Class Newsfeed

pattern/web/__init__.py:2458–2497

# http://www.feedparser.org/

class Newsfeed(SearchEngine):

    def __init__(self, license=None, throttle=1.0, language=None):
        SearchEngine.__init__(self, license, throttle, language)

    def search(self, query, type=NEWS, start=1, count=10, sort=LATEST, size=SMALL, cached=True, **kwargs):
        """ Returns a list of results from the given RSS or Atom newsfeed URL.
        """
        if type != NEWS:
            raise SearchEngineTypeError
        if not query or start < 1 or count < 1:
            return Results(query, query, NEWS)
        # 1) Construct request URL.
        # 2) Parse RSS/Atom response.
        kwargs.setdefault("unicode", True)
        kwargs.setdefault("throttle", self.throttle)
        tags = kwargs.pop("tags", [])
        data = URL(query).download(cached=cached, **kwargs)
        data = feedparser.parse(bytestring(data))
        results = Results(query, query, NEWS)
        results.total = None
        for x in data["entries"][:count]:
            s = "\n\n".join([v.get("value") for v in x.get("content", [])]) or x.get("summary")
            r = Result(url=None)
            r.id       = self.format(x.get("id"))
            r.url      = self.format(x.get("link"))
            r.title    = self.format(x.get("title"))
            r.text     = self.format(s)
            r.date     = self.format(x.get("updated"))
            r.author   = self.format(x.get("author"))
            r.language = self.format(x.get("content") and \
                                x.get("content")[0].get("language") or \
                                               data.get("language"))
            for tag in tags:
                # Parse custom tags.
                # Newsfeed.search(tags=["dc:identifier"]) => Result.dc_identifier.
                tag = tag.replace(":", "_")
                r[tag] = self.format(x.get(tag))
            results.append(r)
        return results

feeds = {
          "Nature": "http://feeds.nature.com/nature/rss/current",