Uh oh!
There was an error while loading. Please reload this page.
[SPARK-5123][SQL] Expose only one version of the data type APIs (i.e. remove Java-specific APIs) - #3925
[SPARK-5123][SQL] Expose only one version of the data type APIs (i.e. remove Java-specific APIs)#3925rxin wants to merge 10 commits into
Conversation
…ve the Java-specific API).
SparkQA
commented
Jan 7, 2015
Test build #25141 has started for PR 3925 at commit
|
SparkQA
commented
Jan 7, 2015
Test build #25141 has finished for PR 3925 at commit
|
AmplabJenkins
commented
Jan 7, 2015
Test FAILed. |
pwendell
commented
Jan 7, 2015
Is this something where we need to make sure to include upgrade details in the release notes? |
rxin
commented
Jan 7, 2015
Yes - is there a way to label these things in JIRA? |
SparkQA
commented
Jan 7, 2015
Test build #25142 has started for PR 3925 at commit
|
rxin
commented
Jan 7, 2015
cc @marmbrus, @yhuai for SQL changes, and @mengxr, @jkbradley for MLlib changes ... |
SparkQA
commented
Jan 7, 2015
Test build #25142 has finished for PR 3925 at commit
|
AmplabJenkins
commented
Jan 7, 2015
Test FAILed. |
SparkQA
commented
Jan 7, 2015
Test build #25145 has started for PR 3925 at commit
|
SparkQA
commented
Jan 7, 2015
Test build #25145 has finished for PR 3925 at commit
|
AmplabJenkins
commented
Jan 7, 2015
Test FAILed. |
There was a problem hiding this comment.
Make ISO8601GMT this as thread local? or leave a TODO for future improvement.
There was a problem hiding this comment.
What do you mean? I don't think I changed this. Simply copied it from a file that was deleted.
chenghao-intel
commented
Jan 7, 2015
That's a very cool idea to make a unified |
There was a problem hiding this comment.
Are scala users expected to use DataTypes?
yhuai
commented
Jan 7, 2015
SQL changes look good to me. |
… into sql.types package.
SparkQA
commented
Jan 7, 2015
Test build #25169 has started for PR 3925 at commit
|
SparkQA
commented
Jan 7, 2015
Test build #25170 has started for PR 3925 at commit
|
SparkQA
commented
Jan 7, 2015
Test build #25169 has finished for PR 3925 at commit
|
AmplabJenkins
commented
Jan 7, 2015
Test FAILed. |
SparkQA
commented
Jan 7, 2015
Test build #25170 has finished for PR 3925 at commit
|
AmplabJenkins
commented
Jan 7, 2015
Test FAILed. |
SparkQA
commented
Jan 7, 2015
Test build #25173 has started for PR 3925 at commit
|
SparkQA
commented
Jan 7, 2015
Test build #25173 has finished for PR 3925 at commit
|
AmplabJenkins
commented
Jan 7, 2015
Test FAILed. |
There was a problem hiding this comment.
Should the "Preserve the original order of fields." comment in the apply() method be moved to the doc?
SparkQA
commented
Jan 8, 2015
Test build #25184 has started for PR 3925 at commit
|
jkbradley
commented
Jan 8, 2015
@rxin MLlib changes look fine to me, and FWIW the other parts did too. My only remaining comments are ordering imports, but I'll leave those out for now. LGTM |
SparkQA
commented
Jan 8, 2015
Test build #25184 has finished for PR 3925 at commit
|
AmplabJenkins
commented
Jan 8, 2015
Test PASSed. |
SparkQA
commented
Jan 8, 2015
Test build #25193 has started for PR 3925 at commit
|
SparkQA
commented
Jan 8, 2015
Test build #25193 has finished for PR 3925 at commit
|
AmplabJenkins
commented
Jan 8, 2015
Test PASSed. |
There was a problem hiding this comment.
Most of the code in Spark SQL uses JavaConversions... I don't think performance is an issue here or anything.
SparkQA
commented
Jan 8, 2015
Test build #25260 has started for PR 3925 at commit
|
SparkQA
commented
Jan 8, 2015
Test build #25260 has finished for PR 3925 at commit
|
AmplabJenkins
commented
Jan 8, 2015
Test PASSed. |
rxin
commented
Jan 8, 2015
See #3958 |
Having two versions of the data type APIs (one for Java, one for Scala) requires downstream libraries to also have two versions of the APIs if the library wants to support both Java and Scala. I took a look at the Scala version of the data type APIs - it can actually work out pretty well for Java out of the box. As part of the PR, I created a sql.types package and moved all type definitions there. I then removed the Java specific data type API along with a lot of the conversion code. This subsumes #3925 Author: Reynold Xin <rxin@databricks.com> Closes#3958 from rxin/SPARK-5123-datatype-2 and squashes the following commits: 66505cc [Reynold Xin] [SPARK-5123] Expose only one version of the data type APIs (i.e. remove the Java-specific API).
There was a problem hiding this comment.
Since this is a java source file, should we move it to sql/catalyst/src/main/java ?
Having two versions of the data type APIs (one for Java, one for Scala) requires downstream libraries to also have two versions of the APIs if the library wants to support both Java and Scala. I took a look at the Scala version of the data type APIs - it can actually work out pretty well for Java out of the box.
As part of the PR, I created a sql.types package and moved all type definitions there. I then removed the Java specific data type API along with a lot of the conversion code.