[
https://issues.apache.org/jira/browse/SPARK-58419?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ganesha S updated SPARK-58419:
------------------------------
Description:
h3. What
Add support for the ANSI SQL {{UNNEST}} collection derived table in the FROM
clause:
{code:sql}
UNNEST ( expression [ , ... ] ) [ WITH ORDINALITY ] [ table_alias ]
{code}
{{UNNEST}} expands one or more arrays into a relation, producing one row per
element. When several arrays are supplied they are expanded in parallel: the
number of output rows equals the length of the longest array, and shorter
arrays are padded with NULLs. A NULL array is treated as an empty array and
contributes no rows. {{WITH ORDINALITY}} appends a trailing 1-based BIGINT
position column. Correlated arrays (e.g. {{{}FROM t, LATERAL UNNEST(t.arr){}}})
are supported through the existing LATERAL machinery, and \{{LEFT JOIN LATERAL
UNNEST(...) ON true}} preserves outer rows whose array is empty or NULL.
h3. Why
{{UNNEST}} is part of the SQL:2016 standard and is available in PostgreSQL,
Trino/Presto, and BigQuery (and, via FLATTEN, Snowflake). Spark currently has
no equivalent ANSI syntax: users must rewrite `{{{}UNNEST(...){}}} to
\{{LATERAL VIEW EXPLODE}}` or the DataFrame {{explode}} API. This is a
recurring pain point for the whole user base when porting queries from other
engines. This change closes
a genuine standards-compliance gap; array expansion is an everyday operation,
so the reach is broad.
was:
h3. What
Add support for the ANSI SQL {{UNNEST}} collection derived table in the FROM
clause:
{code:sql}
UNNEST ( expression [ , ... ] ) [ WITH ORDINALITY ] [ table_alias ]
{code}
{{UNNEST}} expands one or more arrays into a relation, producing one row per
element. When several arrays are supplied they are expanded in parallel: the
number of output rows equals the length of the longest array, and shorter
arrays are padded with NULLs. A NULL array is treated as an empty array and
contributes no rows. {{WITH ORDINALITY}} appends a trailing 1-based BIGINT
position column. Correlated arrays (e.g. {{{}FROM t, LATERAL UNNEST(t.arr){}}})
are
supported through the existing LATERAL machinery, and {{LEFT JOIN LATERAL
UNNEST(...) ON true}} preserves outer rows whose array is empty or NULL.
h3. Why
{{UNNEST}} is part of the SQL:2016 standard and is available in PostgreSQL,
Trino/Presto, and BigQuery (and, via FLATTEN, Snowflake). Spark currently has
no equivalent ANSI syntax: users must rewrite {{UNNEST(...)}} to {{LATERAL VIEW
EXPLODE}} or the DataFrame {{explode}} API. This is a recurring, whole-user-base
migration pain point when porting queries from other engines. This change closes
a genuine standards-compliance gap; array expansion is an everyday operation, so
the reach is broad.
> [SQL] Support ANSI SQL UNNEST in the FROM clause
> ------------------------------------------------
>
> Key: SPARK-58419
> URL: https://issues.apache.org/jira/browse/SPARK-58419
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 4.2.0
> Reporter: Ganesha S
> Priority: Major
>
> h3. What
> Add support for the ANSI SQL {{UNNEST}} collection derived table in the FROM
> clause:
> {code:sql}
> UNNEST ( expression [ , ... ] ) [ WITH ORDINALITY ] [ table_alias ]
> {code}
> {{UNNEST}} expands one or more arrays into a relation, producing one row per
> element. When several arrays are supplied they are expanded in parallel: the
> number of output rows equals the length of the longest array, and shorter
> arrays are padded with NULLs. A NULL array is treated as an empty array and
> contributes no rows. {{WITH ORDINALITY}} appends a trailing 1-based BIGINT
> position column. Correlated arrays (e.g. {{{}FROM t, LATERAL
> UNNEST(t.arr){}}}) are supported through the existing LATERAL machinery, and
> \{{LEFT JOIN LATERAL UNNEST(...) ON true}} preserves outer rows whose array
> is empty or NULL.
> h3. Why
> {{UNNEST}} is part of the SQL:2016 standard and is available in PostgreSQL,
> Trino/Presto, and BigQuery (and, via FLATTEN, Snowflake). Spark currently has
> no equivalent ANSI syntax: users must rewrite `{{{}UNNEST(...){}}} to
> \{{LATERAL VIEW EXPLODE}}` or the DataFrame {{explode}} API. This is a
> recurring pain point for the whole user base when porting queries from other
> engines. This change closes
> a genuine standards-compliance gap; array expansion is an everyday operation,
> so the reach is broad.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
