Repository navigation
Replies: 1 comment 1 reply
|
This is in texmath: raw ∀ and ∃ are |
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
pandoc <<< '$\forall x(Fx \lor \lnot Fx)$'Yields:
However
pandoc <<< '$∀x(Fx ∨ ¬Fx)$'Yields:
The important difference is that in the first one, we have
<mo>∀</mo>and in the second one we have<mi>∀</mi>It’s smart enough to treat the LaTeX-ese\forallas an operator, but not the unicode ∀, which is treated as an identifier. (Similar differences are seen for\existsversus ∃.)The difference isn't very obvious with the usual settings and styles, but if you start applying custom css rules to
<mi>and<mo>tags, the difference becomes important.I confess I'm not sure whether pandoc uses its own MathML conversion routines, or relies on a third party library, but I think it should treat ∀ and
\forallthe same, if possible.I also think the difference between ∨ and ¬ (regardless of whether the unicode or LaTeX is used), where it's always smart enough to recognize ∨ as an operator, but still treats ¬ as an identifier, feels inconsistent, but I'm not sure if there are pre-existing standards from unicode or mathml specifications, or similar, dictating such things.
Relatedly, is there a way to add or control the mathvariant attribute for regular identifiers in generated MathML? Without it, it seems to be difficult to render regular alphabetic characters using non-math fonts (to match the body text, for example).
All reactions