Amos Bird has posted comments on this change.

Change subject: IMPALA-1654: Support general predicates in most partition DDL 
operations.
......................................................................


Patch Set 3:

(14 comments)

http://gerrit.cloudera.org:8080/#/c/1563/3/be/src/service/query-exec-state.cc
File be/src/service/query-exec-state.cc:

PS3, Line 533: request_result_set_.reset(new vector<TResultRow>);
             :     request_result_set_->assign(
             :         ddl_resp->result_set.rows.begin(), 
ddl_resp->result_set.rows.end());
> These two statement can be combined into one using the constructor of std::
Done


http://gerrit.cloudera.org:8080/#/c/1563/3/common/thrift/CatalogService.thrift
File common/thrift/CatalogService.thrift:

Line 145:   // by COMPUTE STATS and ALTER TABLE.
> All ALTER TABLE, or just some types of alterations?
for all alter table stmts.


http://gerrit.cloudera.org:8080/#/c/1563/3/fe/src/main/java/com/cloudera/impala/analysis/AlterTableSetCachedStmt.java
File fe/src/main/java/com/cloudera/impala/analysis/AlterTableSetCachedStmt.java:

Line 74:         if (parts == null) {
> or parts is not null but is an empty list
Done


Line 75:           throw new AnalysisException("No partition matches the 
partition expr");
> Should we allow this to not be an error, now?
Done


http://gerrit.cloudera.org:8080/#/c/1563/3/fe/src/main/java/com/cloudera/impala/analysis/AnalysisContext.java
File fe/src/main/java/com/cloudera/impala/analysis/AnalysisContext.java:

Line 440:             analysisResult_.isComputeStatsStmt() ||
> Why did these lines need to be added here? Were these statements not being 
After partitionspec slotrefs being registered, it is needed.


http://gerrit.cloudera.org:8080/#/c/1563/3/fe/src/main/java/com/cloudera/impala/analysis/PartitionSpec.java
File fe/src/main/java/com/cloudera/impala/analysis/PartitionSpec.java:

Line 45:   private List<Long> partitionIds_;
> Why do you need partitionIds_? Can you get that from calling getId() each H
Done


Line 55:   private Boolean partitionShouldExist_;
> I know this wasn't commented on before, but can you adda comment here expla
see its getter and setter method. Should I move those comments?


Line 58:   private boolean isKV_ = false;
> ", because only they require partition specs in a particular KV form. See P
your comment got truncated.


Line 68:     this.partitionKeyValues_ = Lists.newArrayList();
> Can you initialize these member variables inline at their declaration?
Done


Line 115:     // Only HDFS tables are partitioned.
> Do we really not allow S3 tables to be partitioned?
Well I don't know. It came from the trunk.


Line 232:             "Partition expr is not applicable. It should be bounded 
by " +
> I find this error message opaque. What does "bounded" mean in this sense?
Done
"bounded" means fully bounded by a slotref, that is, a conjunct contains 
exactly one. I've changed the errmsg.


Line 237:     List<HdfsPartition> matchingPartitions;
> Can we just use partitions_ in place of matchingPartitions?
Done


http://gerrit.cloudera.org:8080/#/c/1563/3/fe/src/main/java/com/cloudera/impala/planner/HdfsPartitionPruner.java
File fe/src/main/java/com/cloudera/impala/planner/HdfsPartitionPruner.java:

Line 93:       Analyzer analyzer, List<Expr> conjuncts, boolean alterPartition)
> Please add a comment explaining what this parameter is used for.
Done


http://gerrit.cloudera.org:8080/#/c/1563/3/fe/src/test/java/com/cloudera/impala/analysis/ParserTest.java
File fe/src/test/java/com/cloudera/impala/analysis/ParserTest.java:

> This also deserves some end-to-end tests.
Yeah. Where should I put the tests? Is there a wiki that I ca n refer to?


-- 
To view, visit http://gerrit.cloudera.org:8080/1563
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I2c9162fcf9d227b8daf4c2e761d57bab4e26408f
Gerrit-PatchSet: 3
Gerrit-Project: Impala
Gerrit-Branch: cdh5-trunk
Gerrit-Owner: Amos Bird <[email protected]>
Gerrit-Reviewer: Alex Behm <[email protected]>
Gerrit-Reviewer: Amos Bird <[email protected]>
Gerrit-Reviewer: Dimitris Tsirogiannis <[email protected]>
Gerrit-Reviewer: Jim Apple <[email protected]>
Gerrit-HasComments: Yes

Reply via email to