python-zope.testbrowser 4.0.2-1 / usr / share / pyshared / zope / testbrowser / fixed-bugs.txt

==========
Fixed Bugs
==========

This file includes tests for bugs that were found and then fixed that don't fit
into the more documentation-centric sections above.

    >>> from zope.testbrowser.ftests.wsgitestapp import WSGITestApplication
    >>> from zope.testbrowser.wsgi import Browser
    >>> wsgi_app = WSGITestApplication()

Unicode URLs
============

Unicode URLs or headers cause the entire constructed request to be unicode, and
(as of Python 2.4.4) Cookie.SimpleCookie checks the type of the input against
type(""), so it handles the value inappropriately, causing exceptions that
ended with::

      File "/home/benji/Python-2.4.4/lib/python2.4/Cookie.py", line 623, in load
        self.update(rawdata)
    ValueError: dictionary update sequence element #0 has length 1; 2 is required

As a work-around, unicode strings passed to Browser.open() are now converted to
ASCII before being passed on, as well as the key and value passed to
Browser.addHeader().

The tests below failed before the change was put in place.

    >>> browser = Browser(wsgi_app=wsgi_app)
    >>> browser.addHeader('Cookie', 'test')
    >>> browser.open(u'http://localhost/@@/testbrowser/simple.html')

    >>> browser = Browser(wsgi_app=wsgi_app)
    >>> browser.addHeader(u'Cookie', 'test')
    >>> browser.open('http://localhost/@@/testbrowser/simple.html')


Spaces in URL
=============

When URLs have spaces in them, they're handled correctly (before the bug was
fixed, you'd get "ValueError: too many values to unpack"):

    >>> browser.open('http://localhost/@@/testbrowser/navigate.html')
    >>> browser.getLink('Spaces in the URL').click()


.goBack() Truncation
====================

The .goBack() method used to truncate the .contents.

    >>> browser.open('http://localhost/@@/testbrowser/navigate.html')
    >>> actual_length = len(browser.contents)

    >>> browser.open('http://localhost/@@/testbrowser/navigate.html')
    >>> browser.open('http://localhost/@@/testbrowser/simple.html')
    >>> browser.goBack()
    >>> len(browser.contents) == actual_length
    True


Labeled Radio Buttons
=====================

The .getControl() method was sometimes unable to find radio buttons by label.

    >>> # import mechanize._form; mechanize._form._show_debug_messages()
    >>> browser.open('http://localhost/@@/testbrowser/radio.html')
    >>> browser.getControl('One').optionValue
    '1'
    >>> browser.getControl('Two').optionValue
    '2'
    >>> browser.getControl('Three').optionValue
    '3'


Fragment URLs
=============

Earlier versions of mechanize used to incorrectly follow links containing
fragments. We upgraded our dependency to a newer version of mechanize and make
sure this regression doesn't come back:

    >>> browser.open('http://localhost/@@/testbrowser/fragment.html#asdf')
    >>> browser.url
    'http://localhost/@@/testbrowser/fragment.html#asdf'
    >>> browser.getLink('Follow me')
    <Link text='Follow me' url='http://localhost/@@/testbrowser/fragment.html#foo'>
    >>> browser.getLink('Follow me').click()


Textareas with HTML/XML
=======================

    >>> browser.open('http://localhost/@@/testbrowser/textarea.html')
    >>> browser.getControl('Text Area').value
    '<block>\r\n  <feed/>\r\n    &\r\n</block>'


.click() with non-200 status
============================

The problem was that with the below controls testbrowser forgot to do
after-processing after an exception in mechanize.
That means ``_stop_timer()`` and ``_changed()`` were not executed if an exception
was raised. Not calling ``_changed()`` resulted in not refreshing ``contents``.
The ``contents`` property gets cached on any first access and should be reset
on any navigation.
The problem is that e.g. a simple 403 status raises an exception.

This is how it works with a simple open():

    >>> browser.handleErrors=False

    >>> browser.open('http://localhost/set_status.html')
    >>> print browser.contents
    Everything fine

    >>> browser.open('http://localhost/set_status.html?status=403')
    Traceback (most recent call last):
    ...
    HTTPError: HTTP Error 403: Forbidden

    >>> print browser.contents
    Just set a status of 403


These are the various controls:

A link:

    >>> browser.open('http://localhost/@@/testbrowser/status_lead.html')
    >>> print browser.contents
    <html>...

    >>> browser.getLink('403').click()
    Traceback (most recent call last):
    ...
    HTTPError: HTTP Error 403: Forbidden

    >>> print browser.contents
    Just set a status of 403

A submit button:

    >>> browser.open('http://localhost/@@/testbrowser/status_lead.html')
    >>> print browser.contents
    <html>...

    >>> browser.getControl(name='status').value = '404'
    >>> browser.getControl('Submit This').click()
    Traceback (most recent call last):
    ...
    HTTPError: HTTP Error 404: Not Found

    >>> print browser.contents
    Just set a status of 404

A submit image control:

    >>> browser.open('http://localhost/@@/testbrowser/status_lead.html')
    >>> print browser.contents
    <html>...

    >>> browser.getControl(name='status').value = '403'
    >>> browser.getControl(name='image-value').click()
    Traceback (most recent call last):
    ...
    HTTPError: HTTP Error 403: Forbidden

    >>> print browser.contents
    Just set a status of 403

A javascript-ish form submit:

    >>> browser.open('http://localhost/@@/testbrowser/status_lead.html')
    >>> print browser.contents
    <html>...

    >>> browser.getControl(name='status').value = '404'
    >>> browser.getForm(name='theform').submit()
    Traceback (most recent call last):
    ...
    HTTPError: HTTP Error 404: Not Found

    >>> print browser.contents
    Just set a status of 404

A non-javascript-ish form submit:

    >>> browser.open('http://localhost/@@/testbrowser/status_lead.html')
    >>> print browser.contents
    <html>...

    >>> browser.getControl(name='status').value = '403'
    >>> browser.getForm(name='theform').submit(name='submit-value')
    Traceback (most recent call last):
    ...
    HTTPError: HTTP Error 403: Forbidden

    >>> print browser.contents
    Just set a status of 403