Re: [PATCH v2] travis-ci: build documentation
- From
Junio C Hamano <gitster@pobox.com>
- Date
- Apr 29, 2016, 21:05 UTC
- Message-ID
- <xmqqziscw2g7.fsf@gitster.mtv.corp.google.com>
- In-Reply-To
- <D3ACDA07-DBD4-44B5-8D0A-531F2EC3BD65@gmail.com>
Lars Schneider <larsxschneider@gmail.com> writes:
Show 10 quoted lines
>> This could be less of maintenance if we'd check with a "larger as" operator >> such as >> >> test_file_count_more_than html 200 >> >> using an arbitrary slightly smaller number. > > Well, I was thinking about testing against something like > $(find . -type f -name "git*.txt" | wc -l) but it the end > all of this is not really meaningful I think...
Either is too much, I would say--I have too much faith in the exit status from "make doc", I guess.
What would _REALLY_ be nice is a check that lets us catch an error like this deliberate breakage:
diff --git a/Documentation/diff-options.txt b/Documentation/diff-options.txt
index 4b0318e..a684f2d 100644
--- a/Documentation/diff-options.txt
+++ b/Documentation/diff-options.txt
@@ -560,4 +560,4 @@ endif::git-format-patch[]
Do not show any source or destination prefix. For more detailed explanation on these common options, see also
-linkgit:gitdiffcore[7].
+linkgit:gitdiffnore[7].We just get a dangling link in the result without an error exit from "make doc"; neither "test -s" nor "size is about what we expect" would catch such a breakage, though.
Other things that might be of interest are
make check-builtins
make check-docsbut I am not sure if the latter built target is up to date (it has a whiltelist that needs to stay current). We rarely add new commands these days, so it is easy to forget what these build targets try to check, which makes them good candidates to be thrown into the set of automated tests.