Adding column name rewrite for the identifiers in the format of [table].[column] #5734
Conversation
pinot-broker/src/main/java/org/apache/pinot/broker/requesthandler/BaseBrokerRequestHandler.java
Outdated
Show resolved
Hide resolved
pinot-broker/src/main/java/org/apache/pinot/broker/requesthandler/BaseBrokerRequestHandler.java
Outdated
Show resolved
Hide resolved
pinot-broker/src/main/java/org/apache/pinot/broker/requesthandler/BaseBrokerRequestHandler.java
Outdated
Show resolved
Hide resolved
pinot-broker/src/main/java/org/apache/pinot/broker/requesthandler/BaseBrokerRequestHandler.java
Outdated
Show resolved
Hide resolved
...on-tests/src/test/java/org/apache/pinot/integration/tests/OfflineClusterIntegrationTest.java
Outdated
Show resolved
Hide resolved
...on-tests/src/test/java/org/apache/pinot/integration/tests/OfflineClusterIntegrationTest.java
Outdated
Show resolved
Hide resolved
xiangfu0
force-pushed
the
enable_table_cache_and_check_on_alias_table_column_name
branch
from
2c1102e to
7310a1c
Compare
…e_name].[column_name]
xiangfu0
force-pushed
the
enable_table_cache_and_check_on_alias_table_column_name
branch
from
7310a1c to
1acea4b
Compare
| if (tableNameSplits.length != 2) { | ||
| return; | ||
| } | ||
| String[] tableNameSplits = StringUtils.split(tableName, ".", 2); |
There was a problem hiding this comment.
Does the calcite parser not support getting columns names from [table].[column] format? If so, we should just use that, instead of post processing here? We have been adding small string manipulations per query incrementally, and I fear it will add up to have performance significance soon.
There was a problem hiding this comment.
Calcite doesn't support parse it as [table].[column], we will get the whole as identifier.
Then we need to handle that.
There was a problem hiding this comment.
If Calcite doesn't support, then should we support it? The logic seems brittle anyways.
There was a problem hiding this comment.
We should. The usage is for other system integrations like tableau
|
@mayankshriv Any more comments here? The extra overhead for existing queries is the string split on dot for table and column names. |
xiangfu0
deleted the
enable_table_cache_and_check_on_alias_table_column_name
branch
3 participants
Description
Adding column name rewrite for the format of [table].[column].
This is required as many external integrations like Tableau uses [table].[column] as identifier, and send them through JDBC client.
E.g.
SELECT myTable.colA FROM myTable LIMIT 10will be rewrite toSELECT colA FROM myTable LIMIT 10.Release Notes
Adding column name rewrite for the format of [table].[column]