Instantiating mvec segfault

[esoma]

Presumably you're not supposed to do this, but trying to instantiate an mvec will (most of the time) cause a segfault.

Python 3.9.2 (default, Mar 24 2021, 00:25:35)
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import glm
>>> glm.version
'PyGLM (DEFAULT) version 2.5.0'
>>> glm.mvec2()
Segmentation fault (core dumped)

Behavior also happens on windows.

[Zuzu-Typ]

Hey there (:

Whoopsie, yeah, that should not happen.

Indeed, it wasn't supposed to be possible to instantiate an mvec type, as it normally only holds a reference to data, instead of containing data itself.

I don't remember exactly why I decided to provide direct access to the mvec classes (e.g. by using glm.mvec2).
My best guess is that I wanted to provide the user with the class so they could compare some objects type with it, to test whether or not it's an mvec.

Somehow I must have either forgotten to test it, or broken it later on.

Anyway, it should be fixed now.

When instantiating an mvec type now, a vector is created internally.
Alternatively, you can pass a vector to reference.
Example:

import glm

v = glm.vec3(1,2,3)

mv = glm.mvec3(v)

mv.x = 9

print(v) # prints vec3( 9, 2, 3 )

mv2 = glm.mvec2() # same as glm.mvec2(glm.vec2())

print(mv2) # prints mvec2( 0, 0 )

Cheers
--Zuzu_Typ--