I don’t think RSS is dead. But we can see how many people is using social networking sites to get their information. For this reason I was publishing the entries of my blogs using services like IFTTT and dlvr.it. They are easy to use and they work pretty well. Nevertheless, one is always wondering if we could prepare our own programs to manage these publications and learn something new on the way.
I started with Facebook publishing but I’m presenting here a program for Twitterpublishing: we only need to publish the title and the link (and, maybe, some introductory text).
I found the project twitter as an starting point. It has implemented an important part of the work. We can install it using pip:
fernand0@here:~$ sudo pip install twitter
It needs `BeautifulSoup` and maybe some other modules. If they are not available in our system we will get the adequate ‘complaints’.
Now we can execute it.
This step is useful in order to do the authentication steps in Twitter and getting the oauth token. Our program will not deal with this part and it will be smaller and more simple.
Not so long ago it was possible to send tweets with just the username and passowrd but Twitter decided to start using a more sofisticated systema based in OAuth.
The program launches a browser for authentication and then giving our app the adequate permissions. This generates the tokens and other information needed to interact with Twitter. They will be stored at `~/.twitter_oauth` (in a Unix-like system, I’d be happy to know about other systems) that we will reuse in our own application.
The program is quite simple, it can be downloaded from rssToTwitter.py V.2014-12-07) (link to the commented version, the program has been updated to correct bugs and add features).
We will start reading the configuration:
config = ConfigParser.ConfigParser() config.read([os.path.expanduser('~/.rssBlogs')]) rssFeed = config.get("Blog1", "rssFeed") twitterAc = config.get("Blog1", "twitterAc")
This configuration file must contain a section for each blog (this program uses only the configuration for the first one). Each section will contain the RSS feed, the name of the Twitter account and the name of the Facebook account (it can be empty if it won’t be used. For example, for this blog it would be:
It also needs the Twitter configuration:
config.read([os.path.expanduser('~/.rssTwitter')]) CONSUMER_KEY = config.get("appKeys", "CONSUMER_KEY") CONSUMER_SECRET = config.get("appKeys", "CONSUMER_SECRET") TOKEN_KEY = config.get(twitterAc, "TOKEN_KEY") TOKEN_SECRET = config.get(twitterAc, "TOKEN_SECRET")
We can use the ones that have been generated before (we can copy them from the app); in my system it is at: `/usr/local/lib/python2.7/dist-packages/twitter/cmdline.py` and the tokens are stored at `~/.twitter_oauth`
The configuration file is as follows:
Notice that you can configure as many Twitter accounts as needed. The name of the second section is the same as the one used in the previous configuration file.
Now we can read the RSS feed in order to extract the required data:
feed = feedparser.parse(rssFeed) i = 0 # It will publish the last added item soup = BeautifulSoup(feed.entries[i].title) theTitle = soup.get_text() theLink = feed.entries[i].link For this, we will use `feedparser` in order to download the RSS feed and process it. We are chosing the first entry (position 0), that will be the last one published. For Twitter we just need the title and the link. We use BeautifulSoup for processing the title, in order to avoid the tags (para evitar las posibles etiquetas que pueda contener (CSS, HTLL entities, ...) And finally, we will build the tweet: statusTxt = "Publicado: "+theTitle+" "+theLink
We can now proceed to the steps of identification, authentication and publishing:
t = Twitter( auth=OAuth(TOKEN_KEY, TOKEN_SECRET, CONSUMER_KEY, CONSUMER_SECRET)) t.statuses.update(status=statusTxt)
This entry was originally published (in Spanish) at: Publicar en Twitter las entradas de este sitio.