Physics Filesystem

[FRex]

I've added example code of sf::InputStream that uses PhysicsFS to load from disc or some archive(it's all transparent and abstracted away by physics).
Can someone please double check formatting and that I didn't break the main Sources page or something?

https://github.com/SFML/SFML/wiki/Sources

https://github.com/SFML/SFML/wiki/Source%3A-PhysicsFS-Input-Stream

This will be helpful next time someone molests Laurent about SFML needing a filesystem..

[Tank]

Looking all good.

[Laurent]

To be extremely picky, I would add open/close functions to the stream class, rather than giving it directly a PHYSFS_FILE*. This way it is perfectly encapsulated, and has the same interface as other file streams (like std::ifstream).

[FRex]

To be extremely picky

Done, hope I didn't introduce a mistake now..

[Laurent]

Looks good, but open should return a boolean

[eXpl0it3r]

Nice!
I always wanted to look into that, but never got around to do it.

[FRex]

Looks good, but open should return a boolean

Done too. Really, it's just to show that it's possible and easy not to be best designed API ever.

[Laurent]

The code is so small, there's no reason to have it badly designed

[Tank]

Sorry for this useless post, but: lol

[FRex]

Sorry for this useless post, but: lol

LoL.

The code is so small, there's no reason to have it badly designed

void main(){}

Thanks for re-edit, whoever did it. My style of coding isn't very friendly.

Turns out we forgot to null out file handle when we close so it didn't really look like it closed(we held invalid pointer). Fixed.

[FRex]

I've updated it a tiny bit, added C++ to ``` to make code colored, split it into hpp and cpp, added sf::NonCopyable as a base to make it more secure, etc.

[GRAHAMO]

Decent, need to accomplish something like this with mine.