fixed bug 6028

[crazyo]

The problem was that hyphen-minus was converted to minus signs without differentiating those in math expressions and those in texts. I fixed it by adding an optional parameter nonMath so that hyphens in texts won't be converted.

[tacaswell]

@mdboom @zblz

[mdboom]

Thanks for investigating the issue.

I think actually, for the non-math context, get_unicode_index should always return ord(symbol) and do nothing else. The fact that it could fall through to return tex2uni[symbol.strip("\\")] is also a bug. So maybe add:

if non_math:
    return ord(symbol)

to the top and leave the rest of get_unicode_index as is.

Also, to be PEP8-compliant, the new keyword argument should be spelled non_math, not nonMath.

[crazyo]

@mdboom Hi thank you for your response. I have made changes accordingly :-)

[mdboom]

👍 from me, but I'll leave this up for a little while in case others have comments.

[zblz]

Looks good to me, even though the code might be clearer if a math flag instead of non_math was used. Double negatives like non_math=False make understanding the code a little bit more difficult.

[crazyo]

@zblz I've made the changes :-)

[tacaswell]

@crazyo Can you open a new PR to add a simple test for this?

[tacaswell]

backported to v2.x as ebfbc4a

[crazyo]

yes, will do.