limits separated from integral sign about latexml HOT 5 CLOSED

The postprocessor tries to mockup unattached sub/superscripts
in MathML when parsing has failed, but it doesn't actually attach
them to the anything. The reason it has failed to parse is due to the
zoo of meanings that "|" can have. In your example, it probably
is punctuation (but possibly "divides"); I haven't managed
to get a grammar sorted out that recognizes this case,
unless you've explicitly told it that | is punctuation.
(there are macros in latexml.sty that support this).

I'll have to figure out if it's cheaper to fix the postprocessor
or the parser! :>

from latexml.

zoo of meanings that "|" can have. In your example, it probably
is punctuation (but possibly "divides"); I haven't managed
to get a grammar sorted out that recognizes this case,
unless you've explicitly told it that | is punctuation.
(there are macros in latexml.sty that support this).

In this case it's "given" as in conditional probability, so it's analogous to a comma or semicolon separating function arguments. I'll have a look at using those macros in the paper.

I'll have to figure out if it's cheaper to fix the postprocessor
or the parser! :>

Thanks!

from latexml.

Yes, using \lxPunct{|} does fix the output. Thanks again.

from latexml.

I have a (possibly?) related observation: if the integral is surrounded by mismatching parentheses -- as is necessary when typesetting multi-line formulas -- the limits are also unattached. Here is an example:

\begin{align*}
   X &= Y \times \left[ A   \right. \\ 
    & \qquad \left. + (1/2) \int_0^\infty  B(s) ds \right]
\end{align*}

The following gives the same result:

\begin{align*}
   X &= Y \times [ A  \\ 
    & \qquad  + (1/2) \int_0^\infty  B(s) ds ]
\end{align*}

--p

from latexml.

I guess I'm going to consider that the fix to ticket #144 also fixes
this one....
At least it fixes the symptom; improving parsing will be a long-term
task in LaTeXML anyway! :>

from latexml.