Function ParseResponse

thirdparty/clientform/clientform.py:1012–1072

Parse HTTP response and return a list of HTMLForm instances. The return value of urllib2.urlopen can be conveniently passed to this function as the response parameter. ClientForm.ParseError is raised on parse errors. response: file-like object (supporting read() method) with a met

(response, *args, **kwds)

                        )

def ParseResponse(response, *args, **kwds):
    """Parse HTTP response and return a list of HTMLForm instances.

    The return value of urllib2.urlopen can be conveniently passed to this
    function as the response parameter.

    ClientForm.ParseError is raised on parse errors.

    response: file-like object (supporting read() method) with a method
     geturl(), returning the URI of the HTTP response
    select_default: for multiple-selection SELECT controls and RADIO controls,
     pick the first item as the default if none are selected in the HTML
    form_parser_class: class to instantiate and use to pass
    request_class: class to return from .click() method (default is
     _urllib.request.Request)
    entitydefs: mapping like {"&": "&", ...} containing HTML entity
     definitions (a sensible default is used)
    encoding: character encoding used for encoding numeric character references
     when matching link text.  ClientForm does not attempt to find the encoding
     in a META HTTP-EQUIV attribute in the document itself (mechanize, for
     example, does do that and will pass the correct value to ClientForm using
     this parameter).

    backwards_compat: boolean that determines whether the returned HTMLForm
     objects are backwards-compatible with old code.  If backwards_compat is
     true:

     - ClientForm 0.1 code will continue to work as before.

     - Label searches that do not specify a nr (number or count) will always
       get the first match, even if other controls match.  If
       backwards_compat is False, label searches that have ambiguous results
       will raise an AmbiguityError.

     - Item label matching is done by strict string comparison rather than
       substring matching.

     - De-selecting individual list items is allowed even if the Item is
       disabled.

    The backwards_compat argument will be deprecated in a future release.

    Pass a true value for select_default if you want the behaviour specified by
    RFC 1866 (the HTML 2.0 standard), which is to select the first item in a
    RADIO or multiple-selection SELECT control if none were selected in the
    HTML.  Most browsers (including Microsoft Internet Explorer (IE) and
    Netscape Navigator) instead leave all items unselected in these cases.  The
    W3C HTML 4.0 standard leaves this behaviour undefined in the case of
    multiple-selection SELECT controls, but insists that at least one RADIO
    button should be checked at all times, in contradiction to browser
    behaviour.

    There is a choice of parsers.  ClientForm.XHTMLCompatibleFormParser (uses
    HTMLParser.HTMLParser) works best for XHTML, ClientForm.FormParser (uses
    sgmllib.SGMLParser) (the default) works better for ordinary grubby HTML.
    Note that HTMLParser is only available in Python 2.2 and later.  You can
    pass your own class in here as a hack to work around bad HTML, but at your
    own risk: there is no well-defined interface.