if the URL is well formed for url_regexp in url_finders: if url_regexp.match(a_feed): the_good_url = url_regexp.match(a_feed).group(0).replace("\n", "") try: # launch a new thread for the RSS feed thread = threading.Thread(None, self.process, \ None, (the_good_url,)) thread.start() list_of_threads.append(thread) except: pass break # wait for all threads are done for th in list_of_threads: th.join() def process(self, the_good_url): """Request the URL Executed in a thread. SQLite objects created in a thread can only be used in that same thread ! """ # Protect this part of code. self.locker.acquire() self.conn = sqlite3.connect("./var/feed.db", isolation_level = None) self.c = self.conn.cursor() # Add the articles in the base. self.add_into_sqlite(feedparser.parse(the_good_url)) self.conn.commit() self.c.close() # Release this part of code. self.locker.release() def add_into_sqlite(self, a_feed): """ Add the articles of the feed 'a_feed' in the SQLite base. """ for article in a_feed['entries']: try: description = article.description.encode('utf-8') except Exception, e: description = "" sha256_hash = hashlib.sha256() sha256_hash.update(article.link.encode('utf-8')) article_id = sha256_hash.hexdigest() try: self.c.execute('insert into rss_feed values (?,?,?,?,?,?)', (\ datetime(*article.updated_parsed[:6]), \ article.title.encode('utf-8'), \ article.link.encode('utf-8'), \ description, \ a_feed.feed.title.encode('utf-8'), \ a_feed.feed.link.encode('utf-8'))) except sqlite3.IntegrityError: pass if __name__ == "__main__": # Point of entry in execution mode feed_getter = FeedGetter() feed_getter.retrieve_feed()