No longer wraps 'interval' with parentheses

[dongyoungy]

• Add extra unit tests to test tpc-h + scramble
• Also fix #305

[codecov-io]

Codecov Report

Merging #306 into master will increase coverage by 0.41%. The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #306      +/-   ##
==========================================
+ Coverage   70.15%   70.55%   +0.41%     
==========================================
  Files         164      164              
  Lines       10702    10702              
  Branches     1719     1719              
==========================================
+ Hits         7507     7550      +43     
+ Misses       2765     2733      -32     
+ Partials      430      419      -11

Impacted Files	Coverage Δ
...java/org/verdictdb/sqlwriter/SelectQueryToSql.java	80.98% <ø> (+3.27%)	:arrow_up:
...va/org/verdictdb/connection/InMemoryAggregate.java	92.73% <100%> (ø)	:arrow_up:
.../core/scrambling/FastConvergeScramblingMethod.java	93.46% <0%> (ø)	:arrow_up:
...ictdb/core/querying/ola/AsyncAggExecutionNode.java	68.16% <0%> (+0.6%)	:arrow_up:
...main/java/org/verdictdb/sqlsyntax/MysqlSyntax.java	85.08% <0%> (+1.5%)	:arrow_up:
...querying/simplifier/ConsolidatedExecutionNode.java	83.34% <0%> (+1.76%)	:arrow_up:
...n/java/org/verdictdb/sqlsyntax/RedshiftSyntax.java	74.08% <0%> (+1.86%)	:arrow_up:
...core/querying/ola/SelectAsyncAggExecutionNode.java	89.29% <0%> (+2.39%)	:arrow_up:
...ain/java/org/verdictdb/sqlsyntax/ImpalaSyntax.java	67.57% <0%> (+2.71%)	:arrow_up:
.../verdictdb/coordinator/SelectQueryCoordinator.java	85.82% <0%> (+2.84%)	:arrow_up:
... and 6 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update d99b998...38b22d8. Read the comment docs.

[pyongjoo]

Can you check the build failure in the master? https://circleci.com/gh/mozafari/verdictdb/2700

[dongyoungy]

It has been caused by redshift unit tests, which happen to share the same schema for ScrambleMeta:

com.fasterxml.jackson.databind.exc.UnrecognizedPropertyException: Unrecognized field "hashColumn" (class org.verdictdb.core.scrambling.ScrambleMeta), not marked as ignorable (10 known properties: "aggregationBlockColumn", "aggregationBlockCount", "originalTableName", "tableName", "schemaName", "originalSchemaName", "numberOfTiers", "tierColumn", "cumulativeDistributions", "method"])
 at [Source: {"schemaName":"verdictdb_tpch_query_test_g0kijsbq","tableName":"orders_scramble","originalSchemaName":"verdictdb_tpch_query_test_g0kijsbq","originalTableName":"orders","aggregationBlockColumn":"verdictdbblock","aggregationBlockCount":1,"tierColumn":"verdictdbtier","numberOfTiers":1,"method":"uniform","hashColumn":null,"cumulativeDistributions":{"0":[1.0]}}; line: 1, column: 320] (through reference chain: org.verdictdb.core.scrambling.ScrambleMeta["hashColumn"])

I guess added hashColumn field in ScrambleMeta in the ongoing work on supporting count distinct affected the test in master branch.

I think It should get cleared once you push your latest work into master, yet I will try to have those failed tests to use different randomized schema for meta.

[pyongjoo]

I see. Then, I guess there’s no need to temporarily fix it. I’ll just merge mine once I fix my errors.

[dongyoungy]

It looks like some Presto unit tests in pyverdict job also have same problem: https://circleci.com/gh/mozafari/verdictdb/2741

[pyongjoo]

I feel like there was a temporary network error in CircleCI or so. Let me double-check.

[pyongjoo]

Nope, it seems to be a regular bug. I'll post here again after fixing it.

[pyongjoo]

All tests are passing now for the code in the yongjoo-count-distinct branch.