Names version 2.0
Copyright 1994 by Michael Harvey
Permission granted for unrestricted noncommercial use.


 ---------------------------------------------------------------------

This is a complete rewrite of Names.  It has been simplified considerably.
All code is now ANSI C, and should be portable to almost anything.  The
DOS console I/O is also gone, replaced by ANSI sequences.  The generator
is more flexible now.  Options can be specified on the command line, and
multiple data files can be combined.  Captilization, pluralization, and
"fancy" names are smarter now.  Finally, it now offers "forever" mode for
use as input to a pipe.

 ----------------------------------------------------------------------

USAGE: names [-acCdFhlpsv] [-f outfile] [infile...]
  -a       suppress ansi control codes

		Normally, names uses ANSI sequences to clear the screen
		and position the cursor.  Turn this off if your terminal
		doesn't understand ANSI codes.  When off, text scrolls.

  -c       suppress capitalization

		Names normally capitalizes the first letter of each name
		part (except Middles and Suffixes).

  -C       force capitalization

		This forces names to capitalize every word in the output.

  -d       debug (-D full debug)

		There are several debug levels (not all implemented).
		Each -d increments the debug level by one; "-d -d" would
		specify debug level 2.  The -D option forces level 99.
		Debugging automatically disables ANSI.

  -f file  specify output file, default "names.out"

		Specifies the output file to which names are saved
		in interactive mode.  If no name is specified the
		file defaults to "names.out".  If the file does not
		exist it will be created.  (It will not be created
		unless something is actually saved).  If the file
		exists already, new names will be appended; it will
		not be overwritten.

  -F       forever (default when reading from pipe)

		Forces a non-interactive mode in which names are
		generated forever in an endless loop.  This would be
		useful for feeding a continuous stream of names to
		another program, for instance.

  -h       help

		Displays usage info.

  -l       convert all data to lowercase

		Converts all input data to lowercase as it is read.

  -p       suppress pluralization

  -s       simple output

		Suppresses "fancy" mode.  When names has two nouns,
		for example, "troll" and "rock", it might produce
		output such as "Trollrock", "Trolls of the Rock",
		"Rock Trolls", 'Rock Troll Inn", etc.  Simple mode
		would cause it to simple produce "Trollrock".

		This option has no effect of Prefix/Middle/Suffix
		processing.

  -v       version info

  infile   input data file

		The input data file.  If no file is specified,
		names will read from standard input (and will
		automatically enter "forever" mode).  If a file is
		specified, input data will be read from it.  If
		multiple files are specified, each will be read and
		the data combined internally.  This makes it possible
		to have one file of prefixes and another of suffixes,
		for example.

		Names assumes a default file suffix of .ele on data
		files.  Given a filename 'foo', it will first attempt
		to open 'foo'; if that fails, it will attempt to open
		'foo.ele'.  If both forms fail, it will print a message
		to standard error and continue.

 -------------------------------------------------------------------

HOW IT WORKS

Names takes name-elements from a data file and combines them to
form a name, much like forming a compound word.  (If you have ever
read the appendix to the Lord of the Rings, Tolkein used the same
method to form his elvish names).

 -------------------------------------------------------------------

FILE FORMAT

Data files can have up to six different sections.  Each section
specifies are particular type of name-element.  Sections begin
with "#sec" on a line by itself, where "sec" is the abbreviation
for the section type.  Valid types are:

#PRE	Prefixes
#MID	Middles
#SUF	Suffixes
#NOUN	Nouns
#ADJ	Adjectives
#NADJ	Noun/Adjectives

#sec specifiers are not case sensitive.  If an unknown section is
encountered, it will be ignored.

There are two different forms names uses to generate output.  The
first is prefix/suffix mode.  In this method Names will concatentate
a prefix, and optional middle, and a suffix to produce a word.
The #mid section is optional; #pre and #suf are required.

The second is noun/adjective mode.  In "simple" mode (the -s option)
names will simple concatenate them into a single word.  In "fancy"
mode (the default) names will combine the two in a variety of ways.
Noun/adjective mode requires at the minimum either a #noun or #nadj
section.

#NADJ is a special section containing elements which can be used either
as a noun or an adjective.  Listing an element here is equivalent to
listing it in both the #NOUN and #ADJ sections.  Examples of words
which should be specified as #NADJ are "stone", "north", "dark", etc.
An easy way to determine this is to test it in adjective-noun
combinations.  For example, "dark" can be used as an adjective
(dark stone) or a noun (utter dark).  "darkish" should be an adjective
(darkish stone) as it does not work as a noun (utter darkish).  "Darkness"
is not an adjective (darkness stone) but *is* a noun (utter darkness).
There is a lot of crossover; most nouns can be used as adjectives.  By
distinguishing, however, you can force Names to use an element in a
particular way.

 --------------------------------------------------------------------

You are free to use, modify, and distribute this program without
restriction.  Just leave the copyright notice intact.

Mike Harvey
mike@cs.pdx.edu