[Scons-users] Question over test.must_contain

[RW]

sorry that should have read
"must_contains always deals with text and not binary data which may be
incorrect"

On 4 July 2017 at 14:06, RW <garlicbready at googlemail.com> wrote:

> Hi,
> I've noticed another python3 bug / fix
>
> python3 runtest.py test/Docbook/basic/xinclude/xinclude.py
>
> 1/1 (100.00%) /usr/bin/python3 -tt test/Docbook/basic/xinclude/xinclude.py
> Traceback (most recent call last):
>   File "test/Docbook/basic/xinclude/xinclude.py", line 47, in <module>
>     test.must_contain(test.workpath('manual_xi.xml'),'<para>This is an
> included text.')
>   File "/home/ric/scons-temp/scons_patched/QMTest/TestCommon.py", line
> 278, in must_contain
>     contains = find(file_contents, required)
>   File "/home/ric/scons-temp/scons_patched/QMTest/TestCommon.py", line
> 275, in find
>     return o.index(l)
> TypeError: a bytes-like object is required, not 'str'
>
> There's a couple of different ways to fix this
>
> # Option1
> the simplest is to just change
>
> ```
> test.must_contain(test.workpath('manual_xi.xml'),'<para>This is an
> included text.')
> to
> test.must_contain(test.workpath('manual_xi.xml'),'<para>This is an
> included text.', mode='r')
> ```
>
> which what the other tests seem to do
>
> # Option2
> To maintain consistency between python versions, another possible option
> (maybe at the same time) is to alter the
> must_contain function within QMTest/TestCommon.py
> However with this approach your assuming must_contains always deals with
> binary and not text data which may be incorrect
>
> ```
> import codecs
> file_contents = codecs.latin_1_decode(file_contents)[0]
> ```
>
> Would both options or just option 1 be the best way?
>
> Many Thanks
> Richard
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist4.pair.net/pipermail/scons-users/attachments/20170704/0cbd443e/attachment.html>