override default healthcheck support

[momo-p]

I'm currently using this crate to communicate with my manticore server, even though it's compatible wit MySQL but for some reason it's not support SELECT version() which make the default healthcheck always throw error.

So I'm adding a feature to override the default healthcheck function.

[robjtede]

Is there any reason not to just change the current health check query to SELECT 1?

[momo-p]

Is there any reason not to just change the current health check query to SELECT 1?

Maybe I over-engineered it a little bit, but I think it could be for future-proofing.

Like in case query log enabled for debugging and you don't want it full of SELECT 1 in the logs and your database support it you could swap it with the .ping() call

[robjtede]

Even better, would be nice if ping returned Result<(), Error> then it would be a no-brainer to switch to.

For now I'm tempted to change it to SELECT 1.

[momo-p]

the current ping() method in the mysql crate return bool so I believe it would took a few more step.

what I really concern at that time (sorry I wrote those at 3AM :D) is that give it a little more flexibility, who knows if some future sql-like database in the future support COM_PING or their parser accept SELECT 1;, so I add it just-in-case needed

[robjtede]

Pushed a change that:

• uses more standard type naming
• changes the signature of HealthCheckFn to a function pointer
  • this is equivalent in the fns it accepts to what you'd written before with the &'static bound on health_check_fns param to with_custom_health_check but avoids the Arc overhead

[robjtede]

@outersky curious what you think. I'm hesitant but seems that my PR blackbeam/rust-mysql-simple#355 to change the .ping() signature is slow going so might be a nice QoL change, for now at least. (My feeling is that we necessarily push enough major versions that it could be removed without to much hassle in the future.)

[robjtede]

Also for the CI fail: We'll rebase this after #37 is merged.

[robjtede]

superseded by #38