[sgt/halibut] / doc / running.but

\versionid $Id$

\C{running} Running Halibut

\I{running Halibut}In the simplest case, running Halibut is very
easy. You provide a set of input files on its \i{command line}, and
it produces a set of output files.

\c $ halibut intro.but gettingstarted.but reference.but index.but
\e   bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb

This will generate a large set of \i{output files}:

\b \i\c{output.txt} will be a \i{plain text} version of the input
document.

\b \i\c{output.hlp} and \i\c{output.cnt} will be a \i{Windows Help}
version of the same thing. (Most of the text is in \c{output.hlp};
\c{output.cnt} contains additional contents data used by the Windows
help topic selector. If you lose the latter, the former should still
be usable, but it will look less modern.)

\lcont{
Note that Halibut does not require any external software such as a
\i{Help compiler}. It \e{directly} generates Windows Help files, and
therefore it doesn't need to be run on Windows to do so: it can
generate them even when run from an automated script on a Unix
machine.
}

\b \c{output.1} will be a Unix \i{\cw{man} page}.

\b The set of files \c{*.html} will contain an \i{HTML} version of
the document. If you have configured Halibut to generate more than
one HTML file (the default), then the file \c{Contents.html} will be
the topmost one that users should be directed to initially. If you
have configured Halibut to generate a single file, it will be called
\c{Manual.html}.

\b \c{output.info}, and some additional files \c{output.info-1},
\c{output.info-2} etc., will be files suitable for use with GNU
\c{info}.

\b \c{output.pdf} will be a printable PDF manual.

\b \c{output.ps} will be a printable PostScript manual.

\H{running-options} \ii{Command-line options}

Halibut supports command-line options in case you don't want to use
all of Halibut's \i{output formats}, or you want to configure the
names of your \i{output files}, or you want to supply additional
configuration on the command line. The supported options are listed
here.

Firstly, there are options which indicate which of the output
formats you want Halibut to generate:

\dt \i\cw{--text}[\cw{=}\e{filename}]

\dd Specifies that you want to generate plain text output. You can
optionally specify a file name (e.g. \c{--text=myfile.txt}), in
which case Halibut will change the name of the output file as well.

\dt \i\cw{--html}[\cw{=}\e{filename}]

\dd Specifies that you want to generate HTML output. You can
optionally specify a file name (e.g. \c{--html=myfile.html}), in
which case Halibut will change the name of the output file as well.
Specifying a file name here will also cause the HTML to be output in
\e{only} one file, instead of the usual behaviour of producing
multiple files with links between them. If you want to produce
multiple files and configure their names, you will need to use the
more complete file name configuration directives given in
\k{output-html-file} (although you may want to do so on the command
line, using the \c{-C} option).

\dt \i\cw{--xhtml}[\cw{=}\e{filename}]

\dd Synonym for \c{--html}.

\dt \i\cw{--winhelp}[\cw{=}\e{filename}]

\dd Specifies that you want to generate Windows Help output. You can
optionally specify a file name (e.g. \c{--winhelp=myfile.hlp}), in
which case Halibut will change the name of the output file as well.

\lcont{

Your output file name should end with \c{.hlp}; if it doesn't,
Halibut will append it. Halibut will also generate a contents file
(ending in \c{.cnt}) alongside the file name you specify.

}

\dt \i\cw{--whlp}[\cw{=}\e{filename}]

\dd Synonym for \c{--winhelp}.

\dt \i\cw{--hlp}[\cw{=}\e{filename}]

\dd Synonym for \c{--winhelp}.

\dt \i\cw{--man}[\cw{=}\e{filename}]

\dd Specifies that you want to generate \cw{man} page output. You
can optionally specify a file name (e.g. \c{--man=myfile.5}), in
which case Halibut will change the name of the output file as well.

\dt \i\cw{--info}[\cw{=}\e{filename}]

\dd Specifies that you want to generate GNU \c{info} output. You can
optionally specify a file name (e.g. \c{--info=myfile.info}), in
which case Halibut will change the name of the output file as well.

\lcont{

Unless the \c{info} output format is configured not to (see
\k{output-info}), Halibut will divide the \c{info} output into many
small files. The extra files will have numeric suffixes on their
names; so, for example, \c{output.info} might be accompanied by
additional files \c{output.info-1}, \c{output.info-2} and so on.

}

\dt \i\cw{--pdf}[\cw{=}\e{filename}]

\dd Specifies that you want to generate PDF output. You
can optionally specify a file name (e.g. \c{--pdf=myfile.pdf}), in
which case Halibut will change the name of the output file as well.

\dt \i\cw{--ps}[\cw{=}\e{filename}]

\dd Specifies that you want to generate PostScript output. You
can optionally specify a file name (e.g. \c{--ps=myfile.ps}), in
which case Halibut will change the name of the output file as well.

If you do not specify any of the above options, Halibut will simply
produce \e{all} of its output formats.

Also, there is an option which allows you to specify an arbitrary
\i\c{\\cfg} configuration directive (see \k{input-config}):

\dt \i\cw{-C}\e{config-directive}\cw{:}\e{value}[\cw{:}\e{value}...]

\dd The text following \c{-C} is expected to be a colon-separated
list of strings. (If you need a literal colon, you can escape it
with a backslash: \c{\\:}. If you need a literal \e{backslash}, you
can do the same: \c{\\\\}.) These strings are used as the parts of a
\c{\\cfg} directive. So, for example, the option

\lcont{

\c -Ctext-section-align:2:leftplus

will translate into the configuration directive

\c \cfg{text-section-align}{2}{leftplus}

(Note that your shell may also take an interest in backslashes,
particularly under Unix. You may find that the backslash with which
you escape a colon must be doubled in order to make the shell pass
it to Halibut at all, and to pass a doubled backslash to Halibut you
might have to type four backslashes on your shell command line. This
is not part of Halibut's own behaviour, and it cannot do anything
about it.)

Configuration directives created in this way take effect after all
other input has been processed. (In most cases, this has the effect of
overriding any other instances of the directive in the input.)

}

The options which set the output file names actually work by
implicitly generating these configuration directives. When you
specify \c{--text=myfile.txt}, for example, Halibut treats it
identically to \c{--text -Ctext-filename:myfile.txt}. The Windows
Help and man page formats work similarly. HTML is slightly
different, since it also arranges for single-file output if you pass
a filename to \c{--html}; so \c{--html=myfile.html} is equivalent to
\c{--html -Chtml-single-filename:myfile.html -Chtml-leaf-level:0}.
(See \k{output} for explanations of all these configuration
directives.)

In addition to these, there are also a few other options:

\dt \i\cw{--input-charset}\cw{=}\e{charset}

\dd Changes the default assumed \i{character set} for all input files from
ASCII to something else. (\cw{-Cinput-charset} cannot be used for
this, as \cw{-C} directives are processed after all other input, so
wouldn't affect any files.)

\lcont{

Any \cw{\\cfg\{input-charset\}} directives within input files override
this option.

See \k{input-config} for more information about the input character set.

}

\dt \I{character sets, enumerating}\i\cw{--list-charsets}

\dd List character sets known to Halibut.

\dt \i\cw{--help}

\dd Print a brief help message and exit immediately. (Don't confuse
this with \c{--winhelp}!)

\dt \i\cw{--version}

\dd Print information about Halibut's version number and exit
immediately.

\dt \i\cw{--licence}

\dd Display Halibut's licence (see also \k{licence}) and exit
immediately.

\dt \cw{--license}

\dd US English alternative spelling of \c{--licence}.

\dt \i\cw{--precise}

\dd Report column numbers as well as line numbers when reporting
errors in the Halibut input files.