Introduce jsonb, a structured format for storing json.

The new format accepts exactly the same data as the json type. However, it is
stored in a format that does not require reparsing the orgiginal text in order
to process it, making it much more suitable for indexing and other operations.
Insignificant whitespace is discarded, and the order of object keys is not
preserved. Neither are duplicate object keys kept - the later value for a given
key is the only one stored.

The new type has all the functions and operators that the json type has,
with the exception of the json generation functions (to_json, json_agg etc.)
and with identical semantics. In addition, there are operator classes for
hash and btree indexing, and two classes for GIN indexing, that have no
equivalent in the json type.

This feature grew out of previous work by Oleg Bartunov and Teodor Sigaev, which
was intended to provide similar facilities to a nested hstore type, but which
in the end proved to have some significant compatibility issues.

Authors: Oleg Bartunov,  Teodor Sigaev, Peter Geoghegan and Andrew Dunstan.
Review: Andres Freund

1 files changed, 295 insertions, 0 deletions

diff --git a/src/backend/utils/adt/jsonb_op.c b/src/backend/utils/adt/jsonb_op.c
new file mode 100644
index 00000000000..d6b1855c195
--- /dev/null
+++ b/src/backend/utils/adt/jsonb_op.c
@@ -0,0 +1,295 @@
+/*-------------------------------------------------------------------------
+ *
+ * jsonb_op.c
+ *	 Special operators for jsonb only, used by various index access methods
+ *
+ * Copyright (c) 2014, PostgreSQL Global Development Group
+ *
+ *
+ * IDENTIFICATION
+ *	  src/backend/utils/adt/jsonb_op.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+
+#include "miscadmin.h"
+#include "utils/jsonb.h"
+
+Datum
+jsonb_exists(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jb = PG_GETARG_JSONB(0);
+	text	   *key = PG_GETARG_TEXT_PP(1);
+	JsonbValue	kval;
+	JsonbValue *v = NULL;
+
+	/*
+	 * We only match Object keys (which are naturally always Strings), or
+	 * string elements in arrays.  In particular, we do not match non-string
+	 * scalar elements.  Existence of a key/element is only considered at the
+	 * top level.  No recursion occurs.
+	 */
+	kval.type = jbvString;
+	kval.string.val = VARDATA_ANY(key);
+	kval.string.len = VARSIZE_ANY_EXHDR(key);
+
+	v = findJsonbValueFromSuperHeader(VARDATA(jb),
+									  JB_FOBJECT | JB_FARRAY,
+									  NULL,
+									  &kval);
+
+	PG_RETURN_BOOL(v != NULL);
+}
+
+Datum
+jsonb_exists_any(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jb = PG_GETARG_JSONB(0);
+	ArrayType  *keys = PG_GETARG_ARRAYTYPE_P(1);
+	JsonbValue *arrKey = arrayToJsonbSortedArray(keys);
+	uint32	   *plowbound = NULL,
+				lowbound = 0;
+	int			i;
+
+	if (arrKey == NULL || arrKey->object.nPairs == 0)
+		PG_RETURN_BOOL(false);
+
+	if (JB_ROOT_IS_OBJECT(jb))
+		plowbound = &lowbound;
+
+	/*
+	 * We exploit the fact that the pairs list is already sorted into strictly
+	 * increasing order to narrow the findJsonbValueFromSuperHeader search;
+	 * each search can start one entry past the previous "found" entry, or at
+	 * the lower bound of the last search.
+	 */
+	for (i = 0; i < arrKey->array.nElems; i++)
+	{
+		if (findJsonbValueFromSuperHeader(VARDATA(jb),
+										  JB_FOBJECT | JB_FARRAY,
+										  plowbound,
+										  arrKey->array.elems + i) != NULL)
+			PG_RETURN_BOOL(true);
+	}
+
+	PG_RETURN_BOOL(false);
+}
+
+Datum
+jsonb_exists_all(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jb = PG_GETARG_JSONB(0);
+	ArrayType  *keys = PG_GETARG_ARRAYTYPE_P(1);
+	JsonbValue *arrKey = arrayToJsonbSortedArray(keys);
+	uint32	   *plowbound = NULL;
+	uint32		lowbound = 0;
+	int			i;
+
+	if (arrKey == NULL || arrKey->array.nElems == 0)
+		PG_RETURN_BOOL(true);
+
+	if (JB_ROOT_IS_OBJECT(jb))
+		plowbound = &lowbound;
+
+	/*
+	 * We exploit the fact that the pairs list is already sorted into strictly
+	 * increasing order to narrow the findJsonbValueFromSuperHeader search;
+	 * each search can start one entry past the previous "found" entry, or at
+	 * the lower bound of the last search.
+	 */
+	for (i = 0; i < arrKey->array.nElems; i++)
+	{
+		if (findJsonbValueFromSuperHeader(VARDATA(jb),
+										  JB_FOBJECT | JB_FARRAY,
+										  plowbound,
+										  arrKey->array.elems + i) == NULL)
+			PG_RETURN_BOOL(false);
+	}
+
+	PG_RETURN_BOOL(true);
+}
+
+Datum
+jsonb_contains(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *val = PG_GETARG_JSONB(0);
+	Jsonb	   *tmpl = PG_GETARG_JSONB(1);
+
+	JsonbIterator *it1, *it2;
+
+	if (JB_ROOT_COUNT(val) < JB_ROOT_COUNT(tmpl) ||
+		JB_ROOT_IS_OBJECT(val) != JB_ROOT_IS_OBJECT(tmpl))
+		PG_RETURN_BOOL(false);
+
+	it1 = JsonbIteratorInit(VARDATA(val));
+	it2 = JsonbIteratorInit(VARDATA(tmpl));
+
+	PG_RETURN_BOOL(JsonbDeepContains(&it1, &it2));
+}
+
+Datum
+jsonb_contained(PG_FUNCTION_ARGS)
+{
+	/* Commutator of "contains" */
+	Jsonb	   *tmpl = PG_GETARG_JSONB(0);
+	Jsonb	   *val = PG_GETARG_JSONB(1);
+
+	JsonbIterator *it1, *it2;
+
+	if (JB_ROOT_COUNT(val) < JB_ROOT_COUNT(tmpl) ||
+		JB_ROOT_IS_OBJECT(val) != JB_ROOT_IS_OBJECT(tmpl))
+		PG_RETURN_BOOL(false);
+
+	it1 = JsonbIteratorInit(VARDATA(val));
+	it2 = JsonbIteratorInit(VARDATA(tmpl));
+
+	PG_RETURN_BOOL(JsonbDeepContains(&it1, &it2));
+}
+
+Datum
+jsonb_ne(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jba = PG_GETARG_JSONB(0);
+	Jsonb	   *jbb = PG_GETARG_JSONB(1);
+	bool		res;
+
+	res = (compareJsonbSuperHeaderValue(VARDATA(jba), VARDATA(jbb)) != 0);
+
+	PG_FREE_IF_COPY(jba, 0);
+	PG_FREE_IF_COPY(jbb, 1);
+	PG_RETURN_BOOL(res);
+}
+
+/*
+ * B-Tree operator class operators, support function
+ */
+Datum
+jsonb_lt(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jba = PG_GETARG_JSONB(0);
+	Jsonb	   *jbb = PG_GETARG_JSONB(1);
+	bool		res;
+
+	res = (compareJsonbSuperHeaderValue(VARDATA(jba), VARDATA(jbb)) < 0);
+
+	PG_FREE_IF_COPY(jba, 0);
+	PG_FREE_IF_COPY(jbb, 1);
+	PG_RETURN_BOOL(res);
+}
+
+Datum
+jsonb_gt(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jba = PG_GETARG_JSONB(0);
+	Jsonb	   *jbb = PG_GETARG_JSONB(1);
+	bool		res;
+
+	res = (compareJsonbSuperHeaderValue(VARDATA(jba), VARDATA(jbb)) > 0);
+
+	PG_FREE_IF_COPY(jba, 0);
+	PG_FREE_IF_COPY(jbb, 1);
+	PG_RETURN_BOOL(res);
+}
+
+Datum
+jsonb_le(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jba = PG_GETARG_JSONB(0);
+	Jsonb	   *jbb = PG_GETARG_JSONB(1);
+	bool		res;
+
+	res = (compareJsonbSuperHeaderValue(VARDATA(jba), VARDATA(jbb)) <= 0);
+
+	PG_FREE_IF_COPY(jba, 0);
+	PG_FREE_IF_COPY(jbb, 1);
+	PG_RETURN_BOOL(res);
+}
+
+Datum
+jsonb_ge(PG_FUNCTION_ARGS)
+{
+
+	Jsonb	   *jba = PG_GETARG_JSONB(0);
+	Jsonb	   *jbb = PG_GETARG_JSONB(1);
+	bool		res;
+
+	res = (compareJsonbSuperHeaderValue(VARDATA(jba), VARDATA(jbb)) >= 0);
+
+	PG_FREE_IF_COPY(jba, 0);
+	PG_FREE_IF_COPY(jbb, 1);
+	PG_RETURN_BOOL(res);
+}
+
+Datum
+jsonb_eq(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jba = PG_GETARG_JSONB(0);
+	Jsonb	   *jbb = PG_GETARG_JSONB(1);
+	bool		res;
+
+	res = (compareJsonbSuperHeaderValue(VARDATA(jba), VARDATA(jbb)) == 0);
+
+	PG_FREE_IF_COPY(jba, 0);
+	PG_FREE_IF_COPY(jbb, 1);
+	PG_RETURN_BOOL(res);
+}
+
+Datum
+jsonb_cmp(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jba = PG_GETARG_JSONB(0);
+	Jsonb	   *jbb = PG_GETARG_JSONB(1);
+	int			res;
+
+	res = compareJsonbSuperHeaderValue(VARDATA(jba), VARDATA(jbb));
+
+	PG_FREE_IF_COPY(jba, 0);
+	PG_FREE_IF_COPY(jbb, 1);
+	PG_RETURN_INT32(res);
+}
+
+/*
+ * Hash operator class jsonb hashing function
+ */
+Datum
+jsonb_hash(PG_FUNCTION_ARGS)
+{
+	Jsonb	   *jb = PG_GETARG_JSONB(0);
+	JsonbIterator *it;
+	int32		r;
+	JsonbValue	v;
+	uint32		hash = 0;
+
+	if (JB_ROOT_COUNT(jb) == 0)
+		PG_RETURN_INT32(0);
+
+	it = JsonbIteratorInit(VARDATA(jb));
+
+	while ((r = JsonbIteratorNext(&it, &v, false)) != WJB_DONE)
+	{
+		switch (r)
+		{
+			/* Rotation is left to JsonbHashScalarValue() */
+			case WJB_BEGIN_ARRAY:
+				hash ^= JB_FARRAY;
+				break;
+			case WJB_BEGIN_OBJECT:
+				hash ^= JB_FOBJECT;
+				break;
+			case WJB_KEY:
+			case WJB_VALUE:
+			case WJB_ELEM:
+				JsonbHashScalarValue(&v, &hash);
+				break;
+			case WJB_END_ARRAY:
+			case WJB_END_OBJECT:
+				break;
+			default:
+				elog(ERROR, "invalid JsonbIteratorNext rc: %d", r);
+		}
+	}
+
+	PG_FREE_IF_COPY(jb, 0);
+	PG_RETURN_INT32(hash);
+}