adriangb commented on code in PR #22633: URL: https://github.com/apache/datafusion/pull/22633#discussion_r3338134692
########## benchmarks/sql_benchmarks/clickbench/init/load-partitioned.sql: ########## @@ -0,0 +1,3 @@ +CREATE EXTERNAL TABLE hits_raw STORED AS PARQUET LOCATION 'data/hits_partitioned/'; Review Comment: I think https://github.com/apache/datafusion/pull/22660#discussion_r3338103032 also applies here ########## benchmarks/sql_benchmarks/clickbench/init/load-single.sql: ########## @@ -0,0 +1,3 @@ +CREATE EXTERNAL TABLE hits_raw STORED AS PARQUET LOCATION 'data/hits.parquet'; Review Comment: ```suggestion CREATE EXTERNAL TABLE hits_raw STORED AS PARQUET LOCATION '${DATA_DIR:-data}/hits.parquet'; ``` ########## benchmarks/sql_benchmarks/clickbench/init/load-partitioned.sql: ########## @@ -0,0 +1,3 @@ +CREATE EXTERNAL TABLE hits_raw STORED AS PARQUET LOCATION 'data/hits_partitioned/'; Review Comment: ```suggestion CREATE EXTERNAL TABLE hits_raw STORED AS PARQUET LOCATION '${DATA_DIR:-data}/hits_partitioned/'; ``` ########## benchmarks/sql_benchmarks/clickbench/init/set_config.sql: ########## @@ -0,0 +1,7 @@ +# ClickBench partitioned dataset was written by an ancient version of PyArrow that +# wrote strings with the wrong logical type. To read it correctly, we must +# automatically convert binary to string. + +SET datafusion.execution.parquet.binary_as_string = true; +SET datafusion.execution.parquet.pushdown_filters = ${PUSHDOWN_FILTERS:-false}; +SET datafusion.execution.parquet.reorder_filters = ${REORDER_FILTERS:-false}; Review Comment: Are these needed here? Is it just to match existing env vars supported by the binary? Because these can already be overridden via env vars. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
