Python compat.urlopen方法代碼示例

# 需要導入模塊: from nltk import compat [as 別名]
# 或者: from nltk.compat import urlopen [as 別名]
def _open(resource_url):
    """
    Helper function that returns an open file object for a resource,
    given its resource URL.  If the given resource URL uses the "nltk:"
    protocol, or uses no protocol, then use ``nltk.data.find`` to find
    its path, and open it with the given mode; if the resource URL
    uses the 'file' protocol, then open the file with the given mode;
    otherwise, delegate to ``urllib2.urlopen``.

    :type resource_url: str
    :param resource_url: A URL specifying where the resource should be
        loaded from.  The default protocol is "nltk:", which searches
        for the file in the the NLTK data package.
    """
    resource_url = normalize_resource_url(resource_url)
    protocol, path_ = split_resource_url(resource_url)

    if protocol is None or protocol.lower() == 'nltk':
        return find(path_, path + ['']).open()
    elif protocol.lower() == 'file':
        # urllib might not use mode='rb', so handle this one ourselves:
        return find(path_, ['']).open()
    else:
        return urlopen(resource_url)

######################################################################
# Lazy Resource Loader
######################################################################

# We shouldn't apply @python_2_unicode_compatible
# decorator to LazyLoader, this is resource.__class__ responsibility. 

# 需要導入模塊: from nltk import compat [as 別名]
# 或者: from nltk.compat import urlopen [as 別名]
def _update_index(self, url=None):
        """A helper function that ensures that self._index is
        up-to-date.  If the index is older than self.INDEX_TIMEOUT,
        then download it again."""
        # Check if the index is aleady up-to-date.  If so, do nothing.
        if not (self._index is None or url is not None or
                time.time()-self._index_timestamp > self.INDEX_TIMEOUT):
            return

        # If a URL was specified, then update our URL.
        self._url = url or self._url

        # Download the index file.
        self._index = nltk.internals.ElementWrapper(
            ElementTree.parse(compat.urlopen(self._url)).getroot())
        self._index_timestamp = time.time()

        # Build a dictionary of packages.
        packages = [Package.fromxml(p) for p in
                    self._index.findall('packages/package')]
        self._packages = dict((p.id, p) for p in packages)

        # Build a dictionary of collections.
        collections = [Collection.fromxml(c) for c in
                       self._index.findall('collections/collection')]
        self._collections = dict((c.id, c) for c in collections)

        # Replace identifiers with actual children in collection.children.
        for collection in self._collections.values():
            for i, child_id in enumerate(collection.children):
                if child_id in self._packages:
                    collection.children[i] = self._packages[child_id]
                elif child_id in self._collections:
                    collection.children[i] = self._collections[child_id]
                else:
                    print('removing collection member with no package: {}'.format(child_id))
                    del collection.children[i]

        # Fill in collection.packages for each collection.
        for collection in self._collections.values():
            packages = {}
            queue = [collection]
            for child in queue:
                if isinstance(child, Collection):
                    queue.extend(child.children)
                else:
                    packages[child.id] = child
            collection.packages = packages.values()

        # Flush the status cache
        self._status_cache.clear()