fix(redshift): parser issue for TOP N DISTINCT

[fivetran-kwoodbeck]

Redshift supports SELECT TOP N DISTINCT ... where TOP N precedes DISTINCT. The base parser checks for DISTINCT immediately after SELECT before consuming TOP N

The fix adds two overrides to RedshiftParser. _parse_limit detects a DISTINCT token immediately following TOP N, consumes it, and temporarily stashes an exp.Distinct() on the exp.Limit node as a carrier. Redshift's _parse_select_query then promotes that value up to exp.Select.distinct and removes it from the limit once the exp.Select is fully built.

I don't see a more elegant way to handle it because distinct is resolved as a local variable in the base _parse_select_query before _parse_limit is ever called. There's no obvious way to reach back, so the limit node acts as a handoff point between the two.

[github-actions]

SQLGlot Integration Test Results

Comparing:

• this branch (sqlglot:fix/redshift-top_n_distinct, sqlglot version: fix/redshift-top_n_distinct)
• baseline (main, sqlglot version: 0.0.1.dev1)

By Dialect

dialect	main	sqlglot:fix/redshift-top_n_distinct	transitions	links
bigquery -> bigquery	24647/24652 passed (100.0%)	23497/23497 passed (100.0%)	No change	full result / delta
bigquery -> duckdb	867/1154 passed (75.1%)	0/0 passed (0.0%)	Results not found	full result / delta
duckdb -> duckdb	5823/5823 passed (100.0%)	0/0 passed (0.0%)	Results not found	full result / delta
snowflake -> duckdb	1129/1935 passed (58.3%)	0/0 passed (0.0%)	Results not found	full result / delta
snowflake -> snowflake	65133/65133 passed (100.0%)	63027/63027 passed (100.0%)	No change	full result / delta
databricks -> databricks	1370/1370 passed (100.0%)	1370/1370 passed (100.0%)	No change	full result / delta
postgres -> postgres	6042/6042 passed (100.0%)	6042/6042 passed (100.0%)	No change	full result / delta
redshift -> redshift	7101/7101 passed (100.0%)	7101/7101 passed (100.0%)	No change	full result / delta

Overall

main: 113210 total, 112112 passed (pass rate: 99.0%), sqlglot version: 0.0.1.dev1

sqlglot:fix/redshift-top_n_distinct: 101037 total, 101037 passed (pass rate: 100.0%), sqlglot version: fix/redshift-top_n_distinct

Transitions:
No change

Dialect pair changes: 0 previous results not found, 3 current results not found

✅ 42 test(s) passed

[geooo109]

@fivetran-kwoodbeck Yeah, PR description makes sense.

Is it possible to put the fix in the base parser for this ? and avoid doing this post change on the AST. When we parse the SELECT this should be fine.

I think TOP is used in redshift, t-sql, oracle, snowflake, and teradata. So, if we add the logic for this in the base parser and cover these dialects (avoid breaking anything) we can push it there.

[fivetran-kwoodbeck]

@geooo109 I didn't realize other dialects supported it, then yes, it makes sense to put it into the base parser. That will be cleaner overall, will update.