Bert Frees
Bert Frees
What would be good is to include some short explanation at the top of the working file itself.
Some things that should be included - [ ] How to work with the working file - [ ] How to create the database: see https://github.com/liblouis/liblouis/pull/382#issuecomment-325946477, but the "chunked_text" column...
To do: - [ ] Remove more global variables. - [ ] Add the "const" keyword where appropriate, in order to indicate which function arguments are "inputs" (read-only). - [...
Pt. 1 merged in a94304afbb0ecf52969f9359d87afb15fe4455f9
> I don't understand the first requirement There are also patterns that cancel hyphenation points. You should read up on how the Knuth-Liang hyphenation algorithm works. There used to be...
The answer is the same as in issue https://github.com/liblouis/liblouis/issues/1311: you are not getting the results that you are expecting because the `endmode` opcode is intended to be used in combination...
Fair point. We could keep using Github for specific services like CI while moving away from it in general. E.g. SourceHut has CI for Linux only. But AppVeyor is of...
So basically `\x1234`, `\y12345` and `\z12345678` would change to `\u1234`, `\u12345` and `\u12345678`?
I'm not totally convinced that we need external code to do the parsing actually. I do think it's nice to conform to a standard, but the parsing is really so...
> parsing is really so trivial that I think our own code is also fine I was referring to the escaping part, not the UTF-8 specific things. But you're right,...