Re: [PATCH RESEND] t4015: improve coverage of function context test

["René Scharfe" <l.s.r@web.de>]

Am 18.12.19 um 21:12 schrieb Jeff King:
> On Wed, Dec 18, 2019 at 07:05:54PM +0100, René Scharfe wrote:
>
>> Include an actual function line in the test files to check if context is
>> expanded to include the whole function, and add an ignored change before
>> function context to check if that one stays hidden, while the originally
>> ignored change within function context is shown.
>>
>> [...]
>>  test_expect_success 'combine --ignore-blank-lines with --function-context' '
>> -	test_write_lines 1 "" 2 3 4 5 >a &&
>> -	test_write_lines 1    2 3 4   >b &&
>> +	test_write_lines    1 2 3 "" function 1 2 3 4 5 "" 6 7 8 9 >a &&
>> +	test_write_lines "" 1 2 3 "" function 1 2 3 4 5    6 7 8   >b &&
>
> I'm a little mixed on this one. This _is_ a much better test of how the
> two features should be have together. But I think the reason the
> original was so short was that it was added when fixing a bug where we'd
> iterate off the beginning of list of lines, which now no longer happens.

That fix, b777f3fd61 (xdiff: clamp function context indices in post-image,
2019-07-23), should no longer be necessary, but I didn't check that
thoroughly.  Since we still have it (and I don't intend to remove it,
better keep that extra safety), it makes sense to keep the specific test.

> Maybe we should cover both cases in two separate tests?

That's easy enough to do.  The hardest part is coming up with a name, but
simply counting up should do the trick here.

> I'd also suggest using "a b c" for the first three lines to avoid
> confusion (I don't think it's important that they're the same as the
> lines inside the "function").

Good point.  That turns the last line into a function line, though, which
is unnecessary and confused me a bit, but I think it's a net win.

René