jnturton commented on code in PR #2689: URL: https://github.com/apache/drill/pull/2689#discussion_r1003065011
########## contrib/udfs/src/test/java/org/apache/drill/exec/udfs/TestDateUtils.java: ########## @@ -0,0 +1,77 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you under the Apache License, Version 2.0 (the + * "License"); you may not use this file except in compliance + * with the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package org.apache.drill.exec.udfs; + +import org.junit.Test; + +import java.time.LocalDate; +import java.time.LocalDateTime; + +import static org.junit.Assert.assertEquals; + +public class TestDateUtils { + + @Test + public void testDateFromString() { + LocalDate testDate = LocalDate.of(2022, 3,14); + LocalDate badDate = LocalDate.of(1970, 1, 1); + assertEquals(testDate, DateUtilFunctions.getDateFromString("2022-03-14")); + assertEquals(testDate, DateUtilFunctions.getDateFromString("3/14/2022")); + assertEquals(testDate, DateUtilFunctions.getDateFromString("14/03/2022", true)); + assertEquals(testDate, DateUtilFunctions.getDateFromString("2022/3/14")); + + // Test bad dates + assertEquals(badDate, DateUtilFunctions.getDateFromString(null)); + assertEquals(badDate, DateUtilFunctions.getDateFromString("1975-13-56")); + assertEquals(badDate, DateUtilFunctions.getDateFromString("1975-1s")); Review Comment: Footnote, since this case comes up quite often. Those of us who've done data analytics with Pandas _did_ get used to floating point NaN being used as a "sentinel" for missing or invalid data but we should recognise this for what it was: a performance hack that entered Pandas from its Numpy foundation. It resulted in pain: automatic casting of ints to floats even though precision loss could happen because hardware and C integer types have no NaN value, the subsequent proliferation of different flavours of null - np.nan, Python None, and pd.NaT, special code to implement `skipna` for making NaNs behave like nulls rather than according to the IEEE float rules, etc. A one-line Drill query reveals the square peg and round hole relationship between IEEE 754 NaN and ANSI SQL null. ``` apache drill> select cast('NaN' as float) = cast('NaN' as float), null = null; EXPR$0 true EXPR$1 null ``` And indeed, [here is Wes McKinney talking about moving away from that approach](https://wesmckinney.com/blog/bitmaps-vs-sentinel-values/)[1]. Here's a relevant excerpt for us, one hop over in the SQL world where null is a standardised first class citizen. > From the perspective of databases and data warehousing, reserving certain values to mark a null (or NA) is widely considered unacceptable. NaN is valid data, as is INT32_MIN and other common values used as sentinels. [1] Also see https://pandas.pydata.org/pandas-docs/stable/user_guide/missing_data.html#missing-data-na -- 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: dev-unsubscr...@drill.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
