Alphanumeric, number first issue

[Mottie]

I've found another issue with sorting alphanumeric numbers with the number first. Add this to the unit-test.html file:

wrapTest(
    ['5D','1A','2D','33A','5E','33K','33D','5S','2C','5C','5F','1D','2M'],
    ['1A','1D','2C','2D','2M','5C','5D','5E','5F','5S','33A','33D','33K'],
    'alphanumeric - number first'
);

And the resulting sort will look like this:

['1D','1A','2D','2C','2M','5F','5D','5E','5C','5S','33D','33A','33K']

[Mottie]

Well, I've narrowed the problem down to this regex:

re = /(^-?[0-9]+(\.?[0-9]*)[df]?e?[0-9]?$|^0x[0-9a-f]+$|[0-9]+)/gi

I'm not sure what the [df]?e? in this regex does, but it's causing the values of D, E and F to return 0.

If you remove the [df]?e? it will sort properly and it doesn't appear that any other tests fail.

[overset]

the d, f, e were for trying to tokenize decimal, float and scientific notation, i.e. '5e2' being 500. I'll see if I can make that regex a little "smarter" by maybe using a lookahead past the datatype identifier. I might just drop trying to tokenize on [0-9](d|f|e) all together. I'll add your test case to the runner as well. For your case just dropping the [df]?e?[0-9]? will work.

[overset]

fixed as of @Mottie 's 0ea168c commit