revision: mark non-user-given objects instead

Currently, list-objects.c incorrectly treats all root trees of commits
as USER_GIVEN. Also, it would be easier to mark objects that are
non-user-given instead of user-given, since the places in the code
where we access an object through a reference are more obvious than
the places where we access an object that was given by the user.

Resolve these two problems by introducing a flag NOT_USER_GIVEN that
marks blobs and trees that are non-user-given, replacing USER_GIVEN.
(Only blobs and trees are marked because this mark is only used when
filtering objects, and filtering of other types of objects is not
supported yet.)

This fixes a bug in that git rev-list behaved differently from git
pack-objects. pack-objects would *not* filter objects given explicitly
on the command line and rev-list would filter. This was because the two
commands used a different function to add objects to the rev_info
struct. This seems to have been an oversight, and pack-objects has the
correct behavior, so I added a test to make sure that rev-list now
behaves properly.

Signed-off-by: Matthew DeVore <matvore@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

4 files changed, 39 insertions, 16 deletions

diff --git a/list-objects.c b/list-objects.c
index 243192af53..7a1a0929db 100644
--- a/list-objects.c
+++ b/list-objects.c
@@ -53,7 +53,7 @@ static void process_blob(struct traversal_context *ctx,
 
 	pathlen = path->len;
 	strbuf_addstr(path, name);
-	if (!(obj->flags & USER_GIVEN) && ctx->filter_fn)
+	if ((obj->flags & NOT_USER_GIVEN) && ctx->filter_fn)
 		r = ctx->filter_fn(LOFS_BLOB, obj,
 				   path->buf, &path->buf[pathlen],
 				   ctx->filter_data);
@@ -120,17 +120,19 @@ static void process_tree_contents(struct traversal_context *ctx,
 				continue;
 		}
 
-		if (S_ISDIR(entry.mode))
-			process_tree(ctx,
-				     lookup_tree(the_repository, entry.oid),
-				     base, entry.path);
+		if (S_ISDIR(entry.mode)) {
+			struct tree *t = lookup_tree(the_repository, entry.oid);
+			t->object.flags |= NOT_USER_GIVEN;
+			process_tree(ctx, t, base, entry.path);
+		}
 		else if (S_ISGITLINK(entry.mode))
 			process_gitlink(ctx, entry.oid->hash,
 					base, entry.path);
-		else
-			process_blob(ctx,
-				     lookup_blob(the_repository, entry.oid),
-				     base, entry.path);
+		else {
+			struct blob *b = lookup_blob(the_repository, entry.oid);
+			b->object.flags |= NOT_USER_GIVEN;
+			process_blob(ctx, b, base, entry.path);
+		}
 	}
 }
 
@@ -171,7 +173,7 @@ static void process_tree(struct traversal_context *ctx,
 	}
 
 	strbuf_addstr(base, name);
-	if (!(obj->flags & USER_GIVEN) && ctx->filter_fn)
+	if ((obj->flags & NOT_USER_GIVEN) && ctx->filter_fn)
 		r = ctx->filter_fn(LOFS_BEGIN_TREE, obj,
 				   base->buf, &base->buf[baselen],
 				   ctx->filter_data);
@@ -185,7 +187,7 @@ static void process_tree(struct traversal_context *ctx,
 	if (!failed_parse)
 		process_tree_contents(ctx, tree, base);
 
-	if (!(obj->flags & USER_GIVEN) && ctx->filter_fn) {
+	if ((obj->flags & NOT_USER_GIVEN) && ctx->filter_fn) {
 		r = ctx->filter_fn(LOFS_END_TREE, obj,
 				   base->buf, &base->buf[baselen],
 				   ctx->filter_data);
@@ -301,8 +303,11 @@ static void do_traverse(struct traversal_context *ctx)
 		 * an uninteresting boundary commit may not have its tree
 		 * parsed yet, but we are not going to show them anyway
 		 */
-		if (get_commit_tree(commit))
-			add_pending_tree(ctx->revs, get_commit_tree(commit));
+		if (get_commit_tree(commit)) {
+			struct tree *tree = get_commit_tree(commit);
+			tree->object.flags |= NOT_USER_GIVEN;
+			add_pending_tree(ctx->revs, tree);
+		}
 		ctx->show_commit(commit, ctx->show_data);
 
 		if (ctx->revs->tree_blobs_in_commit_order)
diff --git a/revision.c b/revision.c
index 0627494378..6d355b43c3 100644
--- a/revision.c
+++ b/revision.c
@@ -175,7 +175,6 @@ static void add_pending_object_with_path(struct rev_info *revs,
 		strbuf_release(&buf);
 		return; /* do not add the commit itself */
 	}
-	obj->flags |= USER_GIVEN;
 	add_object_array_with_path(obj, name, &revs->pending, mode, path);
 }
 
diff --git a/revision.h b/revision.h
index 5118aaaa92..4dc45bb9ad 100644
--- a/revision.h
+++ b/revision.h
@@ -20,9 +20,16 @@
 #define SYMMETRIC_LEFT	(1u<<8)
 #define PATCHSAME	(1u<<9)
 #define BOTTOM		(1u<<10)
-#define USER_GIVEN	(1u<<25) /* given directly by the user */
+/*
+ * Indicates object was reached by traversal. i.e. not given by user on
+ * command-line or stdin.
+ * NEEDSWORK: NOT_USER_GIVEN doesn't apply to commits because we only support
+ * filtering trees and blobs, but it may be useful to support filtering commits
+ * in the future.
+ */
+#define NOT_USER_GIVEN	(1u<<25)
 #define TRACK_LINEAR	(1u<<26)
-#define ALL_REV_FLAGS	(((1u<<11)-1) | USER_GIVEN | TRACK_LINEAR)
+#define ALL_REV_FLAGS	(((1u<<11)-1) | NOT_USER_GIVEN | TRACK_LINEAR)
 
 #define DECORATE_SHORT_REFS	1
 #define DECORATE_FULL_REFS	2
diff --git a/t/t6112-rev-list-filters-objects.sh b/t/t6112-rev-list-filters-objects.sh
index 6d69e6a0aa..0d4d43894b 100755
--- a/t/t6112-rev-list-filters-objects.sh
+++ b/t/t6112-rev-list-filters-objects.sh
@@ -30,6 +30,18 @@ test_expect_success 'verify blob:none omits all 5 blobs' '
 	test_cmp observed expected
 '
 
+test_expect_success 'specify blob explicitly prevents filtering' '
+	file_3=$(git -C r1 ls-files -s file.3 |
+		 awk -f print_2.awk) &&
+
+	file_4=$(git -C r1 ls-files -s file.4 |
+		 awk -f print_2.awk) &&
+
+	git -C r1 rev-list --objects --filter=blob:none HEAD $file_3 >observed &&
+	grep -q "$file_3" observed &&
+	test_must_fail grep -q "$file_4" observed
+'
+
 test_expect_success 'verify emitted+omitted == all' '
 	git -C r1 rev-list HEAD --objects \
 		| awk -f print_1.awk \