How can I use a custom compiler for my LaTeX document in LaTeX Lab?

Question

I'm using LaTeX, and recently LaTeX Lab, an online LaTeX interface to Google docs, was released.

For the most part this online LaTeX compiler is great, and I can get a document such as this to compile:

\documentclass[12pt]{article}
\usepackage[utf8]{inputenc}
\usepackage{amsmath}
\title{\LaTeX without extra usepackages}
\date{}
\begin{document}
  \maketitle
  I would like to know how to use \LaTeX{} with extra packages.

  % This is a comment; it is not shown in the final output.
  % The following shows a little of the typesetting power of LaTeX
  \begin{align}
    E &= mc^2                              \\
    m &= \frac{m_0}{\sqrt{1-\frac{v^2}{c^2}}}
  \end{align}
\end{document}

However, larger documents using more packages don't compile, even though they do on local installations of LaTeX. An example is below:

% 
% This document illustrates using LaTeX. LaTeX2HTML and DVIPDFM together
% for graphics and hyperlinks.
% 
%
\documentclass[12pt,a4paper,dvipdfm]{article}
\usepackage[dvips]{graphicx}
\usepackage[usenames,dvipsnames]{color} 
\usepackage{html}
%
% The lines below to use HYPERREF need to be commented out to get numbering to work 
% in the HTML version. Once commented out, run LaTeX several times, and then LaTeX2HTML:
%
%\usepackage[dvipdfm, latex2html,
%                           colorlinks, linkcolor=blue, urlcolor=blue, citecolor=blue,
%                           bookmarks, bookmarksopen, bookmarksnumbered,
%                           pdfauthor={Charles Clayton}]{hyperref}
%
% The command below simplifies the process of including an image. It assumes that the images
% are in a sub-folder called "graphics" in the folder that the .tex document lives in.
% It also assumes a bounding box file created e.g. with EBB exists in the subdirectory.
% If the graphic is (for example) called graph.jpg, then the bounding box file should be graph.bb
% The first parameter is the name of the image, the second the text for the ALT tag.
% Note that for some reason it goes wrong if you use \begin{htmlonly}...\end{htmlonly}
% instead of \html. This "feature" might however change in subsequent versions.
%
\newcommand{\img}[2]{
\begin{makeimage}
\end{makeimage}
\centering
\latex{\includegraphics{graphics/#1}}
\html{\htmladdimg[ALT="#2"]{../graphics/#1}}}
%
% The next two commands are just modified versions of \img
% The first shows how to put a box round things in both the HTML and PDF versions
% The seconds shows how to scale the image.
%
\newcommand{\boxedimg}[2]{
\begin{makeimage}
\end{makeimage}
\centering
\latex{\fbox{\includegraphics{graphics/#1}}}
\html{\htmladdimg[ALT="#2" BORDER="1"]{../graphics/#1}}}
%
\newcommand{\scaledimg}[2]{
\begin{makeimage}
\end{makeimage}
\centering
\latex{\includegraphics[scale=0.5]{graphics/#1}}
\html{\htmladdimg[ALT="#2" WIDTH="5cm"]{../graphics/#1}}}
%
%
%
\title{Example Document}
\author{Charles Clayton}

\begin{document}
\maketitle

\abstract{
This document includes graphics and hyperlinks and should run correctly through 
both LaTeX2HTML and DVIPDFM. See the document 
\htmladdnormallink{Getting \LaTeX\ and \LaTeX2HTML\ and DVIPDFM to Work Together}
{http://members.tripod.com/charlesclayton/index.htm} for more information.}

\tableofcontents\section{Hyperlinks}

These links should `work' (i.e. be clickable) in both the PDF and HTML documents.
\begin{itemize}
\item A link to an external web page: \htmladdnormallink{Amazon.com}{http://www.amazon.com}
\item A link to the \htmlref{next section}{sec:Gr}
\item A link to the next section, using a standard \verb+\ref+: Section \ref{sec:Gr}. 
To get section numbers you must use the \verb+show_section_numbers+ command line 
option with \LaTeX2HTML. Also, the line that uses the \verb+hyperref+ package must 
be commented out (due to a bug) and LaTeX re-run several times before running \LaTeX2HTML.
\item It does not seem possible to get a single link to e.g. `Section 2', as 
\verb+\ref+ does not work within \verb+\htmlref+ and similar commands. 
One can get nearly there by combining an \verb+\htmlref+ 
and a \verb+\ref+: \htmlref{Section}{sec:Gr} \ref{sec:Gr}.
\item This is an ordinary \verb+\cite+: \cite{ref1}. Note that \verb+\htmlcite+ does not
 seem to work in the PDF version (unlike \verb+\htmlref+ which works in both)
\item This is a reference to equation \ref{eqn1}
\item This is a reference to the \htmlref{same equation}{eqn1} done with \verb+\htmlref+
\end{itemize}

\section{Graphics}
\label{sec:Gr}
The figures in this document demonstrate including a JPEG image in both the PDF and HTML
versions. For details of how to use EPS graphics, see the document
\htmladdnormallink{Getting \LaTeX\ and \LaTeX2HTML\ and DVIPDFM to Work Together}
{http://members.tripod.com/charlesclayton/index.htm}. Notice that
\begin{itemize}
\item the ALT tag is set in the HTML versions;
\item if the \verb+\caption+ comes before the \verb+\includegraphics+, 
the caption will be above the image, otherwise it will be below the image. 
This is only the case in the PDF document, however.
\end{itemize}

\begin{figure}
\img{graphic.jpg}{A graph of some statistical distributions}
\caption{A graphic}
\end{figure}

\begin{figure}
\boxedimg{graphic.jpg}{A boxed graph}
\caption{A boxed graphic}
\end{figure}

\begin{figure}
\caption{A scaled graphic. In the PDF version, the caption will be on top of the image}
\scaledimg{graphic.jpg}{A scaled graph}
\end{figure}

\section{Other Remarks}
\begin{itemize}
\item Note that the PDF version has working bookmarks;
\item The author field is also set in the PDF document information;
\item Note the use of a ALT tag in the equation below. 
This provides meaningful alternative text if the equation is not displayed or a tool tip
\end{itemize}
\begin{equation}
\htmlimage{ALT="A fundamental equation"}
\label{eqn1}
e^{i\pi}=-1
\end{equation}


\begin{thebibliography}{99}
\bibitem{ref1} A test bibliography entry.
\end{thebibliography}

\end{document}

I would like to use LaTeX Labs to access, edit, and compile latex documents online. Is using a private compiler off of my own server a solution to this problem? Is there an easier way? If not, can anyone point me in the correct direction to set this up? The main tutorial page for this is blank. (Update: Now there is some description there.)

There are other tutorial pages as well:

• The Common LaTeX Service Interface (Home page)
• Common LaTeX Service Interface (CLSI) (Github Wiki)
• CLSI Request Format (Github Wiki)

Answer 1

You can find some documentation on bringing up a CLSI server instance on the CLSI project site: http://code.google.com/p/common-latex-service-interface/wiki/InstallAndConfig

For directions on using a custom CLSI environment with LaTeX Lab see the following page: code.google.com/p/latex-lab/wiki/UsingPrivateCompiler

I would only recommend bringing up a CLSI instance if you're planning to expose a custom, perhaps more powerful, LaTeX compiler, to a set of users. For getting around missing packages i recommend following DW's response.

There will be an option in the near future to use a local TeX Live/MikTeX installation, this way you'll be able to use a custom LaTeX environment without bringing up a CLSI server.

Answer 2

Rather than use the CLSI, if you are interested in including a few extra packages you can upload them as style documents

Files -> Upload Files

Then you can include them in your compilation

Compiler -> Project resources

For example I uploaded the following two files:

http://www.math.toronto.edu/~drorbn/VasBib/html.sty

http://lx2.saas.hku.hk/Conference/IascAsian05/hkustasc/template/graphicx.sty

and was able to compile the Tex document in the question above.