Use 'printf %s $x' notation in t5401
PATCH 1/7 Use 'printf %s $x' notation in t5401
We only care about getting what should be an empty string and sending it to a file, without a trailing LF, so the empty string translates into a 0 byte file. Earlier when I originally wrote these lines Mac OS X allowed the format string of printf to be the empty string, but more recent versions appear to have been 'improved' with error messages if the format is not given.
This may cause problems if we ever wind up with changes to the hook tests. A minor cleanup makes the test more safe on all systems, by conforming to accepted printf conventions.
--- t/t5401-update-hooks.sh | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)
t/t5401-update-hooks.sh +4 -4
@@ -25,7 +25,7 @@ test_expect_success setup ' cat >victim/.git/hooks/pre-receive <<'EOF' #!/bin/sh -printf "$@" >>$GIT_DIR/pre-receive.args +printf %s "$@" >>$GIT_DIR/pre-receive.args cat - >$GIT_DIR/pre-receive.stdin echo STDOUT pre-receive echo STDERR pre-receive >&2
@@ -35,7 +35,7 @@ chmod u+x victim/.git/hooks/pre-receive cat >victim/.git/hooks/update <<'EOF' #!/bin/sh echo "$@" >>$GIT_DIR/update.args -read x; printf "$x" >$GIT_DIR/update.stdin +read x; printf %s "$x" >$GIT_DIR/update.stdin echo STDOUT update $1 echo STDERR update $1 >&2 test "$1" = refs/heads/master || exit
@@ -44,7 +44,7 @@ chmod u+x victim/.git/hooks/update cat >victim/.git/hooks/post-receive <<'EOF' #!/bin/sh -printf "$@" >>$GIT_DIR/post-receive.args +printf %s "$@" >>$GIT_DIR/post-receive.args cat - >$GIT_DIR/post-receive.stdin echo STDOUT post-receive echo STDERR post-receive >&2
@@ -54,7 +54,7 @@ chmod u+x victim/.git/hooks/post-receive cat >victim/.git/hooks/post-update <<'EOF' #!/bin/sh echo "$@" >>$GIT_DIR/post-update.args -read x; printf "$x" >$GIT_DIR/post-update.stdin +read x; printf %s "$x" >$GIT_DIR/post-update.stdin echo STDOUT post-update echo STDERR post-update >&2 EOF
Signature · 1 line
1.5.4.rc5.1126.g6ba14
Quoted 6 lines · Shawn O. Pearce
We only care about getting what should be an empty string and sending it to a file, without a trailing LF, so the empty string translates into a 0 byte file. Earlier when I originally wrote these lines Mac OS X allowed the format string of printf to be the empty string, but more recent versions appear to have been 'improved' with error messages if the format is not given.
That sounds like a bug to me.
t/t5401-update-hooks.sh test_expect_success setup ' · line 28
Quoted 10 lines
t/t5401-update-hooks.sh +1
-1@@ -25,7 +25,7 @@ test_expect_success setup ' cat >victim/.git/hooks/pre-receive <<'EOF' #!/bin/sh -printf "$@" >>$GIT_DIR/pre-receive.args +printf %s "$@" >>$GIT_DIR/pre-receive.args
Do you really mean this? "printf %s 1 2 3" writes "123" without terminating LF. You seem to check only for size but to be reusable you might want to use something like:
printf '%s\n' "$@"
Likewise for post-receive.args.
Quoted 7 lines · Junio C Hamano
Quoted 6 lines · Shawn O. Pearce
t/t5401-update-hooks.sh +0
-0@@ -25,7 +25,7 @@ test_expect_success setup '
Quoted 8 lines
t/t5401-update-hooks.sh test_expect_success setup ' · line 28
Quoted 4 lines
t/t5401-update-hooks.sh +1
-1cat >victim/.git/hooks/pre-receive <<'EOF' #!/bin/sh -printf "$@" >>$GIT_DIR/pre-receive.args +printf %s "$@" >>$GIT_DIR/pre-receive.argsDo you really mean this? "printf %s 1 2 3" writes "123" without terminating LF. You seem to check only for size but to be reusable you might want to use something like:
The only thing we care about is was there args or not to the hook. I probably could do that test differently, like say:
echo $# >$GIT_DIR/pre-receive.args
and then test that the file contains "0" instead of looking for it to be empty. Not sure why I didn't write it that way in the beginning.
Quoted 1 line
printf '%s\n' "$@"
Eh. Since all we care is that the argument count is 0 we probably should be looking at $# and calling it a day.
Signature · 1 line
Shawn.
Quoted 11 lines · Shawn O. Pearce
Quoted 4 lines · Junio C Hamano
...
Do you really mean this? "printf %s 1 2 3" writes "123" without terminating LF. You seem to check only for size but to be reusable you might want to use something like:The only thing we care about is was there args or not to the hook. I probably could do that test differently, like say:
echo $# >$GIT_DIR/pre-receive.args
Quoted 7 lines
and then test that the file contains "0" instead of looking for it to be empty. Not sure why I didn't write it that way in the beginning.
Quoted 1 line
printf '%s\n' "$@"Eh. Since all we care is that the argument count is 0 we probably should be looking at $# and calling it a day.
I'll apply as-is, but once you break the code that calls pre-receive hook in such a way that it sends unexpected arguments, you'd thank me for suggesting to print the arguments separately while debugging the problem with the test script.
Quoted 10 lines · Junio C Hamano
Quoted 4 lines · Shawn O. Pearce
Quoted 1 line
printf '%s\n' "$@"Eh. Since all we care is that the argument count is 0 we probably should be looking at $# and calling it a day.
I'll apply as-is, but once you break the code that calls pre-receive hook in such a way that it sends unexpected arguments, you'd thank me for suggesting to print the arguments separately while debugging the problem with the test script.
Well, I don't really think it matters. I doubt we'll break that code that invokes the pre-receive or post-receive hooks. And as those hooks are defined to never take arguments if we did suddenly start passing arguments we want the test to break. When tests break sometimes its not obvious why they have broken.
Case in point, the "remote: " prefixing caused by the sideband demultiplexer really confused me for about 15 minutes. I could not figure out why the last test in this script was failing. Until I pulled my head out of the sand and looked at the "actual" file and saw those prefixes. At which point it was very obvious.
:)
We haven't touched that script in almost a year. I poked at it only to fix a stupid error from my printf implementation, and because of the new "remote:" prefix. Which since it is a change in behavior may cause problems for anyone who is trying to screen-scrape the stderr of git-push. Gawd I hope nobody's scripts are that convoluted.
Signature · 1 line
Shawn.