Split CollateClause into separate raw and analyzed node types.

CollateClause is now used only in raw grammar output, and CollateExpr after
parse analysis.  This is for clarity and to avoid carrying collation names
in post-analysis parse trees: that's both wasteful and possibly misleading,
since the collation's name could be changed while the parsetree still
exists.

Also, clean up assorted infelicities and omissions in processing of the
node type.

src/include/nodes/primnodes.h

@@ -653,18 +653,6 @@ typedef struct RelabelType
 	int			location;		/* token location, or -1 if unknown */
 } RelabelType;
 
-/*
- * CollateClause - COLLATE
- */
-typedef struct CollateClause
-{
-	Expr		xpr;
-	Expr	   *arg;			/* original expression */
-	List	   *collnames;		/* assigned collation */
-	Oid			collOid;		/* resolved collation OID */
-	int			location;		/* token location, or -1 if unknown */
-} CollateClause;
-
 /* ----------------
  * CoerceViaIO
  *
@@ -730,6 +718,18 @@ typedef struct ConvertRowtypeExpr
 	int			location;		/* token location, or -1 if unknown */
 } ConvertRowtypeExpr;
 
+/*----------
+ * CollateExpr - COLLATE
+ *----------
+ */
+typedef struct CollateExpr
+{
+	Expr		xpr;
+	Expr	   *arg;			/* input expression */
+	Oid			collOid;		/* collation's OID */
+	int			location;		/* token location, or -1 if unknown */
+} CollateExpr;
+
 /*----------
  * CaseExpr - a CASE expression
  *