> On June 11, 2015, 11:20 p.m., Aman Sinha wrote:
> > exec/java-exec/src/main/java/org/apache/drill/exec/planner/sql/parser/UnsupportedOperatorsVisitor.java,
> >  line 120
> > <https://reviews.apache.org/r/35365/diff/1/?file=982877#file982877line120>
> >
> >     Instead of 'disabled' we should be consistent with the error messages 
> > and say 'The window function <name> is not currently supported <rest of the 
> > message>'

A new message:
"The window function " + functionName + " is not supported\n"


> On June 11, 2015, 11:20 p.m., Aman Sinha wrote:
> > exec/java-exec/src/main/java/org/apache/drill/exec/planner/sql/parser/UnsupportedOperatorsVisitor.java,
> >  line 134
> > <https://reviews.apache.org/r/35365/diff/1/?file=982877#file982877line134>
> >
> >     Change to 'DISTINCT for window aggregate functions is not currently 
> > supported...'  (since DISTINCT only applies to window aggregates not to 
> > rank, dense_rank etc.)

Update according to the comment


> On June 11, 2015, 11:20 p.m., Aman Sinha wrote:
> > exec/java-exec/src/main/java/org/apache/drill/exec/planner/sql/parser/UnsupportedOperatorsVisitor.java,
> >  line 158
> > <https://reviews.apache.org/r/35365/diff/1/?file=982877#file982877line158>
> >
> >     Instead of using string comparisons, we should use the methods 
> > SqlWindow.isCurrentRow(), SqlWindow.isUnboundedPreceding() etc.  or the 
> > enum Bound.

Update according to the comment


> On June 11, 2015, 11:20 p.m., Aman Sinha wrote:
> > exec/java-exec/src/test/java/org/apache/drill/exec/TestWindowFunctions.java,
> >  line 105
> > <https://reviews.apache.org/r/35365/diff/1/?file=982878#file982878line105>
> >
> >     On the latest master branch, the window.enable option is now set to 
> > true, so you can remove these from the unit tests.

Update according to the comment


> On June 11, 2015, 11:20 p.m., Aman Sinha wrote:
> > exec/java-exec/src/test/java/org/apache/drill/exec/TestWindowFunctions.java,
> >  line 148
> > <https://reviews.apache.org/r/35365/diff/1/?file=982878#file982878line148>
> >
> >     For conciseness, it would be better to have a single test which has a 
> > query string which takes various types of windows functions that are 
> > unsupported (lead, lag etc.) and just runs them one by one.  This will 
> > reduce the duplication. In any case, in a subsequent release we will be 
> > supporting these functions.

Sounds concise. But in these cases, the correct behavior is "seeing an expected 
exception", which junit would validate this.

For sure, we can use a while loop to wrap try-catch to take over junit's job. 
But would that presents some confusion ?


- Sean Hsuan-Yi


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/35365/#review87639
-----------------------------------------------------------


On June 11, 2015, 8:33 p.m., Sean Hsuan-Yi Chu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/35365/
> -----------------------------------------------------------
> 
> (Updated June 11, 2015, 8:33 p.m.)
> 
> 
> Review request for drill, Aman Sinha and Jinfeng Ni.
> 
> 
> Bugs: DRILL-3182, DRILL-3188 and DRILL-3195
>     https://issues.apache.org/jira/browse/DRILL-3182
>     https://issues.apache.org/jira/browse/DRILL-3188
>     https://issues.apache.org/jira/browse/DRILL-3195
> 
> 
> Repository: drill-git
> 
> 
> Description
> -------
> 
> Window function with DISTINCT qualifier is disabled; Disable unsupported 
> window frames; Window functions NTILE, LAG, LEAD, FIRST_VALUE, LAST_VALUE are 
> disabled
> 
> 
> Diffs
> -----
> 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/planner/sql/parser/UnsupportedOperatorsVisitor.java
>  b92de3b 
>   exec/java-exec/src/test/java/org/apache/drill/exec/TestWindowFunctions.java 
> fc75d73 
> 
> Diff: https://reviews.apache.org/r/35365/diff/
> 
> 
> Testing
> -------
> 
> All the required tests
> 
> 
> Thanks,
> 
> Sean Hsuan-Yi Chu
> 
>

Reply via email to