The data-reload pipeline exposes intermediate tables via Hive, which we want to use for analytics purposes
and data quality checks.
DDL should be kept under version control at https://gitlab.wikimedia.org/repos/wikidata-platform/analytics
CREATE TABLE IF NOT EXISTS {props.rdf_data_table} (
`context` string,
`subject` string,
`predicate` string,
`object` string
)
PARTITIONED BY (
`date` string,
`wiki` string
)
STORED AS PARQUET
LOCATION '{data_path}/{props.rdf_data_location}';
CREATE TABLE IF NOT EXISTS {props.rdf_data_subgraphs_table} (
`subject` string,
`predicate` string,
`object` string,
`context` string
)
PARTITIONED BY (
`snapshot` string,
`wiki` string,
`scope` string
)
STORED AS PARQUET
LOCATION '{data_path}/{props.rdf_data_subgraphs_location}'
;