Junio C Hamano <gitster@xxxxxxxxx> writes:
>> + test_grep "^trace: prepare_cmd:.*" output
>> +'
>
> If you run
>
> $ cd t && sh ./t0014-alias.sh -d && cat trash*.t0014-alias/output
>
> you'll see two "prepare_cmd" logged, because you added one to
> prepare_shell_cmd() and another to prepare_cmd().
To avoid similar issues, perhaps the test can try to see a bit more
into the output? We already know that we cannot expect the output
to be identical everywhere, but we can do something along the lines
of the attached patch, or is the expected pattern below a bit too
specific?
t/t0014-alias.sh | 10 +++++++++-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git c/t/t0014-alias.sh w/t/t0014-alias.sh
index b7affbe93a..a3c133be96 100755
--- c/t/t0014-alias.sh
+++ w/t/t0014-alias.sh
@@ -45,9 +45,17 @@ test_expect_success 'run-command formats empty args properly' '
'
test_expect_success 'tracing a shell alias with arguments shows trace of prepared command' '
+ cat >expect <<-EOF &&
+ trace: exec: git-echo argument
+ trace: run_command: git-echo argument
+ trace: run_command: ${SQ}echo \$*${SQ} argument
+ trace: prepare_cmd: SHELL -c ${SQ}echo \$* "\$@"${SQ} ${SQ}echo \$*${SQ} argument
+ EOF
git config alias.echo "!echo \$*" &&
env GIT_TRACE=1 git echo argument 2>output &&
- test_grep "^trace: prepare_cmd:.*" output
+ # redact platform differences
+ sed -e "s/^\(trace: prepare_cmd:\) .* -c /\1 SHELL -c /" output >actual &&
+ test_cmp expect actual
'
test_done
