Hi folks,
I've finally decided to seriously work on getting my dissertation
available on the web in HTML format. The only problem is that the thing
is now nearly 3 years old, and it includes some pretty low level TeX/LaTeX
hacking. (Believe it or not, there's actually still a good reason why I
need to get this on the web so long after the fact.) In addition to
several different generic style files that l2h doesn't grok, I have to
teach it about the stuff I added. I pulled over the most current l2h
(98.1p1), and have verified that all my tools are the required
revisisions, but I'm still having all sorts of problems even getting this
to run at all, much less produce correct output. The disseration was done
with latex209, though it still works fine under 2e running in
compatability mode.
One problem I've already noticed and dealt with is that I use lscape.sty,
which provides the landscape environment for rotating the postscript output.
l2h does not know about this style file, so it's just doing the best it
can. In several cases I have figures of the form
\begin{landscape}
\begin{figure}[p]
\centerline{\psfig{figure=ruwt-chapter/drt-stripe.epsi,width=\dissheightc}}
\caption[Striping in the Dirty Map]{Striping in the dirty map.}
\label{fig:drt-stripe}
\end{figure}
\begin{figure}[p]
\centerline{\psfig{figure=ruwt-chapter/vm-stripe.epsi,width=\dissheightb}}
\caption[Striping in the Restored Image]{Striping in the restored image.}
\label{fig:vm-stripe}
\end{figure}
\end{landscape}
This works fine for the normal latexing, but the presence of the landscape
environment around both figures somehow causes l2h to group the two figures
together, and they won't (by definition) fit on the same page so that I
get a figure lost error when it tries to latex the file images.tex. If
I put each figure into its own landscape environment, both latex and
l2h work fine. Is this a known problem? If it's not, and the above outline
isn't sufficient, I can provide more detailed debugging information.
A more difficult problem that I will have to deal with is that I've defined
my own set of custom caption commands. One of them is identical to the
normal \caption, except that it does not increment the figure counter or
make an entry in the list of figures or list of tables. (It's used in
situations where you have two or more related figures. The standard
\caption is used in the first, and the stealth \captcont is used for the
continuations.) I have a tagged caption which takes an extra tag argument
(typically things like "a", "b" or "c"), and a continuation form of that as
well. This allows me to do things like
\tcaption[Point spread functions for snapshot case]{PSFs for snapshot}{a}
...
\tcaptcont{cont.}{b}
A single entry goes into the list of figures for, say, Figure 24. The
first subfigure gets labelled Fig. 24a with its own caption, the second
gets labelled Fig. 24b. This is all very handy, but I need to teach l2h how
to handle it. I could really use at least an outline of what I need to do
in the latex2html style file. I speak perl quite well, but don't have an
adequate understand of the structure of l2h. I've also got my own custom
progfig environment, which is exactly like a figure except that it
generates a list of programs at the start of the document instead of a list
of figures. I need to explain this to l2h as well, and could stand a word
or two on how to go about it. Sorry for the newbie questions, but this
documentation on this seems a bit sparse at best.
The actual implementation of the extended caption commands also seems to
be a problem. The follow snippet
\catcode`\@=11
\def\captcont{\@dblarg{\@captcont\@captype}}
\long\def\@captcont#1[#2]#3{
\par
\begingroup
\@parboxrestore
\normalsize
\@makecaption{\csname fnum@#1\endcsname}{\ignorespaces #3}\par
\endgroup}
from the captioning commands produces the message
%%%%%%%,,Arguments to @ are too complex ...
It will not be processed unless used in another environment
which is passed to LaTeX whole for processing.
Am I bumping up against some limit of l2h? Am I just hosed because it's
TeX and too low level for l2h to deal with? Will the messages go away and
things work OK once I declare these as captioning commands to l2h? I can
certainly provide full source to anyone who cares enough to look at it.
But the above should probably be sufficient for debugging and/or enough to
outline what I'll need to explain to l2h.
The next problem I am having is with \newif. Again, I know that this is a
TeXism, but in principle should l2h be able to handle it? I define a
couple of ifs right after the documentstyle declaration, and each one
generates some scary looking error messages. In particular, the test file
\documentstyle{report}
\newif\ifpsdraft
\begin{document}
%
A very simple document
%
\end{document}
generates (in addition to the normal messages)
%%%%%,Backslash found where operator expected at (eval 426) line 2, near "]\"
(Missing operator before \?)
*** sub do_cmd_psdrafttrue failed:
syntax error at (eval 426) line 2, near "]\"
Backslash found where operator expected at (eval 427) line 2, near "]\"
(Missing operator before \?)
*** sub do_cmd_psdraftfalse failed:
syntax error at (eval 427) line 2, near "]\"
I'm having some other problems as well, but which I suspect to be interactions
with the above, so I'm trying to take this one step at a time and fix the
clearly defined problems first. Any help or suggestions, people?
Many thanks!
Dan Briggs
[EMAIL PROTECTED]
