Re: Review Request 17302: HIVE-6012: restore backward compatibility of arithmetic operations
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/#review36295 --- ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFBaseNumeric.java https://reviews.apache.org/r/17302/#comment67251 what would this be for two decimals? - Sergey Shelukhin On Feb. 7, 2014, 12:13 a.m., Jason Dere wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/ --- (Updated Feb. 7, 2014, 12:13 a.m.) Review request for hive. Bugs: HIVE-6012 https://issues.apache.org/jira/browse/HIVE-6012 Repository: hive-git Description --- Change return types back to Hive-0.12 behavior for the following math operators: int / int - decimal, used to be double in Hive-0.12 float + int - double, used to be float in Hive-0.12 float + float - double, used to be float in Hive-0.12 float * int - double, used to be float in Hive-0.12 float * float - double, used to be float in Hive-0.12 Diffs - common/src/java/org/apache/hadoop/hive/conf/HiveConf.java 89c9349 common/src/java/org/apache/hive/common/HiveCompat.java PRE-CREATION conf/hive-default.xml.template 420d959 ql/src/java/org/apache/hadoop/hive/ql/exec/FunctionRegistry.java 48b7ee1 ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDF.java 1813b64 ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFBaseNumeric.java 28b5ffd ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFOPDivide.java 9a04e81 ql/src/test/org/apache/hadoop/hive/ql/exec/TestExpressionEvaluator.java 65181e4 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPDivide.java e894b54 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMinus.java d7285d9 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMod.java ef17eb5 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMultiply.java 56472f0 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPNumeric.java PRE-CREATION ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPPlus.java 6193257 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFPosMod.java 702e3e7 ql/src/test/queries/clientpositive/ansi_sql_arithmetic.q PRE-CREATION ql/src/test/results/clientpositive/ansi_sql_arithmetic.q.out PRE-CREATION ql/src/test/results/clientpositive/decimal_udf.q.out 74ae554 ql/src/test/results/clientpositive/ql_rewrite_gbtoidx.q.out 2f36c4e ql/src/test/results/clientpositive/udf_pmod.q.out 7766304 ql/src/test/results/clientpositive/vectorization_15.q.out 8c6a6d8 ql/src/test/results/clientpositive/vectorization_5.q.out abcafd2 ql/src/test/results/clientpositive/vectorization_short_regress.q.out 305d336 ql/src/test/results/clientpositive/windowing_expressions.q.out e08d0bc Diff: https://reviews.apache.org/r/17302/diff/ Testing --- Thanks, Jason Dere
Re: Review Request 17302: HIVE-6012: restore backward compatibility of arithmetic operations
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/#review36298 --- ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFBaseNumeric.java https://reviews.apache.org/r/17302/#comment67253 This method (deriveResultApproxTypeInfo) would not get called for 2 decimals - deriveResultExactTypeInfo() would be called. - Jason Dere On Feb. 7, 2014, 12:13 a.m., Jason Dere wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/ --- (Updated Feb. 7, 2014, 12:13 a.m.) Review request for hive. Bugs: HIVE-6012 https://issues.apache.org/jira/browse/HIVE-6012 Repository: hive-git Description --- Change return types back to Hive-0.12 behavior for the following math operators: int / int - decimal, used to be double in Hive-0.12 float + int - double, used to be float in Hive-0.12 float + float - double, used to be float in Hive-0.12 float * int - double, used to be float in Hive-0.12 float * float - double, used to be float in Hive-0.12 Diffs - common/src/java/org/apache/hadoop/hive/conf/HiveConf.java 89c9349 common/src/java/org/apache/hive/common/HiveCompat.java PRE-CREATION conf/hive-default.xml.template 420d959 ql/src/java/org/apache/hadoop/hive/ql/exec/FunctionRegistry.java 48b7ee1 ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDF.java 1813b64 ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFBaseNumeric.java 28b5ffd ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFOPDivide.java 9a04e81 ql/src/test/org/apache/hadoop/hive/ql/exec/TestExpressionEvaluator.java 65181e4 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPDivide.java e894b54 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMinus.java d7285d9 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMod.java ef17eb5 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMultiply.java 56472f0 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPNumeric.java PRE-CREATION ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPPlus.java 6193257 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFPosMod.java 702e3e7 ql/src/test/queries/clientpositive/ansi_sql_arithmetic.q PRE-CREATION ql/src/test/results/clientpositive/ansi_sql_arithmetic.q.out PRE-CREATION ql/src/test/results/clientpositive/decimal_udf.q.out 74ae554 ql/src/test/results/clientpositive/ql_rewrite_gbtoidx.q.out 2f36c4e ql/src/test/results/clientpositive/udf_pmod.q.out 7766304 ql/src/test/results/clientpositive/vectorization_15.q.out 8c6a6d8 ql/src/test/results/clientpositive/vectorization_5.q.out abcafd2 ql/src/test/results/clientpositive/vectorization_short_regress.q.out 305d336 ql/src/test/results/clientpositive/windowing_expressions.q.out e08d0bc Diff: https://reviews.apache.org/r/17302/diff/ Testing --- Thanks, Jason Dere
Re: Review Request 17302: HIVE-6012: restore backward compatibility of arithmetic operations
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/ --- (Updated March 6, 2014, 1:32 a.m.) Review request for hive. Changes --- patch v5, rebased with trunk Bugs: HIVE-6012 https://issues.apache.org/jira/browse/HIVE-6012 Repository: hive-git Description --- Change return types back to Hive-0.12 behavior for the following math operators: int / int - decimal, used to be double in Hive-0.12 float + int - double, used to be float in Hive-0.12 float + float - double, used to be float in Hive-0.12 float * int - double, used to be float in Hive-0.12 float * float - double, used to be float in Hive-0.12 Diffs (updated) - common/src/java/org/apache/hadoop/hive/conf/HiveConf.java 5dbd1a5 common/src/java/org/apache/hive/common/HiveCompat.java PRE-CREATION conf/hive-default.xml.template a52be2b ql/src/java/org/apache/hadoop/hive/ql/exec/FunctionRegistry.java b9a5d38 ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDF.java 1813b64 ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFBaseNumeric.java 28b5ffd ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFOPDivide.java 96f970e ql/src/test/org/apache/hadoop/hive/ql/exec/TestExpressionEvaluator.java 65181e4 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPDivide.java e894b54 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMinus.java d7285d9 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMod.java ef17eb5 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMultiply.java 56472f0 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPNumeric.java PRE-CREATION ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPPlus.java 6193257 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFPosMod.java 702e3e7 ql/src/test/queries/clientpositive/ansi_sql_arithmetic.q PRE-CREATION ql/src/test/results/clientpositive/ansi_sql_arithmetic.q.out PRE-CREATION ql/src/test/results/clientpositive/decimal_udf.q.out 09c410a ql/src/test/results/clientpositive/ql_rewrite_gbtoidx.q.out 6f948ed ql/src/test/results/clientpositive/udf_pmod.q.out 7766304 ql/src/test/results/clientpositive/vectorization_15.q.out 8c6a6d8 ql/src/test/results/clientpositive/vectorization_5.q.out abcafd2 ql/src/test/results/clientpositive/vectorization_short_regress.q.out 8c22391 ql/src/test/results/clientpositive/windowing_expressions.q.out 604c56c Diff: https://reviews.apache.org/r/17302/diff/ Testing --- Thanks, Jason Dere
Re: Review Request 17302: HIVE-6012: restore backward compatibility of arithmetic operations
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/ --- (Updated Feb. 7, 2014, 12:13 a.m.) Review request for hive. Changes --- - Change config setting (hive.compat) used to revert/enable SQL compliant int/int. - Fix unit test failures. Bugs: HIVE-6012 https://issues.apache.org/jira/browse/HIVE-6012 Repository: hive-git Description --- Change return types back to Hive-0.12 behavior for the following math operators: int / int - decimal, used to be double in Hive-0.12 float + int - double, used to be float in Hive-0.12 float + float - double, used to be float in Hive-0.12 float * int - double, used to be float in Hive-0.12 float * float - double, used to be float in Hive-0.12 Diffs (updated) - common/src/java/org/apache/hadoop/hive/conf/HiveConf.java 89c9349 common/src/java/org/apache/hive/common/HiveCompat.java PRE-CREATION conf/hive-default.xml.template 420d959 ql/src/java/org/apache/hadoop/hive/ql/exec/FunctionRegistry.java 48b7ee1 ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDF.java 1813b64 ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFBaseNumeric.java 28b5ffd ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFOPDivide.java 9a04e81 ql/src/test/org/apache/hadoop/hive/ql/exec/TestExpressionEvaluator.java 65181e4 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPDivide.java e894b54 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMinus.java d7285d9 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMod.java ef17eb5 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMultiply.java 56472f0 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPNumeric.java PRE-CREATION ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPPlus.java 6193257 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFPosMod.java 702e3e7 ql/src/test/queries/clientpositive/ansi_sql_arithmetic.q PRE-CREATION ql/src/test/results/clientpositive/ansi_sql_arithmetic.q.out PRE-CREATION ql/src/test/results/clientpositive/decimal_udf.q.out 74ae554 ql/src/test/results/clientpositive/ql_rewrite_gbtoidx.q.out 2f36c4e ql/src/test/results/clientpositive/udf_pmod.q.out 7766304 ql/src/test/results/clientpositive/vectorization_15.q.out 8c6a6d8 ql/src/test/results/clientpositive/vectorization_5.q.out abcafd2 ql/src/test/results/clientpositive/vectorization_short_regress.q.out 305d336 ql/src/test/results/clientpositive/windowing_expressions.q.out e08d0bc Diff: https://reviews.apache.org/r/17302/diff/ Testing --- Thanks, Jason Dere
Re: Review Request 17302: HIVE-6012: restore backward compatibility of arithmetic operations
On Jan. 24, 2014, 4:10 a.m., Mohammad Islam wrote: Very well unit-tested. Just to cover the full test: it will be much better if you can add a .q test something like decimal_udf.q? Basically the same q file after setting new config to backward compat mode. the result could be different. Good point, I will add a .q test to enable/disable the config setting with actual query results. - Jason --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/#review32697 --- On Jan. 24, 2014, 2:54 a.m., Jason Dere wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/ --- (Updated Jan. 24, 2014, 2:54 a.m.) Review request for hive. Bugs: HIVE-6012 https://issues.apache.org/jira/browse/HIVE-6012 Repository: hive-git Description --- Change return types back to Hive-0.12 behavior for the following math operators: int / int - decimal, used to be double in Hive-0.12 float + int - double, used to be float in Hive-0.12 float + float - double, used to be float in Hive-0.12 float * int - double, used to be float in Hive-0.12 float * float - double, used to be float in Hive-0.12 Diffs - common/src/java/org/apache/hadoop/hive/conf/HiveConf.java 371cb0f ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFBaseNumeric.java 28b5ffd ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFOPDivide.java 9a04e81 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPDivide.java e894b54 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMinus.java d7285d9 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMod.java ef17eb5 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMultiply.java 56472f0 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPNumeric.java PRE-CREATION ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPPlus.java 6193257 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFPosMod.java 702e3e7 Diff: https://reviews.apache.org/r/17302/diff/ Testing --- Thanks, Jason Dere
Re: Review Request 17302: HIVE-6012: restore backward compatibility of arithmetic operations
--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/#review32697 --- Very well unit-tested. Just to cover the full test: it will be much better if you can add a .q test something like decimal_udf.q? Basically the same q file after setting new config to backward compat mode. the result could be different. - Mohammad Islam On Jan. 24, 2014, 2:54 a.m., Jason Dere wrote: --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/17302/ --- (Updated Jan. 24, 2014, 2:54 a.m.) Review request for hive. Bugs: HIVE-6012 https://issues.apache.org/jira/browse/HIVE-6012 Repository: hive-git Description --- Change return types back to Hive-0.12 behavior for the following math operators: int / int - decimal, used to be double in Hive-0.12 float + int - double, used to be float in Hive-0.12 float + float - double, used to be float in Hive-0.12 float * int - double, used to be float in Hive-0.12 float * float - double, used to be float in Hive-0.12 Diffs - common/src/java/org/apache/hadoop/hive/conf/HiveConf.java 371cb0f ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFBaseNumeric.java 28b5ffd ql/src/java/org/apache/hadoop/hive/ql/udf/generic/GenericUDFOPDivide.java 9a04e81 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPDivide.java e894b54 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMinus.java d7285d9 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMod.java ef17eb5 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPMultiply.java 56472f0 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPNumeric.java PRE-CREATION ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFOPPlus.java 6193257 ql/src/test/org/apache/hadoop/hive/ql/udf/generic/TestGenericUDFPosMod.java 702e3e7 Diff: https://reviews.apache.org/r/17302/diff/ Testing --- Thanks, Jason Dere
