Support all SQL:2011 options for window frame clauses.

This patch adds the ability to use "RANGE offset PRECEDING/FOLLOWING"
frame boundaries in window functions.  We'd punted on that back in the
original patch to add window functions, because it was not clear how to
do it in a reasonably data-type-extensible fashion.  That problem is
resolved here by adding the ability for btree operator classes to provide
an "in_range" support function that defines how to add or subtract the
RANGE offset value.  Factoring it this way also allows the operator class
to avoid overflow problems near the ends of the datatype's range, if it
wishes to expend effort on that.  (In the committed patch, the integer
opclasses handle that issue, but it did not seem worth the trouble to
avoid overflow failures for datetime types.)

The patch includes in_range support for the integer_ops opfamily
(int2/int4/int8) as well as the standard datetime types.  Support for
other numeric types has been requested, but that seems like suitable
material for a follow-on patch.

In addition, the patch adds GROUPS mode which counts the offset in
ORDER-BY peer groups rather than rows, and it adds the frame_exclusion
options specified by SQL:2011.  As far as I can see, we are now fully
up to spec on window framing options.

Existing behaviors remain unchanged, except that I changed the errcode
for a couple of existing error reports to meet the SQL spec's expectation
that negative "offset" values should be reported as SQLSTATE 22013.

Internally and in relevant parts of the documentation, we now consistently
use the terminology "offset PRECEDING/FOLLOWING" rather than "value
PRECEDING/FOLLOWING", since the term "value" is confusingly vague.

Oliver Ford, reviewed and whacked around some by me

Discussion: https://postgr.es/m/CAGMVOdu9sivPAxbNN0X+q19Sfv9edEPv=HibOJhB14TJv_RCQg@mail.gmail.com

1 files changed, 107 insertions, 0 deletions

diff --git a/src/backend/utils/adt/date.c b/src/backend/utils/adt/date.c
index 747ef497897..eea29044146 100644
--- a/src/backend/utils/adt/date.c
+++ b/src/backend/utils/adt/date.c
@@ -1011,6 +1011,34 @@ timestamptz_cmp_date(PG_FUNCTION_ARGS)
 	PG_RETURN_INT32(timestamptz_cmp_internal(dt1, dt2));
 }
 
+/*
+ * in_range support function for date.
+ *
+ * We implement this by promoting the dates to timestamp (without time zone)
+ * and then using the timestamp-and-interval in_range function.
+ */
+Datum
+in_range_date_interval(PG_FUNCTION_ARGS)
+{
+	DateADT		val = PG_GETARG_DATEADT(0);
+	DateADT		base = PG_GETARG_DATEADT(1);
+	Interval   *offset = PG_GETARG_INTERVAL_P(2);
+	bool		sub = PG_GETARG_BOOL(3);
+	bool		less = PG_GETARG_BOOL(4);
+	Timestamp	valStamp;
+	Timestamp	baseStamp;
+
+	valStamp = date2timestamp(val);
+	baseStamp = date2timestamp(base);
+
+	return DirectFunctionCall5(in_range_timestamp_interval,
+							   TimestampGetDatum(valStamp),
+							   TimestampGetDatum(baseStamp),
+							   IntervalPGetDatum(offset),
+							   BoolGetDatum(sub),
+							   BoolGetDatum(less));
+}
+
 
 /* Add an interval to a date, giving a new date.
  * Must handle both positive and negative intervals.
@@ -1842,6 +1870,45 @@ time_mi_interval(PG_FUNCTION_ARGS)
 	PG_RETURN_TIMEADT(result);
 }
 
+/*
+ * in_range support function for time.
+ */
+Datum
+in_range_time_interval(PG_FUNCTION_ARGS)
+{
+	TimeADT		val = PG_GETARG_TIMEADT(0);
+	TimeADT		base = PG_GETARG_TIMEADT(1);
+	Interval   *offset = PG_GETARG_INTERVAL_P(2);
+	bool		sub = PG_GETARG_BOOL(3);
+	bool		less = PG_GETARG_BOOL(4);
+	TimeADT		sum;
+
+	/*
+	 * Like time_pl_interval/time_mi_interval, we disregard the month and day
+	 * fields of the offset.  So our test for negative should too.
+	 */
+	if (offset->time < 0)
+		ereport(ERROR,
+				(errcode(ERRCODE_INVALID_PRECEDING_FOLLOWING_SIZE),
+				 errmsg("invalid preceding or following size in window function")));
+
+	/*
+	 * We can't use time_pl_interval/time_mi_interval here, because their
+	 * wraparound behavior would give wrong (or at least undesirable) answers.
+	 * Fortunately the equivalent non-wrapping behavior is trivial, especially
+	 * since we don't worry about integer overflow.
+	 */
+	if (sub)
+		sum = base - offset->time;
+	else
+		sum = base + offset->time;
+
+	if (less)
+		PG_RETURN_BOOL(val <= sum);
+	else
+		PG_RETURN_BOOL(val >= sum);
+}
+
 
 /* time_part()
  * Extract specified field from time type.
@@ -2305,6 +2372,46 @@ timetz_mi_interval(PG_FUNCTION_ARGS)
 	PG_RETURN_TIMETZADT_P(result);
 }
 
+/*
+ * in_range support function for timetz.
+ */
+Datum
+in_range_timetz_interval(PG_FUNCTION_ARGS)
+{
+	TimeTzADT  *val = PG_GETARG_TIMETZADT_P(0);
+	TimeTzADT  *base = PG_GETARG_TIMETZADT_P(1);
+	Interval   *offset = PG_GETARG_INTERVAL_P(2);
+	bool		sub = PG_GETARG_BOOL(3);
+	bool		less = PG_GETARG_BOOL(4);
+	TimeTzADT	sum;
+
+	/*
+	 * Like timetz_pl_interval/timetz_mi_interval, we disregard the month and
+	 * day fields of the offset.  So our test for negative should too.
+	 */
+	if (offset->time < 0)
+		ereport(ERROR,
+				(errcode(ERRCODE_INVALID_PRECEDING_FOLLOWING_SIZE),
+				 errmsg("invalid preceding or following size in window function")));
+
+	/*
+	 * We can't use timetz_pl_interval/timetz_mi_interval here, because their
+	 * wraparound behavior would give wrong (or at least undesirable) answers.
+	 * Fortunately the equivalent non-wrapping behavior is trivial, especially
+	 * since we don't worry about integer overflow.
+	 */
+	if (sub)
+		sum.time = base->time - offset->time;
+	else
+		sum.time = base->time + offset->time;
+	sum.zone = base->zone;
+
+	if (less)
+		PG_RETURN_BOOL(timetz_cmp_internal(val, &sum) <= 0);
+	else
+		PG_RETURN_BOOL(timetz_cmp_internal(val, &sum) >= 0);
+}
+
 /* overlaps_timetz() --- implements the SQL OVERLAPS operator.
  *
  * Algorithm is per SQL spec.  This is much harder than you'd think