Overview of generation of the WCSLIB manual using doxygen
---------------------------------------------------------
Documentation is extracted from the C header files using a flex scanner,
doxextr (see ./doxextr.l).  doxextr relies on strict adherence to a set
of formatting rules in order to minimize the amount of doxygen markup
required in the C source files.

doxextr extracts information from the header files and applies doxygen
markup to it.  This is passed through a generic sed filter, doxexr.sed,
and possibly also a specific sed filter, e.g. prj.sed for prj.h, which
applies additional doxygen markup, mainly for fonts and formulae.  The
makefile rule that drives this process writes the output to a set of
files with names of the form *.dox, e.g. prj.dox.

Additional doxygen markup may also be provided via files in this
directory with names of the form *_extras.dox, e.g. prj_extras.dox.
These are maintained manually.  Similarly, the main page and explanatory
sections of the manual are maintained in mainpage.dox.

The makefile rule concatenates ../C/wcslib.h and all of the .dox files
in this directory onto ./wcslib.h for input to doxygen which is told
via its configuration file, Doxyfile, to ignore ../C/wcslib.h.  This
subterfuge is necessary to avoid spurious file references in the
generated latex manual.


Author: Mark Calabretta, Australia Telescope National Facility, CSIRO.
http://www.atnf.csiro.au/people/Mark.Calabretta
$Id: README,v 6.2 2018/10/20 10:03:14 mcalabre Exp $
