Jameson Miller <jameson.mille...@gmail.com> writes:

> Teach the status command more flexibility in how ignored files are
> reported. Currently, the reporting of ignored files and untracked
> files are linked. You cannot control how ignored files are reported
> independently of how untracked files are reported (i.e. `all` vs
> `normal`). This makes it impossible to show untracked files with the
> `all` option, but show ignored files with the `normal` option.
> 
> This work 1) adds the ability to control the reporting of ignored
> files independently of untracked files and 2) introduces the concept
> of status reporting ignored paths that explicitly match an ignored
> pattern. ...
>
> When status is set to report matching ignored files, it has the
> following behavior. Ignored files and directories that explicitly
> match an exclude pattern are reported. If an ignored directory matches
> an exclude pattern, then the path of the directory is returned. If a
> directory does not match an exclude pattern, but all of its contents
> are ignored, then the contained files are reported instead of the
> directory.

Thanks for an updated log message.  Very nicely explained.

> +static void handle_ignored_arg(struct wt_status *s)
> +{
> +     if (!ignored_arg)
> +             ; /* default already initialized */
> +     else if (!strcmp(ignored_arg, "traditional"))
> +             s->show_ignored_mode = SHOW_TRADITIONAL_IGNORED;
> +     else if (!strcmp(ignored_arg, "no"))
> +             s->show_ignored_mode = SHOW_NO_IGNORED;
> +     else if (!strcmp(ignored_arg, "matching"))
> +             s->show_ignored_mode = SHOW_MATCHING_IGNORED;
> +     else
> +             die(_("Invalid ignored mode '%s'"), ignored_arg);
> +}
>  
>  static void handle_untracked_files_arg(struct wt_status *s)
>  {
> @@ -1363,8 +1376,10 @@ int cmd_status(int argc, const char **argv, const char 
> *prefix)
>                 N_("mode"),
>                 N_("show untracked files, optional modes: all, normal, no. 
> (Default: all)"),
>                 PARSE_OPT_OPTARG, NULL, (intptr_t)"all" },
> -             OPT_BOOL(0, "ignored", &show_ignored_in_status,
> -                      N_("show ignored files")),
> +             { OPTION_STRING, 0, "ignored", &ignored_arg,
> +               N_("mode"),
> +               N_("show ignored files, optional modes: traditional, 
> matching, no. (Default: traditional)"),
> +               PARSE_OPT_OPTARG, NULL, (intptr_t)"traditional" },

We used to use "show_ignored_in_status" variable that can be either
0 (when no --ignored is given, or an explicit --no-ignored is given,
on the command line) or 1 (when --ignored is given on the command
line).

We still allow "--ignored" without value for backward compatibility,
but 

    $ git status -uall --ignored \*.c

may trigger die() from handle_ignored_arg().  I wonder if this is
something we care about.

        ... goes and digs ...

OK, because of PARSE_OPT_OPTARG, the above is safe; "*.c" is not
mistaken as the value given to the --ignored option.  It also means
that

    $ git status -uno --ignored matching

would not mean what a naïve reader may expect and does not error
out, even though

    $ git status -uno --ignored=matching

would.  Which is something we eventually might care about, but that
is how parse-options PARSE_OPT_OPTARG works, and I consider "fixing"
it is totally out of the scope of this series (e.g. the next option
"--ignore-submodules" below shares exactly the same issue).

>               { OPTION_STRING, 0, "ignore-submodules", &ignore_submodule_arg, 
> N_("when"),
>                 N_("ignore changes to submodules, optional when: all, dirty, 
> untracked. (Default: all)"),
>                 PARSE_OPT_OPTARG, NULL, (intptr_t)"all" },

> diff --git a/wt-status.c b/wt-status.c
> index 6f730ee8f2..8301c84946 100644
> --- a/wt-status.c
> +++ b/wt-status.c
> @@ -660,10 +660,15 @@ static void wt_status_collect_untracked(struct 
> wt_status *s)
>       if (s->show_untracked_files != SHOW_ALL_UNTRACKED_FILES)
>               dir.flags |=
>                       DIR_SHOW_OTHER_DIRECTORIES | DIR_HIDE_EMPTY_DIRECTORIES;
> -     if (s->show_ignored_files)
> +     if (s->show_ignored_mode) {

Now we no longer use show_ignored_files that was a boolean yes/no,
and instead use an enum show_ignored_mode, we'd better spell this
out like so:

        if (s->show_ignored_mode == SHOW_NO_IGNORED) {

>               dir.flags |= DIR_SHOW_IGNORED_TOO;
> -     else
> +
> +             if (s->show_ignored_mode == SHOW_MATCHING_IGNORED)
> +                     dir.flags |= DIR_SHOW_IGNORED_TOO_MODE_MATCHING;

Mental note: the old "show_ignored_files is true" case is now split
into two, i.e. traditional vs matching.  When matching is used,
dir.flags gets a new bit, i.e. DIR_SHOW_IGNORED_TOO_MODE_MATCHING.
Lack of that bit means dir.c helpers should behave as before.

> +     } else {
>               dir.untracked = the_index.untracked;
> +     }
> +
>       setup_standard_excludes(&dir);
>  
>       fill_directory(&dir, &the_index, &s->pathspec);
> @@ -1621,7 +1626,7 @@ static void wt_longstatus_print(struct wt_status *s)
>       }
>       if (s->show_untracked_files) {
>               wt_longstatus_print_other(s, &s->untracked, _("Untracked 
> files"), "add");
> -             if (s->show_ignored_files)
> +             if (s->show_ignored_mode)

Likewise, i.e.

        if (s->show_ignored_mode == SHOW_NO_IGNORED)

>                       wt_longstatus_print_other(s, &s->ignored, _("Ignored 
> files"), "add -f");
>               if (advice_status_u_option && 2000 < s->untracked_in_ms) {
>                       status_printf_ln(s, GIT_COLOR_NORMAL, "%s", "");

> diff --git a/dir.c b/dir.c
> index 1d17b800cf..b9af87eca9 100644
> --- a/dir.c
> +++ b/dir.c
> @@ -1389,6 +1389,30 @@ static enum path_treatment treat_directory(struct 
> dir_struct *dir,
>       case index_nonexistent:
>               if (dir->flags & DIR_SHOW_OTHER_DIRECTORIES)
>                       break;
> +             if (exclude &&
> +                     (dir->flags & DIR_SHOW_IGNORED_TOO) &&
> +                     (dir->flags & DIR_SHOW_IGNORED_TOO_MODE_MATCHING)) {

It somehow feels superfluous to check both TOO and TOO_MODE_MATCHING
here (IOW, wouldn't the latter be superset of the former?  Is there
a valid case where the latter is set but not the former?), but it is
perhaps OK as belt-and-suspenders sanity.

> +                     /*
> +                      * This is an excluded directory and we are
> +                      * showing ignored paths that match an exclude
> +                      * pattern.  (e.g. show directory as ignored
> +                      * only if it matches an exclude pattern).
> +                      * This path will either be 'path_excluded`
> +                      * (if we are showing empty directories or if
> +                      * the directory is not empty), or will be
> +                      * 'path_none' (empty directory, and we are
> +                      * not showing empty directories).
> +                      */

The only caller of treat_directory() calls us when the path we are
seeing (sans trailing '/' that was added immediately before we were
called) passed "is_excluded()" check in the caller, so at this
point, the directory in question is being excluded because it
matched an exclude pattern.  IOW, if we are a/b/c, a pattern such as
"a/b/*" may have said we are excluded and made the call to this
function.  Do we need to ensure that a pattern like "a/*" is not
what excluded us?  Or IGNORED_TOO_MODE_MATCHING also should show
"a/b/c" when a pattern "a/*" matched it as "ignored because it
matched an exclude pattern"?

> +                     if (!(dir->flags & DIR_HIDE_EMPTY_DIRECTORIES))
> +                             return path_excluded;
> +
> +                     if (read_directory_recursive(dir, istate, dirname, len,
> +                                                  untracked, 1, 1, pathspec) 
> == path_excluded)
> +                             return path_excluded;
> +
> +                     return path_none;
> +             }

... I guess we won't get into such a situation because this codepath
will return only _excluded or _none when we are handing "a/b" which
is already excluded "a/*", and does not cause the caller to recurse
into "a/b/c" in the first place.  So the updated code looks fine.

>               if (!(dir->flags & DIR_NO_GITLINKS)) {
>                       unsigned char sha1[20];
>                       if (resolve_gitlink_ref(dirname, "HEAD", sha1) == 0)
> diff --git a/dir.h b/dir.h
> index e3717055d1..57b0943dae 100644
> --- a/dir.h
> +++ b/dir.h
> @@ -152,7 +152,8 @@ struct dir_struct {
>               DIR_COLLECT_IGNORED = 1<<4,
>               DIR_SHOW_IGNORED_TOO = 1<<5,
>               DIR_COLLECT_KILLED_ONLY = 1<<6,
> -             DIR_KEEP_UNTRACKED_CONTENTS = 1<<7
> +             DIR_KEEP_UNTRACKED_CONTENTS = 1<<7,
> +             DIR_SHOW_IGNORED_TOO_MODE_MATCHING = 1<<8
>       } flags;
>       struct dir_entry **entries;
>       struct dir_entry **ignored;

Reply via email to