[PATCH 2/2] Setup working tree in describe

[PATCH 2/2] Setup working tree in describe

[Sebastian Staudt]

This ensures the given working tree is used for --dirty and --broken.

Signed-off-by: Sebastian Staudt <koraktor@gmail.com>
---
 builtin/describe.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/builtin/describe.c b/builtin/describe.c
index cc118448ee..ba1a0b199b 100644
--- a/builtin/describe.c
+++ b/builtin/describe.c
@@ -601,6 +601,8 @@ int cmd_describe(int argc, const char **argv,
const char *prefix)
     if (!hashmap_get_size(&names) && !always)
         die(_("No names found, cannot describe anything."));

+    setup_work_tree();
+
     if (argc == 0) {
         if (broken) {
             struct child_process cp = CHILD_PROCESS_INIT;
-- 
2.20.1

Re: [PATCH 2/2] Setup working tree in describe

[Duy Nguyen]

On Sat, Jan 26, 2019 at 5:44 PM Sebastian Staudt <koraktor@gmail.com> wrote:
>
> This ensures the given working tree is used for --dirty and --broken.
>
> Signed-off-by: Sebastian Staudt <koraktor@gmail.com>
> ---
>  builtin/describe.c | 2 ++
>  1 file changed, 2 insertions(+)
>
> diff --git a/builtin/describe.c b/builtin/describe.c
> index cc118448ee..ba1a0b199b 100644
> --- a/builtin/describe.c
> +++ b/builtin/describe.c
> @@ -601,6 +601,8 @@ int cmd_describe(int argc, const char **argv,
> const char *prefix)
>      if (!hashmap_get_size(&names) && !always)
>          die(_("No names found, cannot describe anything."));
>
> +    setup_work_tree();

This forces worktree's presence in all cases and will die() if
worktree is not available. You need to check if broken or dirty is set
and only call this function in that case.

Though in my opinion it's better to call before we need it in the "if
(broke)" and "else if (dirty)" code blocks. That way you don't even
need to check if it's "dirty" or "broken". Does "broken" really need
this though? If it runs "git diff-index" separately, that command
should handle this setup_work_tree() already, or we may need to fix it
there, not here.

> +
>      if (argc == 0) {
>          if (broken) {
>              struct child_process cp = CHILD_PROCESS_INIT;
> --
> 2.20.1



-- 
Duy

Re: [PATCH 2/2] Setup working tree in describe

[Sebastian Staudt]

Am Sa., 26. Jan. 2019 um 12:01 Uhr schrieb Duy Nguyen <pclouds@gmail.com>:
>
> On Sat, Jan 26, 2019 at 5:44 PM Sebastian Staudt <koraktor@gmail.com> wrote:
> >
> > This ensures the given working tree is used for --dirty and --broken.
> >
> > Signed-off-by: Sebastian Staudt <koraktor@gmail.com>
> > ---
> >  builtin/describe.c | 2 ++
> >  1 file changed, 2 insertions(+)
> >
> > diff --git a/builtin/describe.c b/builtin/describe.c
> > index cc118448ee..ba1a0b199b 100644
> > --- a/builtin/describe.c
> > +++ b/builtin/describe.c
> > @@ -601,6 +601,8 @@ int cmd_describe(int argc, const char **argv,
> > const char *prefix)
> >      if (!hashmap_get_size(&names) && !always)
> >          die(_("No names found, cannot describe anything."));
> >
> > +    setup_work_tree();
>
> This forces worktree's presence in all cases and will die() if
> worktree is not available. You need to check if broken or dirty is set
> and only call this function in that case.
>
> Though in my opinion it's better to call before we need it in the "if
> (broke)" and "else if (dirty)" code blocks. That way you don't even
> need to check if it's "dirty" or "broken". Does "broken" really need
> this though? If it runs "git diff-index" separately, that command
> should handle this setup_work_tree() already, or we may need to fix it
> there, not here.
>

Thanks for your feedback.
Are you sure that it will fail without a working tree?
Is it even possible to have *no* working tree?

I already tested this with some real life examples, e.g.

    git --git-dir /some/path/.git describe

From inside and outside of other repositories.
I didn‘t hit any errors so far.

> > +
> >      if (argc == 0) {
> >          if (broken) {
> >              struct child_process cp = CHILD_PROCESS_INIT;
> > --
> > 2.20.1
>
>
>
> --
> Duy

Best regards,
    Sebastian

Am Sa., 26. Jan. 2019 um 12:01 Uhr schrieb Duy Nguyen <pclouds@gmail.com>:
>
> On Sat, Jan 26, 2019 at 5:44 PM Sebastian Staudt <koraktor@gmail.com> wrote:
> >
> > This ensures the given working tree is used for --dirty and --broken.
> >
> > Signed-off-by: Sebastian Staudt <koraktor@gmail.com>
> > ---
> >  builtin/describe.c | 2 ++
> >  1 file changed, 2 insertions(+)
> >
> > diff --git a/builtin/describe.c b/builtin/describe.c
> > index cc118448ee..ba1a0b199b 100644
> > --- a/builtin/describe.c
> > +++ b/builtin/describe.c
> > @@ -601,6 +601,8 @@ int cmd_describe(int argc, const char **argv,
> > const char *prefix)
> >      if (!hashmap_get_size(&names) && !always)
> >          die(_("No names found, cannot describe anything."));
> >
> > +    setup_work_tree();
>
> This forces worktree's presence in all cases and will die() if
> worktree is not available. You need to check if broken or dirty is set
> and only call this function in that case.
>
> Though in my opinion it's better to call before we need it in the "if
> (broke)" and "else if (dirty)" code blocks. That way you don't even
> need to check if it's "dirty" or "broken". Does "broken" really need
> this though? If it runs "git diff-index" separately, that command
> should handle this setup_work_tree() already, or we may need to fix it
> there, not here.
>
> > +
> >      if (argc == 0) {
> >          if (broken) {
> >              struct child_process cp = CHILD_PROCESS_INIT;
> > --
> > 2.20.1
>
>
>
> --
> Duy

Re: [PATCH 2/2] Setup working tree in describe

[Jeff King]

On Sat, Jan 26, 2019 at 03:07:54PM +0100, Sebastian Staudt wrote:

> Are you sure that it will fail without a working tree?
> Is it even possible to have *no* working tree?
> 
> I already tested this with some real life examples, e.g.
> 
>     git --git-dir /some/path/.git describe
> 
> From inside and outside of other repositories.
> I didn‘t hit any errors so far.

Try:

  git clone --bare . bare.git
  cd bare.git
  git describe

That works with current versions of Git, but yields "fatal: this
operation must be run in a work tree" with your patch.

-Peff

PS Your patches seem whitespace-damaged. You might want to look into
   using git-send-email.