drag_float and drag_int don't respect the min and max values about pyimgui

[jeremyherbert]

Found the problem: all of the min and max arguments are swapped when passing through Cython. This is the same for all drag_float and drag_int types (and maybe more? I only checked these ones). I'll try to create a pull request when I get the time, I don't have a build environment set up to test the changes so it might take a bit.

core.pyx: (around line 3084)

cdef float[2] inout_values = [value0, value1]
    return cimgui.DragFloat2(
        _bytes(label), <float*>&inout_values,
        change_speed, max_value, min_value, _bytes(display_format), power
    ), (inout_values[0], inout_values[1])

cimgui.pxd: (around line 595)

bool DragFloat(  # ✓
            const char* label, float* v,
            # note: optional
            float v_speed, float v_min, float v_max,
            const char* display_format, float power
    ) except +

[swistakm]

@jeremyherbert this is a very good catch. Thanks for PR too :)

[swistakm]

Fix released as #54. I have additionally reversed the order of min/max arguments (non-keyword) in drag widgets to be more consistent with C++ API.

[jeremyherbert]

Cool, thanks for the quick work. I was thinking about changing the order of arguments, but I didn’t want to break backwards compatibility.

Although in hindsight, if this is the first time the bug has showed up, I might be the only one using this feature ;)

[swistakm]

Released as 0.1.0