Fix interval parsing logic and precedence

[sarahyurick]

I originally opened an issue for this on the DataFusion side: apache/datafusion#3944, but found the error to lie in how the queries were being parsed.

These changes allow something like "SELECT col FROM test WHERE d3_date > d1_date + INTERVAL '5 days' AND d2_date > d1_date + INTERVAL '3 days'" to be parsed correctly. Previously, the '5 days' was getting separated from the INTERVAL keyword and lumped into the AND operator.

[coveralls]

Pull Request Test Coverage Report for Build 3491640784

• 82 of 85 (96.47%) changed or added relevant lines in 2 files are covered.
• 883 unchanged lines in 9 files lost coverage.
• Overall coverage increased (+0.2%) to 86.341%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
src/parser.rs	13	14	92.86%
tests/sqlparser_common.rs	69	71	97.18%

Files with Coverage Reduction	New Missed Lines	%
src/dialect/mod.rs	1	88.46%
src/dialect/snowflake.rs	1	90.0%
tests/sqlparser_mysql.rs	2	99.68%
src/ast/data_type.rs	13	89.71%
src/ast/ddl.rs	16	81.31%
tests/sqlparser_postgres.rs	24	97.27%
tests/sqlparser_common.rs	80	97.16%
src/ast/mod.rs	274	77.47%
src/parser.rs	472	83.85%

Totals
Change from base Build 3413823174:	0.2%
Covered Lines:	11877
Relevant Lines:	13756

---

💛 - Coveralls

[alamb]

I wonder if you investigated following a similar pattern to NOT which adjusts the precedence based on peeking in the token stream?

https://github.com/sqlparser-rs/sqlparser-rs/blob/87b4a168cbcd83439c4e8729dd992d150eebd15c/src/parser.rs#L1648-L1660

I couldn't find any example of a similar partten in the codebase

https://github.com/search?q=repo%3Asqlparser-rs%2Fsqlparser-rs%20parse_infix&type=code

What do you think @AugustoFKL ?

[AugustoFKL]

@alamb I think we may need to refactor the expression parsing...

With the current structure, we end up with no such good expansion of expressions to parse. And there's still the (big) problem with precedence analysis.

But, for now, and considering the scope of this PR, I think that's OK. Doing it for NOT would be better in another PR, right?

[alamb]

Thank you for the contribution @sarahyurick

[sarahyurick]

Thanks @alamb @AugustoFKL ! Let me know what you think. Should I update anything with the precedence logic, or just keep as is for now?

[AugustoFKL]

lgtm

@sarahyurick IMHO that's ok

[alamb]

Thanks again @sarahyurick and @AugustoFKL