path: root/fonts/halftone/pixkit.post

$! Archive name: DRA1:[TEX.PIX]PIXKIT.POST;
$!
$! This file is a ``DCL archive``, a VAX/VMS DCL command procedure, analogous
$! to a ``shell archive`` on UNIX systems. It is intended primarily for
$! sending a set of files across a network in a format which allows them
$! to be restored. On VMS systems, the archive is restored simply by typing:
$!
$!        $ @archive-name
$!
$! The format of the archive is as follows:
$!
$!  1. As you start reading the file, lines which start with a dollar ($)
$!     are DCL commands.
$!  2. On DCL commands, an exclaimation mark (!) marks the beginning of
$!     a comment. In the DCL archive, this only occurs as the second character
$!     on a line.
$!  3. Each file to be created is preceded by a ``CREATE`` command. The
$!     text following the word `CREATE' is the name of the file to be created.
$!     If your system allows it, you should create a new version of the file.
$!  4. Filenames consist of two parts, the name itself and the extension.
$!     Either part may contain zero to 39 characters, which must be dollar,
$!     underscore (_), hyphen (-) or alphanumeric; a hyphen may not be the
$!     first character of a name.
$!  5. The `CREATE' command is followed by a line containing `$DECK' only.
$!     Lines of text following the `$DECK' should be copied verbatim to
$!     the output file.
$!  6. The end of the output file is delimited by a line containing only
$!     `$EOD'.
$!
$! Note that lines starting with a dollar may occur within `$DECK'..`$EOD'.
$!
$! The following two lines contain the ASCII character sequence from
$! exclaination mark to the end of the upper case letters, followed by
$! the remaining characters up to #126:
$! !"#$%^&()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ
$! [\]^_`abcdefghijklmnopqrstuvwxyz{|}~
$!
$! This archive was created using DCLAR version 0.00--0
$!                          on 15-JAN-1988 14:52:26.56.
$!                          under VAX/VMS V4.5    
$! DCLAR was written by Dr. Adrian F. Clark (``Alien``), Essex University.
$!
$WRITE SYS$OUTPUT "creating BLURB.TXT" 
$CREATE BLURB.TXT
$DECK
This DCL archive contains all the required bits and pieces for preparing
pictures for plotting by TeX. When executed, the archive creates
the following files:

   o BLURB.TXT     this file
   o HALFTONE.TEX  user-level documentation (for LaTeX)
   o PICTURE.TEX   picture from TeXPIC used by HALFTONE.TEX
   o TEXPIC.FOR    the software. Note that there are two versions
                   of TEXPIC in this file, one for VMS and another
                   which is a little more portable.
   o HALFTONE.MF   Don Knuth's halftone font

If you intend to produce pictures of size >= 256 x 256, you'll need to
expand TeX's memory (see my article in TUGboat vol 8 no 3 for more
details). I can supply a TeX change file for VAX/VMS which effects this
change (and provides an editor interface) either by e-MAIL or on tape.
Contact me for more details. 

If you want to use this software, but you're not a TeX guru...find
someone who is!  He/she will be able to generate convert the Metafont
source for the halftone font (HALFTONE.MF) to a form acceptible to
TeX, and put it in the right directories for you to use.  I can supply
outline help, but the font generation process tends to be horribly
site-dependent.  Sorry.

If you have any trouble with the software, please let me know (it's not
quite the same as the version I use myself). I'm particularly keen
to receive feedback from non-VMS sites.

                           `Alien'   (ALIEN @UK.AC.ESSEX.ESE on JANET)
$EOD
$WRITE SYS$OUTPUT "creating HALFTONE.TEX" 
$CREATE HALFTONE.TEX
$DECK
% --- ------------------------------------------------------------------
% --- halftone.tex
% --- Format me with LaTeX, print me with an inch of top and left margin
% --- ------------------------------------------------------------------

\font\halftone=halftone

\def\|#1|{{\tt#1}}
\def\mmax{M_{\hbox{\scriptsize max}}}
\def\SYSDEP#1{\marginpar{\scriptsize site\\ specific}}
\def\TL{{\tt\TeX lase}}
\def\TV{{\tt\TeX view}}
\def\VAX{\leavevmode\hbox{V\kern-.12em A\kern-.1em X}}
\def\VMS{\leavevmode\hbox{V\kern-.06em MS}}

\def\mc{\small}
\def\FORTRAN{{\mc FORTRAN}}

\nofiles
\documentstyle[11pt]{article}
\advance\textheight by 21mm

\title{Halftone Output From \TeX\footnote{Taken from the {\sl \VAX/\VMS\ \TeX\
       User's Guide}}}
\author{Dr. Adrian F. Clark\\
        Department of Electronic Systems Engineering\\
        University of Essex}

\begin{document}
\maketitle

\section{Introduction}

This note describes how image data can be converted to a form suitable
for \TeX\ and then typeset in documents. Pictures are prepared for \TeX\
in a somewhat different way to normal text---the hard work is done outside
the \TeX\ input file; incorporating the image into the \TeX\ document is
then quite straightforward.

The image data must be converted from the binary representation used in Image
Processing and Pattern Recognition work to a series of characters which \TeX\
can map into pseudo grey-levels via a special font. A user-callable \FORTRAN\
\|SUBROUTINE| called \|TEXPIC| is available for this purpose. \|TEXPIC|
converts a picture held in memory to a text file containing \TeX\ input. (You
can look at this file if you like, but it is unlikely to make much sense.)
There are a few additional routines which can be used with \|TEXPIC| to produce
particular effects: for example, the picture may be plotted as a positive or
negative image. 

The special font for setting pictures is
defined to have some 65~grey levels. This does not necessarily imply that all
the grey levels are distinguishable, nor even that they are strictly monotonic
on laser printers. However, the quality of the output is certainly adequate for
run-of-the-mill technical reports. 

An example of an image typeset with \|TEXPIC| and \TeX\ is shown in
Figure~1. This is the ``girl'' picture from the image database
distributed by the Signal and Image Processing Institute of the University of
Southern California, which is widely used to demonstrate image processing
techniques. This representation is some $64 \times 64$~pixels. 

\begin{figure}
   \centering
   \fbox{\input picture\relax}
   \caption{Example of an Image Typeset using \|TEXPIC| and \TeX}
\end{figure}

\section{The \|TEXPIC| Routine}

To generate a file containing the image data in a form suitable for input to
\TeX, you simply invoke \|TEXPIC| from a program. \|TEXPIC| is called as
follows:

\begin{verbatim}
    CALL TEXPIC( ARRAY, M, N, FILE )
\end{verbatim}

\noindent where \|ARRAY| is a \|M| $\times$ \|N| \|REAL| array containing the
picture to be plotted and \|FILE| is a quoted string or \|CHARACTER| variable
containing the name of the file to receive the data. If you omit a filetype
(``extension'') from \|FILE|, \|TEXPIC| will use \|.TEX|. You can call
\|TEXPIC| as many times as you want from a program; a separate output file will
be generated for each picture.

Due to problems with the controlling micro-processor in the standard \|LN03|
printer, pictures with $\|N| > \approx 64$ will come out with white bands
across them, rendering them effectively useless. However, this problem has been
overcome on the \|LN03+| device, which can print images up to a full \|A4|
page in size. 


\section{Associated Routines}

\|TEXPIC| normally maximises the contrast of pictures by determining the range
of grey levels in the image and scaling the output to make use of them. This
behaviour can be overridden by invoking the routine \|ZRANGE| before \|TEXPIC|.
\|ZRANGE| requires two \|REAL| arguments, the minimum and maximum grey level
values to be output: 

\begin{verbatim}
   CALL ZRANGE( ZMIN, ZMAX )
   CALL TEXPIC( ARRAY, M, N, FILE )
\end{verbatim}

\noindent where \|ZMIN| $\le$ \|ZMAX|.
\|TEXPIC| then uses the supplied extrema, rather than values
determined from the image, for scaling the data. Values which lie {\em
outside\/} these extrema are clipped. Hence, if \|ZRANGE| is used with values
which are outside the values of the pixels in \|ARRAY|, the contrast of the
data generated by \|TEXPIC| is {\em reduced.\/} The converse is also true. The
extrema set up by \|ZRANGE| remain in effect until another call is made to
\|ZRANGE|, superseding the previous limits, or \|ZAUTO| is invoked, which
restores the default behaviour. 

\|TEXPIC| normally plots positive pictures---i.e.,\ pixels which contain
low values come out darker than pixels with high values. It can also produce
negatives, however; you do this by invoking the routine \|DONEG| before
\|TEXPIC|. Normal, positive picture-plotting is restored by calling \|DOPOS|.
As with \|ZRANGE| and \|ZAUTO|, \|DOPOS| and \|DONEG| remain in effect until
explicitly cancelled.
You can, of course, use the routines in conjunction to produce any
required effect.

There is a limit, $\mmax$, to the number of pixels that \|TEXPIC| can fit
across a page of output. For
pictures with $\|M| \le \mmax$, the data are written out exactly as supplied.
However, when pictures have $\|M| > \mmax$, the image is interpolated so
that the data written out form a $\mmax$-pixel image.
The same interpolation factor is used in both dimensions so that the aspect
ratio of the picture remains correct. The interpolation technique used is
a context-sensitive three-point bilinear method due to {\sc
Smith}.\footnote{There is also a version which uses standard four-point
bilinear interpolation.} This produces marginally better results in the
vicinity of edges in the picture.

The default value of $\mmax$ is~256. You can alter this (for example, when
producing pictures for a document with the text set in narrow columns) by
invoking \|TEXMAX| before \|TEXPIC|, as follows:

\begin{verbatim}
   CALL TEXMAX( MV )
\end{verbatim}

\noindent This sets $\mmax$ to \|MV|. Obviously, $\|MV| \ge 2$.

\|TEXPIC| and the support routines are all to be found in the algorithm
library, \verb|ESE$LIB:ALG|. \SYSDEP{Location of \\TEXPIC| and utility.}
There is also a stand-alone program which you can use for images stored in the
standard formats on disc. This program is \verb"ESE$PROGRAMS:V159". 


\section{Inserting the Picture into Your \TeX\ Document}

This is quite easy, although there are a few steps in the process.
The procedure you follow depends on which macro package you are using with
\TeX. Most people use ``plain \TeX,'' the default macro package, but \LaTeX\
is also in widespread use. If you use a customised \TeX\ but don't know exactly
what is different about it, try the plain \TeX\ procedure.

Whichever version of \TeX\ you use, there are a few things you should know.
\TeX\ must hold the whole image in memory, so the maximum image size is
obviously limited to \TeX's memory capacity. Since \TeX\ was designed for
typesetting rather than image hard-copy, its memory capacity isn't
large enough to hold dozens of pictures, although it should be adequate
for reasonable use.

\|TEXPIC| tells \TeX\ to use a special font, a {\em halftone\/} font, for
setting the picture. You have to tell \TeX\ to load this font. To do this,
you type the following command at the top of your document:

\begin{verbatim}                                          
   \font\halftone=halftone
\end{verbatim}

\noindent This works with both plain \TeX\ and \LaTeX.

We shall now consider the procedure for telling \TeX\ to typeset your picture.
In the examples that follow, it is assumed that the picture produced by
\|TEXPIC| is in the file \|PICTURE.TEX|.

\subsection{Procedure for Use with Plain \TeX}
The first thing you must do is to create an input file for \TeX\ which contains
the document to be typeset. At the point where you want the picture to
appear, you type:

\begin{verbatim}
   \centerline{\input picture}
\end{verbatim}

\noindent For \TeX{}perts, the image is set as a single \verb|\hbox|.
  
The picture can be put into a ``floating'' insert, which will cause the
picture to be held in \TeX's memory until there is enough free space on
the page to hold it. For example, to set the picture in a \verb"\midinsert",
the relevant \TeX\ input would be:

\begin{verbatim}
   \midinsert
      \centerline{\input picture}
   \endinsert
\end{verbatim}

\noindent This command sequence should be typed between paragraphs,
when \TeX\ is in `vertical mode.' Typing \verb"\goodbreak" immediately
before the \verb"\midinsert" may help \TeX's page-breaking mechanism. (In
practise, you'll probably want to add a title to the picture, too.)

There is one other thing you might want to do and that is to draw a border
around the picture. You would do this by defining a \TeX\ macro, \|border|,
as follows:

\begin{verbatim}
   \def\border#1{\vbox{\hrule\hbox{\vrule\kern3pt
      \vbox{\kern3pt#1\kern3pt}\kern3pt\vrule}\hrule}}
\end{verbatim}

\noindent You would then set the picture with

\begin{verbatim}
   \centerline{\border{\input picture}}
\end{verbatim}

\noindent instead of simply \verb"\centerline{\input picture}".

\subsection{Procedure for Use with \LaTeX}
The most sensible way to plot pictures with \LaTeX\ is in the \|figure|
environment ({\em not\/} the \|picture| environment). This generates
a ``floating'' figure, which usually surfaces at the top of the following
page of output. The \verb|\centering| declaration causes the picture to
be centred in the page.

\begin{verbatim}
   \begin{figure}
      \centering
      \mbox{\input picture\relax}
      \caption{Figure Title}
   \end{figure}
\end{verbatim}

\noindent The \verb|\relax| following the filename in the \verb|\mbox| command
is required for \LaTeX\ to know where the filename ends. To draw a border
around your picture, you simply replace the \verb|\mbox| command with
\verb|\fbox|. 

\section{Producing the Output}

Having inserted the appropriate commands into your document, you run \TeX\ in
the normal way. If it exits with a ``memory capacity exceeded'' error message,
you've either tried to set images which are too big or you're holding too
many in memory at once.  The only solution to this problem is to re-format your
document. \LaTeX\ users may be able to cure the problem by strategically
inserting \verb"\clearpage" commands into the manuscript.

You run \TV\ or \TL\ on the \TeX\ output file as normal, although the \TV\
display bears little relationship to the grey levels produced by \TL. Note that
you must use \TL\ version~10 to send output to the \|LN03|, since earlier
versions cannot read the halftone font file. (When you invoke \TL\, it outputs
a short introductory message which includes its version number.) 

The current version of \TeX\ 2.0 has a larger-than-usual memory capacity,
\SYSDEP{Size of \TeX's memory.}
large enough for four $256 \times 256$ images to be held in memory.

\end{document}
$EOD
$WRITE SYS$OUTPUT "creating PICTURE.TEX" 
$CREATE PICTURE.TEX
$DECK
 \hbox{\vbox{\halftone\offinterlineskip % machine-generated by TEXPIC.
 \def\BHT{\hbox\bgroup\ignorespaces}
 \catcode`\^=12 \catcode`\_=12 \catcode`\.=\active \let.=\egroup
 \catcode`\,=\active \let,=\BHT \catcode`\/=0 \catcode`\\=12
 ,ccdggfcbabbcfeeddeccehhlmkkklljdejceggiagjfgbecbbadiiXNQLEHDGHJL.
 ,dcdggfbbbaadeddcddbdfghkmkkkklljjicfhiiaiiggbddac`fkcQLQLEGGGHKL.
 ,ecbfffcbbabdeedecddfgfglmkkkjkllmgc`ghgcjhfdbebaaahk\NLQLGGHGKMN.
 ,cccfgfdaaaadddddceffdcekmkkkj_PI?GME;FNckfgfbbb`_djgVLLQMGIIGJNM.
 ,dddggfbbbbbeedcdcedcccdjmkkk\GHEMHEJH><86Fcfdcc`_glaQOMPLHIHFJNL.
 ,ddcggfcbaabdedddcdbbbbdjmkfHH?GQLQNLNNLB<@=NedbaaikZMLMQKGJHGJNM.
 ,dbdffebaabbddeddddccabbjlgWSLLFKIRQLJMTKMHKDKda_dkhUILNQJIKJILPN.
 ,dbeggfdbbbbdfddddcccbbchdJFGISKTMPWRQLWYVOKGDBaahlaTRQUWOKMRNQUT.
 ,ebdggecbaabdddcedccbcbaUHAIKNPTUNROORQYXYUWVPIOdij[TWWYZTSSWWY\\.
 ,edegfebabbadeedddccbbbRRVOJLHCJOOPSVT^WY]XTTTLDakbQKONNRMOST[\^\.
 ,dcdfgeabcbaceddcdbccbZ[VWYJJMWMTRLS]]bZ_a`Z^QTQ9l^HHMNONJRYYYZ\\.
 ,ecdgffcaaabdecdcdccb_d\YVQUWQ^a_Yf`^^lkha^_^SLWB\YHHLNONGPXXVWZ[.
 ,ebdgfebaabbdeddccbcd^bYZQ[W]VZbjaijgdgonhhiiWJJEWOJJLNOLHQXVVTY\.
 ,ebegffcbabbdecdbcbba^X`a\WC`hjdbekleknnopjkifXGRRMHIMLPNIQVVWUX\.
 ,ecdgffbaaabdedcdbba^_ef_`iegcfglnneihlnonmlii\KQULGHKJOONUTVUTY[.
 ,fcdggebbaacdddcddbb^_gbhllfiknhnoinlmlllmmlih`Q]SKILJLPQTXTUVTVZ.
 ,ecdgfecbbabdedddddba\ehjljiklmlmnnmilkghcjgicb\YUJJJMRYTQYXUVWWZ.
 ,dcdgfeab`bbedddcdcb\a]flmjijldecnommjf]jcTf\bdbbELLNSU^TPWYWWWWX.
 ,fcfggeababbedddcdbc]^g`gm_fa]hgaemkgjhgj_UhIb^f^SMJQXY]TQYXZYVYZ.
 ,fbeffebabbbdecdcdcbb`ZSbg]b\Ya_b^gjijgjibajWdY]ZLNJRXZ]SNY][YWW[.
 ,fbeffdcbaabdddddcbabhgi_b`]^bbhdb`dgfdjaagjjbV[ZLMLWXV\QR[^^\XY[.
 ,ebdffdcabbbededdccc`mlnmjecfihhijjikkdj^efije_]_NNPWXYZRU^]^[WUZ.
 ,fcegfdbbbbcdddddccbbinonnjhgkinkmnlnmiidegfkggjZVWVYY\\UW]^\ZVU\.
 ,ebdffecbacceeeddcccbginnijjjkkkjklnnnijjiihmjjfdRUV^^^aY\_^]]][^.
 ,gdeggdbbbbcdededddbcggYNVWX\]ii_`b_cmljlkkhkibi\NKQYY\_SW_ZZ^[^].
 ,fdegfebabbceeeeddddcdh]diZYZYYfjgacllihnkmjbifhSNKNTY_^RV^^^^]^].
 ,fbfggeccdceeddddeddcceghggaZ\be_fefjljhjjmnkjegQKKIOY^_RT_^^^\\\.
 ,fcegffdbcbdeddedeecbchnZm]dNVgggjmpohfeeiglljkZNJHFPW^^RT^\^YY[^.
 ,fcfhfdbccbcfeeeddddcdcZgibUO[]RcWelifcegijigflLLJIIOU\_VT_]\ZY\^.
 ,edfgfebccddeeeedeeedcbTYXRUP^\XTbdd\`a_ggekebPQLLJHMOX^VW][]\YY\.
 ,edfgfedcdcdfeeffeededdTQOOVU_\WOQTW[_afkflhaSOQJLMKNRZ^UY[[_^YV].
 ,fdfgfeddccdeefeeeeefbeTOKYSY`aWJQVY]afeklihbSNMJJKLMT\^UY]\_b\Z^.
 ,fdgfgfdccddfeffdfffeedTTTaQZbd^TSUW]bg`_fgfQUPLLLJKOQX]UX]\\]Z[^.
 ,fdgggedccdcfefeefededc\Zaa_jij^^XY\`de^[X\VPRORPLKJLNU\VX^[\[VY^.
 ,fdfgffcdcbdfedefefeeee_bg\W^Z[[_ab_ciadXUNVVYWWUQLJMOW\TY\YZZYZ_.
 ,fdfggfdcecdeeefeeefedfebaidVY\]abbdfidaZXNWTWVVTQJKNOXXTZ[Z\]ZY\.
 ,fdfggedddddffeefeedeejj_[Y`KHYlok`degfa]`QYWYWVTSNORW\\T]]]a^WVZ.
 ,fcgggfddddeffeeffeeeegie^YaYV]ecZ]bfgda_`PZ[ZZ[WUQNRV^[Z_bb`]ZZ^.
 ,fdghgfededffffffeeedc[jh`\]]_acaa_fggcfhaXURTVYUMLKMR\_Y_cba]Z]a.
 ,ffhhhgedddfggffffgfee_cil]UTQWX]^bfhhjhi`XRSRPURKHJKQ[\UZ]]_\[\\.
 ,gfgjihfeffgiggfgggfegnlimo[VUY\abdhjkiiiaMQTUPTSJILPRVWRTVUWVSRR.
 ,ggjjjiggfgfhhhggggffjljjmojbabdgjllljhjj_GRUVQURLJPURXWRQSRQQPPQ.
 ,fikjjjhgfgfghghghhfhmnjgjnfgZlnnmlkjiiiiaRGPWQSRKKVWV[ZRTVSURSRQ.
 ,chkjjjggggffghghhgfjnnmlmdliiY^lljjjjiiib`d[^_TQKNWWV][RWVVVUTSS.
 ,]`gkkjihghhhjjjjjjijmnoldfijhkXY\gijid_]fjihffe`QMWUV[ZQUVTSTTRT.
 ,\^]ejllkkjjjllmmlmlllnfcdfghijk\]][\\[gkmhjhhgfda]`^X_]TYWVYYUUW.
 ,ab[\cegihgffhjjkkkkibecddfhfljjofgfhikiiiljigfiijkklg\[UZYY]]ZYZ.
 ,d```a``cfeffijjjj`^edegdfgjfkijnogfhhhgiihijjjklmnoobnYQWYY\[ZYV.
 ,hifiie__egggijj]cbfeecegfflgfnjinofggiiihhhfgiljjndhdhbRXZZ][XXX.
 ,hklllic]ageeffNYcfgfgffggfkifijihmjhhhiiiijjjijlmmmiZQESY\^][[[[.
 ,hghlmkib^cdeYY^dchhffgeggfhmhiiignjihihhijijjkllnoni`YXFDYZ\WWYZ.
 ,a]bkmkigdjmc^`bbdigfggfhgggoiijjkhljjhijijjjklmnonheddaXRHK]YUXY.
 ,c^`knihighl[__adgjhffghiiiiimjjikkjjiikikkkllmooihhgfgec[XOHc_ac.
 ,gbbjnllmkjk[^_baiiggehiijiiholnomlklknmikklmmoohgghgfiifea_[W^^_.
 ,kghllnooona^___gcliijkddfhiijommihnnmnhipomnoniggghhhhhijjced]ZZ.
 ,nonmlmopppd``a`lgakifhghijjjjjllmjlklkkjanlkiggihiiiiiihlkiZha]Z.
 ,pponlloppo`a`bdllkgfddgijgebblkmilimmkUACG8@iiihghiiihhnmiifWg\W.
 ,pponllmopmabacdlllkkkkjie^bcbfajominXlk`=L=3BEbfgiihhgjnjijj`W[U.
 ,ponmlllnolbbbadkmllkklkfdhbhdghimmkiaO?HO>G8?CELhjihghkoiklje[YY.
 ,onnmlmmmocdcbacjlmmlmmlcgiijgefkjopooonNA>7=BBDC<jhghlooillki`QX.
 ,nmnmnmmmobcbbbbjlmmmmjjkljlgiielilio\[IZWPRS?D4BDWghilokilllkg\[.
 ,mmlkllklneddbbbjmmmllfhjkillkgjjb\TYpKTVO;JJ<7I;3FkRmoohlnnnlkcT.
 ,lllllllnnc`dbcbblmmmmlgijiggfhhkR@TCYI[HghUWS?@DCFBBEA?mnoonmkf\.
 ,pponmmnnjb\kjfdekklkkijkjijhiigiILAP<LIL^VNSHGKLBABJ>BGbooonnmia.
 }}%
$EOD
$WRITE SYS$OUTPUT "creating TEXPIC.FOR" 
$CREATE TEXPIC.FOR
$DECK
C --- -----------------------------------------------------------------
C ---                          T e X P I C
C --- -----------------------------------------------------------------
C
C --- These routines are used to generate an input file to TeX which,
C --- when processed, gives a halftone representation of the grey-level
C --- image. TEXPIC is the main picture-plotting routine: it requires a
C --- M x N pixel array ARRAY. The TeX commands are written to the file
C --- FN. These routines are taken from a general-purpose library of
C --- image processing software developed by the author, which explains
C --- why most of the variables in the COMMON blocks are not used. The
C --- Fortran is also machine-generated, which may account for some odd
C --- line breaks in the code. (No line, even in comments, has more
C --- than 72 characters in it.)
C
C --- The image representation used here conforms to the one adopted by
C --- the Numerical Algorithms Group (NAG) for their ``Image Processing
C --- Algorithm Library'' IPAL, although the coding does not  (it  does
C --- not allow a sub-region to be plotted).
C
C --- There are two versions of TEXPIC in this file: the first is VAX-
C --- specific, while the second should be fairly portable. Note that
C --- both routines declare the BLOCK DATA module ALGINI as EXTERNAL;
C --- this usually forces the linker to build it into executable files.
C
C --- TEXPIC's support routines are:
C
C ---    TEXMAX  set the maximum pixel width across the page
C ---    ZRANGE  fix the contrast for subsequent TEXPIC calls
C ---    ZAUTO   subsequent pictures have their contrast determined
C ---            from the data
C ---    ZSAME   subsequent pictures are plotted with the same
C ---            contrast as the previous one
C ---    DOPOS   subsequent pix have low pixel values plotted black
C ---    DONEG   subsequent pix have low pixel values plotted white
C ---    MINMAX  determines the range of the data
C ---    ALGERR  outputs error messages
C ---    ABANDN  VAX-specific ^C trap routine
C ---    ALGINI  block data module
C
C --- Details of the invocations are given in the comments associated
C --- with each routine. There is also a separate document which gives
C --- user-level documentation and examples. This is available as part
C --- of the ``VAX/VMS TeX User's Guide'', written by the author, or as
C --- a separate document.
C
C --- As supplied, TEXPIC uses a three-point contextual bilinear method
C --- to interpolate between pixels. The results it produces should be
C --- marginally better than using standard four-point interpolation;
C --- however, the author can detect no difference. If you'd prefer to
C --- use four-point interpolation, the line to change is marked in the
C --- TEXPIC source code.
C
C --- Since you get TEXPIC free of charge, there is no formal guarantee
C --- given by Essex University OR the author that the software works
C --- or that the documentation agrees with the code. Nevertheless, the
C --- author would be pleased to hear of any problems.
C
C --- TEXPIC and associated routines were written by:
C
C ---    Dr. Adrian F. Clark  (``Alien'')
C --- of Department of Electronic Systems Engineering
C ---    University of Essex
C ---    Wivenhoe Park
C ---    Colchester
C ---    Essex C04 3SQ
C ---    United Kingdom
C ---    Tel: Colchester (0206) 872432 (direct)
C ---    JANET: user ALIEN @UK.AC.ESSEX.ESE
C
C --- If you write, please mark the envelope with ``TeX''.
C
C --- Acknowledgements in any published work that uses TEXPIC would be
C --- appreciated.
C
C --- ENJOY!
C
C --- -----------------------------------------------------------------
      SUBROUTINE TEXPIC( ARRAY, M, N, FN )
C --- -----------------------------------------------------------------
C
C --- TEXPIC version 0.1 was written by Alien in Fortran-77.
C
C --- This  routine  writes out the M x N image ARRAY into the file FN
C --- in a form which is suitable for insertion into a  TeX  document.
C --- If FN has no filetype (``extension''), .TEX is used.
C
C ---    By  default,  the range of the data is determined and used to
C --- maximise  the  contrast  of  the  output  image.  This  can   be
C --- overridden  by  pre-setting the range of data values with a call
C --- to ZRANGE. ZAUTO  restores  the  default  behaviour.  Similarly,
C --- TEXPIC  will  produce  negated  images  on  output  if DONEG has
C --- previously been invoked.  DOPOS  sets  it  to  produce  positive
C --- pictures again.
C
C --- USAGE:    CALL TEXPIC( ARRAY, M, N, FN )
C
C --- PARAMETERS
C ---     ARRAY  REAL           image to be output to the file
C ---     M      INTEGER        first dimension of ARRAY
C ---     N      INTEGER        second dimension of ARRAY
C ---     FN     CHARACTER*(*)  name  of  file  to  which ARRAY will
C ---                           be written
C
C --- RESTRICTIONS
C --- If N is greater than MMAX, the image will be sub-sampled in both
C --- directions  to  make  the result  MMAX x MMAX. The interpolation
C --- technique used is due to P.R. Smith (Ultramicroscopy vol  6,  pp
C --- 201--204, 1981).
C
C --- COMMONS
C --- /ALG/, /ALGTEX/
C
C --- SUBPROGRAMS INVOKED
C --- MINMAX, LIB$GET_LUN, LIB$FREE_LUN
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
      INTEGER LEVELS, MINIDX, CMAX
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
      PARAMETER( LEVELS=65, MINIDX=48, CMAX=132 )
C
      CHARACTER*(*) FN
      INTEGER M, N
      REAL ARRAY(M,N)
C
      CHARACTER*(CMAX) C
      CHARACTER*6 RUTNAM
      INTEGER NMAX, I, J, IC, IV, LUN, IOS, ILO, JLO, IHI, JHI
      INTEGER LIB$GET_LUN, LIB$FREE_LUN
      LOGICAL POS
      REAL RANGE, INC
      REAL X, Y, DX, DY, DX1, DY1, VAL
C
      INTEGER*2 CHAN
      LOGICAL ABFLAG
      INTEGER GLUN, PLUN, ELUN, POLHIS
      LOGICAL XFIX, YFIX, ZFIX, RFIX, TFIX, LOGX, LOGY, LOGZ, LOGR
      LOGICAL VRBOSE, NEG, MIDORG, FTNRML
      REAL XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX, TMIN, TMAX,
     & FTFWD, FTREV
      REAL TOL
      INTEGER MMAX
C
      COMMON /ALG_ABANDN/ ABFLAG, CHAN
      SAVE /ALG_ABANDN/
      COMMON /ALG/ XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX,
     & TMIN, TMAX, TOL, FTFWD, FTREV, XFIX, YFIX, ZFIX, RFIX, TFIX,
     & LOGX, LOGY, LOGZ, LOGR, NEG, GLUN, PLUN, ELUN, POLHIS, MIDORG,
     & FTNRML, VRBOSE
      SAVE /ALG/
      COMMON/ALGTEX/ MMAX
      SAVE /ALGTEX/
      EXTERNAL ALGINI
      DATA RUTNAM/'TEXPIC'/
C
C --- Find the range of the data if we're in auto mode; otherwise,
C --- use the specified ranges.
C
      IF( .NOT. ZFIX ) THEN
         CALL MINMAX( ARRAY, M, N, ZMIN, ZMAX )
         IF( ZMIN .EQ. ZMAX ) ZMAX = ZMIN + 1
      END IF
C
C --- We can't print more than MMAX columns across the TeX output.
C --- If the user passes an array bigger than this, we'll interpolate
C --- it down to MMAX.
C
      IF( M .GT. MMAX ) THEN
         INC = FLOAT(M) / FLOAT(MMAX)
         NMAX = NINT( FLOAT(N) / INC )
      ELSE
         INC = 0
      END IF
C
C --- Get a free channel number and open the output file.
C
      IOS = LIB$GET_LUN( LUN )
      IF( .NOT. IOS ) CALL EXIT( IOS )
C
      OPEN( UNIT=LUN, FILE=FN, STATUS='NEW', RECL=CMAX+1, IOSTAT=IOS,
     & DEFAULTFILE='.TEX', CARRIAGECONTROL='LIST' )
      IF( IOS .EQ. 0 ) THEN
C
C --- Calculate the scaling factor.
C
         POS = .NOT. NEG
         RANGE = FLOAT(LEVELS-1) / (ZMAX-ZMIN)
C
C --- Output the introduction.
C
         WRITE( LUN, 100 )
C
C --- Output the image without interpolation if INC is zero.
C
         IF( ABS(INC) .LT. TOL ) THEN
            DO 2 J = 1, N
      IF( ABFLAG ) GO TO 5
               IC = 1
               C(1:1) = ','
               DO 1 I = 1, M
                  VAL = ARRAY(I,J)
               IF( VAL .LT. ZMIN ) VAL = ZMIN
               IF( VAL .GT. ZMAX ) VAL = ZMAX
               IV = NINT((VAL-ZMIN) * RANGE)
               IF( POS ) IV = (LEVELS-1) - IV
               IC = IC + 1
               C(IC:IC) = CHAR( IV + MINIDX )
               IF( IC .GE. CMAX-1 ) THEN
                  WRITE( LUN, 200 ) C(1:IC)
                  IC = 1
                  C(1:1) = ' '
               END IF
    1          CONTINUE
               IF( IC .GT. 0 ) WRITE(LUN, 300) C(1:IC)
    2       CONTINUE
         ELSE
C
C --- Interpolate the output.
C
            Y = 1
            DO 4 J = 1, NMAX
      IF( ABFLAG ) GO TO 5
               DY = Y - INT(Y)
               DY1 = 1 - DY
               JLO = MOD( INT(Y-1), N ) + 1
               JHI = MOD( JLO, N ) + 1
               X = 1
               IC = 1
               C(1:1) = ','
               DO 3 I = 1, MMAX
                  DX = X - INT(X)
                  DX1 = 1 - DX
                  ILO = MOD( INT(X)-1, M ) + 1
                  IHI = MOD( ILO, M ) + 1
C
C --- Smith's three-point contextual bilinear interpolation.
C
                  IF( ABS(ARRAY(ILO,JLO)-ARRAY(IHI,JHI)) .GT.
     &               ABS(ARRAY(IHI,JLO)-ARRAY(ILO,JHI)) ) THEN
                     VAL = (DX-DY)*ARRAY(IHI,JLO) + DX1*ARRAY(ILO,JLO) +
     &                  DY*ARRAY(IHI,JHI)
                  ELSE
                     VAL = (DX1-DY)*ARRAY(ILO,JLO) + DX*ARRAY(IHI,JLO) +
     &                  DY*ARRAY(ILO,JHI)
                  END IF
                  X = X + INC
               IF( VAL .LT. ZMIN ) VAL = ZMIN
               IF( VAL .GT. ZMAX ) VAL = ZMAX
               IV = NINT((VAL-ZMIN) * RANGE)
               IF( POS ) IV = (LEVELS-1) - IV
               IC = IC + 1
               C(IC:IC) = CHAR( IV + MINIDX )
               IF( IC .GE. CMAX-1 ) THEN
                  WRITE( LUN, 200 ) C(1:IC)
                  IC = 1
                  C(1:1) = ' '
               END IF
    3          CONTINUE
               IF( IC .GT. 0 ) WRITE(LUN, 300) C(1:IC)
C
               Y = Y + INC
    4       CONTINUE
         END IF
C
C --- Close off the file.
C
    5    CONTINUE
         WRITE( LUN, 400 )
         CLOSE( UNIT=LUN )
      ELSE
         CALL ALGERR( RUTNAM, 'Cannot open specified output file:',
     &    FN )
      END IF
C
C --- Release the channel.
C
      IOS = LIB$FREE_LUN( LUN )
      IF( .NOT. IOS ) CALL EXIT( IOS )
C
      RETURN
  100 FORMAT(' \hbox{\vbox{\halftone\offinterlineskip ',
     & '% machine-generated by TEXPIC.'/
     &  ' \def\BHT{\hbox\bgroup\ignorespaces}'/
     &  ' \catcode`\^=12 \catcode`\_=12 \catcode`\.=\active',
     &  ' \let.=\egroup'/ ' \catcode`\,=\active \let,=\BHT',
     &  ' \catcode`\/=0 \catcode`\\=12')
  200 FORMAT(1X,A,'%')
  300 FORMAT(1X,A,'.')
  400 FORMAT(' }}%')
      END
C --- -----------------------------------------------------------------
      SUBROUTINE TEXPIC( ARRAY, M, N, FN )
C --- -----------------------------------------------------------------
C
C --- TEXPIC version 0.1 was written by Alien in Fortran-77.
C
C --- This  routine  writes out the M x N image ARRAY into the file FN
C --- in a form which is suitable for insertion into a  TeX  document.
C
C ---    By  default,  the range of the data is determined and used to
C --- maximise  the  contrast  of  the  output  image.  This  can   be
C --- overridden  by  pre-setting the range of data values with a call
C --- to ZRANGE. ZAUTO  restores  the  default  behaviour.  Similarly,
C --- TEXPIC  will  produce  negated  images  on  output  if DONEG has
C --- previously been invoked.  DOPOS  sets  it  to  produce  positive
C --- pictures again.
C
C --- USAGE:    CALL TEXPIC( ARRAY, M, N, FN )
C
C --- PARAMETERS
C ---     ARRAY  REAL           image to be output to the file
C ---     M      INTEGER        first dimension of ARRAY
C ---     N      INTEGER        second dimension of ARRAY
C ---     FN     CHARACTER*(*)  name  of  file  to  which ARRAY will
C ---                           be written
C
C --- RESTRICTIONS
C --- If N is greater than MMAX, the image will be sub-sampled in both
C --- directions  to  make  the  result MMAX x MMAX. The interpolation
C --- technique used is due to P.R. Smith (Ultramicroscopy vol  6,  pp
C --- 201--204, 1981).
C
C --- COMMONS
C --- /ALG/, /ALGTEX/
C
C --- SUBPROGRAMS INVOKED
C --- MINMAX
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT CHARACTER*1 (A-Z)
C
      INTEGER LEVELS, MINIDX, CMAX
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
      PARAMETER( LEVELS=65, MINIDX=48, CMAX=132 )
C
      CHARACTER*(*) FN
      INTEGER M, N
      REAL ARRAY(M,N)
C
      CHARACTER*(CMAX) C
      CHARACTER*6 RUTNAM
      INTEGER NMAX, I, J, IC, IV, LUN, IOS, ILO, JLO, IHI, JHI
      LOGICAL POS
      REAL RANGE, INC
      REAL X, Y, DX, DY, DX1, DY1, VAL
C
      LOGICAL ABANDN
      INTEGER GLUN, PLUN, ELUN, POLHIS
      LOGICAL XFIX, YFIX, ZFIX, RFIX, TFIX, LOGX, LOGY, LOGZ, LOGR
      LOGICAL VRBOSE, NEG, MIDORG, FTNRML
      REAL XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX, TMIN, TMAX,
     & FTFWD, FTREV
      REAL TOL
      INTEGER MMAX
C
      COMMON /ALG/ XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX,
     & TMIN, TMAX, TOL, FTFWD, FTREV, XFIX, YFIX, ZFIX, RFIX, TFIX,
     & LOGX, LOGY, LOGZ, LOGR, NEG, GLUN, PLUN, ELUN, POLHIS, MIDORG,
     & FTNRML, VRBOSE
      SAVE /ALG/
      COMMON/ALGTEX/ MMAX
      SAVE /ALGTEX/
      EXTERNAL ALGINI
      DATA RUTNAM/'TEXPIC'/
C
C --- Find the range of the data if we're in auto mode; otherwise,
C --- use the specified ranges.
C
      IF( .NOT. ZFIX ) THEN
         CALL MINMAX( ARRAY, M, N, ZMIN, ZMAX )
         IF( ZMIN .EQ. ZMAX ) ZMAX = ZMIN + 1
      END IF
C
C --- We can't print more than MMAX columns across the TeX output.
C --- If the user passes an array bigger than this, we'll interpolate
C --- it down to MMAX.
C
      IF( M .GT. MMAX ) THEN
         INC = FLOAT(M) / FLOAT(MMAX)
         NMAX = NINT( FLOAT(N) / INC )
      ELSE
         INC = 0
      END IF
C
C --- We always open the output file on channel 7 (fix me!).
C
      LUN = 7
      OPEN( UNIT=LUN, FILE=FN, STATUS='NEW', RECL=CMAX+1, IOSTAT=IOS )
      IF( IOS .EQ. 0 ) THEN
C
C --- Calculate the scaling factor.
C
         POS = .NOT. NEG
         RANGE = FLOAT(LEVELS-1) / (ZMAX-ZMIN)
C
C --- Output the introduction.
C
         WRITE( LUN, 100 )
C
C --- Output the image without interpolation if INC is zero.
C
         IF( ABS(INC) .LT. TOL ) THEN
            DO 2 J = 1, N
      IF( ABANDN(0) ) GO TO 5
               IC = 1
               C(1:1) = ','
               DO 1 I = 1, M
                  VAL = ARRAY(I,J)
               IF( VAL .LT. ZMIN ) VAL = ZMIN
               IF( VAL .GT. ZMAX ) VAL = ZMAX
               IV = NINT((VAL-ZMIN) * RANGE)
               IF( POS ) IV = (LEVELS-1) - IV
               IC = IC + 1
               C(IC:IC) = CHAR( IV + MINIDX )
               IF( IC .GE. CMAX-1 ) THEN
                  WRITE( LUN, 200 ) C(1:IC)
                  IC = 1
                  C(1:1) = ' '
               END IF
    1          CONTINUE
               IF( IC .GT. 0 ) WRITE(LUN, 300) C(1:IC)
    2       CONTINUE
         ELSE
C
C --- Interpolate the output.
C
            Y = 1
            DO 4 J = 1, NMAX
      IF( ABANDN(0) ) GO TO 5
               DY = Y - INT(Y)
               DY1 = 1 - DY
               JLO = MOD( INT(Y-1), N ) + 1
               JHI = MOD( JLO, N ) + 1
               X = 1
               IC = 1
               C(1:1) = ','
               DO 3 I = 1, MMAX
                  DX = X - INT(X)
                  DX1 = 1 - DX
                  ILO = MOD( INT(X)-1, M ) + 1
                  IHI = MOD( ILO, M ) + 1
C
C --- Smith's three-point contextual bilinear interpolation.
C
                  IF( ABS(ARRAY(ILO,JLO)-ARRAY(IHI,JHI)) .GT.
     &               ABS(ARRAY(IHI,JLO)-ARRAY(ILO,JHI)) ) THEN
                     VAL = (DX-DY)*ARRAY(IHI,JLO) + DX1*ARRAY(ILO,JLO) +
     &                  DY*ARRAY(IHI,JHI)
                  ELSE
                     VAL = (DX1-DY)*ARRAY(ILO,JLO) + DX*ARRAY(IHI,JLO) +
     &                  DY*ARRAY(ILO,JHI)
                  END IF
                  X = X + INC
               IF( VAL .LT. ZMIN ) VAL = ZMIN
               IF( VAL .GT. ZMAX ) VAL = ZMAX
               IV = NINT((VAL-ZMIN) * RANGE)
               IF( POS ) IV = (LEVELS-1) - IV
               IC = IC + 1
               C(IC:IC) = CHAR( IV + MINIDX )
               IF( IC .GE. CMAX-1 ) THEN
                  WRITE( LUN, 200 ) C(1:IC)
                  IC = 1
                  C(1:1) = ' '
               END IF
    3          CONTINUE
               IF( IC .GT. 0 ) WRITE(LUN, 300) C(1:IC)
C
               Y = Y + INC
    4       CONTINUE
         END IF
C
C --- Close off the file.
C
    5    CONTINUE
         WRITE( LUN, 400 )
         CLOSE( UNIT=LUN )
      ELSE
         CALL ALGERR( RUTNAM, 'Cannot open specified output file:',
     &    FN )
      END IF
C
C
      RETURN
  100 FORMAT(' \hbox{\vbox{\halftone\offinterlineskip ',
     & '% machine-generated by TEXPIC.'/
     &  ' \def\BHT{\hbox\bgroup\ignorespaces}'/
     &  ' \catcode`\^=12 \catcode`\_=12 \catcode`\.=\active',
     &  ' \let.=\egroup'/ ' \catcode`\,=\active \let,=\BHT',
     &  ' \catcode`\/=0 \catcode`\\=12')
  200 FORMAT(1X,A,'%')
  300 FORMAT(1X,A,'.')
  400 FORMAT(' }}%')
      END
C --- -----------------------------------------------------------------
      SUBROUTINE TEXMAX( MV )
C --- -----------------------------------------------------------------
C
C --- TEXMAX version 0.0 was written by Alien in Fortran-77.
C
C --- This  routine sets the maximum number of pixels across a picture
C --- which TEXPIC will output to a file. Pictures  which  have  their
C --- first  dimension greater than MMAX are interpolated down to MMAX
C --- pixels.
C
C --- USAGE:  CALL TEXMAX( MMAX )
C
C --- PARAMETERS
C ---     MMAX  INTEGER  maximum number of pixels  to  be  plotted  by
C ---                    TEXPIC
C
C --- RESTRICTIONS
C --- none
C
C --- COMMONS
C --- /ALGTEX/
C
C --- SUBPROGRAMS INVOKED
C --- none
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
C
      INTEGER MV
C
      CHARACTER*(MAXICH) BUF
C
      INTEGER MMAX
C
      COMMON/ALGTEX/ MMAX
      SAVE /ALGTEX/
C
      IF( MV .GE. 2 ) THEN
         MMAX = MV
      ELSE
         WRITE( BUF, 100 ) MV
         CALL ALGERR( 'TEXMAX',
     &    'Too few pixels selected across page:'//BUF,
     &    'You must have two or more pixels across the page' )
      END IF
C
      RETURN
  100 FORMAT(I11)
      END
C --- ------------------------------------------------------------------
      SUBROUTINE ZRANGE( ZVMIN, ZVMAX )
C --- ------------------------------------------------------------------
C
C --- ZRANGE version 0.0 was written by Alien in Fortran-77.
C
C --- This  routine  fixes  the  range  of  the  Z-axis for subsequent
C --- graphical plots.
C
C --- USAGE:  CALL ZRANGE( ZMIN, ZMAX )
C
C --- PARAMETERS
C ---     ZMIN  REAL  minimum value to appear on the Z-axis
C ---     ZMAX  REAL  maximum value to appear on the Z-axis
C
C --- RESTRICTIONS
C --- ZMIN must be smaller than ZMAX.
C --- Note that the range of values actually produced on graphs may be
C --- slightly greater than those specified.
C
C --- SUBPROGRAMS INVOKED
C --- ALGERR
C
C --- COMMONS
C --- ZFIX, ZMIN, ZMAX in /ALG/
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
C
      REAL ZVMIN, ZVMAX
C
      INTEGER GLUN, PLUN, ELUN, POLHIS
      LOGICAL XFIX, YFIX, ZFIX, RFIX, TFIX, LOGX, LOGY, LOGZ, LOGR
      LOGICAL VRBOSE, NEG, MIDORG, FTNRML
      REAL XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX, TMIN, TMAX,
     & FTFWD, FTREV
      REAL TOL
C
      COMMON /ALG/ XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX,
     & TMIN, TMAX, TOL, FTFWD, FTREV, XFIX, YFIX, ZFIX, RFIX, TFIX,
     & LOGX, LOGY, LOGZ, LOGR, NEG, GLUN, PLUN, ELUN, POLHIS, MIDORG,
     & FTNRML, VRBOSE
      SAVE /ALG/
C
      IF( ZVMIN .LT. ZVMAX ) THEN
         ZMIN = ZVMIN
         ZMAX = ZVMAX
         ZFIX = .TRUE.
      ELSE
         CALL ALGERR('ZRANGE','Zmin was not smaller than Zmax.',' ')
      END IF
C
      RETURN
      END
C --- ------------------------------------------------------------------
      SUBROUTINE ZAUTO
C --- ------------------------------------------------------------------
C
C --- ZAUTO version 0.0 was written by Alien in Fortran-77.
C
C --- This routine causes  the  Z-axis  of  subsequent  graphical plots
C --- to be scaled according to the data being plotted.
C
C --- USAGE:  CALL ZAUTO
C
C --- PARAMETERS
C --- none
C
C --- RESTRICTIONS
C --- none
C
C --- SUBPROGRAMS INVOKED
C --- none
C
C --- COMMONS
C --- ZFIX in /ALG/
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
      INTEGER GLUN, PLUN, ELUN, POLHIS
      LOGICAL XFIX, YFIX, ZFIX, RFIX, TFIX, LOGX, LOGY, LOGZ, LOGR
      LOGICAL VRBOSE, NEG, MIDORG, FTNRML
      REAL XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX, TMIN, TMAX,
     & FTFWD, FTREV
      REAL TOL
C
      COMMON /ALG/ XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX,
     & TMIN, TMAX, TOL, FTFWD, FTREV, XFIX, YFIX, ZFIX, RFIX, TFIX,
     & LOGX, LOGY, LOGZ, LOGR, NEG, GLUN, PLUN, ELUN, POLHIS, MIDORG,
     & FTNRML, VRBOSE
      SAVE /ALG/
C
      ZFIX = .FALSE.
C
      RETURN
      END
C --- ------------------------------------------------------------------
      SUBROUTINE ZSAME
C --- ------------------------------------------------------------------
C ---                         ZSAME version 0.0
C
C ---  Written by Alien in Fortran-77. 
C
C ---  This  routine  causes  the  range  of  the  Z-axis on subsequent
C ---  graphical plots to be the same as those  used  on  the  previous
C ---  invocation. 
C
C ---  USAGE: CALL ZSAME 
C
C ---  PARAMETERS 
C ---  none 
C
C ---  RESTRICTIONS 
C ---  none 
C
C ---  SUBPROGRAMS INVOKED 
C ---  |ALGERR| 
C
C ---  COMMONS 
C ---  |ZFIX|, |ZMIN|, |ZMAX| in |/ALG/| 
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
C
      INTEGER GLUN, PLUN, ELUN, POLHIS
      LOGICAL XFIX, YFIX, ZFIX, RFIX, TFIX, LOGX, LOGY, LOGZ, LOGR
      LOGICAL VRBOSE, NEG, MIDORG, FTNRML
      REAL XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX, TMIN, TMAX,
     & FTFWD, FTREV
      REAL TOL
C
      COMMON /ALG/ XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX,
     & TMIN, TMAX, TOL, FTFWD, FTREV, XFIX, YFIX, ZFIX, RFIX, TFIX,
     & LOGX, LOGY, LOGZ, LOGR, NEG, GLUN, PLUN, ELUN, POLHIS, MIDORG,
     & FTNRML, VRBOSE
      SAVE /ALG/
C
      IF( ZMIN .LT. ZMAX ) THEN
         ZFIX = .TRUE.
      ELSE
         CALL ALGERR('ZSAME','Zmin was not smaller than Zmax.',
     &    'Call ignored.' )
      END IF
C
      RETURN
      END
C --- -----------------------------------------------------------------
      SUBROUTINE DONEG
C --- -----------------------------------------------------------------
C
C --- DONEG version 0.0 was written by Alien in Fortran-77.
C
C --- This routine causes subsequent grey-level pictures to  be  drawn
C --- with negative contrast.
C
C --- USAGE:  CALL DONEG
C
C --- PARAMETERS
C --- none
C
C --- RESTRICTIONS
C --- none
C
C --- SUBPROGRAMS INVOKED
C --- none
C
C --- COMMONS
C --- NEG in /ALG/
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
      INTEGER GLUN, PLUN, ELUN, POLHIS
      LOGICAL XFIX, YFIX, ZFIX, RFIX, TFIX, LOGX, LOGY, LOGZ, LOGR
      LOGICAL VRBOSE, NEG, MIDORG, FTNRML
      REAL XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX, TMIN, TMAX,
     & FTFWD, FTREV
      REAL TOL
C
      COMMON /ALG/ XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX,
     & TMIN, TMAX, TOL, FTFWD, FTREV, XFIX, YFIX, ZFIX, RFIX, TFIX,
     & LOGX, LOGY, LOGZ, LOGR, NEG, GLUN, PLUN, ELUN, POLHIS, MIDORG,
     & FTNRML, VRBOSE
      SAVE /ALG/
C
C
      NEG = .TRUE.
C
      RETURN
      END
C --- -----------------------------------------------------------------
      SUBROUTINE DOPOS
C --- -----------------------------------------------------------------
C
C --- DOPOS version 0.0 was written by Alien in Fortran-77.
C
C --- This routine causes subsequent grey-level pictures to  be  drawn
C --- with negative contrast.
C
C --- USAGE:  CALL DOPOS
C
C --- PARAMETERS
C --- none
C
C --- RESTRICTIONS
C --- none
C
C --- SUBPROGRAMS INVOKED
C --- none
C
C --- COMMONS
C --- NEG in /ALG/
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
      INTEGER GLUN, PLUN, ELUN, POLHIS
      LOGICAL XFIX, YFIX, ZFIX, RFIX, TFIX, LOGX, LOGY, LOGZ, LOGR
      LOGICAL VRBOSE, NEG, MIDORG, FTNRML
      REAL XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX, TMIN, TMAX,
     & FTFWD, FTREV
      REAL TOL
C
      COMMON /ALG/ XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX,
     & TMIN, TMAX, TOL, FTFWD, FTREV, XFIX, YFIX, ZFIX, RFIX, TFIX,
     & LOGX, LOGY, LOGZ, LOGR, NEG, GLUN, PLUN, ELUN, POLHIS, MIDORG,
     & FTNRML, VRBOSE
      SAVE /ALG/
C
C
      NEG = .FALSE.
C
      RETURN
      END
C --- ------------------------------------------------------------------
      SUBROUTINE MINMAX( ARRAY, M, N, LOWEST, HIEST )
C --- ------------------------------------------------------------------
C
C --- MINMAX version 0.0 was written by Alien in Fortran-77.
C
C --- This  routine  determines  the smallest and largest values of an
C --- array. If an interrupt is detected during the estimation of  the
C --- limits of the data, the currently-detected limits are returned.
C
C --- USAGE: CALL MINMAX( ARRAY, M, N, MIN, MAX )
C
C --- PARAMETERS
C ---     ARRAY  REAL     array of which the limits are to be
C ---                     determined
C ---     M      INTEGER  first dimension of ARRAY
C ---     N      INTEGER  second dimension of ARRAY
C ---     MIN    REAL     minimum value found in ARRAY (returned)
C ---     MAX    REAL     maximum value found in ARRAY (returned)
C
C --- RESTRICTIONS
C --- If  interrupts are to be detected, interrupt detection must have
C --- been enabled  by  the  calling  program---see  ABANDN  for  more
C --- details.
C
C --- SUBPROGRAMS INVOKED
C --- ABANDN
C
C --- COMMONS
C --- none
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
C
      INTEGER M, N
      REAL ARRAY(M*N), LOWEST, HIEST
C
      INTEGER I, J, JM
      REAL VAL, LO, HI
      INTEGER*2 CHAN
      LOGICAL ABFLAG
C
      COMMON /ALG_ABANDN/ ABFLAG, CHAN
      SAVE /ALG_ABANDN/
      EXTERNAL ALGINI
C
C --- We have declared ARRAY as a 1-D array, but will still access
C --- it via two DO-loops. This is so that ABANDN is only invoked once
C --- per "row" of ARRAY.
C
      LO = ARRAY(1)
      HI = LO
C
      DO 1 J = 1, N
         JM = (J-1) * M
      IF( ABFLAG ) GO TO 2
         DO 1 I = 1, M
            VAL = ARRAY(I+JM)
            IF( VAL .LT. LO ) LO = VAL
            IF( VAL .GT. HI ) HI = VAL
    1 CONTINUE
C
    2 CONTINUE
      LOWEST = LO
      HIEST = HI
C
      RETURN
      END
C --- ------------------------------------------------------------------
      SUBROUTINE ALGERR( NAME, MESS, EXTRA )
C --- ------------------------------------------------------------------
C
C --- ALGERR version 0.0 was written by Alien in Fortran-77.
C
C --- This routine reports errors generated by other routines. NAME is
C --- the  name  of the invoking routine while MESS and EXTRA form the
C --- message to be reported to the user. MESS  is  the  text  of  the
C --- message.  EXTRA,  if non-blank, contains extra information about
C --- the error; leading and trailing blanks are  removed  from  EXTRA
C --- before  it is output. The current version of this routine simply
C --- outputs the message text on the error output channel  --  future
C --- versions will be more sophisticated.
C
C --- USAGE: CALL ALGERR( NAME, MESS, EXTRA )
C
C --- PARAMETERS
C ---     NAME  CHARACTER*(*)  name of the invoking routine
C ---     MESS  CHARACTER*(*)  message to be output
C ---     EXTRA CHARACTER*(*)  additional text for the message
C
C --- RESTRICTIONS
C --- The  total  length  of  the message must be less than the output
C --- line length (usually 132 characters)  --  this  means  that  the
C --- lengths of NAME and MESS, when added together, must come to less
C --- then 110 characters.
C
C --- SUBPROGRAMS INVOKED
C --- none
C
C --- COMMONS
C --- /ALG/
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
      CHARACTER*1 BLANK
      PARAMETER( BLANK=' ' )
C
      CHARACTER*(*) NAME, MESS, EXTRA
C
      INTEGER NEXTRA, FC, LC, NMESS, LM
C
      INTEGER GLUN, PLUN, ELUN, POLHIS
      LOGICAL XFIX, YFIX, ZFIX, RFIX, TFIX, LOGX, LOGY, LOGZ, LOGR
      LOGICAL VRBOSE, NEG, MIDORG, FTNRML
      REAL XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX, TMIN, TMAX,
     & FTFWD, FTREV
      REAL TOL
C
      COMMON /ALG/ XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX,
     & TMIN, TMAX, TOL, FTFWD, FTREV, XFIX, YFIX, ZFIX, RFIX, TFIX,
     & LOGX, LOGY, LOGZ, LOGR, NEG, GLUN, PLUN, ELUN, POLHIS, MIDORG,
     & FTNRML, VRBOSE
      SAVE /ALG/
      EXTERNAL ALGINI
C
      NMESS = LEN( MESS )
      DO 1 LM = NMESS, 1, -1
         IF( MESS(LM:LM) .NE. BLANK ) GO TO 2
    1 CONTINUE
      LM = 1
    2 CONTINUE
C
      IF( EXTRA .EQ. BLANK ) THEN
         WRITE( ELUN, 100 ) NAME, MESS(1:LM)
      ELSE
         NEXTRA = LEN( EXTRA )
         DO 3 FC = 1, NEXTRA
            IF( EXTRA(FC:FC) .NE. BLANK ) GO TO 4
    3    CONTINUE
    4    CONTINUE
         DO 5 LC = NEXTRA, FC, -1
            IF( EXTRA(LC:LC) .NE. BLANK ) GO TO 6
    5    CONTINUE
    6    CONTINUE
         WRITE( ELUN, 100 ) NAME, MESS(1:LM)
         WRITE( ELUN, 101 ) EXTRA(FC:LC)
      END IF
C
      RETURN
  100 FORMAT(1X,A,': error -- ',A)
  101 FORMAT(10X,A)
      END
C --- ------------------------------------------------------------------
      LOGICAL FUNCTION ABANDN( OP )
C --- ------------------------------------------------------------------
C
C --- ABANDN version 0.1 was written by Alien in Fortran-77.
C
C --- This  LOGICAL  function  is  used to detect whether the user has
C --- tried to interrupt execution.
C
C ---    The method of specifying an interrupt varies from  system  to
C --- system,   but   is  typically  by  typing  a  control  character
C --- (control-C on the VAX). ABANDN is used  with  OP  =  0  to  TEST
C --- whether the user has signalled an interrupt -- the value TRUE is
C --- returned  as  the  value  of  the  function if this is the case.
C --- ABANDN is used with OP = 1 to SET or CLEAR the  interrupt  trap;
C --- this  must  be  done  by  the  calling program. Note that ABANDN
C --- returns the value TRUE if an error  occurred  while  setting  or
C --- clearing the interrupt trap.
C
C --- USAGE:   <logical variable> = ABANDN( OP )
C
C --- PARAMETERS
C ---     OP  INTEGER  operation 0 ==> test, 1 ==> set
C
C --- RESTRICTIONS
C --- Interrupts  will  not  be trapped before the first invocation of
C --- ABANDN(1).
C --- After  the  user  has   signalled   an   interrupt,   subsequent
C --- invocations  of  ABANDN(0)  will  return TRUE until ABANDN(1) is
C --- used to clear it.
C --- If the user generates interrupts very quickly (for  example,  by
C --- letting  the  ^C  auto-repeat), they may be delivered so quickly
C --- that ABANDN does not manage to reset its trap in time;  in  this
C --- case, it will actually interrupt program execution.
C --- This version of ABANDN requires VAX/VMS 3.0 or later.
C
C --- SUBPROGRAMS INVOKED
C --- ALG_ABANDN_AST (condition handler)
C
C --- COMMONS
C --- /ALG_ABANDN/
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
      INCLUDE '($IODEF)'
C
      INTEGER ABSET, ABTEST
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
      PARAMETER( ABSET=1, ABTEST=0 )
C
      INTEGER OP
C
      CHARACTER*(MAXICH) CODE
      INTEGER F
      INTEGER IOS, VAL, SYS$GETDVI, SYS$ASSIGN, SYS$QIOW, SYS$DASSGN
      INTEGER DVIBLK(4)/ '40004'X, 0, 0, 0/
      INTEGER*2 IOSB(4)
      LOGICAL FIRST/.TRUE./
      EXTERNAL ALG_ABANDN_AST
      INTEGER*2 CHAN
      LOGICAL ABFLAG
C
      COMMON /ALG_ABANDN/ ABFLAG, CHAN
      SAVE /ALG_ABANDN/
C
C --- Branch according to the value of OP.
C
      IF( OP .EQ. ABTEST ) THEN
         ABANDN = ABFLAG
      ELSE IF( OP .EQ. ABSET ) THEN
         ABANDN = .FALSE.
         IF( FIRST ) THEN
C
C --- Check that we're using a terminal.
C
            DVIBLK(2) = %LOC( VAL )
            IOS = SYS$GETDVI( ,, 'TT', DVIBLK, ,,, )
            IF( .NOT. IOS ) CALL EXIT( IOS )
            IF( VAL .EQ. '42'X ) THEN
C
C --- If we're using a terminal, assign a channel to the device and set
C --- the trap; failure from any of the system services is taken as a
C --- fatal error. We will only return a failure code to the user if
C --- IOSB(1) indicates an error.
C
               IOS = SYS$ASSIGN( 'TT', CHAN,, )
               IF( .NOT. IOS ) CALL EXIT( IOS )
               IOS = SYS$QIOW(, %VAL(CHAN), %VAL(IO$_SETMODE.OR.IO$M_CTR
     &LCAST),
     &            IOSB, ,,   ALG_ABANDN_AST, ,,,, )
               IF( .NOT. IOS ) CALL EXIT( IOS )
               IF( IOSB(1) ) THEN
                  FIRST = .FALSE.
               ELSE
C
C --- We didn't succeed in setting the interrupt trap, heaven knows why!
C --- Set the function to return an error code, then close the channel
C --- we
C --- have so carefully opened.
C
                  ABANDN = .TRUE.
                  IOS = SYS$DASSGN( %VAL(CHAN) )
                  IF( .NOT. IOS ) CALL EXIT( IOS )
               END IF
            END IF
         END IF
         ABFLAG = .FALSE.
      ELSE
         STOP 'ABANDN: Illegal argument value.'
      END IF
C
      RETURN
  100 FORMAT(I11)
      END
C --- ------------------------------------------------------------------
      SUBROUTINE ALG_ABANDN_AST
C --- ------------------------------------------------------------------
C
C --- This routine is the condition handler which is used in the
C --- implementation of ABANDN for VAX/VMS. It is called by the system
C --- when the user types ^C at his terminal; its main purpose is to set
C --- ABFLAG in common /ALG_ABANDN/, to be tested by ABANDN(ABTEST).
C --- However, because VMS ^C condition handlers are one-shot affairs,
C --- we must also re-impose the trap. This is done by invoking the
C --- simple routine ALG_ABANDN_RESET_TRAP.
C
      INTEGER*2 CHAN
      LOGICAL ABFLAG
C
      COMMON /ALG_ABANDN/ ABFLAG, CHAN
      SAVE /ALG_ABANDN/
C
C --- Set ABFLAG.
C
      ABFLAG = .TRUE.
C
C --- And reset the trap.
C
      CALL ALG_ABANDN_RESET_TRAP
C
      RETURN
      END
C --- ------------------------------------------------------------------
      SUBROUTINE ALG_ABANDN_RESET_TRAP
C --- ------------------------------------------------------------------
C
C --- This routine resets the ^C trap for ALG_ABANDN_AST, because the
C --- Fortran compiler will not allow ALG_ABANDN_AST to be used in the
C --- $QIOW call inside its own code.
C
      INCLUDE '($IODEF)'
      INTEGER*2 IOSB(4)
      INTEGER IOS, SYS$QIOW
      EXTERNAL ALG_ABANDN_AST
C
      INTEGER*2 CHAN
      LOGICAL ABFLAG
C
      COMMON /ALG_ABANDN/ ABFLAG, CHAN
      SAVE /ALG_ABANDN/
C
      IOS = SYS$QIOW(, %VAL(CHAN), %VAL(IO$_SETMODE.OR.IO$M_CTRLCAST),
     &   IOSB, ,,   ALG_ABANDN_AST, ,,,, )
      IF( .NOT. IOS ) CALL EXIT( IOS )
C
      RETURN
      END
C --- ------------------------------------------------------------------
           B L O C K    D A T A    A L G I N I
C --- ------------------------------------------------------------------
C
C --- Copyright (C) Alien 1987 -- All rights reserved.
C
      IMPLICIT NONE
C
C --- The following definitions are used to allow expressions to be
C --- typed for the initial values of variables.
C
      INTEGER GRFGRF, GRFHIS, GRFPOL
      INTEGER MAXICH, MAXRCH
      INTEGER MINLUN, MAXLUN
      INTEGER MAXOUT, MAXARR, MAXAFT, MAXCH
      INTEGER MAXSTV, NOUSEM
      REAL PI
      INTEGER ORDRED, RNDOM, STRAIT
C
      PARAMETER( GRFGRF=0, GRFHIS=1, GRFPOL=2 )
      PARAMETER( MAXICH=11, MAXRCH=11 )
      PARAMETER( MINLUN=1, MAXLUN=99 )
      PARAMETER( MAXOUT=133, MAXARR=512, MAXAFT=(9*MAXARR)/4+1 )
      PARAMETER( MAXCH=16 )
      PARAMETER( MAXSTV=3, NOUSEM=-1 )
      PARAMETER( PI=3.1415926535897932384626433 )
      PARAMETER( ORDRED=1, RNDOM=2, STRAIT=0 )
      INTEGER MOUT1, MSTAT
      PARAMETER (MOUT1=MAXOUT-1, MSTAT=MAXSTV+1)
C
C --- Include all the common variables and blocks.
C
      INTEGER GLUN, PLUN, ELUN, POLHIS
      LOGICAL XFIX, YFIX, ZFIX, RFIX, TFIX, LOGX, LOGY, LOGZ, LOGR
      LOGICAL VRBOSE, NEG, MIDORG, FTNRML
      REAL XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX, TMIN, TMAX,
     & FTFWD, FTREV
      REAL TOL
      INTEGER*2 CHAN
      LOGICAL ABFLAG
      CHARACTER*1 BLANK, VMARK, HMARK, TICK
      CHARACTER*(MAXOUT) LPBUF
      INTEGER LPHT, LPWID
      LOGICAL FF
      REAL ASPECT
      LOGICAL USEM, EXACT, KNOWEM(0:MAXSTV)
      REAL MINS(0:MAXSTV), MAXS(0:MAXSTV), MEANS(0:MAXSTV),
     &   SDS(0:MAXSTV)
      INTEGER MMAX
      INTEGER METHOD
C
      COMMON /ALG/ XMIN, XMAX, YMIN, YMAX, ZMIN, ZMAX, RMIN, RMAX,
     & TMIN, TMAX, TOL, FTFWD, FTREV, XFIX, YFIX, ZFIX, RFIX, TFIX,
     & LOGX, LOGY, LOGZ, LOGR, NEG, GLUN, PLUN, ELUN, POLHIS, MIDORG,
     & FTNRML, VRBOSE
      SAVE /ALG/
      COMMON /ALG_ABANDN/ ABFLAG, CHAN
      SAVE /ALG_ABANDN/
      COMMON /LPC/ LPBUF, BLANK, VMARK, HMARK, TICK
      COMMON /LPN/ ASPECT, LPHT, LPWID, FF
      SAVE /LPC/, /LPN/
      COMMON /STAT/ MINS, MAXS, MEANS, SDS, USEM, KNOWEM, EXACT
      SAVE /STAT/
      COMMON/ALGTEX/ MMAX
      SAVE /ALGTEX/
      COMMON /V80/ METHOD
C
C --- / A L G /
C
      DATA LOGX/.FALSE./, LOGY/.FALSE./, LOGZ/.FALSE./, LOGR/.FALSE./
      DATA XMIN/0.0/, XMAX/0.0/, YMIN/0.0/, YMAX/0.0/, ZMIN/0.0/,
     & ZMAX/0.0/, RMIN/0.0/, RMAX/0.0/, TMIN/0.0/, TMAX/0.0/
      DATA XFIX/.FALSE./, YFIX/.FALSE./, ZFIX/.FALSE./
      DATA RFIX/.FALSE./, TFIX/.FALSE./
      DATA VRBOSE/.FALSE./, NEG/.FALSE./
      DATA MIDORG/.TRUE./, FTNRML/.FALSE./
      DATA GLUN/6/, PLUN/6/, ELUN/6/, POLHIS/GRFGRF/
      DATA TOL/1.0E-8/
      DATA FTFWD/0.0/, FTREV/0.0/
C
C --- / A B A N D N /
C
      DATA ABFLAG/.FALSE./
C
C --- / A L G T E X /
C
      DATA MMAX/256/
C
C --- / L P C /
C
      DATA BLANK/' '/, VMARK/'|'/, HMARK/'-'/, TICK/'+'/
C
C --- / L P N /
C
      DATA FF/.TRUE./
      DATA LPHT/62/, LPWID/MOUT1/
      DATA ASPECT/ 0.604 /
C
C --- / S T A T /
C
      DATA USEM/NOUSEM/, KNOWEM/MSTAT*.FALSE./, EXACT/.FALSE./
      DATA MEANS/MSTAT*0.0/, SDS/MSTAT*0.0/
      DATA MINS/MSTAT*0.0/, MAXS/MSTAT*0.0/
C
C --- / V 8 0 /
C
      DATA METHOD/STRAIT/
C
      END
$EOD
$WRITE SYS$OUTPUT "creating HALFTONE.MF" 
$CREATE HALFTONE.MF
$DECK
% halftone font with 65 levels of gray, characters "0" (white) to "p" (black)

pair p[]; % the pixels in order (first p0 becomes black, then p1, etc)
p0=(1,1);
p8=(2,0);
p16=(1,0);
p24=(0,0);
p32=(3,-1);
p40=(2,-1);
p48=(1,-1);
p56=(2,-2);
transform r; r=identity rotatedaround ((1.5,1.5),90);

for i=0 step 8 until 56:
 p[i+2]=p[i] transformed r;
 p[i+6]=p[i+2] transformed r;
 p[i+5]=p[i+6] transformed r;
 p[i+1]=p[i] shifted (4,4);
 p[i+3]=p[i+2] shifted (4,4);
 p[i+7]=p[i+6] shifted (4,4);
 p[i+4]=p[i+5] shifted (4,4);
 endfor

for i=32 step 1 until 63:
 p[i] := (xpart p[i] mod 8, ypart p[i] mod 8);
endfor

mode_setup;
designsize:=64/pt; % that's 64 pixels

def makebox(text t)= enddef; % shut off boxes

picture prevchar;

prevchar=nullpicture;
for i=0 upto 64:
 beginchar(i+ASCII"0",designsize/8,designsize/8,0);
 currentpicture:=prevchar;
 if i>0: fill unitsquare shifted p[i-1]; fi
 prevchar:=currentpicture;
 endchar;
 endfor

font_quad=designsize/8;
end
$EOD