Source code for mechanize._mechanize

"""Stateful programmatic WWW navigation, after Perl's WWW::Mechanize.

Copyright 2003-2006 John J. Lee <jjl@pobox.com>
Copyright 2003 Andy Lester (original Perl code)

This code is free software; you can redistribute it and/or modify it
under the terms of the BSD or ZPL 2.1 licenses (see the file LICENSE
included with the distribution).

"""
from __future__ import absolute_import

import copy
import os
import re

from . import _request, _response, _rfc3986, _sockettimeout, _urllib2_fork
from ._clientcookie import Cookie
from ._headersutil import normalize_header_name
from ._html import Factory
from ._useragent import UserAgentBase
from .polyglot import pathname2url, HTTPError, is_string, iteritems
from ._response import make_response


class BrowserStateError(Exception):
    pass


class LinkNotFoundError(Exception):
    pass


class FormNotFoundError(Exception):
    pass


def sanepathname2url(path):
    urlpath = pathname2url(path)
    if os.name == "nt" and urlpath.startswith("///"):
        urlpath = urlpath[2:]
    # XXX don't ask me about the mac...
    return urlpath


[docs]class History: """ Though this will become public, the implied interface is not yet stable. """ def __init__(self): self._history = [] # LIFO def add(self, request, response): self._history.append((request, response)) def back(self, n, _response): response = _response # XXX move Browser._response into this class? while n > 0 or response is None: try: request, response = self._history.pop() except IndexError: raise BrowserStateError("already at start of history") n -= 1 return request, response def clear(self): del self._history[:] def close(self): for request, response in self._history: if response is not None: response.close() del self._history[:] def __copy__(self): ans = self.__class__() ans._history = self._history[:] return ans
class HTTPRefererProcessor(_urllib2_fork.BaseHandler): def http_request(self, request): # See RFC 2616 14.36. The only times we know the source of the # request URI has a URI associated with it are redirect, and # Browser.click() / Browser.submit() / Browser.follow_link(). # Otherwise, it's the user's job to add any Referer header before # .open()ing. if hasattr(request, "redirect_dict"): request = self.parent._add_referer_header( request, origin_request=False) return request https_request = http_request
[docs]class Browser(UserAgentBase): """Browser-like class with support for history, forms and links. :class:`BrowserStateError` is raised whenever the browser is in the wrong state to complete the requested operation - e.g., when :meth:`back()` is called when the browser history is empty, or when :meth:`follow_link()` is called when the current response does not contain HTML data. Public attributes: request: current request (:class:`mechanize.Request`) form: currently selected form (see :meth:`select_form()`) :param history: object implementing the :class:`mechanize.History` interface. Note this interface is still experimental and may change in future. This object is owned by the browser instance and must not be shared among browsers. :param request_class: Request class to use. Defaults to :class:`mechanize.Request` :param content_parser: A function that is responsible for parsing received html/xhtml content. See the builtin :func:`mechanize._html.content_parser()` function for details on the interface this function must support. :param factory_class: HTML Factory class to use. Defaults to :class:`mechanize.Factory` """ handler_classes = copy.copy(UserAgentBase.handler_classes) handler_classes["_referer"] = HTTPRefererProcessor default_features = copy.copy(UserAgentBase.default_features) default_features.append("_referer") def __init__( self, history=None, request_class=None, content_parser=None, factory_class=Factory, allow_xhtml=False, ): """ Only named arguments should be passed to this constructor. """ self._handle_referer = True if history is None: history = History() self._history = history if request_class is None: request_class = _request.Request factory = factory_class(allow_xhtml=allow_xhtml) factory.set_request_class(request_class) if content_parser is not None: factory.set_content_parser(content_parser) self._factory = factory self.request_class = request_class self.request = None self._set_response(None, False) # do this last to avoid __getattr__ problems UserAgentBase.__init__(self) def __copy__(self): ''' Clone this browser instance. The clone will share the same, thread-safe cookie jar, and have all the same handlers/settings, but will not share any other state, making it safe to use in another thread. ''' ans = self.__class__() self._copy_state(ans) ans._handle_referer = self._handle_referer for attr in ('_response_type_finder', '_encoding_finder', '_content_parser'): setattr(ans._factory, attr, getattr(self._factory, attr)) ans.request_class = self.request_class ans._history = copy.copy(self._history) return ans def close(self): UserAgentBase.close(self) if self._response is not None: self._response.close() if self._history is not None: self._history.close() self._history = None # make use after .close easy to spot self.form = None self.request = self._response = None self.request = self.response = self.set_response = None self.geturl = self.reload = self.back = None self.clear_history = self.set_cookie = self.links = self.forms = None self.viewing_html = self.encoding = self.title = None self.select_form = self.click = self.submit = self.click_link = None self.follow_link = self.find_link = None
[docs] def set_handle_referer(self, handle): """Set whether to add Referer header to each request.""" self._set_handler("_referer", handle) self._handle_referer = bool(handle)
def _add_referer_header(self, request, origin_request=True): if self.request is None: return request scheme = request.get_type() original_scheme = self.request.get_type() if scheme not in ["http", "https"]: return request if not origin_request and not self.request.has_header("Referer"): return request if (self._handle_referer and original_scheme in ["http", "https"] and not (original_scheme == "https" and scheme != "https")): # strip URL fragment (RFC 2616 14.36) parts = _rfc3986.urlsplit(self.request.get_full_url()) parts = parts[:-1] + (None, ) referer = _rfc3986.urlunsplit(parts) request.add_unredirected_header("Referer", referer) return request
[docs] def open_novisit(self, url_or_request, data=None, timeout=_sockettimeout._GLOBAL_DEFAULT_TIMEOUT): """Open a URL without visiting it. Browser state (including request, response, history, forms and links) is left unchanged by calling this function. The interface is the same as for :meth:`open()`. This is useful for things like fetching images. See also :meth:`retrieve()` """ return self._mech_open( url_or_request, data, visit=False, timeout=timeout)
[docs] def open(self, url_or_request, data=None, timeout=_sockettimeout._GLOBAL_DEFAULT_TIMEOUT): ''' Open a URL. Loads the page so that you can subsequently use :meth:`forms()`, :meth:`links()`, etc. on it. :param url_or_request: Either a URL or a :class:`mechanize.Request` :param dict data: data to send with a POST request :param timeout: Timeout in seconds :return: A :class:`mechanize.Response` object ''' return self._mech_open(url_or_request, data, timeout=timeout)
def _mech_open(self, url, data=None, update_history=True, visit=None, timeout=_sockettimeout._GLOBAL_DEFAULT_TIMEOUT): try: url.get_full_url except AttributeError: # string URL -- convert to absolute URL if required scheme, authority = _rfc3986.urlsplit(url)[:2] if scheme is None: # relative URL if self._response is None: raise BrowserStateError("can't fetch relative reference: " "not viewing any document") url = _rfc3986.urljoin(self._response.geturl(), url) request = self._request(url, data, visit, timeout) visit = request.visit if visit is None: visit = True if visit: self._visit_request(request, update_history) success = True try: response = UserAgentBase.open(self, request, data) except HTTPError as error: success = False if error.fp is None: # not a response raise response = error # except (IOError, socket.error, OSError) as error: # Yes, urllib2 really does raise all these :-(( # See test_urllib2.py for examples of socket.gaierror and OSError, # plus note that FTPHandler raises IOError. # XXX I don't seem to have an example of exactly socket.error being # raised, only socket.gaierror... # I don't want to start fixing these here, though, since this is a # subclass of OpenerDirector, and it would break old code. Even in # Python core, a fix would need some backwards-compat. hack to be # acceptable. # raise if visit: self._set_response(response, False) response = copy.copy(self._response) elif response is not None: response = _response.upgrade_response(response) if not success: raise response return response def __str__(self): text = [] text.append("<%s " % self.__class__.__name__) if self._response: text.append("visiting %s" % self._response.geturl()) else: text.append("(not visiting a URL)") if self.form: text.append("\n selected form:\n %s\n" % str(self.form)) text.append(">") return "".join(text)
[docs] def response(self): """Return a copy of the current response. The returned object has the same interface as the object returned by :meth:`.open()` """ return copy.copy(self._response)
def open_local_file(self, filename): path = sanepathname2url(os.path.abspath(filename)) url = 'file://' + path return self.open(url)
[docs] def set_response(self, response): """Replace current response with (a copy of) response. response may be None. This is intended mostly for HTML-preprocessing. """ self._set_response(response, True)
def _set_response(self, response, close_current): # sanity check, necessary but far from sufficient if not (response is None or (hasattr(response, "info") and hasattr(response, "geturl") and hasattr(response, "read"))): raise ValueError("not a response object") self.form = None if response is not None: response = _response.upgrade_response(response) if close_current and self._response is not None: self._response.close() self._response = response self._factory.set_response(response)
[docs] def visit_response(self, response, request=None): """Visit the response, as if it had been :meth:`open()` ed. Unlike :meth:`set_response()`, this updates history rather than replacing the current response. """ if request is None: request = _request.Request(response.geturl()) self._visit_request(request, True) self._set_response(response, False)
def _visit_request(self, request, update_history): if self._response is not None: self._response.close() if self.request is not None and update_history: self._history.add(self.request, self._response) self._response = None # we want self.request to be assigned even if UserAgentBase.open # fails self.request = request
[docs] def set_html(self, html, url="http://example.com/"): """Set the response to dummy with given HTML, and URL if given. Allows you to then parse that HTML, especially to extract forms information. If no URL was given then the default is "example.com". """ response = make_response(html, [("Content-type", "text/html")], url) self._set_response(response, True)
[docs] def geturl(self): """Get URL of current document.""" if self._response is None: raise BrowserStateError("not viewing any document") return self._response.geturl()
[docs] def reload(self): """Reload current document, and return response object.""" if self.request is None: raise BrowserStateError("no URL has yet been .open()ed") if self._response is not None: self._response.close() return self._mech_open(self.request, update_history=False)
[docs] def back(self, n=1): """Go back n steps in history, and return response object. n: go back this number of steps (default 1 step) """ if self._response is not None: self._response.close() self.request, response = self._history.back(n, self._response) self.set_response(response) if not response.read_complete: return self.reload() return copy.copy(response)
def clear_history(self): self._history.clear() @property def cookiejar(self): ' Return the current cookiejar (:class:`mechanize.CookieJar`) or None ' try: return self._ua_handlers["_cookies"].cookiejar except Exception: pass
[docs] def set_header(self, header, value=None): ''' Convenience method to set a header value in `self.addheaders` so that the header is sent out with all requests automatically. :param header: The header name, e.g. User-Agent :param value: The header value. If set to None the header is removed. ''' found = False header = normalize_header_name(header) q = header.lower() remove = [] for i, (k, v) in enumerate(tuple(self.addheaders)): if k.lower() == q: if value: self.addheaders[i] = (header, value) found = True else: remove.append(i) if not found: self.addheaders.append((header, value)) if remove: for i in reversed(remove): del self.addheaders[i]
[docs] def forms(self): """Return iterable over forms. The returned form objects implement the :class:`mechanize.HTMLForm` interface. """ if not self.viewing_html(): raise BrowserStateError("not viewing HTML") return self._factory.forms()
[docs] def global_form(self): """Return the global form object, or None if the factory implementation did not supply one. The "global" form object contains all controls that are not descendants of any FORM element. The returned form object implements the :class:`mechanize.HTMLForm` interface. This is a separate method since the global form is not regarded as part of the sequence of forms in the document -- mostly for backwards-compatibility. """ if not self.viewing_html(): raise BrowserStateError("not viewing HTML") return self._factory.global_form
[docs] def viewing_html(self): """Return whether the current response contains HTML data.""" if self._response is None: raise BrowserStateError("not viewing any document") return self._factory.is_html
def encoding(self): if self._response is None: raise BrowserStateError("not viewing any document") return self._factory.encoding
[docs] def title(self): ' Return title, or None if there is no title element in the document. ' if not self.viewing_html(): raise BrowserStateError("not viewing HTML") return self._factory.title
[docs] def select_form(self, name=None, predicate=None, nr=None, **attrs): """Select an HTML form for input. This is a bit like giving a form the "input focus" in a browser. If a form is selected, the Browser object supports the HTMLForm interface, so you can call methods like :meth:`set_value()`, :meth:`set()`, and :meth:`click()`. Another way to select a form is to assign to the .form attribute. The form assigned should be one of the objects returned by the :meth:`forms()` method. If no matching form is found, :class:`mechanize.FormNotFoundError` is raised. If `name` is specified, then the form must have the indicated name. If `predicate` is specified, then the form must match that function. The predicate function is passed the :class:`mechanize.HTMLForm` as its single argument, and should return a boolean value indicating whether the form matched. `nr`, if supplied, is the sequence number of the form (where 0 is the first). Note that control 0 is the first form matching all the other arguments (if supplied); it is not necessarily the first control in the form. The "global form" (consisting of all form controls not contained in any FORM element) is considered not to be part of this sequence and to have no name, so will not be matched unless both name and nr are None. You can also match on any HTML attribute of the `<form>` tag by passing in the attribute name and value as keyword arguments. To convert HTML attributes into syntactically valid python keyword arguments, the following simple rule is used. The python keyword argument name is converted to an HTML attribute name by: Replacing all underscores with hyphens and removing any trailing underscores. You can pass in strings, functions or regular expression objects as the values to match. For example: .. code-block:: python # Match form with the exact action specified br.select_form(action='http://foo.com/submit.php') # Match form with a class attribute that contains 'login' br.select_form(class_=lambda x: 'login' in x) # Match form with a data-form-type attribute that matches a regex br.select_form(data_form_type=re.compile(r'a|b')) """ if not self.viewing_html(): raise BrowserStateError("not viewing HTML") if name is None and predicate is None and nr is None and not attrs: raise ValueError( "at least one argument must be supplied to specify form") global_form = self._factory.global_form if nr is None and name is None and predicate is not None and predicate( global_form): self.form = global_form return def attr_selector(q): if is_string(q): return lambda x: x == q if callable(q): return q return lambda x: q.match(x) is not None attrsq = {aname.rstrip('_').replace('_', '-'): attr_selector(v) for aname, v in iteritems(attrs)} def form_attrs_match(form_attrs): for aname, q in iteritems(attrsq): val = form_attrs.get(aname) if val is None or not q(val): return False return True orig_nr = nr for form in self.forms(): if name is not None and name != form.name: continue if predicate is not None and not predicate(form): continue if nr: nr -= 1 continue if attrs and not form_attrs_match(form.attrs): continue self.form = form break # success else: # failure description = [] if name is not None: description.append("name '%s'" % name) if predicate is not None: description.append("predicate %s" % predicate) if orig_nr is not None: description.append("nr %d" % orig_nr) if attrs: for k, v in iteritems(attrs): description.append('%s = %r' % (k, v)) description = ", ".join(description) raise FormNotFoundError("no form matching " + description)
[docs] def click(self, *args, **kwds): """See :meth:`mechanize.HTMLForm.click()` for documentation.""" if not self.viewing_html(): raise BrowserStateError("not viewing HTML") request = self.form.click(*args, **kwds) return self._add_referer_header(request)
[docs] def submit(self, *args, **kwds): """Submit current form. Arguments are as for :meth:`mechanize.HTMLForm.click()`. Return value is same as for :meth:`open()`. """ return self.open(self.click(*args, **kwds))
def __getattr__(self, name): # pass through _form.HTMLForm methods and attributes form = self.__dict__.get("form") if form is None: raise AttributeError( "%s instance has no attribute %s (perhaps you forgot to " ".select_form()?)" % (self.__class__, name)) return getattr(form, name) def __getitem__(self, name): if self.form is None: raise BrowserStateError('No form selected') return self.form[name] def __setitem__(self, name, val): if self.form is None: raise BrowserStateError('No form selected') self.form[name] = val def _filter_links(self, links, text=None, text_regex=None, name=None, name_regex=None, url=None, url_regex=None, tag=None, predicate=None, nr=0): if not self.viewing_html(): raise BrowserStateError("not viewing HTML") orig_nr = nr for link in links: if url is not None and url != link.url: continue if url_regex is not None and not re.search(url_regex, link.url): continue if (text is not None and (link.text is None or text != link.text)): continue if ( text_regex is not None and ( link.text is None or not re.search( text_regex, link.text))): continue if name is not None and name != dict(link.attrs).get("name"): continue if name_regex is not None: link_name = dict(link.attrs).get("name") if link_name is None or not re.search(name_regex, link_name): continue if tag is not None and tag != link.tag: continue if predicate is not None and not predicate(link): continue if nr: nr -= 1 continue yield link nr = orig_nr