git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH 11/25] t7400: fix broken "submodule add/reconfigure --force" test

From
Eric Sunshine <sunshine@sunshineco.com>
Date
Jul 16, 2018, 23:05 UTC
Message-ID
<20180716230535.GA49654@flurp.local>
In-Reply-To
<xmqq7elusvz7.fsf@gitster-ct.c.googlers.com>
On Mon, Jul 16, 2018 at 02:37:32PM -0700, Junio C Hamano wrote:
Show 22 quoted lines
> Eric Sunshine <sunshine@sunshineco.com> writes:
> > On Mon, Jul 16, 2018 at 11:51 AM Johannes Schindelin
> > <Johannes.Schindelin@gmx.de> wrote:
> >> On Mon, 16 Jul 2018, Johannes Schindelin wrote:
> >> > > -           git submodule add --force bogus-url submod &&
> >> > > +           git submodule add --force /bogus-url submod &&
> >> > This breaks on Windows because of the absolute Unix-y path having to be
> >> > translated to a Windows path:
> >> > I could imagine that using "$(pwd)/bogus-url" (which will generate a
> >> > Windows-y absolute path on Windows) would work, though.
> >>
> >> Yes, this works indeed, see the patch below. Could you please squash it in
> >> if you send another iteration of your patch series? Junio, could you
> >> please add this as a SQUASH??? commit so that `pu` is fixed on Windows?
> >
> > So, this SQUASH looks like the correct way forward. Hopefully, Junio
> > can just squash it so I don't have to flood the list again with this
> > long series.
> 
> The topic already has another dependent topic so rerolling or
> squashing would be a bit cumbersome.  I'll see what I could do but
> it may not be until tomorrow's pushout.

No problem. Here's Dscho's fix in the form of a proper patch if that makes it easier for you (it just needs his sign-off):

--- >8 ---
From: Johannes Schindelin <Johannes.Schindelin@gmx.de>
Subject: [PATCH] t7400: make "submodule add/reconfigure --force" work on
 Windows

On Windows, the "Unixy" path /bogus-url gets translated automatically to a Windows-style path (such as C:/git-sdk/bogus_url). This is normally not problem, since it's still a valid and usable path in that form, however, this test wants to do a comparison against the original path. $(pwd) was invented exactly for this case, so use it to make the path comparison work.

[es: commit message]
Signed-off-by: Eric Sunshine <sunshine@sunshineco.com>
---
 t/t7400-submodule-basic.sh | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/t/t7400-submodule-basic.sh b/t/t7400-submodule-basic.sh
index 76cf522a08..bfb09dd566 100755
--- a/t/t7400-submodule-basic.sh
+++ b/t/t7400-submodule-basic.sh
@@ -171,10 +171,11 @@ test_expect_success 'submodule add to .gitignored path with --force' '
 test_expect_success 'submodule add to reconfigure existing submodule with --force' '
 	(
 		cd addtest-ignore &&
-		git submodule add --force /bogus-url submod &&
+		bogus_url="$(pwd)/bogus-url" &&
+		git submodule add --force "$bogus_url" submod &&
 		git submodule add --force -b initial "$submodurl" submod-branch &&
-		test "/bogus-url" = "$(git config -f .gitmodules submodule.submod.url)" &&
-		test "/bogus-url" = "$(git config submodule.submod.url)" &&
+		test "$bogus_url" = "$(git config -f .gitmodules submodule.submod.url)" &&
+		test "$bogus_url" = "$(git config submodule.submod.url)" &&
 		# Restore the url
 		git submodule add --force "$submodurl" submod &&
 		test "$submodurl" = "$(git config -f .gitmodules submodule.submod.url)" &&
-- 
2.18.0.233.g985f88cf7e
Previous: Junio C HamanoNext: Johannes Schindelin
Message 20 of 52 in “fix buggy tests, modernize tests, fix broken &&-chains”
  1. 00/25 fix buggy tests, modernize tests, fix broken &&-chainsEric Sunshine, Jul 2, 2018
  2. 01/25 t: use test_might_fail() instead of manipulating exit code manuallyEric Sunshine, Jul 2, 2018
  3. Stefan BellerJul 2, 2018
  4. Eric SunshineJul 2, 2018
  5. Junio C HamanoJul 3, 2018
  6. 04/25 t: drop unnecessary terminating semicolon in subshellEric Sunshine, Jul 2, 2018
  7. Junio C HamanoJul 3, 2018
  8. 05/25 t/lib-submodule-update: fix "absorbing" testEric Sunshine, Jul 2, 2018
  9. 03/25 t: use sane_unset() rather than 'unset' with broken &&-chainEric Sunshine, Jul 2, 2018
  10. 02/25 t: use test_write_lines() instead of series of 'echo' commandsEric Sunshine, Jul 2, 2018
  11. Stefan BellerJul 2, 2018
  12. 08/25 t5505: modernize and simplify hard-to-digest testEric Sunshine, Jul 2, 2018
  13. 07/25 t5406: use write_script() instead of birthing shell script manuallyEric Sunshine, Jul 2, 2018
  14. 10/25 t7201: drop pointless "exit 0" at end of subshellEric Sunshine, Jul 2, 2018
  15. 11/25 t7400: fix broken "submodule add/reconfigure --force" testEric Sunshine, Jul 2, 2018
  16. Johannes SchindelinJul 16, 2018
  17. Johannes SchindelinJul 16, 2018
  18. Eric SunshineJul 16, 2018
  19. Junio C HamanoJul 16, 2018
  20. Eric SunshineJul 16, 2018
  21. Johannes SchindelinJul 17, 2018
  22. Junio C HamanoJul 17, 2018
  23. 12/25 t7810: use test_expect_code() instead of hand-rolled comparisonEric Sunshine, Jul 2, 2018
  24. Stefan BellerJul 2, 2018
  25. Eric SunshineJul 2, 2018
  26. 15/25 t0000-t0999: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  27. 16/25 t1000-t1999: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  28. 17/25 t2000-t2999: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  29. 25/25 t9119: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  30. 24/25 t9000-t9999: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  31. 23/25 t7000-t7999: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  32. 22/25 t6000-t6999: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  33. 21/25 t5000-t5999: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  34. SZEDER GáborJul 12, 2018
  35. Eric SunshineJul 12, 2018
  36. Jeff KingJul 12, 2018
  37. Junio C HamanoJul 12, 2018
  38. Junio C HamanoJul 12, 2018
  39. Eric SunshineJul 12, 2018
  40. Junio C HamanoJul 12, 2018
  41. Eric SunshineJul 12, 2018
  42. SZEDER GáborJul 17, 2018
  43. 20/25 t4000-t4999: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  44. 19/25 t3030: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  45. 18/25 t3000-t3999: fix broken &&-chainsEric Sunshine, Jul 2, 2018
  46. 14/25 t9814: simplify convoluted check that command correctly errors outEric Sunshine, Jul 2, 2018
  47. 09/25 t6036: fix broken "merge fails but has appropriate contents" testsEric Sunshine, Jul 2, 2018
  48. 06/25 t5405: use test_must_fail() instead of checking exit code manuallyEric Sunshine, Jul 2, 2018
  49. 13/25 t9001: fix broken "invoke hook" testEric Sunshine, Jul 2, 2018
  50. Stefan BellerJul 2, 2018
  51. Eric SunshineJul 2, 2018
  52. Junio C HamanoJul 3, 2018

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.