Snowflake: TIMESTAMP precision regression

For Snowflake time formats support precision parameter ([from Snowflake documentation](https://docs.snowflake.com/en/sql-reference/data-types-datetime#timestamp)):
> All timestamp variations, as well as the TIMESTAMP alias, support an optional precision parameter for fractional seconds (for example, TIMESTAMP(3)). Timestamp precision can range from 0 (seconds) to 9 (nanoseconds). The default precision is 9.

Note: this comes from using Python wrapper sqloxide

Query that fails:
```sql
create table hello(test timestamp_ntz(6))
```

Error:
```
sql parser error: Expected: ',' or ')' after column definition, found: ( at Line: 1, Column: 38
```

removing precision from `timestamp_ntz` passes through

Unfortunately I cannot directly link which version (not visible throughintroduced this regression, but it seems to be recent

Start by reproducing the failing `create table hello(test timestamp_ntz(6))` query through the Python sqloxide wrapper and trace how Snowflake timestamp types are parsed. Done means timestamp variations with precision values from 0 to 9 parse successfully without regressing unparameterized timestamps.

Written by the indexing model from the issue text.