Re: [PATCH v2] parallel-checkout: avoid dash local bug in tests

Am 05.06.21 um 21:56 schrieb Ævar Arnfjörð Bjarmason:

On Sat, Jun 05 2021, René Scharfe wrote:

quoted

The parallel checkout tests fail when run with /bin/dash on MacOS 11.4,
reporting the following error:

   ./t2080-parallel-checkout-basics.sh: 33: local: 0: bad variable name

That's because wc's output contains leading spaces and this version of
dash erroneously expands the variable declaration as "local workers= 0",
i.e. it tries to set the "workers" variable to the empty string and also
declare a variable named "0", which not a valid name.  This is a known
dash bug (https://bugs.launchpad.net/ubuntu/+source/dash/+bug/139097).

Work around it by passing the command output directly to test instead of
storing it in a variable first.  While at it, let grep count the number
of lines instead of piping its output to wc, which is a bit shorter and
more efficient.

Helped-by: Matheus Tavares Bernardino [off-list ref]
Signed-off-by: René Scharfe <redacted>
---
Changes since v1:
- Explain the root cause.
- Get rid of the local variable "workers".
- Adjust title accordingly.
- Still use grep -c, though.
- Remove input redirection.

 t/lib-parallel-checkout.sh | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/t/lib-parallel-checkout.sh b/t/lib-parallel-checkout.sh
index 21f5759732..66350d5207 100644
--- a/t/lib-parallel-checkout.sh
+++ b/t/lib-parallel-checkout.sh

@@ -27,8 +27,7 @@ test_checkout_workers () {
 	rm -f "$trace_file" &&
 	GIT_TRACE2="$(pwd)/$trace_file" "$@" 2>&8 &&

-	local workers=$(grep "child_start\[..*\] git checkout--worker" "$trace_file" | wc -l) &&
-	test $workers -eq $expected_workers &&
+	test $(grep -c "child_start\[..*\] git checkout--worker" "$trace_file") -eq $expected_workers &&
 	rm "$trace_file"
 } 8>&2 2>&4

I'd find this thing much clearer if the v2 just narrowly focused on
avoiding the "local", and thus demonstrated the non-portable shell
issue,

I was not aiming for a minimal fix and I don't think the patch above is
too complex, but you're right that at this point in the release cycle a
duct-tape-style patch would be better.

and perhaps with something like:

	diff --git a/t/check-non-portable-shell.pl b/t/check-non-portable-shell.pl
	index fd3303552be..aad6f3e2bf1 100755
	--- a/t/check-non-portable-shell.pl
	+++ b/t/check-non-portable-shell.pl
	@@ -48,6 +48,7 @@ sub err {
	 	/\bexport\s+[A-Za-z0-9_]*=/ and err '"export FOO=bar" is not portable (use FOO=bar && export FOO)';
	 	/^\s*([A-Z0-9_]+=(\w*|(["']).*?\3)\s+)+(\w+)/ and exists($func{$4}) and
	 		err '"FOO=bar shell_func" assignment extends beyond "shell_func"';
	+	/\blocal\b \S+=\$\(.*?\|\s*\bwc -l\)/ and err 'whitespace handling in local=$(... | wc -l) differs in some dash versions';

Any command can output whitespace, it's not limited to wc -l.  So a
better rule might be to always quote command substitutions in local
variable declarations (local foo="$(...)").  Or to disallow assignments
with local altogether, like we already do for export, but that might be
a bit much.

	 	$line = '';
	 	# this resets our $. for each file
	 	close ARGV if eof;


The let's do grep -c while we're at it part of this IMO just adds
confusion while skimming future portability issues with --grep=dash or
--grep=POSIX in the future, and looking at the history in v1 it's just
there because in v1 the root cause wasn't fully understood.

True, I was still gripping the "use grep -c instead of grep | wc -l"
hammer.  I better rewatch https://www.youtube.com/watch?v=Yv4tI6939q0

If we're doing a general cleanup of that pattern it would seem to be
better to search-replace this with the rest of them in another commit:

    $ git grep '\$\(grep.*\| wc -l' -- t | wc -l
    27