Re: [PATCH v1 3/5] fsmonitor: add test cases for fsmonitor extension

[Ben Peart <peartben@gmail.com>]

On 5/16/2017 12:59 AM, Junio C Hamano wrote:
> Ben Peart <peartben@gmail.com> writes:
>
>> Add test cases that ensure status results are correct when using the new
>> fsmonitor extension.  Test untracked, modified, and new files by
>> ensuring the results are identical to when not using the extension.
>>
>> Add a test to ensure updates to the index properly mark corresponding
>> entries in the index extension as dirty so that the status is correct
>> after commands that modify the index but don't trigger changes in the
>> working directory.
>>
>> Add a test that verifies that if the fsmonitor extension doesn't tell
>> git about a change, it doesn't discover it on its own.  This ensures
>> git is honoring the extension and that we get the performance benefits
>> desired.
>>
>> Signed-off-by: Ben Peart <benpeart@microsoft.com>
>> ---
>>  t/t7519-status-fsmonitor.sh | 134 ++++++++++++++++++++++++++++++++++++++++++++
>>  1 file changed, 134 insertions(+)
>>  create mode 100644 t/t7519-status-fsmonitor.sh
>
> Please make this executable.
>

Sorry, long time Windows developer so I forgot this extra step.  Fixed 
for next roll.

>> diff --git a/t/t7519-status-fsmonitor.sh b/t/t7519-status-fsmonitor.sh
>> new file mode 100644
>> index 0000000000..2d63efc27b
>> --- /dev/null
>> +++ b/t/t7519-status-fsmonitor.sh
>> @@ -0,0 +1,134 @@
>> ...
>> +# Ensure commands that call refresh_index() to move the index back in time
>> +# properly invalidate the fsmonitor cache
>> +...
>> +	git status >output &&
>> +	git -c core.fsmonitor=false status >expect &&
>> +	test_i18ncmp expect output
>> +'
>
> Hmm. I wonder if we can somehow detect the case where we got the
> correct and expected result only because fsmonitor was not in
> effect, even though the test requested it to be used?  Not limited
> to this particular test piece, but applies to all of them in this
> file.
>

I have tested this manually by editing the test hook proc to output 
invalid results and ensured that the test failed as a result but adding 
that to the test script was kind of ugly (all tests end up getting 
duplicated - one ensuring success, one ensuring failure).

On further reflection, a better idea is to have the test hook proc 
output a marker file that can be tested for existence.  If it exists, 
the hook was used to update the results, if it doesn't exist, then the 
hook proc wasn't used.  A much cleaner solution that doesn't require 
duplicating the tests.