On Sun 25th Jul, 2010 at 17:13, Lars Hellström seems to have written:
Dr. Clea F. Rees skrev:
tex
I presume you mean fontinst? Going
tex msam.etx
has never been supposed to work, even though
latex msam.etx
should work.
I did mean tex, but not tex msam.etx. I was using msam.etx indirectly
by calling it from another file e.g. tex yirm-drv.tex where this file
loads fontinst and calls msam.etx.
seems to choke on the encoding file msam.etx distributed with
fontinst. This appears to be because it interprets a line of commentary
literally.
Yes, the problem seems to be that both \metrics and \encoding get \let to \fi
(by \inputmtx and \inputetx respectively), so either one will be seen as
beginning the body of an encoding or metrics file. This has however been the
case throughout the 1.x series of fontinst, so apparently it is the markup
that is buggy.
Commenting this line out solves the problem:
--- msam.etx 2006-01-12 23:45:36.000000000 +0000
+++ msam.etx 2010-07-22 12:02:41.000000000 +0100
@@ -23,7 +23,7 @@
like
\begin{quote}
\textellipsis
- \verb|,\metrics\generalpltomtx{msam10}{msam10}{pl}{msam},msam10,|%
+% \verb|,\metrics\generalpltomtx{msam10}{msam10}{pl}{msam},msam10,|%
\textellipsis
\end{quote}
in the \textit{file list} argument of \verb+\installfont+, or generate
However, I'm sure this isn't the best way to do it. I assume there is a
way of getting fontinst to regard this part of the commentary as a
comment so that it is still typeset as part of the encoding's
documentation. I'm just not sure how to do that so I did this just to
get it working.
Since it is the \metrics token that is troublesome, a work-around is to split
it between two \verb's, like so:
\textellipsis\verb|,\m|%
\verb|etrics\generalpltomtx{msam10}{msam10}{pl}{msam},msam10,|%
\textellipsis
However, the trick being described has since been rolled into a proper
feature, so the comment should really say
\textellipsis\verb|,msam10 encoding msam,|\textellipsis
I'll have to fix that (and msbm.etx, which has the same bug).
Although I don't experience the same issue with msbm.etx which seems
odd now you point this out.
Thanks very much,
cfr
