The primary consideration is security. When adding a very big new
extension like MathML, it is very tempting to cut corners, and not
truly understand every corner of the specification and build a parser
that *truly* understands what it reads, and isn't just checking
syntax blindly. I'm happy to review PRs on the htmlpurifier
repo, but the more obviously secure the PR is, the faster
this will happen.
Excerpts from manuel's message of 2018-09-21 04:02:06 -0700:
> My name is Manuel, lead plugin developer at MathType.
> Actually we are working in a new HTML Purifier module to implement MathML
> specification. We have quite experience working with MathML specification,
> and we expect to have the new module ready soon.
> We are very interested to add this new module to HTML Purifier core: there
> are some platforms, like Moodle, using HTML Purifier to sanitize HTML
> output. If this new module is added to HTML Purifier, it would be easy for
> this platforms to inherit the changes. In addition: since MathML is a W3
> standard I can't see any reason, besides MathML spec is quite complicated,
> to don't have this MathML module in it's core.
> So: what is the process to contribute to HTML Purifer code? It's enough to
> create a pull request against the GitHub repository? An issue in the GitHub
> tracker is needed? There are some additional steps?
> Thanks in advance,
> MathType 7 is out! Check the new version at wiris.com/mathtype