Implement parsing of Postgres regex match operators

[returnString]

Hey all,

This PR adds support for parsing Postgres' POSIX regex matching operators. These are similar in nature to LIKE et al but with full support for regex patterns rather than being restricted to wildcards, so you can use select * from my_table where name ~ '^pattern$'.

I've restricted the parsing to only run when using the Postgres dialect, but I've not given lexing the same treatment as that doesn't seem to be the done thing (e.g. PGBitwiseXor is lexed unconditionally). If this is actually desired, please let me know!

I'm not entirely happy with the token names, but given the overlap (e.g. ~ is both a unary and bitwise op now) I didn't want to attach names that were overly specific either.

[coveralls]

Pull Request Test Coverage Report for Build 749129485

• 65 of 73 (89.04%) changed or added relevant lines in 4 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage decreased (-0.01%) to 90.041%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
src/parser.rs	8	11	72.73%
src/tokenizer.rs	43	48	89.58%

Totals
Change from base Build 676410191:	-0.01%
Covered Lines:	5524
Relevant Lines:	6135

---

💛 - Coveralls

[Dandandan]

I think this is looking good @returnString . Can you fix or ignore the linting errors?

[alamb]

Possibly made redundant by #328

[alamb]

Closing this PR as #328 was merged