Compositor Tests: Reorganize folders to match node groupings

This addresses issue #120949 to move compositor tests to reflect the
grouping used when adding a new node.

This PR only moves the relevant single tests and their renders into
matching directories. Folders such as 'multi-node setups' and
'pixel nodes' were not changed.

Pull Request: https://projects.blender.org/blender/blender/pulls/139757

tests/python/CMakeLists.txt

@@ -907,11 +907,11 @@ endif()
 
 if(TEST_SRC_DIR_EXISTS)
   set(compositor_tests
-    color
-    converter
-    filter
     input
     output
+    color
+    filter
+    utilities    
     vector
 
     pixel_nodes
@@ -919,7 +919,7 @@ if(TEST_SRC_DIR_EXISTS)
   )
 
   if(WITH_LIBMV)
-    list(APPEND compositor_tests distort matte anisotropic_filtering)
+    list(APPEND compositor_tests keying mask tracking transform anisotropic_filtering)
   endif()
 
   foreach(comp_test ${compositor_tests})
@@ -942,11 +942,11 @@ endif()
 
 if(WITH_GPU_COMPOSITOR_TESTS AND TEST_SRC_DIR_EXISTS)
   set(compositor_tests
-    color
-    converter
-    filter
     input
     output
+    color
+    filter
+    utilities    
     vector
 
     pixel_nodes
@@ -954,7 +954,7 @@ if(WITH_GPU_COMPOSITOR_TESTS AND TEST_SRC_DIR_EXISTS)
   )
 
   if(WITH_LIBMV)
-    list(APPEND compositor_tests distort matte)
+    list(APPEND compositor_tests keying mask tracking transform)
   endif()
 
   if(WITH_OPENGL_BACKEND)

tests/python/compositor_render_tests.py

@@ -65,7 +65,7 @@ def main():
     if os.path.basename(args.testdir) == 'filter':
         # Temporary change to pass OpenImageDenoise test with both 1.3 and 1.4.
         report.set_fail_threshold(0.05)
-    elif os.path.basename(args.testdir) == 'matte':
+    elif os.path.basename(args.testdir) == 'mask' or os.path.basename(args.testdir) == 'keying':
         # The node_keying_matte.blend test is very sensitive to the exact values in the
         # input image. It makes it hard to precisely match results on different systems
         # (with and without SSE, i.e.), especially when OCIO has different precision for