fix stmtcache invalidation

[ethanpailes]

This patch fixes #841. The meat of the fix lives
in a PR to the pgconn repo. This change just checks
for errors after executing a prepared statement and informs
the underlying stmtcache about them so that it can properly
clean up. We don't try to get fancy with retries or anything
like that, just return the error and allow the application to handle it.

I had to make some changes to to the jackc/pgconn package as well
as this package, so I've retargeted this this repo on
ethanpailes/pgconn for the time being. This PR is not really in a
mergable state until those PRs merge and I rip out the retargeting
commit, but I would say it is ready for review (just review the most recent commit).

Fixes #841

[jackc]

I think it would be possible to call StatementErrored in rows.Close. That would avoid needing to expose ResultReader.Err as I mentioned in the other PR.

Also, I think you could use a replace directive in go.mod to avoid having to rewrite all the import paths while you're working in development.

[ethanpailes]

I think it would be possible to call StatementErrored in rows.Close. That would avoid needing to expose ResultReader.Err as I mentioned in the other PR.

Good idea, I've done that and it works great. (I've updated the most recent patch so you can still review it all as one diff and ignore my big re-targeting diff). I did have to add a backreference to the connection the rows was spawned from though.

Also, I think you could use a replace directive in go.mod to avoid having to rewrite all the import paths while you're working in development.

That's a good idea. I'll keep it in mind for next time!

[ethanpailes]

Just noticed that rows.Next calls rows.Close on error, so we should be covered there as well!

[jackc]

This should be ready once it is pointed back at jackc/pgconn (point it at commit cba610c) and the StatementErrored call signature is changed.

[ethanpailes]

@jackc I think this should be good to go now.

[jackc]

Thanks!

[eh-steve]

🎉