feat: use standard errors on MapClaims validation (v4)

[viniciusgabrielfo]

Not sure if v4 still receives updates

But trying to handle proper the expiration errors in a project that uses jwt/v4 when we want to identify the Token Expired error for example, as is a custom error we need to check by string comparison.

Just adding the proper exported standardized errors on MapClaims Valid().

---

This change can cause issues for users that have already created validations using a CamelCase string comparison when call Parse() like:

token, err := jwt.Parse(tokenString, keyFunc{})
if strings.EqualFold(err, "Token is expired") {
    // handle error
} 

OR

token, err := jwt.Parse(tokenString, keyFunc{})
if err.Error() == "Token is expired" {
    // handle error
} 

Since the standardized expired error uses lower case t and the current custom errors use upper case T

[oxisto]

Not sure if v4 still receives updates

But trying to handle proper the expiration errors in a project that uses jwt/v4 when we want to identify the Token Expired error for example, as is a custom error we need to check by string comparison.

Just adding the proper exported standardized errors on MapClaims Valid().

This change can cause issues for users that have already created validations using a CamelCase string comparison when call Parse() like:

token, err := jwt.Parse(tokenString, keyFunc{})
if strings.EqualFold(err, "Token is expired") {
    // handle error
} 

OR

token, err := jwt.Parse(tokenString, keyFunc{})
if err.Error() == "Token is expired" {
    // handle error
} 

Since the standardized expired error uses lower case t and the current custom errors use upper case T

Thanks for your contribution. Well this is tricky one, our policy states that we only support security updates for v4 but develpment is focused on v5 (and v6, see #335). I strongly suggest to upgrade to v5, which should be more or less harmless.

Additionally, changing the error strings is sort of a breaking change as well, this is why we did not touch it in v4, so I am afraid I would vote no against this PR, sorry :(

@mfridman your thoughts?