Sha256: 4b8ccbb9050c3879695c6834c100ee2d92ff94f99a9e72c8ff0025cd788454de
Contents?: true
Size: 1.25 KB
Versions: 6
Compression:
Stored size: 1.25 KB
Contents
== rspec At the moment there is just one set of tests, in `transform_spec.rb`, which test a tiny piece of a small part of the code. This is the `gsub` call in the `TeXPreprocessor`. I had been having some trouble with instances of $ ... $ not being mapped properly, and I think I've narrowd it down to a few small edge cases -- I am trying to collect these in the wild, as welll as t imagine them. I've organized the files as `a.rb`. `b.rb`, and `c.rb`. The first two pass, the third deos not. The strings that I try the transformer on in `c.rb` are only a few characters long. === Help @mojavelinux, my knowledge of regular expressions is not so great -- could you look at this? Please note that I had changed the output regular expression: ---- TEX_DOLLAR_SUB = '\1\\\(\2\\\)\3' TEX_DOLLAR_SUB2 = '\1latexmath:[\2]\3' ---- using the first, rather than the second, which is your original. I don't think that that is the problem -- it must be in the recognizer, ---- TEX_DOLLAR_RX = /(^|\s|\()\$(.*?)\$($|\s|\)|,|\.)/ ---- We can change back to (2), though I must say I like (1) better since it parallels `\[ ... \]` and is more succinct. Is there any difference in their function or behavior _vis a vis_ Asciidoctor, i.e., in being protected from substitution?
Version data entries
6 entries across 3 versions & 1 rubygems