Add comments explaining moving comments tests.

Signed-off-by: Thomas Del Vecchio <tdelvecchio@janestreet.com>

test/passing/tests/modes_cmts_move.ml

@@ -1,4 +1,20 @@
+(* This comment moves for similar reason to existing behavior as shown below.
+   In particular, the operator ([@@] / [->]) does not carry a source locaion,
+   and the comment attaches to the location of the identifier on the right side
+   of the operator rather than the left due to the parenthesis. Fixing this would
+   require messing with the comment association logic, which is difficult. *)
 type t = A of t1 @@ m1 m2 * t2 @@ m3 m4 * (t3 @ m5 -> t4 @ m6) (* cmt *) @@ m7 m8
+type t = A of (t -> u) (* cmt *) @@ m
+type t = A of ((t -> u) (* cmt *) -> m)
 
+(* This comment attaches to the [f]; the syntax [(f @ mode1)] doesn't have its own
+   location to latch onto. This seems like a rare position to put a comment, so it doesn't
+   seem worth changing the parser to be able to differentiate the following two. *)
 let (* cmt *) (f @ mode1) (arg1 @ mode2) (arg2 @ mode3) : typ = x
+let (* cmt *) (f @ mode) arg = x
+let ((* cmt *) f @ mode) arg = x
+
+(* This comment moves due to existing behavior as shown below. *)
 let (f @ mode1) (arg1 @ mode2) (arg2 @ mode3) (* cmt *) : typ = x
+let f (arg @ mode) (* cmt *) : typ = x
+let f (arg : typ) (* cmt *) : typ = x

test/passing/tests/modes_cmts_move.ml.js-ref

@@ -1,4 +1,20 @@
+(* This comment moves for similar reason to existing behavior as shown below.
+   In particular, the operator ([@@] / [->]) does not carry a source locaion,
+   and the comment attaches to the location of the identifier on the right side
+   of the operator rather than the left due to the parenthesis. Fixing this would
+   require messing with the comment association logic, which is difficult. *)
 type t = A of t1 @@ m1 m2 * t2 @@ m3 m4 * (t3 @ m5 -> t4 @ m6) @@ (* cmt *) m7 m8
+type t = A of (t -> u) @@ (* cmt *) m
+type t = A of ((t -> u) -> (* cmt *) m)
 
+(* This comment attaches to the [f]; the syntax [(f @ mode1)] doesn't have its own
+   location to latch onto. This seems like a rare position to put a comment, so it doesn't
+   seem worth changing the parser to be able to differentiate the following two. *)
 let ((* cmt *) f @ mode1) (arg1 @ mode2) (arg2 @ mode3) : typ = x
+let ((* cmt *) f @ mode) arg = x
+let ((* cmt *) f @ mode) arg = x
+
+(* This comment moves due to existing behavior as shown below. *)
 let (f @ mode1) (arg1 @ mode2) (arg2 @ mode3) : typ = (* cmt *) x
+let f (arg @ mode) : typ = (* cmt *) x
+let f (arg : typ) : typ = (* cmt *) x

test/passing/tests/modes_cmts_move.ml.ref

@@ -1,6 +1,29 @@
+(* This comment moves for similar reason to existing behavior as shown below.
+   In particular, the operator ([@@] / [->]) does not carry a source locaion,
+   and the comment attaches to the location of the identifier on the right
+   side of the operator rather than the left due to the parenthesis. Fixing
+   this would require messing with the comment association logic, which is
+   difficult. *)
 type t =
   | A of t1 @@ m1 m2 * t2 @@ m3 m4 * (t3 @ m5 -> t4 @ m6) @@ (* cmt *) m7 m8
 
+type t = A of (t -> u) @@ (* cmt *) m
+
+type t = A of ((t -> u) -> (* cmt *) m)
+
+(* This comment attaches to the [f]; the syntax [(f @ mode1)] doesn't have
+   its own location to latch onto. This seems like a rare position to put a
+   comment, so it doesn't seem worth changing the parser to be able to
+   differentiate the following two. *)
 let ((* cmt *) f @ mode1) (arg1 @ mode2) (arg2 @ mode3) : typ = x
 
+let ((* cmt *) f @ mode) arg = x
+
+let ((* cmt *) f @ mode) arg = x
+
+(* This comment moves due to existing behavior as shown below. *)
 let (f @ mode1) (arg1 @ mode2) (arg2 @ mode3) : typ = (* cmt *) x
+
+let f (arg @ mode) : typ = (* cmt *) x
+
+let f (arg : typ) : typ = (* cmt *) x