Testing my changes

src/backend/statistics/Makefile

@@ -16,6 +16,9 @@ OBJS = \
 	dependencies.o \
 	extended_stats.o \
 	mcv.o \
-	mvdistinct.o
+	mvdistinct.o \
+	statistics_gram.o
+
+statistics_gram.o: statistics_scanner.c
 
 include $(top_srcdir)/src/backend/common.mk

src/backend/statistics/dependencies.c

@@ -88,6 +88,7 @@ static Selectivity clauselist_apply_dependencies(PlannerInfo *root, List *clause
 												 int ndependencies,
 												 AttrNumber *list_attnums,
 												 Bitmapset **estimatedclauses);
+extern void statistic_scanner_init(const char *query_string);
 
 static void
 generate_dependencies_recurse(DependencyGenerator state, int index,
@@ -648,21 +649,26 @@ statext_dependencies_load(Oid mvoid, bool inh)
 /*
  * pg_dependencies_in		- input routine for type pg_dependencies.
  *
- * pg_dependencies is real enough to be a table column, but it has no operations
- * of its own, and disallows input too
+ * converts the dependencies from the external format in "string" to its
+ * internal format.
  */
 Datum
 pg_dependencies_in(PG_FUNCTION_ARGS)
 {
-	/*
-	 * pg_node_list stores the data in binary form and parsing text input is
-	 * not needed, so disallow this.
-	 */
-	ereport(ERROR,
-			(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
-			 errmsg("cannot accept a value of type %s", "pg_dependencies")));
-
-	PG_RETURN_VOID();			/* keep compiler quiet */
+	char	   *str = PG_GETARG_CSTRING(0);
+	MVDependencies *mvdependencies;
+	int			parse_rc;
+
+	statistic_scanner_init(str);
+	parse_rc = statistic_yyparse();
+	if (parse_rc != 0)
+		ereport(ERROR,
+				(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
+				 errmsg("failed to parse a value of type %s", "pg_dependencies")));
+	statistic_scanner_finish();
+	mvdependencies = mvdependencies_parse_result;
+
+	PG_RETURN_MVNDistinct_P(statext_dependencies_serialize(mvdependencies));
 }
 
 /*

src/backend/statistics/mcv.c

@@ -1469,21 +1469,13 @@ pg_stats_ext_mcvlist_items(PG_FUNCTION_ARGS)
 /*
  * pg_mcv_list_in		- input routine for type pg_mcv_list.
  *
- * pg_mcv_list is real enough to be a table column, but it has no operations
- * of its own, and disallows input too
+ * converts serialized text MCV lists into a byte values by simply
+ * calling byeain().
  */
 Datum
 pg_mcv_list_in(PG_FUNCTION_ARGS)
 {
-	/*
-	 * pg_mcv_list stores the data in binary form and parsing text input is
-	 * not needed, so disallow this.
-	 */
-	ereport(ERROR,
-			(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
-			 errmsg("cannot accept a value of type %s", "pg_mcv_list")));
-
-	PG_RETURN_VOID();			/* keep compiler quiet */
+	PG_RETURN_MCVList_P(byteain(fcinfo));
 }
 
 

src/backend/statistics/meson.build

@@ -6,3 +6,18 @@ backend_sources += files(
   'mcv.c',
   'mvdistinct.c',
 )
+
+stats_scan = custom_target('statsscan',
+  input: 'statistics_scanner.l',
+  output: 'statistics_scanner.c',
+  command: flex_cmd,
+)
+generated_sources += stats_scan
+backend_sources += stats_scan
+
+stats_gram = custom_target('stats_gram',
+  input: 'statistics_gram.y',
+  kwargs: bison_kw,
+)
+generated_sources += stats_gram.to_list()
+backend_sources += stats_gram

src/backend/statistics/mvdistinct.c

@@ -42,6 +42,8 @@ static double estimate_ndistinct(double totalrows, int numrows, int d, int f1);
 static int	n_choose_k(int n, int k);
 static int	num_combinations(int n);
 
+extern void statistic_scanner_init(const char *query_string);
+
 /* size of the struct header fields (magic, type, nitems) */
 #define SizeOfHeader		(3 * sizeof(uint32))
 
@@ -333,17 +335,26 @@ statext_ndistinct_deserialize(bytea *data)
  * pg_ndistinct_in
  *		input routine for type pg_ndistinct
  *
- * pg_ndistinct is real enough to be a table column, but it has no
- * operations of its own, and disallows input (just like pg_node_tree).
+ * converts the distinct from the external format in "string" to its internal
+ * format.
  */
 Datum
 pg_ndistinct_in(PG_FUNCTION_ARGS)
 {
-	ereport(ERROR,
-			(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
-			 errmsg("cannot accept a value of type %s", "pg_ndistinct")));
-
-	PG_RETURN_VOID();			/* keep compiler quiet */
+	char	   *str = PG_GETARG_CSTRING(0);
+	MVNDistinct *mvndistinct;
+	int			parse_rc;
+
+	statistic_scanner_init(str);
+	parse_rc = statistic_yyparse();
+	if (parse_rc != 0)
+		ereport(ERROR,
+				(errcode(ERRCODE_INVALID_TEXT_REPRESENTATION),
+				 errmsg("failed to parse a value of type %s", "pg_ndistinct")));
+	statistic_scanner_finish();
+	mvndistinct = mvndistinct_parse_result;
+
+	PG_RETURN_MVNDistinct_P(statext_ndistinct_serialize(mvndistinct));
 }
 
 /*

src/backend/statistics/statistics_gram.y

@@ -0,0 +1,176 @@
+%{
+#include "postgres.h"
+
+#include "statistics/extended_stats_internal.h"
+#include "statistics/statistics.h"
+
+MVNDistinct *mvndistinct_parse_result;
+MVDependencies *mvdependencies_parse_result;
+/*
+ * Bison doesn't allocate anything that needs to live across parser calls,
+ * so we can easily have it use palloc instead of malloc.  This prevents
+ * memory leaks if we error out during parsing.  Note this only works with
+ * bison >= 2.0.  However, in bison 1.875 the default is to use alloca()
+ * if possible, so there's not really much problem anyhow, at least if
+ * you're building with gcc.
+ */
+#define YYMALLOC palloc
+#define YYFREE   pfree
+
+%}
+
+%expect 0
+%name-prefix="statistic_yy"
+
+
+%union {
+		uint32					uintval;
+		double					doubleval;
+
+		MVNDistinct				*ndistinct;
+		MVNDistinctItem				*ndistinct_item;
+
+		MVDependencies				*dependencies;
+		MVDependency				*dependency;
+
+		Bitmapset				*bitmap;
+		List					*list;
+}
+
+/* Non-keyword tokens */
+%token <uintval> UCONST
+%token <doubleval> DOUBLE
+%token ARROW
+
+%type <ndistinct>	ndistinct
+%type <ndistinct_item>	ndistinct_item
+%type <list>	ndistinct_item_list
+
+%type <dependencies>	dependencies
+%type <list>	dependency_item_list
+%type <dependency>	dependency_item
+
+%type <bitmap>	attrs
+
+%%
+
+extended_statistic:
+	ndistinct { } |
+	dependencies { }
+	;
+
+/*
+ * "ndistinct" rule helps to parse the input string recursively and stores the output into MVNDistinct structure.
+ * Exmple:
+ * 	intput : '{"1, 2": 1,"2, 3": 2, "3, 1", 2}'
+ * 	output : returns MVNDistinct object
+*/
+ndistinct:
+	'{' ndistinct_item_list '}'
+		{
+			$$ = palloc0(MAXALIGN(offsetof(MVNDistinct, items)) +
+						 list_length($2) * sizeof(MVNDistinctItem));
+			mvndistinct_parse_result = $$;
+			$$->magic = STATS_NDISTINCT_MAGIC;
+			$$->type = STATS_NDISTINCT_TYPE_BASIC;
+			$$->nitems = list_length($2);
+
+			ListCell *cell;
+			MVNDistinctItem *pointer = $$->items;
+			foreach (cell, $2)
+			{
+				memcpy(pointer, lfirst(cell), sizeof(MVNDistinctItem));
+				pointer += 1;
+			}
+		}
+	;
+
+ndistinct_item_list:
+	ndistinct_item_list ',' ndistinct_item
+		{
+			$$ = lappend($1, $3);
+		}
+	| ndistinct_item { $$ = lappend(NIL, $1);}
+	;
+
+ndistinct_item:
+	'"' attrs '"' ':' DOUBLE
+		{
+			int attrCount = 0;
+			$$ = (MVNDistinctItem *)palloc0(sizeof(MVNDistinctItem));
+			$$->attributes = build_attnums_array($2, 0, &attrCount);
+			$$->nattributes = attrCount;
+			$$->ndistinct = $5;
+		}
+	;
+
+attrs:
+	attrs ',' UCONST
+		{
+			$$ = bms_add_member($1, $3);
+
+		}
+	| UCONST ',' UCONST
+		{
+			$$ = bms_make_singleton($1);
+			$$ = bms_add_member($$, $3);
+		}
+	;
+
+/*
+ * "dependencies" rule helps to parse the input string recursively and stores the output into MVDependencies structure.
+ * example:
+ *	intput : '{"1 => 2": 1.000000, "2 => 3": 2.000000}'
+ * 	output : returns MVDependencies	object
+*/
+dependencies:
+	'{' dependency_item_list '}'
+		{
+			$$ = palloc0(MAXALIGN(offsetof(MVDependencies, deps)) + list_length($2) * sizeof(MVDependency *));
+			mvdependencies_parse_result = $$;
+
+			$$->magic = STATS_DEPS_MAGIC;
+			$$->type = STATS_DEPS_TYPE_BASIC;
+			$$->ndeps = list_length($2);
+
+			for (int i=0; i<$$->ndeps; i++)
+			{
+				$$->deps[i] = list_nth($2, i);
+			}
+		}
+	;
+
+dependency_item_list:
+	dependency_item_list ',' dependency_item
+		{
+			$$ = lappend($1, $3);
+		}
+	| dependency_item { $$ = lappend(NIL, $1);}
+	;
+
+dependency_item:
+	'"' attrs ARROW UCONST '"' ':' DOUBLE
+	{
+		int attrCount = 0;
+		AttrNumber *ptr = build_attnums_array($2, 0, &attrCount);
+		$$ = (MVDependency *)palloc0(sizeof(MVDependency) + sizeof(AttrNumber) * (attrCount + 1));
+		$$->nattributes = attrCount + 1;
+		$$->degree = $7;
+		for (int i = 0; i < attrCount; i++)
+		{
+			$$->attributes[i] = *(ptr+i);
+		}
+		$$->attributes[$$->nattributes - 1] = $4;
+	}
+	| '"' UCONST ARROW UCONST '"' ':' DOUBLE
+	{
+		$$ = (MVDependency *)palloc0(sizeof(MVDependency) + sizeof(AttrNumber) * 2);
+		$$->nattributes = 2;
+		$$->degree = $7;
+		$$->attributes[0] = $2;
+		$$->attributes[1] = $4;
+	}
+	;
+%%
+
+#include "statistics_scanner.c"