ClientForm is a Python module for handling HTML forms on the client
side, useful for parsing HTML forms, filling them in and returning the
completed forms to the server. It developed from a port of Gisle Aas'
Perl module HTML::Form, from the libwww-perl library, but the
interface is not the same.
Simple working example: @{colorize("".join(open("examples/simple.py").readlines()[2:]))}
A more complicated working example (Note: this example makes use of the ClientForm 0.2 API; refer to the README.html file in the latest 0.1 release for the corresponding code for that version.): @{colorize("".join(open("examples/example.py").readlines()[2:]))}
All of the standard control types are supported: TEXT,
PASSWORD, HIDDEN, TEXTAREA,
ISINDEX, RESET, BUTTON (INPUT
TYPE=BUTTON and the various BUTTON types),
SUBMIT, IMAGE, RADIO,
CHECKBOX, SELECT/OPTION and
FILE (for file upload). Both standard form encodings
(application/x-www-form-urlencoded and
multipart/form-data) are supported.
The module is designed for testing and automation of web interfaces, not for implementing interactive user agents.
Security note: Remember that any passwords you store in
HTMLForm instances will be saved to disk in the clear if you
pickle them (directly or indirectly). The simplest solution to this is to
avoid pickling HTMLForm objects. You could also pickle before
filling in any password, or just set the password to "" before
pickling.
Python 2.0 or above is required. To run the tests, you need the
unittest module (from PyUnit). unittest is a
standard library module with Python 2.1 and above.
For full documentation, see the docstrings in ClientForm.py.
Note: this page describes the 0.2 (stable release) interface. See here for the old 0.1 interface.
ClientForm contains two parsers. See the FAQ entry on XHTML for details.
mxTidy or µTidylib can be useful for dealing with bad HTML.
I think it would be nice to have an implementation of ClientForm based on BeautifulSoup (i.e. all methods and attributes implemented using the BeautifulSoup API), since that module does tolerant HTML parsing with a nice API for doing non-forms stuff. (I'm not about to do this, though. For anybody interested in doing this, note that the ClientForm tests would need making constructor-independent first.)
ClientForm 0.2 includes three minor backwards-incompatible interface changes from version 0.1.
To make upgrading from 0.1 easier, and to allow me to stop supporting
version 0.1 sooner, version 0.2 contains support for operating in a
backwards-compatible mode, under which code written for 0.1 should work without
modification. This is done on a per-HTMLForm basis via the
.backwards_compat attribute, but for convenience the
ParseResponse() and ParseFile() factory functions accept
backwards_compat arguments. These backwards-compatibility
features will be removed in version 0.3. The default is to operate in
backwards-compatible mode. To run with backwards compatible mode turned
OFF (strongly recommended):
@{colorize(r"""
from urllib2 import urlopen
from ClientForm import ParseResponse
forms = ParseResponse(urlopen("http://example.com/"), backwards_compat=False)
# ...
""")}
The backwards-incompatible changes are:
Ambiguous specification of controls or items now results in
AmbiguityError. If you want the old behaviour, explicitly pass
nr=0 to indicate you want the first matching control or item.
Item label matching is now done by substring, not by strict string-equality (but note leading and trailing space is always stripped). (Control label matching is always done by substring.)
Handling of disabled list items has changed. First, note that handling
of disabled list items in 0.1 (and in 0.2's backwards-compatibility mode!) is
buggy: disabled items are successful (ie. disabled item names are sent back to
the server). As a result, there was no distinction to be made between
successful items and selected items. In 0.2, the bug is fixed, so this is no
longer the case, and it is important to note that list controls'
.value attribute contains only the successful item names;
items that are selected but not successful (because disabled) are not
included in .value. Second, disabled list items may no longer be
deselected: AttributeError is raised in 0.2, whereas deselection was allowed in
0.1. The bug in 0.1 and in 0.2's backwards-compatibility mode will not be
fixed, to preserve compatibility and to encourage people to upgrade to the new
0.2 backwards_compat=False behaviour.
Apart from Gisle Aas for allowing the original port from libwww-perl, particular credit is due to Gary Poster and Benji York, and their employer, Zope Corporation, for their contributions which led to ClientForm 0.2 being released. Thanks also to the many people who have contributed bug reports.
For installation instructions, see the INSTALL.txt file included in the distribution.
Stable release There have been three fairly minor backwards-incompatible interface changes since version 0.1 (see above), but by default the code operates in a backwards-compatible mode so that code written for 0.1 should work without changes.
0.2 includes better support for labels, and a simpler interface (all the old methods are still there, but some have been deprecated and a few added).
Old release No longer maintained. I recommend upgrading from 0.1 to 0.2.
There were many interface changes between 0.0 and 0.1, so you should take care if upgrading old code from 0.0.
0.1 includes FILE control support for file upload, handling
of disabled list items, and a redesigned interface.
Ancient release No longer maintained. You don't want this.
The Subversion (SVN) trunk is http://codespeak.net/svn/wwwsearch/ClientForm/trunk, so to check out the source:
svn co http://codespeak.net/svn/wwwsearch/ClientForm/trunk ClientForm
cgi, do this?
No: the cgi module does the server end of the job. It
doesn't know how to parse or fill in a form or how to send it back to the
server.
2.0 or above (ClientForm 0.2; version 0.1 requires Python 1.5.2 or above).
urllib2 required?
No.
urllib2?
Use .click_request_data() instead of .click().
urllib2 do I need?
You don't. It's convenient, though. If you have Python 2.0, you need to upgrade to the version from Python 2.1 (available from www.python.org). Otherwise, you're OK.
ClientForm is dual-licensed: you may pick either the BSD license, or the ZPL 2.1 (both are included in the distribution).
Yes. You must pass
form_parser_class=ClientForm.XHTMLCompatibleFormParser to
ParseResponse() / ParseFile(). Note this parser
is less tolerant of bad HTML than the default,
ClientForm.FormParser
print form is usually all you need.
In your code, things like the HTMLForm.items attribute of
HTMLForm instances can be useful to inspect forms at
runtime. Note that it's possible to use item labels instead of item
names, which can be useful — use the by_label
arguments to the various methods, and the .get_value_by_label() /
.set_value_by_label() methods on ListControl.
'*' characters mean in the string
representations of list controls?
A * next to an item means that item is selected.
Parentheses (foo) around an item mean that item is disabled.
.click*() when that control has non-None value?
Either the control is disabled, or it is not successful for some other reason. 'Successful' (see HTML 4 specification) means that the control will cause data to get sent to the server.
RADIO and multiple-selection SELECT controls?
Because by default, it follows browser behaviour when setting the
initially-selected items in list controls that have no items explicitly
selected in the HTML. Use the select_default argument to
ParseResponse if you want to follow the RFC 1866 rules
instead. Note that browser behaviour violates the HTML 4.01 specification
in the case of RADIO controls.
.click()ing on a button not work for me?
RESET button doesn't do anything, by design
- this is a library for web automation, not an interactive browser.
Even in an interactive browser, clicking on RESET sends
nothing to the server, so there is little point in having
.click() do anything special here.
BUTTON TYPE=BUTTON doesn't do anything
either, also by design. This time, the reason is that that
BUTTON is only in the HTML standard so that one can attach
callbacks to its events. The callbacks are functions in
SCRIPT elements (such as Javascript) embedded in the HTML,
and their execution may result in information getting sent back to the
server. ClientForm, however, knows nothing about these callbacks, so
it can't do anything useful with a click on a BUTTON whose
type is BUTTON.
See the General FAQs page and the next FAQ entry for what to do about this.
INPUT TYPE=HIDDEN field values (for example,
to emulate the effect of JavaScript code)?
As with any control, set the control's readonly attribute
false.
@{colorize(r""" form.find_control("foo").readonly = False # allow changing .value of control foo form.set_all_readonly(False) # allow changing the .value of all controls """)}
The ClientCookie package makes it
easy to get .seek()able response objects, which is
convenient for debugging. See also here for few
relevant tips. Also see General
FAQs.
@{colorize(r""" import bisect def closest_int_value(form, ctrl_name, value): values = map(int, [item.name for item in form.find_control(ctrl_name).items]) return str(values[bisect.bisect(values, value) - 1]) form["distance"] = [closest_int_value(form, "distance", 23)] """)}
I prefer questions and comments to be sent to the mailing list rather than direct to me.
John J. Lee, @(time.strftime("%B %Y", last_modified)).