Added SchemaProxy mutex #163

[daveshanley]

Rendering causes bits to be set on structs in the model. Concurrency causes race issues with reading a writing of these bits, as there is no locking in the model.

Until now. locking prevents concurrent renders that use a shared model from conflicting with one another.

Addresses #163 and pb33f/libopenapi-validator#23

[codecov]

Codecov Report

Patch coverage: 100.00% and no project coverage change.

Comparison is base (dab6346) 99.80% compared to head (776c2fb) 99.80%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #172   +/-   ##
=======================================
  Coverage   99.80%   99.80%           
=======================================
  Files         148      148           
  Lines       10639    10643    +4     
=======================================
+ Hits        10618    10622    +4     
  Misses         21       21           

Flag	Coverage Δ
unittests	99.80% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files Changed	Coverage Δ
datamodel/high/base/schema.go	100.00% <100.00%> (ø)
datamodel/high/base/schema_proxy.go	100.00% <100.00%> (ø)

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[jxsl13]

why is the mutex a pointer?

[daveshanley]

My general rule is to ask myself "why should this NOT be a pointer". I generally 99% lean on pointer.

This is just a case of me defaulting on my coding behavior. If I pass around anything, I move it to a pointer, because I don't want to copy anything.

[jxsl13]

should a pointer to a mutex be copied?
I think that might lead to problems.
Go has a builtin detection for when a struct is copied that contains a mutex. If I'm not mistaken, a pointer mutex is not caught by that detection.

[daveshanley]

I don't know, I haven't seen any literature saying that it's any different from a struct.

https://go.dev/src/sync/mutex.go

The receiver on Mutex is a pointer for Lock and Unlock

[jxsl13]

I am not sure on that one either, just having seen warnings pop up in the terminal when a struct with a mutex was copied.
So I'm wondering if anyone is supposed to copy the struct or not. Once you have multiple instances of the wrapping struct, all of them will refer to the same mutex. Dunno if thats wanted by you or not or if copying of the wrappung struct is wanted or not.
Just giving my two cents :). Do whatever you want with it.

[daveshanley]

I've not seen any warnings pop up and the behavior is as I expect, tests pass accordingly. If any strangeness appears with this functionality, then I'll drop it back to a struct :)