Fixes for prepared statements and literal values

[zachmu]

This PR fixes issues when using bindvars in a couple different places in certain queries. It introduces a new LiteralExpression so that it GMS and Doltgres implementations of literal values can be used interchangeably during planning and analysis.

Relies on dolthub/go-mysql-server#2946

[github-actions]

	Main	PR
covering_index_scan_postgres	341.82/s	338.65/s	-1.0%
index_join_postgres	154.76/s	155.49/s	+0.4%
index_join_scan_postgres	185.03/s	185.65/s	+0.3%
index_scan_postgres	12.64/s	12.55/s	-0.8%
oltp_point_select	2654.26/s	2634.46/s	-0.8%
oltp_read_only	1860.84/s	1842.20/s	-1.1%
select_random_points	117.25/s	116.08/s	-1.0%
select_random_ranges	129.42/s	129.05/s	-0.3%
table_scan_postgres	11.80/s	11.84/s	+0.3%
types_table_scan_postgres	5.48/s	5.57/s	+1.6%

[github-actions]

	Main	PR
Total	42090	42090
Successful	16361	16361
Failures	25729	25729
Partial Successes1	5530	5530

	Main	PR
Successful	38.8715%	38.8715%
Failures	61.1285%	61.1285%

Footnotes

1. These are tests that we're marking as Successful, however they do not match the expected output in some way. This is due to small differences, such as different wording on the error messages, or the column names being incorrect while the data itself is correct. ↩

[Hydrocharged]

LGTM!

[Hydrocharged]

I remember way back you mentioned that we should abandon interface{} and use any now, but I've noticed you using interface{} for a few PRs now. Should we go back to interface{}?

[zachmu]

Superseded by #1437