Table of Contents * Previous Chapter * Next Chapter
This chapter describes how to run miftran.
In order to run miftran, you will need two things (other than miftran itself):
You make your MIF file from Frame using the Save As command. You create your rc
file using your favorite text editor.
Miftran is designed to accept a single MIF file as input, but it can produce multiple output files. Thus you can create a separate output file for each chapter of your document
(or each section, or whatever else you select). This is appropriate for hypertext browsers,
where each chunk is typically considerably smaller than what goes into a MIF file (such
as this entire document).
If your document is saved in multiple MIF files, you can concatenate them all together
into one bug MIF file and run that through miftran. This may or may not work, depending on your MIF files. You'll just have to try it to find out.
The simplest way to use miftran is to create a directory in which to produce the output
files, and to place your rc file into that directory. You can put a simple makefile into the
directory so that you don't have to remember how to run miftran; then you can simply cd
into that directory and type `make' in order to update your HTML files.
Miftran includes only a few
command line options, mainly for debugging. The
debugging command line options (-cmd, -fmt, -lex, -sub, -tran) allow you to control which
phases of the translation are performed by miftran (see
Section 1.1, `The Translation
Process' on page 2). Other options are:
An additional argument not starting with `-' is assumed to be an input filename. A single `-' specifies that input is to come from stdin. Input will also be taken from stdin if no
input file is specified (either on the command line or in the rc file).
There are no command line options to specify output files. The output file or files are
normally specified in the rc file (see
Chapter 3, `The RC File' on page 6), and are processed during the fmt stage of the translation (see
Section 1.1, `The Translation Process'
on page 2). Thus if you use one of the debugging command line options to get early output, it will always come to stdout.
Miftran uses the following
environment variables:
Miftran makes a single pass through the document and produces one or more output
files. In order to properly handle forward cross references, a second pass of some sort
needs to be made. This is handled by an external script, called fixref, which uses a data
file produced by miftran in order to update the cross references in the HTML files. Control over production of the data file is contained in the miftran rc file.
The fixref script is normally included as part of the make procedure as defined in the
Makefile for the HTML documents.
Miftran comes with a set of rc files that can be included by reference in your rc file. This
makes it simpler to create your customized rc file, and makes it easier to incorporate bug
fixes and enhancements to the rc files when a new version of miftran is released.
The standard miftran include directory is
miftran/mtinc, where
miftran
is the directory in which the miftran kit has been installed. You can reference this directory in the
miftran command line by using the -I command line option. Then you can use the
<include> command in your rc file to reference any of the standard rc files from the
include directory.
Table of Contents
* Previous Chapter
* Next Chapter
2.1 Requirements
2.2 Command Line Options
2.3 Environment
2.4 Postprocessing
2.5 The Include Directory