Add support for snowflake exclusive create table options

[balliegojr]

Add support for snowflake exclusive create table options

  [ CLUSTER BY ( <expr> [ , <expr> , ... ] ) ]
  [ ENABLE_SCHEMA_EVOLUTION = { TRUE | FALSE } ]
  [ DATA_RETENTION_TIME_IN_DAYS = <integer> ]
  [ MAX_DATA_EXTENSION_TIME_IN_DAYS = <integer> ]
  [ CHANGE_TRACKING = { TRUE | FALSE } ]
  [ DEFAULT_DDL_COLLATION = '<collation_specification>' ]
  [ COPY GRANTS ]
  [ COMMENT = '<string_literal>' ]
  [ [ WITH ] ROW ACCESS POLICY <policy_name> ON ( <col_name> [ , <col_name> ... ] ) ]
  [ [ WITH ] AGGREGATION POLICY <policy_name> ]
  [ [ WITH ] TAG ( <tag_name> = '<tag_value>' [ , <tag_name> = '<tag_value>' , ... ] ) ]

Since snowflake does not require a strict order in the statement, the create table logic was moved to inside the snowflake dialect

[iffyio]

Suggested change

	.transient(transient)
	.hive_formats(Some(Default::default()));
	.transient(transient);

Is hive_formats needed in this case or it can be set to None?

[balliegojr]

The hive_formats is tested in a generic test against all dialects, since I don't know which dialect uses the hive_format I added the default value in the return struct to keep that test happy.

[iffyio]

general comment: can we include some negative tests as well? examples like incompatible keywords being used, incomplete keyword sequence etc

[coveralls]

Pull Request Test Coverage Report for Build 9423048544

Details

• 410 of 452 (90.71%) changed or added relevant lines in 9 files are covered.
• 1 unchanged line in 1 file lost coverage.
• Overall coverage increased (+0.04%) to 88.878%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
src/ast/mod.rs	16	18	88.89%
tests/sqlparser_postgres.rs	14	17	82.35%
tests/sqlparser_snowflake.rs	192	196	97.96%
src/ast/dml.rs	28	37	75.68%
src/ast/helpers/stmt_create_table.rs	52	62	83.87%
src/dialect/snowflake.rs	92	106	86.79%

Files with Coverage Reduction	New Missed Lines	%
src/ast/dml.rs	1	70.59%

Totals
Change from base Build 9416866271:	0.04%
Covered Lines:	26258
Relevant Lines:	29544

---

💛 - Coveralls

[alamb]

Thanks @balliegojr, I think this is looking really close to me.

A few more comments and I think it will be ready to go

[alamb]

Some examples would be nice here too.

Also perhaps we should implement Display for this enum which would simplify the callsites

[balliegojr]

I have implemented Display for Vec<T> only, I believe the HashMap may have different representation depending on the dialect/use case, but let me know if I should implement Display for the HashMap as well

[alamb]

Thank you @iffyio for the review

[iffyio]

@balliegojr sorry for the delay getting back to this PR
The changes LGTM! only the minor comment regarding the clickhouse_generic test and a conflicts to resolve from main

[lovasoa]

@alamb , I think we can merge this

[alamb]

Epic work @balliegojr -- thank you so much

Thank you for the review @lovasoa

[alamb]

🤔 looks like there are some CI failures,.

[balliegojr]

@alamb I believe I solved the CI issues

[alamb]

Thanks again @iffyio @balliegojr and @lovasoa