Terry J. Reedy added the comment:
Trying it out, I rediscovered that patchcheck has a Windows bug. This time I
reported it ;-) #18439.
--
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue18279
Roundup Robot added the comment:
New changeset ec71fcdcfeac by Terry Jan Reedy in branch '2.7':
Issue #18279: Add tests for idlelib/RstripExtension.py. Original patch by
http://hg.python.org/cpython/rev/ec71fcdcfeac
New changeset 22ce68d98345 by Terry Jan Reedy in branch '3.3':
Issue #18279:
Terry J. Reedy added the comment:
I had to remove the dependency to close this, since the test_text issue #18365
was reopened.
--
stage: patch review - committed/rejected
status: open - closed
___
Python tracker rep...@bugs.python.org
Terry J. Reedy added the comment:
A simple change to RstripExtension.py fixed the marking of unchanged files as
changed. I also removed a useless extra iteration. Having a test makes it
possible to do things like this without breaking what already worked.
--
dependencies: -Idle: mock
Terry J. Reedy added the comment:
Phil (and everyone else): PLEASE submit patches with 4 space indents and no
tabs and no trailing spaces. Even if the code below runs in the CPython
interpreter,
self.undo = mockUndoDelegator() 8 spaces
4 spaces
def get_selection_indices(self):
Todd Rovito added the comment:
As a suggestion I always use the command make patchcheck (before making the
patch) which catches the white space and tab problem plus it fixes other
things. Here is more information on patch check in the developer's guide.
Terry J. Reedy added the comment:
I am aware of patchcheck, but the problem for me is that 'make patchcheck' does
not work on Windows; the doc is wrong on the awkward to type Windows
alternative; and it is usually useless anyway. But I agree that anyone who does
not use a editor configured to
Phil Webster added the comment:
This patch contains mock_idle.py and the rstrip test using the mock text widget
from #18365.
Terry- For some reason, the Text widget always contains a '\n' as the last
character even when there is nothing visible. Doing a text.get('1.0','end')
always has a
Terry J. Reedy added the comment:
I decided that the first commit should be a separate issue that this issue
depends on.
--
assignee: - terry.reedy
dependencies: +Idle: mock Text class and test thereof
___
Python tracker rep...@bugs.python.org
Phil Webster added the comment:
Added to Terry's Text Widget code (in #18226) and created mock_idle.py for the
mock EditorWindow. Todd's FormatParagraph test in the aforementioned issue also
passes with the mock EditorWindow.
--
Added file:
Terry J. Reedy added the comment:
I want to make two separate commits. First add mock Text into mock_tk.py, and
add a new test_text.py. I suggested looking at the tkinter Text test. I turns
out that it only tests a few special search cases. I am guessing that they have
something to do with
Phil Webster added the comment:
Modified the first patch to get rid of mock EditorWindow in favor of the real
thing. Also renamed the test to 'test_rstrip'.
--
Added file: http://bugs.python.org/file30694/test_rstrip.patch
___
Python tracker
New submission from Phil Webster:
This is a single test for RstripExtension.py, following from #15392. I also
added a mock EditorWindow module with a Text widget for testing.
test_rstripextension.py seems to run fine inside IDLE and produces the
following output:
test_do_rstrip
Changes by Phil Webster webster.p...@gmail.com:
--
nosy: +Todd.Rovito
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue18279
___
___
Python-bugs-list
Changes by Phil Webster webster.p...@gmail.com:
--
nosy: +terry.reedy
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue18279
___
___
Python-bugs-list
Changes by Phil Webster webster.p...@gmail.com:
--
keywords: +patch
Added file: http://bugs.python.org/file30664/test_rstripext.patch
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue18279
___
Terry J. Reedy added the comment:
There is a separate issue about killing that warning.
If you leave off '-ugui', you will see the traceback and why I said that
requires('gui') should be wrapped for unittest discovery. As the code is, the
call is executed when the file is imported, possibly
Phil Webster added the comment:
Thank you for the feedback Terry. I'm not seeing the traceback without '-ugui'
either, so I'm going to look into that. I get the same results with
requires('gui') moved inside of setUp, is that what you mean by wrapping?
For mock_ewin I used a real Text widget
18 matches
Mail list logo