Add jsonb <=> bytea casts

Author: Nikita Glukhov

src/backend/utils/adt/jsonb.c

@@ -215,6 +215,23 @@ jsonb_typeof(PG_FUNCTION_ARGS)
 	PG_RETURN_TEXT_P(cstring_to_text(result));
 }
 
+/*
+ * jsonb_from_bytea -- bytea to jsonb conversion with validation
+ */
+Datum
+jsonb_from_bytea(PG_FUNCTION_ARGS)
+{
+	bytea	   *ba = PG_GETARG_BYTEA_P(0);
+	Jsonb	   *jb = (Jsonb *) ba;
+
+	if (!JsonbValidate(VARDATA(jb), VARSIZE(jb) - VARHDRSZ))
+		ereport(ERROR,
+				(errcode(ERRCODE_INVALID_BINARY_REPRESENTATION),
+				 errmsg("incorrect jsonb binary data format")));
+
+	PG_RETURN_JSONB(jb);
+}
+
 /*
  * jsonb_from_cstring
  *

src/backend/utils/adt/jsonb_util.c

@@ -1843,3 +1843,133 @@ uniqueifyJsonbObject(JsonbValue *object)
 		object->val.object.nPairs = res + 1 - object->val.object.pairs;
 	}
 }
+
+bool
+JsonbValidate(const void *data, uint32 size)
+{
+	const JsonbContainer *jbc = data;
+	const JEntry *entries;
+	const char *base;
+	uint32		header;
+	uint32		nentries;
+	uint32		offset;
+	uint32		i;
+	uint32		nkeys = 0;
+
+	check_stack_depth();
+
+	/* check header size */
+	if (size < offsetof(JsonbContainer, children))
+		return false;
+
+	size -= offsetof(JsonbContainer, children);
+
+	/* read header */
+	header = jbc->header;
+	entries = &jbc->children[0];
+	nentries = header & JB_CMASK;
+
+	switch (header & ~JB_CMASK)
+	{
+		case JB_FOBJECT:
+			nkeys = nentries;
+			nentries *= 2;
+			break;
+
+		case JB_FARRAY | JB_FSCALAR:
+			if (nentries != 1)
+				/* scalar pseudo-array must have one element */
+				return false;
+			break;
+
+		case JB_FARRAY:
+			break;
+
+		default:
+			/* invalid container type */
+			return false;
+	}
+
+	/* check JEntry array size */
+	if (size < sizeof(JEntry) * nentries)
+		return false;
+
+	size -= sizeof(JEntry) * nentries;
+
+	base = (const char *) &entries[nentries];
+
+	for (i = 0, offset = 0; i < nentries; i++)
+	{
+		JEntry		entry = entries[i];
+		uint32		offlen = JBE_OFFLENFLD(entry);
+		uint32		length;
+		uint32		nextOffset;
+
+		if (JBE_HAS_OFF(entry))
+		{
+			nextOffset = offlen;
+			length = nextOffset - offset;
+		}
+		else
+		{
+			length = offlen;
+			nextOffset = offset + length;
+		}
+
+		if (nextOffset < offset || nextOffset > size)
+			return false;
+
+		/* check that object key is string */
+		if (i < nkeys && !JBE_ISSTRING(entry))
+			return false;
+
+		switch (entry & JENTRY_TYPEMASK)
+		{
+			case JENTRY_ISSTRING:
+				break;
+
+			case JENTRY_ISNUMERIC:
+				{
+					uint32		padding = INTALIGN(offset) - offset;
+
+					if (length < padding)
+						return false;
+
+					if (length - padding < sizeof(struct varlena))
+						return false;
+
+					/* TODO JsonValidateNumeric(); */
+					break;
+				}
+
+			case JENTRY_ISBOOL_FALSE:
+			case JENTRY_ISBOOL_TRUE:
+			case JENTRY_ISNULL:
+				if (length)
+					return false;
+
+				break;
+
+			case JENTRY_ISCONTAINER:
+				{
+					uint32		padding = INTALIGN(offset) - offset;
+
+					if (length < padding)
+						return false;
+
+					if (!JsonbValidate(&base[offset + padding],
+									   length - padding))
+						return false;
+
+					break;
+				}
+
+			default:
+				return false;
+		}
+
+		offset = nextOffset;
+	}
+
+	return true;
+}

src/include/catalog/pg_cast.h

@@ -391,5 +391,8 @@ DATA(insert ( 1700 1700 1703 i f ));
 /* json to/from jsonb */
 DATA(insert (  114 3802    0 a i ));
 DATA(insert ( 3802	114    0 a i ));
+/* jsonb to/from bytea */
+DATA(insert ( 3802   17    0 e b ));
+DATA(insert (   17 3802 6105 e f ));
 
 #endif							/* PG_CAST_H */

src/include/catalog/pg_proc.h

@@ -4911,6 +4911,8 @@ DATA(insert OID =  3804 (  jsonb_out		PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0
 DESCR("I/O");
 DATA(insert OID =  3803 (  jsonb_send		PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 17 "3802" _null_ _null_ _null_ _null_ _null_ jsonb_send _null_ _null_ _null_ ));
 DESCR("I/O");
+DATA(insert OID =  6105 (  jsonb_from_bytea	PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 3802 "17" _null_ _null_ _null_ _null_ _null_ jsonb_from_bytea _null_ _null_ _null_ ));
+DESCR("bytea to jsonb");
 
 DATA(insert OID = 3263 (  jsonb_object	 PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 3802 "1009" _null_ _null_ _null_ _null_ _null_ jsonb_object _null_ _null_ _null_ ));
 DESCR("map text array of key value pairs to jsonb object");

src/include/utils/jsonb.h

@@ -389,6 +389,7 @@ extern Jsonb *JsonbValueToJsonb(JsonbValue *val);
 extern bool JsonbDeepContains(JsonbIterator **val,
 				  JsonbIterator **mContained);
 extern void JsonbHashScalarValue(const JsonbValue *scalarVal, uint32 *hash);
+extern bool JsonbValidate(const void *data, uint32 size);
 
 /* jsonb.c support functions */
 extern char *JsonbToCString(StringInfo out, JsonbContainer *in,

src/test/regress/expected/jsonb.out

@@ -4260,3 +4260,24 @@ select ts_headline('[]'::jsonb, tsquery('aaa & bbb'));
  []
 (1 row)
 
+-- test jsonb to/from bytea conversion
+SELECT '{"a": 1, "b": [2, true]}'::jsonb::bytea;
+                                                   bytea                                                    
+------------------------------------------------------------------------------------------------------------
+ \x0200002001000080010000000a000010140000506162000020000000008001000200004008000090000000302000000000800200
+(1 row)
+
+SELECT '{"a": 1, "b": [2, true]}'::jsonb::bytea::jsonb;
+          jsonb           
+--------------------------
+ {"a": 1, "b": [2, true]}
+(1 row)
+
+SELECT 'aaaa'::bytea::jsonb;
+ERROR:  incorrect jsonb binary data format
+SELECT count(*) FROM testjsonb WHERE j::bytea::jsonb <> j;
+ count 
+-------
+     0
+(1 row)
+

src/test/regress/expected/opr_sanity.out

@@ -845,6 +845,9 @@ WHERE c.castfunc = p.oid AND
 -- texttoxml(), which does an XML syntax check.
 -- As of 9.1, this finds the cast from pg_node_tree to text, which we
 -- intentionally do not provide a reverse pathway for.
+-- As of 10.0, this finds the cast from jsonb to bytea, because those are
+-- binary-compatible while the reverse goes through jsonb_from_bytea(),
+-- which does a jsonb structure validation.
 SELECT castsource::regtype, casttarget::regtype, castfunc, castcontext
 FROM pg_cast c
 WHERE c.castmethod = 'b' AND
@@ -863,7 +866,8 @@ WHERE c.castmethod = 'b' AND
  xml               | text              |        0 | a
  xml               | character varying |        0 | a
  xml               | character         |        0 | a
-(9 rows)
+ jsonb             | bytea             |        0 | e
+(10 rows)
 
 -- **************** pg_conversion ****************
 -- Look for illegal values in pg_conversion fields.

src/test/regress/sql/jsonb.sql

@@ -1099,3 +1099,9 @@ select ts_headline('english', '{"a": "aaa bbb", "b": {"c": "ccc ddd fff", "c1":
 select ts_headline('null'::jsonb, tsquery('aaa & bbb'));
 select ts_headline('{}'::jsonb, tsquery('aaa & bbb'));
 select ts_headline('[]'::jsonb, tsquery('aaa & bbb'));
+
+-- test jsonb to/from bytea conversion
+SELECT '{"a": 1, "b": [2, true]}'::jsonb::bytea;
+SELECT '{"a": 1, "b": [2, true]}'::jsonb::bytea::jsonb;
+SELECT 'aaaa'::bytea::jsonb;
+SELECT count(*) FROM testjsonb WHERE j::bytea::jsonb <> j;

src/test/regress/sql/opr_sanity.sql

@@ -480,6 +480,10 @@ WHERE c.castfunc = p.oid AND
 -- As of 9.1, this finds the cast from pg_node_tree to text, which we
 -- intentionally do not provide a reverse pathway for.
 
+-- As of 10.0, this finds the cast from jsonb to bytea, because those are
+-- binary-compatible while the reverse goes through jsonb_from_bytea(),
+-- which does a jsonb structure validation.
+
 SELECT castsource::regtype, casttarget::regtype, castfunc, castcontext
 FROM pg_cast c
 WHERE c.castmethod = 'b' AND