Bugfix: Autokeying was non-functional after the Keying Sets refactor

- The RNA wrapping for the generate callback was still wrong, with the primary effect being that C-code calling this had unexpected consequences that were hard to debug.
- Fixed some defective checks that meant that when specifying the RNA-pointers for the Keying Set to use (rather than using the Keying Set's own iterator callback) would never add any info.

1 files changed, 6 insertions, 3 deletions

diff --git a/source/blender/editors/animation/keyingsets.c b/source/blender/editors/animation/keyingsets.c
index b1f1cbbea1e..a93220aeb67 100644
--- a/source/blender/editors/animation/keyingsets.c
+++ b/source/blender/editors/animation/keyingsets.c
@@ -662,11 +662,14 @@ void ANIM_relative_keyingset_add_source (ListBase *dsources, ID *id, StructRNA *
 	tRKS_DSource *ds;
 	
 	/* sanity checks 
-	 *	we must have at least one valid data pointer to use 
+	 *	- we must have somewhere to output the data
+	 *	- we must have both srna+data (and with id too optionally), or id by itself only
 	 */
-	if (ELEM(NULL, dsources, srna) || ((id == data) && (id == NULL)))
+	if (dsources == NULL)
 		return;
-
+	if (ELEM(NULL, srna, data) && (id == NULL))
+		return;
+	
 	/* allocate new elem, and add to the list */
 	ds = MEM_callocN(sizeof(tRKS_DSource), "tRKS_DSource");
 	BLI_addtail(dsources, ds);