stanyao commented on code in PR #54972:
URL: https://github.com/apache/spark/pull/54972#discussion_r3148781723
##########
sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/PlanParserSuite.scala:
##########
@@ -885,6 +885,207 @@ class PlanParserSuite extends AnalysisTest {
stop = 65))
}
+ test("SPARK-55978: TABLESAMPLE SYSTEM and BERNOULLI - basic parsing") {
+ val sql = "select * from t"
+ // SYSTEM produces SampleMethod.System
+ assertEqual(
+ s"$sql tablesample system (43 percent) as x",
+ Sample(0, .43d, withReplacement = false, None,
+ table("t").as("x"), SampleMethod.System).select(star()))
+ // BERNOULLI produces SampleMethod.Bernoulli
+ assertEqual(
+ s"$sql tablesample bernoulli (43 percent) as x",
+ Sample(0, .43d, withReplacement = false, None,
+ table("t").as("x"), SampleMethod.Bernoulli).select(star()))
+ // No qualifier defaults to Bernoulli (backward compat)
+ assertEqual(
+ s"$sql tablesample(43 percent) as x",
+ Sample(0, .43d, withReplacement = false, None,
+ table("t").as("x")).select(star()))
+ }
+
+ test("SPARK-55978: TABLESAMPLE SYSTEM - case insensitivity") {
+ val sql = "select * from t"
+ // Keywords are case-insensitive
+ assertEqual(
+ s"$sql TABLESAMPLE SYSTEM (43 PERCENT) as x",
+ Sample(0, .43d, withReplacement = false, None,
+ table("t").as("x"), SampleMethod.System).select(star()))
+ assertEqual(
+ s"$sql TabLeSaMpLe SyStEm (43 PeRcEnT) as x",
+ Sample(0, .43d, withReplacement = false, None,
+ table("t").as("x"), SampleMethod.System).select(star()))
+ assertEqual(
+ s"$sql TABLESAMPLE BERNOULLI (43 PERCENT) as x",
+ Sample(0, .43d, withReplacement = false, None,
+ table("t").as("x"), SampleMethod.Bernoulli).select(star()))
+ }
+
+ test("SPARK-55978: TABLESAMPLE SYSTEM - boundary fractions") {
+ val sql = "select * from t"
+ // 0 PERCENT
+ assertEqual(
+ s"$sql tablesample system (0 percent) as x",
+ Sample(0, 0d, withReplacement = false, None,
+ table("t").as("x"), SampleMethod.System).select(star()))
+ // 100 PERCENT
+ assertEqual(
+ s"$sql tablesample system (100 percent) as x",
+ Sample(0, 1d, withReplacement = false, None,
+ table("t").as("x"), SampleMethod.System).select(star()))
+ // Fractional percent
+ assertEqual(
+ s"$sql tablesample system (0.1 percent) as x",
+ Sample(0, 0.001d, withReplacement = false, None,
+ table("t").as("x"), SampleMethod.System).select(star()))
+ }
+
+ test("SPARK-55978: TABLESAMPLE SYSTEM - unsupported sample methods") {
+ val sql = "select * from t"
+ // SYSTEM + ROWS -> error
+ checkError(
+ exception = parseException(s"$sql tablesample system (100 rows)"),
+ condition = "UNSUPPORTED_FEATURE.TABLESAMPLE_SYSTEM_SAMPLE_METHOD",
+ sqlState = "0A000",
+ parameters = Map("sampleMethod" -> "ROWS"),
+ context = ExpectedContext(
+ fragment = "tablesample system (100 rows)",
+ start = 16,
+ stop = 44))
+ // SYSTEM + BYTES -> error
+ checkError(
+ exception = parseException(s"$sql tablesample system (300M)"),
+ condition = "UNSUPPORTED_FEATURE.TABLESAMPLE_SYSTEM_SAMPLE_METHOD",
+ sqlState = "0A000",
+ parameters = Map("sampleMethod" -> "BYTES"),
+ context = ExpectedContext(
+ fragment = "tablesample system (300M)",
+ start = 16,
+ stop = 40))
+ // SYSTEM + BUCKET -> error
+ checkError(
+ exception = parseException(s"$sql tablesample system (bucket 4 out of
10)"),
+ condition = "UNSUPPORTED_FEATURE.TABLESAMPLE_SYSTEM_SAMPLE_METHOD",
+ sqlState = "0A000",
+ parameters = Map("sampleMethod" -> "BUCKET"),
+ context = ExpectedContext(
+ fragment = "tablesample system (bucket 4 out of 10)",
+ start = 16,
+ stop = 54))
+ // SYSTEM + BUCKET ON colname -> error
+ checkError(
+ exception = parseException(s"$sql tablesample system (bucket 4 out of 10
on x)"),
+ condition = "UNSUPPORTED_FEATURE.TABLESAMPLE_SYSTEM_SAMPLE_METHOD",
+ sqlState = "0A000",
+ parameters = Map("sampleMethod" -> "BUCKET"),
+ context = ExpectedContext(
+ fragment = "tablesample system (bucket 4 out of 10 on x)",
+ start = 16,
+ stop = 59))
+ // SYSTEM + BUCKET ON function -> error
+ checkError(
+ exception = parseException(s"$sql tablesample system (bucket 3 out of 32
on rand())"),
+ condition = "UNSUPPORTED_FEATURE.TABLESAMPLE_SYSTEM_SAMPLE_METHOD",
+ sqlState = "0A000",
+ parameters = Map("sampleMethod" -> "BUCKET"),
+ context = ExpectedContext(
+ fragment = "tablesample system (bucket 3 out of 32 on rand())",
+ start = 16,
+ stop = 64))
+ }
+
+ test("SPARK-55978: TABLESAMPLE BERNOULLI - REPEATABLE is supported") {
+ assertEqual(
+ "select * from t tablesample bernoulli (43 percent) repeatable (123) as
x",
+ Sample(0, .43d, withReplacement = false, 123L,
+ table("t").as("x"), SampleMethod.Bernoulli).select(star()))
+ }
+
+ test("SPARK-55978: TABLESAMPLE SYSTEM - REPEATABLE not supported") {
+ val sql = "select * from t"
+ checkError(
+ exception = parseException(s"$sql tablesample system (43 percent)
repeatable (123)"),
+ condition = "UNSUPPORTED_FEATURE.TABLESAMPLE_SYSTEM_REPEATABLE",
+ sqlState = "0A000",
+ context = ExpectedContext(
+ fragment = "tablesample system (43 percent) repeatable (123)",
+ start = 16,
+ stop = 63))
+ }
+
+ test("SPARK-55978: TABLESAMPLE SYSTEM - fraction out of range") {
+ val sql = "select * from t"
+ // > 100 PERCENT
+ checkError(
+ exception = parseException(s"$sql tablesample system (150 percent) as
x"),
+ condition = "_LEGACY_ERROR_TEMP_0064",
Review Comment:
This is due to the call chain here -> AnalysisTest.scala ->
AbstractSqlParser.scala -> AstBuilder.scala -> validate() function in
ParserUtils.scala that is shared by both the existing Bernoulli sampling and
the new System sampling. Fixing this will need code changes to 10 call sites
and 20 test cases that are not related to this PR. It's best to be a dedicated
follow up PR to keep this PR focused. I logged a JIRA item to track
[SPARK-56638](https://issues.apache.org/jira/browse/SPARK-56638).
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
