Thu 11 May: TeX Hour: Using LaTeXML to access audit arXiv LaTeX source files

[Jonathan Fine]

Hi Deyan

Thank you for the news and encouragement. Here's a nice example from last
week. It came from looking at a random article.

Here's the example:
===
Warning:not_parsed:>VERTBAR MathParser failed to match rule 'Anything'
at 0910.4716.tex; line 169 col 74 - line 170 col 17
In "$|G:H|=n Warning:not_parsed:>VERTBAR MathParser failed to match rule
'Anything'
at 0910.4716.tex; line 169 col 74 - line 170 col 17
In "$|G:H|=n < \infty $
> |[[VERTBAR]] G[[UNKNOWN]] :[[METARELOP]] H[[UNKNOWN]] |[[VERTBAR]]
=[[RELOP]] n[[UNKNOWN]]
===

I'm been a math student for long enough to know that $|G:H|$ is a notation
for the index of the subgroup H in G. So there's a question here of uniform
and best practice. And it's an example of something Neil Soiffer pointed
out at the arXiv forum, namely the ambiguity of the vertical bar.

I'd be more than pleased if we could achieve a similar understanding for 5%
of the errors produced by LaTeXML when it processes the arXiv. Because that
would be a good start to getting to 25%, which would be a substantial
achievement.

Here's the URL I look at: https://ar5iv.labs.arxiv.org/html/0910.4716

Anyone can inspect a random article by going to
https://ar5iv.labs.arxiv.org/feeling_lucky.

with best wishes

Jonathan

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://tug.org/pipermail/texhax/attachments/20230516/34b69309/attachment.htm>