Class MediaWiki

pattern/web/__init__.py:1616–1754

MEDIAWIKI_REFERENCE = r"\s*\[[0-9]{1,3}\]"

class MediaWiki(SearchEngine):

    def __init__(self, license=None, throttle=5.0, language="en"):
        SearchEngine.__init__(self, license or MEDIAWIKI_LICENSE, throttle, language)

    @property
    def _url(self):
        # Must be overridden in a subclass; see Wikia and Wikipedia.
        return None

    @property
    def MediaWikiArticle(self):
        return MediaWikiArticle
    @property
    def MediaWikiSection(self):
        return MediaWikiSection
    @property
    def MediaWikiTable(self):
        return MediaWikiTable

    def __iter__(self):
        return self.all()

    def all(self, **kwargs):
        """ Returns an iterator over all MediaWikiArticle objects.
            Optional parameters can include those passed to
            MediaWiki.list(), MediaWiki.search() and URL.download().
        """
        for title in self.list(**kwargs):
            yield self.search(title, **kwargs)

    articles = all

    def list(self, namespace=0, start=None, count=100, cached=True, **kwargs):
        """ Returns an iterator over all article titles (for a given namespace id).
        """
        kwargs.setdefault("unicode", True)
        kwargs.setdefault("throttle", self.throttle)
        # Fetch article titles (default) or a custom id.
        id = kwargs.pop("_id", "title")
        # Loop endlessly (= until the last request no longer yields an "apcontinue").
        # See: http://www.mediawiki.org/wiki/API:Allpages
        while start != -1:
            url = URL(self._url, method=GET, query={
                     "action": "query",
                       "list": "allpages",
                "apnamespace": namespace,
                     "apfrom": start or "",
                    "aplimit": min(count, 500),
              "apfilterredir": "nonredirects",
                     "format": "json"
            })
            data = url.download(cached=cached, **kwargs)
            data = json.loads(data)
            for x in data.get("query", {}).get("allpages", {}):
                if x.get(id):
                    yield x[id]
            start = data.get("query-continue", {}).get("allpages", {})