-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/30697/#review77690
-----------------------------------------------------------



exec/java-exec/src/main/java/org/apache/drill/exec/planner/logical/DrillReduceExpressionsRule.java
<https://reviews.apache.org/r/30697/#comment125873>

    You could have a single function createEmptyRelOrEquivalent(cluster, 
traitset, child...) that could be invoked for both FilterRel and CalcRel.



exec/java-exec/src/main/java/org/apache/drill/exec/store/dfs/easy/EasyGroupScan.java
<https://reviews.apache.org/r/30697/#comment125884>

    I feel that the estRowCount should use a formula that is consistent - in 
this case the formula is data/<row size>.  This will reduce chances of 
introducing unintended side effects.  So, instead of using the number of 
chunks, could we scale down the denominator until we get a non-zero estRowCount 
? 
    
    Also, I would be in favor of treating this as a separate jira since it was 
exposed by partition pruning.


- Aman Sinha


On March 24, 2015, 4:53 p.m., Jason Altekruse wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/30697/
> -----------------------------------------------------------
> 
> (Updated March 24, 2015, 4:53 p.m.)
> 
> 
> Review request for drill, Aditya Kishore, Aman Sinha, Jacques Nadeau, Jinfeng 
> Ni, Mehant Baid, and Parth Chandra.
> 
> 
> Bugs: DRILL-2060
>     https://issues.apache.org/jira/browse/DRILL-2060
> 
> 
> Repository: drill-git
> 
> 
> Description
> -------
> 
> THIS REQUEST HAS BEEN BROKEN INTO TWO REVIEWS:
> - I had to make a fairly substandial change to the interpreteed expression 
> evaluation to make the tests work.
> 
> Use a small modification of a rule in optiq and hook up the interpreted 
> expression evaluator to fold constant expressions (including those with Drill 
> UDFs) into literals. Fragment memory limits have been disabled, a more 
> complete refactoring of memory management is planned, the changes made here 
> were to allow the creation of a childAllocator without access to a fragment 
> context. I have added two comments next to the modifiations I made to the 
> optiq rule for our use case.
> 
> 
> Diffs
> -----
> 
>   
> contrib/storage-hive/core/src/main/java/org/apache/drill/exec/planner/sql/HiveUDFOperator.java
>  7524690 
>   exec/java-exec/src/main/java/io/netty/buffer/DrillBuf.java d87fb76 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/expr/fn/DrillFunctionRegistry.java
>  32cf362 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/expr/fn/impl/StringFunctionHelpers.java
>  3b1d7ef 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/expr/fn/interpreter/InterpreterEvaluator.java
>  35c35ec 
>   exec/java-exec/src/main/java/org/apache/drill/exec/ops/BufferManager.java 
> 536f6fd 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/planner/logical/DrillConstExecutor.java
>  PRE-CREATION 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/planner/logical/DrillReduceExpressionsRule.java
>  PRE-CREATION 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/planner/logical/DrillRuleSets.java
>  b1a7189 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/planner/physical/PlannerSettings.java
>  f320157 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/planner/sql/DrillSqlWorker.java
>  710418b 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/server/options/SystemOptionManager.java
>  608fac7 
>   
> exec/java-exec/src/main/java/org/apache/drill/exec/store/dfs/easy/EasyGroupScan.java
>  54cad56 
>   exec/java-exec/src/test/java/org/apache/drill/PlanningBase.java e673230 
>   exec/java-exec/src/test/java/org/apache/drill/TestProjectPushDown.java 
> cbb5c6e 
>   
> exec/java-exec/src/test/java/org/apache/drill/exec/fn/impl/TestAggregateFunctions.java
>  72ad31a 
>   
> exec/java-exec/src/test/java/org/apache/drill/exec/fn/interp/TestConstantFolding.java
>  PRE-CREATION 
>   exec/java-exec/src/test/resources/parquet/alltypes.json PRE-CREATION 
>   pom.xml 05573a1 
> 
> Diff: https://reviews.apache.org/r/30697/diff/
> 
> 
> Testing
> -------
> 
> This is a work in progress, some testing has been done, no full unit test run 
> yet
> 
> 
> Thanks,
> 
> Jason Altekruse
> 
>

Reply via email to