perf improvement : has_eval() in CoreEnforcer.enforce_ex

[etienne-lebrun]

Hello!
In a simple profile calling enforce a few times,looping over ~500 policies, I have noticed that I spend almost 10% of my time in casbin.util.has_eval.
I believe it could be avoided if we reused the has_eval variable computed out of the loop (line 369). It could also probably be cached in the Assertion object stored in self.model["m"][mtype]
https://github.com/casbin/pycasbin/blob/a16bfaa669c37ac1598684e36b0319430ab749e5/casbin/core_enforcer.py#L369
has_eval = util.has_eval(exp_string)
https://github.com/casbin/pycasbin/blob/a16bfaa669c37ac1598684e36b0319430ab749e5/casbin/core_enforcer.py#L388

Let me know if you would like a PR

[casbin-bot]

@ffyuanda @Zxilly @techoner @elfisworking

[leeqvip]

@etienne-lebrun It should reuse has_eval, can you make PR ?

[github-actions]

🎉 This issue has been resolved in version 1.15.4 🎉

The release is available on:

• v1.15.4
• GitHub release

Your semantic-release bot 📦🚀