Hi Philip, On Thu, 10 Oct 2024 at 15:08, Phillip Wood <phillip.wood123@xxxxxxxxx> wrote: > > Hi Chizoba > > On 10/10/2024 07:39, chizobajames21@xxxxxxxxx wrote: > > From: Chizoba ODINAKA <chizobajames21@xxxxxxxxx> > > > > In pipes, the exit code of a chain of commands is determined by > > the downstream command. > > I would perhaps say "final command" rather than "downstream command" as > in a pipeline "cmd1 | cmd2 | cmd3" cmd2 and cmd3 are downstream of cmd1 > but it is the exit code of cmd3 that will be used > Yes, final command is a more appropriate explain. > > In order not to loss the entire result code of tests, > > write output of upstreams into a file. > > We're interested in checking the exit code of git, but not other > commands so it would be helpful to make that clear. Usman's patch [1] > has a good explanation of this. > I just read that sentence again, it obviously needs some clarity. "In order not to miss the exit code of any Git command, avoid using pipe and write output into a file" has more clarity. I will look up on Usman's patch [1], before my next changes. > This patch also changes instances of "grep" to "test_grep" so the commit > message needs to explain the reason for that change which is that it > gives a better debugging experience if the test fails. > I had included that in my "Changes in v2", appended beneath my "Sign-off-by". "Changes in v2: - split multiple commands chain on the same line across multiple line, for easier readability - replace "grep" with "test_grep", for more context in case of a "grep" failure" Maybe it was not so obvious that you didn't notice, or it is not the traditional way of including it. > The patch is looking pretty good, most of the conversions look correct. > I've left a few comments below > > > [1] > https://lore.kernel.org/git/bfff7937cd20737bb5a8791dc7492700b1d7881f.1728315124.git.gitgitgadget@xxxxxxxxx > > > test_expect_success 'replace the author' ' > > - git cat-file commit $HASH2 | grep "author A U Thor" && > > - R=$(git cat-file commit $HASH2 | sed -e "s/A U/O/" | git hash-object -t commit --stdin -w) && > > - git cat-file commit $R | grep "author O Thor" && > > + git cat-file commit $HASH2 >actual && > > + test_grep "author A U Thor" actual && > > + git cat-file commit $HASH2 >actual && > > You don't need to repeat this command now that we are saving the output > of "git cat-file commit $HASH2" > > > + R=$(sed -e "s/A U/O/" actual | git hash-object -t commit --stdin -w) && > > + git cat-file commit $R >actual && > > + test_grep "author O Thor" actual && > > > test_expect_success 'push branch with replacement' ' > > - git cat-file commit $PARA3 | grep "author A U Thor" && > > - S=$(git cat-file commit $PARA3 | sed -e "s/A U/O/" | git hash-object -t commit --stdin -w) && > > - git cat-file commit $S | grep "author O Thor" && > > + git cat-file commit $PARA3 >actual && > > + test_grep "author A U Thor" actual && > > + git cat-file commit $PARA3 >actual && > > We can drop this line for the same reason as above > > > + S=$(sed -e "s/A U/O/" actual | git hash-object -t commit --stdin -w) && > > + git cat-file commit $S >actual && > > + test_grep "author O Thor" actual && > > > @@ -260,14 +291,14 @@ test_expect_success 'fetch branch with replacement' ' > > cd clone_dir && > > git fetch origin refs/heads/tofetch:refs/heads/parallel3 && > > git log --pretty=oneline parallel3 >output.txt && > > - ! grep $PARA3 output.txt && > > + ! test_grep $PARA3 output.txt && > > test_grep will print an error message the pattern does not match. In > this case we expect it not to match so we want to print an error if it > does match. We can do that with > > test_grep ! $PARA3 output.txt && > > > test_expect_success 'index-pack and replacements' ' > > - git --no-replace-objects rev-list --objects HEAD | > > + git --no-replace-objects rev-list --objects HEAD >actual && > > git --no-replace-objects pack-objects test- && > > This command has lost its input - you need to use '<actual' to get it to > read output from "git rev-list". This test itself could probably do a > better job of checking that index-pack does what we expect but that is > outside the scope of this patch. > > > git index-pack test-*.pack > > ' > > Everything that I've not commented on looks correct to me. > Thanks Philip for the review, I will make the needed changes in my next patch. And look into the index-pack proposal in a new patch, since it is outside this scope. > Best Wishes > > Phillip Thanks Chizoba
