Re: [PATCH v3 11/19] dir.c: use a single struct exclude_list per source of excludes

Adam Spiers [off-list ref] writes:

This looks somewhat ugly for two reasons.

 * The abstraction add_exclude() offers to its callers is just to
   let them add one pattern to the list of patterns for the kind
   (here, EXC_CMDL); why should they care about .ary[0] part?  Are
   there cases any sane caller (not the implementation of the
   exclude_list_group machinery e.g. add_excludes_from_... function)
   may want to call it with .ary[1]?  I do not think of any.
   Shouldn't the public API function add_exclude() take a pointer to
   the list group itself?

 * When naming an array of things, we tend to prefer naming it

     type thing[count]

   so that the second element can be called "thing[2]" and not
   "things[2]".  dir.exclude_list_group[EXC_CMDL] reads beter.

This is another example where the caller would wish to be able to say

	add_exclude(arg, "", 0, group);

instead.

On Fri, Jan 04, 2013 at 01:03:59PM -0800, Junio C Hamano wrote:

Adam Spiers [off-list ref] writes:

quoted

diff --git a/builtin/clean.c b/builtin/clean.c
index 0c7b3d0..bd18b88 100644
--- a/builtin/clean.c
+++ b/builtin/clean.c

@@ -97,9 +97,10 @@ int cmd_clean(int argc, const char **argv, const char *prefix)
 	if (!ignored)
 		setup_standard_excludes(&dir);
 
+	add_exclude_list(&dir, EXC_CMDL);
 	for (i = 0; i < exclude_list.nr; i++)
 		add_exclude(exclude_list.items[i].string, "", 0,
-			    &dir.exclude_list[EXC_CMDL]);
+			    &dir.exclude_list_groups[EXC_CMDL].ary[0]);

This looks somewhat ugly for two reasons.

 * The abstraction add_exclude() offers to its callers is just to
   let them add one pattern to the list of patterns for the kind
   (here, EXC_CMDL); why should they care about .ary[0] part?

Because the caller has to decide which exclude list the new exclude
should be added to; that is how it has been for a long time, and I am
not proposing we change that.

There are currently three callers:

  builtin/clean.c:    cmd_clean()
  builtin/ls-files.c: option_parse_exclude()
  dir.c:              add_excludes_from_file_to_list()

The first two add to EXC_CMDL, but the latter could be adding to
numerous different possible lists, e.g.

    - .git/info/exclude (EXC_FILE)
    - core.excludesfile (EXC_FILE)
    - any of the per-directory .gitignore lists (EXC_DIRS)

   Are
   there cases any sane caller (not the implementation of the
   exclude_list_group machinery e.g. add_excludes_from_... function)
   may want to call it with .ary[1]?

Effectively yes, although it is not written like ".ary[1]".  For
example prep_exclude() calls add_excludes_from_file_to_list() for each
new .gitignore file

   Shouldn't the public API function add_exclude() take a pointer to
   the list group itself?

Typically EXC_DIRS and EXC_FILE will contain excludes from multiple
sources, whereas EXC_CMDL will contain excludes from a single source.

Therefore when transitioning to exclude_list_groups, I had to make a
choice whether to keep EXC_CMDL as a singleton list, or split it out
into a separate field in struct dir_struct, e.g.:

	struct exclude_list exclude_list_cmdl;
	struct exclude_list exclude_list[2];
#define EXC_DIRS 0
#define EXC_FILE 1

I decided it was cleaner to use a singleton list, because this
preserved the design that excludes in earlier members of
exclude_list[3] take priority over excludes in later members of
exclude_list[3].  That way, this loop in last_exclude_matching():

	for (i = EXC_CMDL; i <= EXC_FILE; i++) {

still makes sense.

 * When naming an array of things, we tend to prefer naming it

     type thing[count]

   so that the second element can be called "thing[2]" and not
   "things[2]".  dir.exclude_list_group[EXC_CMDL] reads beter.

OK, I will change that.

quoted

diff --git a/builtin/ls-files.c b/builtin/ls-files.c
index ef7f99a..c448e06 100644
--- a/builtin/ls-files.c
+++ b/builtin/ls-files.c

@@ -420,10 +420,10 @@ static int option_parse_z(const struct option *opt,
 static int option_parse_exclude(const struct option *opt,
 				const char *arg, int unset)
 {
-	struct exclude_list *list = opt->value;
+	struct exclude_list_group *group = opt->value;
 
 	exc_given = 1;
-	add_exclude(arg, "", 0, list);
+	add_exclude(arg, "", 0, &group->ary[0]);

This is another example where the caller would wish to be able to say

	add_exclude(arg, "", 0, group);

instead.

Why?  The caller needs to decide which exclude list the exclude should
go on, because that determines matching priority.  Specifying an
exclude_list_group is insufficient information to determine that.