fix: unnest join constraint with alias parsing for BigQuery dialect

[Ziinc]

This PR fixes a parsing bug for the BigQuery Dialect.
The expected outcome is for the following statement to be parsed with join constraints:

select * from t1 join unnest(t1.col) as f1 on  c1 = 123

However, this statement would result in the following error:

ParserError("Expected end of statement, found: c1")

Closer inspection revealed that the following ast was produced for the join table factor

UNNEST { 
  alias: Some(TableAlias { name: Ident { value: "f", quote_style: None }, columns: [] }), 
  array_expr: CompoundIdentifier([Ident { value: "t1", quote_style: None }, Ident { value: "a", quote_style: None }]), 
  with_offset: false, 
  with_offset_alias: Some(Ident { value: "ON", quote_style: None }) 
}

As you can see, the with_offset_alias parses the ON keyword despite the with_offset being false.

To replicate this behaviour, I added a failing test that verifies this behaviour difference

thread 'parse_join_constraint_unnest' panicked at 'called `Result::unwrap()` on an `Err` value: ParserError("Expected end of statement, found: c1")', src/test_utils.rs:86:61
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
test parse_join_constraint_unnest ... FAILED

This fix uses the boolean check to determine if the subsequent optional alias should be parsed.

[alamb]

Looks great -- thank you @Ziinc

[alamb]

@Ziinc there are some CI failures on this PR -- can you please resolve them?

[Ziinc]

@alamb I've fixed the lint error. For the test case, it seems like the inline comments for the tested cases were off, and the last case that I removed should actually raise a ParserError, since an alias + WITH OFFSET alias without a WITH OFFSET clause does not make sense.

[alamb]

Makes sense to me -- thank you @Ziinc

[coveralls]

Pull Request Test Coverage Report for Build 3623351580

• 18 of 19 (94.74%) changed or added relevant lines in 2 files are covered.
• 1044 unchanged lines in 8 files lost coverage.
• Overall coverage increased (+0.03%) to 86.385%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
src/parser.rs	3	4	75.0%

Files with Coverage Reduction	New Missed Lines	%
src/tokenizer.rs	1	90.28%
tests/sqlparser_hive.rs	8	96.55%
tests/sqlparser_postgres.rs	25	97.41%
src/ast/ddl.rs	34	81.25%
src/ast/query.rs	47	84.23%
tests/sqlparser_common.rs	86	97.24%
src/ast/mod.rs	314	77.7%
src/parser.rs	529	83.44%

Totals
Change from base Build 3568725328:	0.03%
Covered Lines:	12449
Relevant Lines:	14411

---

💛 - Coveralls

[alamb]

Thanks again @Ziinc