(fix) Change context manager type to avoid race in threads

[vlmaksimuk]

Fix issue from issue #1132 when we get race condition in multithreading with generator based context manager. The generator is not thread safe!

[sl0thentr0py]

@Fofanko if this really fixes the problem, happy to merge it in, but I'm trying to fully understand where the threading problem actually lies.

In the original stacktrace on the linked issue, the error occurs in _after_cursor_execute, so it seems the ctx_mgr variable on conn is being acted on by multiple threads, and so the conn itself is being used across threads.

From the Connection documentation

The Connection object is not thread-safe. While a Connection can be shared among threads using properly synchronized access, it is still possible that the underlying DBAPI connection may not support shared access between threads. Check the DBAPI documentation for details.

So I believe the actual fix for this would be to wrap the _sentry_sql_span_manager getter/setter in a lock?

[sl0thentr0py]

thx @Fofanko for the PR!
I'm still not convinced this is the actual problem, but it still does get rid of generator magic so I'm merging it in.

[sl0thentr0py]

fyi @Fofanko I'm reverting this since it caused a regression and trying to fix the problem properly in #1388.

[Forty-Bot]

Can you add something like

record_sql_queries = RecordSqlQueries

to the module? This breaks users who were using the previous name.

[sl0thentr0py]

@Forty-Bot this is reverted anyway in one of the newer releases.

[Forty-Bot]

Ah, looks like I was looking at an older branch.