Adds Finds support to is_element_present()

[martinpelikan]

Context
I'm still quite new to the page object design pattern, so this may not normally come up. I have a table of elements that need to be deleted one at a time via a menu/option. My solution is to iterate while there are still elements visible and delete them one at a time (avoiding the stale element reference in doing so). The problem is that Webium's nice is_element_present() method fails for Finds. This is my fix, which seems to work for me, and should not break existing behaviour. The question is, if it should check that all elements are visible, or if only the first is necessary...

Error Encountered

ntp.py:42: in delete_all
    while self.is_element_present('ntp_servers'):
/usr/lib/python3.5/site-packages/webium/base_page.py:29: in is_element_present
    is_displayed() if just_in_dom else wait(lambda: is_displayed(), timeout_seconds=timeout)
/usr/lib/python3.5/site-packages/webium/wait.py:16: in wait
    return wait_lib(*args, **kwargs)
/usr/lib/python3.5/site-packages/waiting/__init__.py:16: in wait
    for x in iterwait(result=result, *args, **kwargs):
/usr/lib/python3.5/site-packages/waiting/__init__.py:40: in iterwait
    result.result = predicate()
/usr/lib/python3.5/site-packages/webium/base_page.py:29: in <lambda>
    is_displayed() if just_in_dom else wait(lambda: is_displayed(), timeout_seconds=timeout)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

    def is_displayed():
        element = getattr(self, element_name, None)
        if not element:
            raise WebiumException('No element "{0}" within container {1}'.format(element_name, self))
>       return element.is_displayed()
E       AttributeError: 'list' object has no attribute 'is_displayed'

[ufranske]

Hi martinpelikan!

Thanks for your PR. It looks good to me. This behavior is kind of tricky but the changes do not make the state of things worse. However before we merge this PR please could you:

1. Add tests to https://github.com/wgnet/webium/blob/master/tests/simple_page/test_is_element_present.py:

One for SamplePage.anchor_list field (https://github.com/martinpelikan/webium/blob/f9a8a654f4dc147f00f79ebdeb8f9fc212f99f9b/tests/simple_page/__init__.py#L54)
Other - for SamplePage.empty_element_list field (https://github.com/martinpelikan/webium/blob/f9a8a654f4dc147f00f79ebdeb8f9fc212f99f9b/tests/simple_page/__init__.py#L56)

2. Add a note about this tricky behavior in case of Finds field to docs: https://github.com/martinpelikan/webium/blob/f9a8a654f4dc147f00f79ebdeb8f9fc212f99f9b/docs/is_element_present.rst

@SUNx2YCH , @Khrol could please take a look as well?

[Khrol]

nested if is not needed.

[ufranske]

Well, all([]) produces True. So, it depends on what you expect from Finds which does not contain any visible elements

[Khrol]

Tests are needed to display what behaviour we are trying to achieve.

[martinpelikan]

Semantically, it might be more clear if I wrote this as if element != []: instead of if element:. I can change it to that if you prefer. This is mostly my habit when checking for empty sequences/iterables.

[martinpelikan]

I was able to spend some time today on addressing your reviews. Many thanks for being so quick to respond and review @Khrol and @ufranske!

Should my commit also include a version bump, or is this a responsibility of the maintainers?

[Khrol]

LGTM 👍