Merge remote-tracking branch 'origin/master' into SPARK-29783

sql/catalyst/src/main/scala/org/apache/spark/sql/internal/SQLConf.scala

@@ -1798,14 +1798,6 @@ object SQLConf {
       .checkValues(StoreAssignmentPolicy.values.map(_.toString))
       .createWithDefault(StoreAssignmentPolicy.ANSI.toString)
 
-  val ANSI_ENABLED = buildConf("spark.sql.ansi.enabled")
-    .doc("When true, Spark tries to conform to the ANSI SQL specification: 1. Spark will " +
-      "throw a runtime exception if an overflow occurs in any operation on integral/decimal " +
-      "field. 2. Spark will forbid using the reserved keywords of ANSI SQL as identifiers in " +
-      "the SQL parser.")
-    .booleanConf
-    .createWithDefault(false)
-
   object IntervalStyle extends Enumeration {
     type IntervalStyle = Value
     val SQL_STANDARD, ISO_8601, MULTI_UNITS = Value
@@ -2554,7 +2546,7 @@ class SQLConf extends Serializable with Logging {
 
   def intervalOutputStyle: IntervalStyle.Value = IntervalStyle.withName(getConf(INTERVAL_STYLE))
 
-  def ansiEnabled: Boolean = getConf(ANSI_ENABLED)
+  def usePostgreSQLDialect: Boolean = getConf(DIALECT) == Dialect.POSTGRESQL.toString
 
   def dialectSparkAnsiEnabled: Boolean = getConf(DIALECT_SPARK_ANSI_ENABLED)
 

sql/core/src/test/resources/sql-tests/inputs/interval.sql

@@ -91,45 +91,169 @@ select justify_days(interval '1 month 59 day -25 hour');
 select justify_hours(interval '1 month 59 day -25 hour');
 select justify_interval(interval '1 month 59 day -25 hour');
 
--- interval output style
-set spark.sql.intervalOutputStyle=SQL_STANDARD;
-SELECT
-  cast(null as interval), -- null
-  interval '0 day', -- 0
-  interval '1 year', -- year only
-  interval '1 month', -- month only
-  interval '1 year 2 month', -- year month only
-  interval '1 day -1 hours',
-  interval '-1 day -1 hours',
-  interval '-1 day 1 hours',
-  interval '-1 days +1 hours',
-  interval '1 years 2 months -3 days 4 hours 5 minutes 6.789 seconds',
-  - interval '1 years 2 months -3 days 4 hours 5 minutes 6.789 seconds';
-
-set spark.sql.intervalOutputStyle=ISO_8601;
+-- interval literal
+select interval 13.123456789 seconds, interval -13.123456789 second;
+select interval 1 year 2 month 3 week 4 day 5 hour 6 minute 7 seconds 8 millisecond 9 microsecond;
+select interval '30' year '25' month '-100' day '40' hour '80' minute '299.889987299' second;
+select interval '0 0:0:0.1' day to second;
+select interval '10-9' year to month;
+select interval '20 15:40:32.99899999' day to hour;
+select interval '20 15:40:32.99899999' day to minute;
+select interval '20 15:40:32.99899999' day to second;
+select interval '15:40:32.99899999' hour to minute;
+select interval '15:40.99899999' hour to second;
+select interval '15:40' hour to second;
+select interval '15:40:32.99899999' hour to second;
+select interval '20 40:32.99899999' minute to second;
+select interval '40:32.99899999' minute to second;
+select interval '40:32' minute to second;
+select interval 30 day day;
+
+-- ns is not supported
+select interval 10 nanoseconds;
+
+-- map + interval test
+select map(1, interval 1 day, 2, interval 3 week);
+
+-- typed interval expression
+select interval 'interval 3 year 1 hour';
+select interval '3 year 1 hour';
+
+-- malformed interval literal
+select interval;
+select interval 1 fake_unit;
+select interval 1 year to month;
+select interval '1' year to second;
+select interval '10-9' year to month '2-1' year to month;
+select interval '10-9' year to month '12:11:10' hour to second;
+select interval '1 15:11' day to minute '12:11:10' hour to second;
+select interval 1 year '2-1' year to month;
+select interval 1 year '12:11:10' hour to second;
+select interval '10-9' year to month '1' year;
+select interval '12:11:10' hour to second '1' year;
+select interval (-30) day;
+select interval (a + 1) day;
+select interval 30 day day day;
+
+-- sum interval values
+-- null
+select sum(cast(null as interval));
+
+-- empty set
+select sum(cast(v as interval)) from VALUES ('1 seconds') t(v) where 1=0;
+
+-- basic interval sum
+select sum(cast(v as interval)) from VALUES ('1 seconds'), ('2 seconds'), (null) t(v);
+select sum(cast(v as interval)) from VALUES ('-1 seconds'), ('2 seconds'), (null) t(v);
+select sum(cast(v as interval)) from VALUES ('-1 seconds'), ('-2 seconds'), (null) t(v);
+select sum(cast(v as interval)) from VALUES ('-1 weeks'), ('2 seconds'), (null) t(v);
+
+-- group by
+select
+    i,
+    sum(cast(v as interval))
+from VALUES (1, '-1 weeks'), (2, '2 seconds'), (3, null), (1, '5 days') t(i, v)
+group by i;
+
+-- having
+select
+    sum(cast(v as interval)) as sv
+from VALUES (1, '-1 weeks'), (2, '2 seconds'), (3, null), (1, '5 days') t(i, v)
+having sv is not null;
+
+-- window
 SELECT
-  cast(null as interval), -- null
-  interval '0 day', -- 0
-  interval '1 year', -- year only
-  interval '1 month', -- month only
-  interval '1 year 2 month', -- year month only
-  interval '-1 day -1 hours',
-  interval '-1 day 1 hours',
-  interval '1 day -1 hours',
-  interval '-1 days +1 hours',
-  interval '1 years 2 months -3 days 4 hours 5 minutes 6.789 seconds',
-  - interval '1 years 2 months -3 days 4 hours 5 minutes 6.789 seconds';
-
-set spark.sql.intervalOutputStyle=MULTI_UNITS;
+    i,
+    sum(cast(v as interval)) OVER (ORDER BY i ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING)
+FROM VALUES(1, '1 seconds'), (1, '2 seconds'), (2, NULL), (2, NULL) t(i,v);
+
+-- average with interval type
+-- null
+select avg(cast(v as interval)) from VALUES (null) t(v);
+
+-- empty set
+select avg(cast(v as interval)) from VALUES ('1 seconds'), ('2 seconds'), (null) t(v) where 1=0;
+
+-- basic interval avg
+select avg(cast(v as interval)) from VALUES ('1 seconds'), ('2 seconds'), (null) t(v);
+select avg(cast(v as interval)) from VALUES ('-1 seconds'), ('2 seconds'), (null) t(v);
+select avg(cast(v as interval)) from VALUES ('-1 seconds'), ('-2 seconds'), (null) t(v);
+select avg(cast(v as interval)) from VALUES ('-1 weeks'), ('2 seconds'), (null) t(v);
+
+-- group by
+select
+    i,
+    avg(cast(v as interval))
+from VALUES (1, '-1 weeks'), (2, '2 seconds'), (3, null), (1, '5 days') t(i, v)
+group by i;
+
+-- having
+select
+    avg(cast(v as interval)) as sv
+from VALUES (1, '-1 weeks'), (2, '2 seconds'), (3, null), (1, '5 days') t(i, v)
+having sv is not null;
+
+-- window
 SELECT
-  cast(null as interval), -- null
-  interval '0 day', -- 0
-  interval '1 year', -- year only
-  interval '1 month', -- month only
-  interval '1 year 2 month', -- year month only
-  interval '-1 day -1 hours',
-  interval '-1 day 1 hours',
-  interval '1 day -1 hours',
-  interval '-1 days +1 hours',
-  interval '1 years 2 months -3 days 4 hours 5 minutes 6.789 seconds',
-  - interval '1 years 2 months -3 days 4 hours 5 minutes 6.789 seconds';
+    i,
+    avg(cast(v as interval)) OVER (ORDER BY i ROWS BETWEEN CURRENT ROW AND UNBOUNDED FOLLOWING)
+FROM VALUES (1,'1 seconds'), (1,'2 seconds'), (2,NULL), (2,NULL) t(i,v);
+
+-- Interval year-month arithmetic
+
+create temporary view interval_arithmetic as
+  select CAST(dateval AS date), CAST(tsval AS timestamp) from values
+    ('2012-01-01', '2012-01-01')
+    as interval_arithmetic(dateval, tsval);
+
+select
+  dateval,
+  dateval - interval '2-2' year to month,
+  dateval - interval '-2-2' year to month,
+  dateval + interval '2-2' year to month,
+  dateval + interval '-2-2' year to month,
+  - interval '2-2' year to month + dateval,
+  interval '2-2' year to month + dateval
+from interval_arithmetic;
+
+select
+  tsval,
+  tsval - interval '2-2' year to month,
+  tsval - interval '-2-2' year to month,
+  tsval + interval '2-2' year to month,
+  tsval + interval '-2-2' year to month,
+  - interval '2-2' year to month + tsval,
+  interval '2-2' year to month + tsval
+from interval_arithmetic;
+
+select
+  interval '2-2' year to month + interval '3-3' year to month,
+  interval '2-2' year to month - interval '3-3' year to month
+from interval_arithmetic;
+
+-- Interval day-time arithmetic
+
+select
+  dateval,
+  dateval - interval '99 11:22:33.123456789' day to second,
+  dateval - interval '-99 11:22:33.123456789' day to second,
+  dateval + interval '99 11:22:33.123456789' day to second,
+  dateval + interval '-99 11:22:33.123456789' day to second,
+  -interval '99 11:22:33.123456789' day to second + dateval,
+  interval '99 11:22:33.123456789' day to second + dateval
+from interval_arithmetic;
+
+select
+  tsval,
+  tsval - interval '99 11:22:33.123456789' day to second,
+  tsval - interval '-99 11:22:33.123456789' day to second,
+  tsval + interval '99 11:22:33.123456789' day to second,
+  tsval + interval '-99 11:22:33.123456789' day to second,
+  -interval '99 11:22:33.123456789' day to second + tsval,
+  interval '99 11:22:33.123456789' day to second + tsval
+from interval_arithmetic;
+
+select
+  interval '99 11:22:33.123456789' day to second + interval '10 9:8:7.123456789' day to second,
+  interval '99 11:22:33.123456789' day to second - interval '10 9:8:7.123456789' day to second
+from interval_arithmetic;