[5.5] bcrypt changed to Hash::make and renamed to str_hash :D

[Dylan-DPC-zz]

Small change as requested in laravel/ideas#695

[m1guelpf]

@Dylan-DPC Maybe you could close this in favour of #20155?

[Dylan-DPC-zz]

no i could merge your changes in this PR?

[m1guelpf]

@Dylan-DPC I'll PR your PR :D

[fernandobandeira]

Maybe hash_make could be used instead? It would be closer to the Hash::make syntax so it could help ppl remember about it...

[m1guelpf]

@fernandobandeira Well, str_hash() is consistent with other str_* functions. I like it 😄

[fernandobandeira]

Yep i get it just a suggestion...

[devcircus]

If the "str_hash" function is going to happen, why not add it alongside bcrypt for now, and change any core references to the new wording.

This way users will be able to continue using bcrypt, but know it is deprecated.

Personally don't see any reason to change at all just because something may be changing in 7.2.

[m1guelpf]

@devcircus I'd say that's Taylor decision. It has been suggested lots of times to trigger E_DEPRECATED_ERROR errors and deprecate instead of removing them, but as far as I know, it hasn't been done yet.

[devcircus]

Definitely Taylor's call. Just would be good to avoid needlessly breaking apps due to some possible future change.

[koenhoeijmakers]

Should just put a @depricated docblock by the bcrypt() helper and create str_hash() alongside it to prevent breaking applications i suppose?

[m1guelpf]

@koenhoeijmakers See my comment above.

[koenhoeijmakers]

Also, i'd suggest calling it hasher(), since any hasher class should implement the HasherContract, but thats my opinion.

@m1guelpf yea i just noticed :)

[joshmanders]

Lets give it some uniqueness and call it hashbrowns($str) because they are tasty.

[taylorotwell]

Don't remove the bcrypt method. I would just add the new method for now and we'll remove bcrypt helper in a later release.

[Dylan-DPC-zz]

@taylorotwell fine i will add the bcrypt method again. Reopen? or create a new PR?

[m1guelpf]

@Dylan-DPC New PR.

[m1guelpf]

@taylorotwell How about deprecating bcrypt()?