Release v2.14.0.3

The following are the changes in this release:

    Added builtin functions to convert between degrees and
    degrees, minutes and seconds under the config("mod")
    round rules:

	d2dms(degs, d, m, s [,rnd]) - given degs, compute d, m, s
	d2dm(degs, d, m [,rnd]) - given degs, compute d, m

	See help/d2dms and help/d2dm.

    Example:

	; print d2dms(360.321,deg=,min=,sec=), deg, min, sec;
	0.321 0 19 15.6

	; print d2dm(360.321,deg=,min=), deg, min;
	0.321 0 19.26

    Added builtin functions to convert between gradians and
    gradians, minutes and seconds under the config("mod")
    round rules:

	g2gms(grads, g, m, s [,rnd]) - given grads, compute g, m, s
	g2gm(grads, g, m [,rnd]) - given grads, compute g, m

	See help/g2gms and help/g2gm.

    Example:

	; print g2gms(400.321,grad=,min=,sec=), grad, min, sec;
	0.321 0 19 15.6

	; print g2gm(400.321,grad=,min=), grad, min;
	0.321 0 19.26

    Added builtin functions to convert between hours and
    hours, minutes and seconds under the config("mod")
    round rules:

	h2hms(hours, h, m, s [,rnd]) - given hours, compute h, m, s
	h2hm(hours, h, m [,rnd]) - given hours, compute h, m

	See help/h2hms and help/h2hm.

    Example:

	; print h2hms(24.321,hour=,min=,sec=), hour, min, sec;
	0.321 0 19 15.6

	; print h2hm(24.321,hour=,min=), hour, min;
	0.321 0 19.26

    Renumbered regression tests 3408 thru 3437, to 9102 thru 9131.

    Updated Added hms.cal resource file to use h2hms() builtin.
    Updated Added dms.cal resource file to use d2dms() builtin.

    Fix minor typo in help/mod SYNOPSIS.
    Fix minor typo in help/quo SYNOPSIS.

    Added a few more examples to help/strcmp.

    Added builtin functions to convert between degrees, minutes and
    seconds and degrees under the config("mod") round rules:

	dms2d(d, m, s [,rnd]) - convert deg, min, and secs to deg
	dm2d(d, m [,rnd]) - convert deg, min to deg

	See help/dms2d and help/dm2d.

    Example:

	; print dms2d(12, 20, 44.16);
	12.3456

	; print dm2d(3601, -25.5594);
	0.57401

    Added builtin functions to convert between gradians, minutes and
    seconds and gradians under the config("mod") round rules:

	gms2g(g, m, s [,rnd]) - convert grad, min, and secs to grad
	gm2g(g, m [,rnd]) - convert grad and min to grad

	See help/g2gms and help/g2gm.

    Example:

	; print gms2g(12, 20, 44.16);
	12.3456

	; print gm2g(4001, -25.5594);
	0.57401

    Added builtin functions to convert between hours, minutes and
    seconds and hours under the config("mod") round rules:

	hms2h(h, m, s [,rnd]) - convert hours, min, and secs to hours
	hm2h(h, m [,rnd]) - convert hours, min to hours

	See help/hms2h and help/hm2h.

    Example:

	; print hms2h(12, 20, 44.16);
	12.3456

	; print hm2h(241, -25.5594);
	0.57401

.gitignore

@@ -0,0 +1,107 @@
+# files and directories created during the building of calc and other Makefile actions
+#
+# NOTE: While many of these might be part of a released calc tarball, they are
+#	not consider development source.  Some other file(s) and/or programs
+#	generate these files.
+#
+.dynamic
+.hsrc
+Makefile.simple
+NOTES
+align32.h
+args.h
+cal/.all
+calc
+calc.1
+calc.usage
+calcerr.c
+calcerr.h
+conf.h
+cscript/.all
+cscript/4dsphere
+cscript/README
+cscript/fproduct
+cscript/mersenne
+cscript/piforever
+cscript/plus
+cscript/powerterm
+cscript/simple
+cscript/square
+custom/.all
+custom/Makefile.simple
+custom/libcustcalc*
+endian
+endian_calc.h
+fposval.h
+have_ban_pragma.h
+have_const.h
+have_fpos.h
+have_fpos_pos.h
+have_getpgid.h
+have_getprid.h
+have_getsid.h
+have_gettime.h
+have_memmv.h
+have_newstr.h
+have_offscl.h
+have_posscl.h
+have_rusage.h
+have_stdlib.h
+have_stdvs
+have_strdup.h
+have_string.h
+have_strlcat.h
+have_strlcpy.h
+have_times.h
+have_uid_t.h
+have_unistd.h
+have_unused.h
+have_urandom.h
+have_ustat.h
+help/.all
+help/COPYING
+help/COPYING-LGPL
+help/binding
+help/bindings
+help/bug
+help/bugs
+help/builtin
+help/change
+help/changes
+help/contrib
+help/copy
+help/cscript
+help/custom_cal
+help/errorcode
+help/errorcodes
+help/full
+help/funclist
+help/funclist.c
+help/libcalc
+help/man
+help/new_custom
+help/question
+help/questions
+help/resource
+help/type
+help/usage
+libcalc*
+libcustcalc*
+longbits
+longbits.h
+sample_many
+sample_rand
+tags
+terminal.h
+ver_calc
+
+# other commonly excluded patterns
+#
+*~
+*.BAK
+core*
+.DS_Store
+*.dSYM/
+*.[oa]
+.*.swp
+*,v

BUGS

@@ -1,12 +1,12 @@
 If you notice something wrong, strange or broken, try rereading:
 
-   README.FIRST
-   HOWTO.INSTALL
-   BUGS (this file)
+    README.FIRST
+    HOWTO.INSTALL
+    BUGS (this file)
 
 If that does not help, cd to the calc source directory and try:
 
-   make check
+    make check
 
 Look at the end of the output, it should say something like:
 
@@ -15,76 +15,119 @@ Look at the end of the output, it should say something like:
 
 If it does not, then something is really broken!
 
+To be sure that your version of calc is up to date, check out:
+
+    http://www.isthe.com/chongo/tech/comp/calc/calc-mirror.html
+
 If you made and modifications to calc beyond the simple Makefile
 configuration, try backing them out and see if things get better.
 
-To be sure that your version of calc is up to date, check out:
-
-	http://www.isthe.com/chongo/tech/comp/calc/calc-download.html
-
-The calc web site is located at:
-
-	http://www.isthe.com/chongo/tech/comp/calc/index.html
+If you have tried all of the above and things still are not right,
+then it may be time to send in a bug report.
 
 =-=
 
-If you have tried all of the above and things still are not right,
-then it may be time to send in a bug report.  You can send bug
-and bug fixes reports to:
+The BEST WAY TO SEND IN A BUG REPORT is via calc GitHub issue URL:
 
-    calc-bugs at asthe dot com
+    https://github.com/lcn2/calc/issues
 
-    [[ NOTE: Replace 'at' with @, 'dot' is with . and remove the spaces ]]
-    [[ NOTE: The EMail address uses 'asthe', the web site URL uses 'isthe' ]]
+If you see an existing issue that matches your problem, look
+over the notes and if needed, add your own observation,
+even if you just add to an existing issue:
 
-Your subject must contain the words:
+    I have this issue too
 
-       calc bug report
+If you don't see your issue addressed, then on the above
+GitHub web page, click on this button:
 
-You may have additional words in your subject line.
+    ((New Issue))
 
-When you send your report, please include the following information:
+=-=
 
-    * a description of the problem
-    * the version of calc you are using (if you cannot get calc
-      to run, then send us the 4 #define lines from version.c)
-    * if you modified calc from an official patch, send me the mods you made
-    * the type of system you were using
-    * the type of compiler you were using
-    * any compiler warnings or errors that you saw
-    * cd to the calc source directory, and type:
+If you prefer to not use GitHub, then you
+may send bug and bug fixes reports to:
 
-	make debug > debug.out 2>&1		(sh, ksh, bash users)
-	make debug >& debug.out			(csh, tcsh users)
+    calc-bugrept-mail at asthe dot com
 
-      and send the contents of the 'debug.out' file.
+    NOTE: Remove spaces and replace 'at' with @, 'dot' with .
 
-Stack traces from core dumps are useful to send as well.
+    NOTE: This replaces the old calc-bugs at asthe dot com address.
 
-Fell free to use the above address to send in big fixes (in the form
+PLEASE put following the SPECIAL PHRASE somewhere in your Email Subject line:
+
+    calc bug report
+
+You may add additional words to your subject line.
+
+Suggestion:
+
+    From time to time, the Email address and Subject SPECIAL PHRASE
+    may change so verify you have the current info by visiting:
+
+	http://www.isthe.com/chongo/tech/comp/calc/calc-bugrept.html
+
+IMPORTANT:
+
+    PLEASE use the above SPECIAL PHRASE somewhere in the Subject line or
+    the mail system won't deliver your Email message.
+
+Keep in mind that the best way to report on a calc bug is
+via the above mentioned calc GitHub issue URL.
+
+Please be patient as we cannot always respond to Email messages quickly.
+
+=-=
+
+When you send your report, via the calc GitHub issue URL or Email,
+please include the following information:
+
+    * A description of the problem
+
+    * Version of calc you are using
+
+	If you cannot compile calc, then look at version.c
+	and report the #define that start with:
+
+	    #define MAJOR_VER
+	    #define MINOR_VER
+	    #define MAJOR_PATCH
+	    #define MINOR_PATCH
+
+    * If you modified calc from an official patch,
+      send us the mods you made
+
+    * Type and version of the operating system
+
+    * Type and version of compiler
+
+    * Send us all compiler warnings or errors you find
+
+    * If calc dumped core, try to send us a core dump stack trace
+
+    * cd to the calc source directory, and send the contents
+      of debug.out produced by this command:
+
+	make debug > debug.out 2>&1
+
+Fell free to use the above address to send in bug fixes (in the form
 of a context diff patch).
 
 =-=
 
-Known bugs:
+Known bugs in calc:
+
+    The output of the alg_config.cal resource file is bogus.
+    We would welcome a replacement for this code.
 
     We are sure some more bugs exist.  When you find them, please let
     us know!  See the above for details on how to report and were to
-    EMail your bug reports and hopefully patches to fix them.
-
-=-=
-
-Problems that have known work-a-rounds:
-
-    * There is a bug in gcc v4.1.0 that causes calc to fail the regression
-      test.  The work-a-round is to compile with gcc v4.1.1 or later.  This
-      problems was observed on Fedora 5.
+    Email your bug reports and hopefully patches to fix them.
 
 =-=
 
 mis-features in calc:
 
-    Some problems are not bugs but rarther mis-features / things that could
+    Some problems are not bugs but rather mis-features / things that could
     work better.  The following is a list of mis-features that should be
     addressed and improved someday.
 
@@ -129,7 +172,9 @@ mis-features in calc:
 
       will not.
 
-## Copyright (C) 1999-2007  Landon Curt Noll
+=-=
+
+## Copyright (C) 1999-2014,2021  Landon Curt Noll
 ##
 ## Calc is open software; you can redistribute it and/or modify it under
 ## the terms of the version 2.1 of the GNU Lesser General Public License
@@ -145,10 +190,6 @@ mis-features in calc:
 ## received a copy with calc; if not, write to Free Software Foundation, Inc.
 ## 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
 ##
-## @(#) $Revision: 30.1 $
-## @(#) $Id: BUGS,v 30.1 2007/03/16 11:09:46 chongo Exp $
-## @(#) $Source: /usr/local/src/cmd/calc/RCS/BUGS,v $
-##
 ## Under source code control:	1994/03/18 14:06:13
 ## File existed as early as:	1994
 ##

CONTRIB-CODE

@@ -0,0 +1,100 @@
+Calc is open source. Contributions of code are welcome.
+
+We welcome and encourage you to send us:
+
+    * calc resource files (cal/*.cal files)
+    * calc shell scripts (cscript/*.calc files)
+    * builtin functions that you have modified or written, i.e.:
+	assocfunc.c comfunc.c func.c func.h
+	listfunc.c matfunc.c qfunc.c zfunc.c
+    * custom functions that you have modified or written
+    * help files modified or written (help/*)
+    * Makefile improvements (Makefile, */Makefile)
+    * other source code modifications (*.c, *.h)
+    * etc. (* */* :) )
+
+In order to consider integrating your code, we need:
+
+    * calc version you are working with (please try use the latest version)
+    * new help files or help file patches, if applicable (documentation)
+    * proposed text for the CHANGES file (brief description of what it does)
+    * regress.cal test patch as needed
+    * your source code and/or source code changes (:-))
+
+If you add functionality to calc, please be sure to modify/patch/add
+Makefiles, help files, cal/regress.cal test code as well.
+Regression test cases are vital to maintaining calc's level
+of correctness and helps us avoid code bug regression.
+
+Please try to generate a patch against the most recent
+version of calc, and if you use GitHub, the top of the
+master branch:
+
+    https://github.com/lcn2/calc
+
+The best way contribute to calc bug is to generate calc
+GitHub pull request:
+
+    https://github.com/lcn2/calc/pulls
+
+Your code needs to be contributed under either the 2.1 of the GNU
+Lesser General Public License (LGPL 2.1) or in the public domain.
+
+If you do not want to use calc GitHub, then send Email to:
+
+    calc-contrib-mail at asthe dot com
+
+    NOTE: Remove spaces and replace 'at' with @, 'dot' with .
+
+You MUST use following SPECIAL PHRASE in your Email Subject line:
+
+    calc contribution
+
+You may add additional words to your subject line.
+
+Suggestion:
+
+    From time to time, the Email address and Subject SPECIAL PHRASE
+    may change so verify you have the current info by visiting:
+
+	http://www.isthe.com/chongo/tech/comp/calc/calc-contrib.html
+
+IMPORTANT: If your Email doesn't contain the above phrase,
+	   then we WILL NOT SEE your Email.
+
+PLEASE BE SURE you have that SPECIAL PHRASE somewhere in the subject line!
+
+Keep in mind that the best way to report on a calc bug is to use
+the above mentioned calc GitHub procedure.
+
+Please be patient as we cannot always respond to Email messages quickly.
+
+=-=
+
+See also the calc wishlist by running the calc command:
+
+    ; help wishlist
+
+=-=
+
+## Copyright (C) 1999,2014,2021  Landon Curt Noll
+##
+## Calc is open software; you can redistribute it and/or modify it under
+## the terms of the version 2.1 of the GNU Lesser General Public License
+## as published by the Free Software Foundation.
+##
+## Calc is distributed in the hope that it will be useful, but WITHOUT
+## ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+## or FITNESS FOR A PARTICULAR PURPOSE.	 See the GNU Lesser General
+## Public License for more details.
+##
+## A copy of version 2.1 of the GNU Lesser General Public License is
+## distributed with calc under the filename COPYING-LGPL.  You should have
+## received a copy with calc; if not, write to Free Software Foundation, Inc.
+## 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+##
+## Under source code control:	1997/03/09 16:33:22
+## File existed as early as:	1997
+##
+## chongo <was here> /\oo/\	http://www.isthe.com/chongo/
+## Share and enjoy!  :-)	http://www.isthe.com/chongo/tech/comp/calc/

COPYING

@@ -6,17 +6,13 @@ This file is Copyrighted
 
     This file is covered under the following Copyright:
 
-	Copyright (C) 1999  Landon Curt Noll
+	Copyright (C) 1999-2021  Landon Curt Noll
 	All rights reserved.
 
 	Everyone is permitted to copy and distribute verbatim copies
 	of this license document, but changing it is not allowed.
 
-    # @(#) $Revision: 30.1 $
-    # @(#) $Id: COPYING,v 30.1 2007/03/16 11:09:46 chongo Exp $
-    # @(#) $Source: /usr/local/src/cmd/calc/RCS/COPYING,v $
-
-=-=
+-=-
 
 Calc is covered by the GNU Lesser General Public License
 --------------------------------------------------------
@@ -53,48 +49,6 @@ Calc is covered by the GNU Lesser General Public License
 	    Boston, MA  02110-1301
 	    USA
 
-    The contact addresses for calc is as follows:
-
-	Web:	http://www.isthe.com/chongo/tech/comp/calc/email.html
-
-    To join the low volume calc mailing list.  Send a EMail message to:
-
-	calc-tester-request at asthe dot com
-
-    Your subject must contain the words:
-
-	calc mailing list subscription
-
-    You may have additional words in your subject line.
-
-    Your message body (not the subject) should consist of:
-
-	 subscribe calc-tester address
-	 end
-	 name your_full_name
-
-    where ``address'' is your EMail address and ``your_full_name'' is
-    your full name.
-
-    Feel free to follow the name line with additional EMail text as desired.
-
-=-=
-
-    Calc bug reports and calc bug fixes should be sent to:
-
-	calc-bugs at asthe dot com
-
-    [[ NOTE: Replace 'at' with @, 'dot' is with . and remove the spaces ]]
-    [[ NOTE: The EMail address uses 'asthe' and the web site URL uses 'isthe' ]]
-
-    Your subject must contain the words:
-
-	   calc bug report
-
-    You may have additional words in your subject line.
-
-=-=
-
 Calc's relationship to the GNU Lesser General Public License
 ------------------------------------------------------------
 
@@ -149,7 +103,7 @@ Calc's relationship to the GNU Lesser General Public License
      except for the exception files explicitly listed in the ``Calc
      copyrights and exception files'' section below.
 
-=-=
+-=-
 
 Calc copyrights and exception files
 -----------------------------------
@@ -165,12 +119,12 @@ Calc copyrights and exception files
 	Copyright (C) year  Ernest Bowen and Landon Curt Noll
 	Copyright (C) year  Ernest Bowen
 	Copyright (C) year  Petteri Kettunen and Landon Curt Noll
+	Copyright (C) year  Christoph Zurnieden
 
     These files are not covered under one of the Copyrights listed above:
 
-	    shs1.c	    shs1.h	 shs.c		shs.h
-	    md5.c	    md5.h	 COPYING	COPYING-LGPL
-	    cal/qtime.cal   cal/screen.cal
+	    sha1.c		sha1.h		COPYING
+	    COPYING-LGPL	cal/qtime.cal	cal/screen.cal
 
     The file COPYING-LGPL, which contains a copy of the version 2.1
     GNU Lesser General Public License, is itself Copyrighted by the
@@ -182,17 +136,14 @@ Calc copyrights and exception files
     top of this file.  It is important to note that you may distribute
     verbatim copies of this file but you may not modify this file.
 
-    Some of these exception files are in the public domain.  The md5.c
-    and md5.h files were "derived from the RSA Data Security, Inc. MD5
-    Message-Digest Algorithm" and are under a copyright that allows these
-    two files to be freely used and distributed.  Other files are under
-    the LGPL but have different authors that those listed above.
+    Some of these exception files are in the public domain.  Other files
+    are under the LGPL but have different authors that those listed above.
 
     In all cases one may use and distribute these exception files freely.
     And because one may freely distribute the LGPL covered files, the
     entire calc source may be freely used and distributed.
 
-=-=
+-=-
 
 General Copyleft and License info
 ---------------------------------
@@ -206,7 +157,7 @@ General Copyleft and License info
 	http://www.gnu.org/copyleft/lesser.html
 	http://www.gnu.org/copyleft/lesser.txt
 
-=-=
+-=-
 
 Why calc did not use the GNU General Public License
 ---------------------------------------------------

COPYING-LGPL

@@ -1,8 +1,8 @@
-		  GNU LESSER GENERAL PUBLIC LICENSE
-		       Version 2.1, February 1999
+                  GNU LESSER GENERAL PUBLIC LICENSE
+                       Version 2.1, February 1999
 
  Copyright (C) 1991, 1999 Free Software Foundation, Inc.
-     51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+ 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
  Everyone is permitted to copy and distribute verbatim copies
  of this license document, but changing it is not allowed.
 
@@ -10,7 +10,7 @@
  as the successor of the GNU Library Public License, version 2, hence
  the version number 2.1.]
 
-			    Preamble
+                            Preamble
 
   The licenses for most software are designed to take away your
 freedom to share and change it.  By contrast, the GNU General Public
@@ -112,7 +112,7 @@ modification follow.  Pay close attention to the difference between a
 former contains code derived from the library, whereas the latter must
 be combined with the library in order to run.
 
-		  GNU LESSER GENERAL PUBLIC LICENSE
+                  GNU LESSER GENERAL PUBLIC LICENSE
    TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
 
   0. This License Agreement applies to any software library or other
@@ -146,7 +146,7 @@ such a program is covered only if its contents constitute a work based
 on the Library (independent of the use of the Library in a tool for
 writing it).  Whether that is true depends on what the Library does
 and what the program that uses the Library does.
-  
+
   1. You may copy and distribute verbatim copies of the Library's
 complete source code as you receive it, in any medium, provided that
 you conspicuously and appropriately publish on each copy an
@@ -432,7 +432,7 @@ decision will be guided by the two goals of preserving the free status
 of all derivatives of our free software and of promoting the sharing
 and reuse of software generally.
 
-			    NO WARRANTY
+                            NO WARRANTY
 
   15. BECAUSE THE LIBRARY IS LICENSED FREE OF CHARGE, THERE IS NO
 WARRANTY FOR THE LIBRARY, TO THE EXTENT PERMITTED BY APPLICABLE LAW.
@@ -455,7 +455,7 @@ FAILURE OF THE LIBRARY TO OPERATE WITH ANY OTHER SOFTWARE), EVEN IF
 SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
 DAMAGES.
 
-		     END OF TERMS AND CONDITIONS
+                     END OF TERMS AND CONDITIONS
 
            How to Apply These Terms to Your New Libraries
 
@@ -485,7 +485,8 @@ convey the exclusion of warranty; and each file should have at least the
 
     You should have received a copy of the GNU Lesser General Public
     License along with this library; if not, write to the Free Software
-    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
+    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301
+    USA
 
 Also add information on how to contact you by electronic and paper mail.
 
@@ -500,5 +501,3 @@ necessary.  Here is a sample; alter the names:
   Ty Coon, President of Vice
 
 That's all there is to it!
-
-

HOWTO.INSTALL

@@ -1,6 +1,18 @@
-Installing calc from the gziped tarball in 4 easy steps:
 
-    0)  If your platform supports i686 RPMs, you may want to go to:
+IMPORTANT: Please see the section at the bottom of this file for
+	   some important information on Makefiles used in calc.
+
+###################################################################
+# IMPORTANT: DO NOT run GNU make in parallel mode!!!		  #
+###################################################################
+# Unfortunately due to the complex dependency issues between	  #
+# Makefile, Makefile.ship and custom/Makefile, parallel GNU make  #
+# is NOT recommended.  Sorry (tm Canada) :)			  #
+###################################################################
+
+Installing calc from the bzip2-ed tarball in 4 easy steps:
+
+    (0)  If your platform supports i686 RPMs, you may want to go to:
 
 	    http://www.isthe.com/chongo/src/calc/
 
@@ -17,23 +29,33 @@ Installing calc from the gziped tarball in 4 easy steps:
 
 	The following 4 steps apply to calc source tree that comes from either:
 
-		gunzip -c calc-*.tar.gz | tar -xvf -
+		bunzip2 -c calc-*.tar.bz2 | tar -xvf -
 
 	or from:
 
 		rpm -ivh calc-*.src.rpm
 		cd /var/tmp
-		gunzip -c /usr/src/redhat/SOURCES/calc-*.tar.gz | tar -xvf -
+		bunzip2 -c /usr/src/redhat/SOURCES/calc-*.tar.bz2 | tar -xvf -
 
-    1)	Look at the makefile, and adjust it to suit your needs.
+    (1)	Look at the makefile, and adjust it to suit your needs.
 
-	The top level Makefile and the custom/Makefile require a GNU
+	The top level Makefile and the custom/Makefile require a modern
 	Make (such as gmake) or an equivalently advanced make.	On many
-	targets, the default make is sufficent.  On FreeBSD for example,
-	one must use gmake instead of make.
+	targets, the default make is sufficient.  On FreeBSD for example,
+	probably want to use gmake instead of make.
 
-	If your target system does not have GNU Make (or equivalent), then
-	you should try using the Makefile.simple and custom/Makefile.simple
+	Some report that under macOS, one has to declare the target
+	to be Darwin.  So for macOS, one might have to do:
+
+	    # for macOS users only, force the target to be Darwin
+	    #
+	    make target=Darwin clobber
+	    make target=Darwin all
+	    make target=Darwin chk
+	    make target=Darwin install
+
+	If your target system does not have a moderm Makefile (such as gmake),
+	then you should try using the Makefile.simple and custom/Makefile.simple
 	files:
 
 	    mv Makefile Makefile.gmake
@@ -46,6 +68,14 @@ Installing calc from the gziped tarball in 4 easy steps:
 	values should work.  If in doubt, follow the 'When in doubt'
 	suggestion.
 
+	If you are using a modern make (such as gmake), you may override
+	any values set in the Mkefile by adding them to Makefile.local
+	using the := directive.  For example:
+
+	    HAVE_STRING_H:= YES
+	    HAVE_TIMES_H:= YES
+	    SED:= /usr/local/bin/nsed
+
 	!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 	! If you are building under Windoz or a Windoz-like environment !
 	! (such as Cygwin or DJGPP), read the README.WINDOWS file.      !
@@ -110,7 +140,7 @@ Installing calc from the gziped tarball in 4 easy steps:
 	    found below, comment out the Linux set and comment in that
 	    set or edit the gcc set or the common cc set as needed.
 
-	You may want to change these Makrfile variables from their defaults:
+	You may want to change these Makefile variables from their defaults:
 
 	    RANLIB
 
@@ -138,11 +168,11 @@ Installing calc from the gziped tarball in 4 easy steps:
 
 	   Adjust other Makefile variables as needed.
 
-    2)	build calc:
+    (2)	build calc:
 
 	The top level Makefile and the custom/Makefile require a GNU
 	Make (such as gmake) or an equivalently advanced make.	On many
-	targets, the default make is sufficent.  On FreeBSD for example,
+	targets, the default make is sufficient.  On FreeBSD for example,
 	one must use gmake instead of make.
 
 	If your target system does not have GNU Make (or equivalent), then
@@ -170,7 +200,7 @@ Installing calc from the gziped tarball in 4 easy steps:
 		make clobber
 		make calc-dynamic-only BLD_TYPE=calc-dynamic-only
 
-    3)	test calc:
+    (3)	test calc:
 
 	make check
 
@@ -181,14 +211,223 @@ Installing calc from the gziped tarball in 4 easy steps:
 
 		make chk
 
-    4)	install calc:
+    (4)	install calc:
 
 	make install
 
-We suggest that you might want to read the README file and look at
-the calc help subsystem.  See the README file for details.
+We suggest that you might want to read the README.FIRST file and look at
+the calc help subsystem.  See also the README.md file.
 
-## Copyright (C) 1999-2007  Landon Curt Noll
+=-=
+
+On calc Makefiles:
+
+How to tell the origin of of a Makefile:
+
+    The "# SRC: ... - ..." comment line near the top
+    of the file indicates the origin of this file.
+    In each segment below, we indicate what the SRC
+    comment like will read.
+
+SHELL= ...
+
+    On some systems, /bin/sh is a rather reduced shell with
+    deprecated behavior.
+
+    If your system has a up to date, bash shell, then
+    you may wish to edit the Makefile to use:
+
+	    SHELL= /bin/bash
+
+    On some systems such as macOS, the bash shell is very
+    far behind to the point where is cannot be depended on.
+    On such systems, the zsh may be a much better alternative
+    shell for this Makefile to use:
+
+	    SHELL= /bin/zsh
+
+Makefile
+
+    # SRC: Makefile - Our calc build environment
+
+    This is our internal Makefile that drives how we build calc.
+    For critical rules in Makefile.ship, there is a corresponding
+    rule in Makefile that calls ${MAKE} -f Makefile.ship with,
+    additional commands, args and Makefile variables.  The ${XVAR}
+    Makefile variable, for example, is passed on each call to
+    ${MAKE} -f Makefile.ship so that the environment and parameters of
+    Makefile.ship may be managed for our build environment.
+
+    This Makefile is used to make the calc rpm.  In addition to these
+    comments, this Makefile differs from the non-rpm bzip2-ed tarball
+    source Makefile in at least the the following ways:
+
+	CCWERR= -Werror
+	USE_READLINE= -DUSE_READLINE
+	READLINE_LIB= -lreadline -lhistory -lncurses
+
+    If the Makefile is not suitable for you, then
+    you may wish to replace it with Makefile.ship:
+
+	mv -f -v Makefile.ship Makefile
+
+    This Makefile assumes you have a modern make command such as
+    the GNU make.  See Makefile.simple comment below if you do not
+    have such a modern make command.
+
+    In packages such as RPMs, and the tar.bz2 source tarball,
+    Makefile contains the contents of Makefile.ship.  If Makefile.ship
+    is missing it has likely moved to replace Makefile.
+
+    In the calc GitHub repo, Makefile is the calc build environment
+    and Makefile.ship is the top level Makefile:
+
+	https://github.com/lcn2/calc
+
+Makefile.ship
+
+    # SRC: Makefile.ship - top level Makefile
+
+    This is the main top level Makefile.
+
+    In calc packages such as RPMs, and the tar.bz2 source tarball,
+    Makefile.ship has been moved into Makefile.
+
+    In the calc GitHub repo, Makefile is the calc build environment
+    and Makefile.ship is the top level Makefile:
+
+	https://github.com/lcn2/calc
+
+    The Makefile.simple rule of Makefile.ship, when forming the
+    Makefile.simple file, removes lines from Makefile.ship between
+    pairs of '#if 0' AND '#endif':
+
+	#if 0
+	lines removed when forming Makefile.simple
+	...
+	lines removed when forming Makefile.simple
+	#endif
+
+    The '#if 0' AND '#endif' MUST be at the start of the line.
+    Any text after the '#if 0' OR '#endif' is ignored.
+
+    While they may look like a CPP directives, they are not.  The
+    inline awk script of the Makefile.simple rule does NOT allow
+    them to nest:
+
+	#if 0 /* DO NOT DO THIS */
+	lines removed when forming Makefile.simple
+	...
+	#if 0 /* DO NOT DO THIS */
+	...
+	#endif /* DO NOT DO THIS */
+	...
+	#endif /* DO NOT DO THIS */
+
+    The custom/Makefile rule of Makefile.ship is used to form
+    the custom/Makefile.  In particular the Makefile.ship lines:
+
+	# start of host target cut
+	... these go in between custom/Makefile.head and custom/Makefile.tail
+	# end of host target cut
+
+Makefile.simple
+
+    # SRC: non-GNU Makefile via make -f Makefile.ship Makefile.simple
+
+    This is a non-GNU or simple Makefile designed for environments
+    that do not have a modern make command.
+
+    If you have a Makefile.simple file, use these commands to
+    form a Makefile:
+
+	if [ -f Makefile ]; then mv -f Makefile Makefile.orig; fi
+	cp Makefile.simple Makefile
+
+    The Makefile.simple rule from Makefile.ship is used to construct
+    this file from the contents of Makefile.ship.
+
+    In calc packages such as RPMs, and the tar.bz2 source the
+    Makefile.simple exists.  In the calc GitHub repo:
+
+	https://github.com/lcn2/calc
+
+    this file does NOT exist and must be made via the Makefile.simple
+    make rule.  Therefore, non-GNU and simple make commands are NOT
+    supported by the calc GitHub repo master branch.  Instead, you
+    need to extract Makefile.simple from one of the calc tar.bz2
+    source tarball source from a calc source mirror:
+
+	http://www.isthe.com/chongo/tech/comp/calc/calc-mirror.html
+
+custom/Makefile.head
+
+    # SRC: custom/Makefile.head
+
+    This file forms the top part of the custom/Makefile.
+
+custom/Makefile.tail
+
+    # SRC: custom/Makefile.tail
+
+    This file forms the bottom part of the custom/Makefile.
+
+custom/Makefile
+
+    # SRC: Makefile via make -f Makefile custom/Makefile
+
+    This is the custom directory Makefile.
+
+    The custom/Makefile rule of Makefile.ship is used to form
+    the custom/Makefile.  In particular the Makefile.ship lines:
+
+	# start of host target cut
+	... these go in between custom/Makefile.head and custom/Makefile.tail
+	# end of host target cut
+
+    NOTE: The clobber rule does not remove this file. - XXX
+
+custom/Makefile.simple
+
+    # SRC: non-GNU Makefile via make -f Makefile custom/Makefile.simple
+
+    This is a non-GNU or simple Makefile for the custom directory
+    that is designed for environments that do not have a modern make
+    command.
+
+    In calc packages such as RPMs, and the tar.bz2 source the
+    Makefile.simple exists.  In the calc GitHub repo:
+
+	https://github.com/lcn2/calc
+
+    this file does NOT exist and must be made via the
+    custom/Makefile.simple make rule.  Therefore, non-GNU and simple
+    make commands are NOT supported by the calc GitHub repo master
+    branch.  Instead, you need to extract custom/Makefile.simple
+    from one of the calc tar.bz2 source tarball source from a calc
+    source mirror:
+
+	http://www.isthe.com/chongo/tech/comp/calc/calc-mirror.html
+
+cal/Makefile
+
+    # SRC: cal/Makefile
+
+    The Makefile for the cal sub-directory.
+
+cscript/Makefile
+
+    # SRC: cscript/Makefile
+
+    The Makefile for the cscript sub-directory.
+
+help/Makefile
+
+    # SRC: help/Makefile
+
+    The Makefile for the help sub-directory.
+
+## Copyright (C) 1999-2007,2021  Landon Curt Noll
 ##
 ## Calc is open software; you can redistribute it and/or modify it under
 ## the terms of the version 2.1 of the GNU Lesser General Public License
@@ -204,10 +443,6 @@ the calc help subsystem.  See the README file for details.
 ## received a copy with calc; if not, write to Free Software Foundation, Inc.
 ## 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
 ##
-## @(#) $Revision: 30.5 $
-## @(#) $Id: HOWTO.INSTALL,v 30.5 2007/09/01 19:54:03 chongo Exp $
-## @(#) $Source: /usr/local/src/cmd/calc/RCS/HOWTO.INSTALL,v $
-##
 ## Under source code control:	1999/09/27 20:48:44
 ## File existed as early as:	1999
 ##

LIBRARY

@@ -638,10 +638,6 @@ need call libcalc_call_me_last() only once.
 ## received a copy with calc; if not, write to Free Software Foundation, Inc.
 ## 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
 ##
-## @(#) $Revision: 30.1 $
-## @(#) $Id: LIBRARY,v 30.1 2007/03/16 11:09:46 chongo Exp $
-## @(#) $Source: /usr/local/src/cmd/calc/RCS/LIBRARY,v $
-##
 ## Under source code control:	1993/07/30 19:44:49
 ## File existed as early as:	1993
 ##

Makefile.local

@@ -0,0 +1 @@
+# Add below to override Makefile values (using :=) as in: HAVE_STRING_H:= YES

QUESTIONS

@@ -0,0 +1,54 @@
+If you have a simple general question about calc, send Email to:
+
+    calc-quest-mail at asthe dot com
+
+    NOTE: Remove spaces and replace 'at' with @, and 'dot' with .
+
+PLEASE put following the SPECIAL PHRASE somewhere in your Email Subject line:
+
+    calc question
+
+You may add additional words to your subject line.
+
+IMPORTANT: If your Email doesn't contain the above phrase,
+	   then we WILL NOT SEE your Email.
+
+PLEASE BE SURE you have that SPECIAL PHRASE somewhere in the subject line!
+
+Suggestion:
+
+    From time to time, the Email address and Subject SPECIAL PHRASE
+    may change so verify you have the current info by visiting:
+
+	http://www.isthe.com/chongo/tech/comp/calc/calc-question.html
+
+Please limit your questions to general questions about calc.  We
+cannot go into great detail in our answers, nor can we do your
+homework, nor can we do much more than answer short general questions
+about calc.
+
+Please be patient as we cannot always respond to Email messages quickly.
+
+=-=
+
+## Copyright (C) 2021  Landon Curt Noll
+##
+## Calc is open software; you can redistribute it and/or modify it under
+## the terms of the version 2.1 of the GNU Lesser General Public License
+## as published by the Free Software Foundation.
+##
+## Calc is distributed in the hope that it will be useful, but WITHOUT
+## ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+## or FITNESS FOR A PARTICULAR PURPOSE.	 See the GNU Lesser General
+## Public License for more details.
+##
+## A copy of version 2.1 of the GNU Lesser General Public License is
+## distributed with calc under the filename COPYING-LGPL.  You should have
+## received a copy with calc; if not, write to Free Software Foundation, Inc.
+## 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+##
+## Under source code control:	2021/02/10 00:15:05
+## File existed as early as:	2021
+##
+## chongo <was here> /\oo/\	http://www.isthe.com/chongo/
+## Share and enjoy!  :-)	http://www.isthe.com/chongo/tech/comp/calc/

README.FIRST

@@ -32,30 +32,30 @@ something that you type in.
 
 For list of help topics:
 
-    > help
+    ; help
 
 For overview of calc overview:
 
-    > help intro
-    > help overview
-    > help command
-    > help define
-    > help statement
-    > help variable
-    > help usage
+    ; help intro
+    ; help overview
+    ; help command
+    ; help define
+    ; help statement
+    ; help variable
+    ; help usage
 
 For list of builtin functions:
 
-    > help builtin
+    ; help builtin
 
 C programmers should note some unexpected differences in the calc syntax:
 
-    > help unexpected
+    ; help unexpected
 
 Calc is shipped with a standard collection of calc resource files.
 For a list of calc standard resource files see:
 
-    > help resource
+    ; help resource
 
 =-=
 
@@ -71,54 +71,33 @@ or run:
 
 for a wish/todo list.  Code contributions are welcome.
 
-=-=
+-=-
 
-To join the calc-tester mailing list.  Send an EMail message to:
+If you you notice something wrong, strange or broken, see the file:
 
-    calc-tester-request at asthe dot com
+    BUGS
 
-    [[ NOTE: Replace 'at' with @, 'dot' is with . and remove the spaces ]]
-    [[ NOTE: The EMail address uses 'asthe' and the web site URL uses 'isthe' ]]
+or run:
 
-Your subject must contain the words:
+    calc help bugs
 
-    calc mailing list subscription
+for information about how to report a bug.
 
-You may have additional words in your subject line.
+-=-
 
-Your message body (not the subject) should consist of:
+If you have a general question about calc, see the file:
 
-     subscribe calc-tester address
-     end
-     name your_full_name
+    QUESTIONS
 
-where ``address'' is your EMail address and ``your_full_name'' is
-your full name.
+or run:
 
-Feel free to follow the name line with additional EMail text as desired.
+    calc help questions
 
-=-=
+for information about how to ask a question.
 
-Send Calc bug and bug fixes to:
+-=-
 
-    calc-bugs at asthe dot com
-
-    [[ NOTE: Replace 'at' with @, 'dot' is with . and remove the spaces ]]
-    [[ NOTE: The EMail address uses 'asthe' and the web site URL uses 'isthe' ]]
-
-but see the BUGS file first.
-
-Your subject must contain the words:
-
-       calc bug report
-
-You may have additional words in your subject line.
-
-The calc web site is located at:
-
-	http://www.isthe.com/chongo/tech/comp/calc/
-
-## Copyright (C) 1999  Landon Curt Noll
+## Copyright (C) 1999,2014,2017,2021  Landon Curt Noll
 ##
 ## Calc is open software; you can redistribute it and/or modify it under
 ## the terms of the version 2.1 of the GNU Lesser General Public License
@@ -134,10 +113,6 @@ The calc web site is located at:
 ## received a copy with calc; if not, write to Free Software Foundation, Inc.
 ## 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
 ##
-## @(#) $Revision: 30.1 $
-## @(#) $Id: README,v 30.1 2007/03/16 11:09:46 chongo Exp $
-## @(#) $Source: /usr/local/src/cmd/calc/RCS/README,v $
-##
 ## Under source code control:	1995/10/25 05:27:59
 ## File existed as early as:	1995
 ##

README.WINDOWS

@@ -10,6 +10,78 @@ NOTE: The main developers do not have access to a Windoz based platform.
       Of course you are welcome to send us any patches that fix your
       Windoz build environment.
 
+
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
+=-= compiling with Windows Subsystem for Linux (WSL) =-Cygwin =-=
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
+
+It has been reported that calc version 2.12.6.4 has been successfully
+compiled, installed and running on Windows 10 on 2018 Jan 21.
+
+We were told:
+
+  "The Windows Subsystem for Linux (WSL) is a new Windows 10 feature that
+   enables you to run native Linux command-line tools directly on Windows"
+
+   https://docs.microsoft.com/cs-cz/windows/wsl/about
+
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
+=-= compiling with Cygwin =-=
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
+
+An effort is being made to allow windows users to compile calc using the
+Cygwin project (http://sources.redhat.com/cygwin/) with the GCC compiler
+and Un*x tools for Windows.
+
+The major porting work for Cygwin was performed by Thomas Jones-Low
+(tjoneslo at softstart dot com).
+
+In March 2009, Michael Penk (mpenk at wuska dot com) reported success in
+installs under Cygwin:
+
+    On my fairly complete Cygwin installs, everything compiles,
+    checks, and installs correctly.  My Cygwin is configured
+    in a very standard way (out of the box, using all of Cygwin's
+    defaults).  The install worked on 5 different machines with
+    Cygwin on them: one XP home, one XP professional, and three
+    Vista professionals.
+
+Using the calc Makefile, he did the following:
+
+    make all target=Cygwin
+    make check
+    make install
+
+He also reports:
+
+    Of course, one should be logged in as an Administrator when
+    one builds and installs calc.
+
+He was compiling calc 2.12.4.0 with Cygwin version 1.5.25-15.
+
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
+=-= If all else fails, for Cygwin =-=
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
+
+Much earlier (2001?) Thomas Jones-Low (tjoneslo at softstart dot com)
+recommended that you generate by hand all of the header files that
+by the Makefile.  This has been done for you via the makefile rule:
+
+	make win32_hsrc
+
+which uses the Makefile variables in win32.mkdef to form these header
+files under win32 directory.
+
+You will find generated versions of these files located in the win32
+sub-directory.  These files may be appropriate for your Cygwin building
+needs.
+
+In particular:
+
+	Just copy the win32/*.[ch] files up into the top level calc
+	source directory, edit them (if needed) and build using the
+	Cygwin GCC compiler and Cygwin build environment.
+
 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 =-= compiling under DJGPP =-=
 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=
@@ -54,7 +126,7 @@ recommends the following settings:
 	CATDIR= /dev/env/DJDIR/man/cat1
 	NROFF= groff
 	CALCPATH= .;./cal;~/.cal;${CALC_SHAREDIR};${CUSTOMCALDIR}
-	CALCRC= ${CALC_SHAREDIR}/startup;~/.calcrc;./.calcinit
+	CALCRC= ./.calcinit;~/.calcrc;${CALC_SHAREDIR}/startup
 	CALCPAGER= less.exe -ci
 	DEBUG= -O2 -gstabs+ -DWINDOZ
 
@@ -68,52 +140,8 @@ Look for Makefile comments of the form:
 Follow those recommendations.  In cases where they conflict with
 the above Makefile list, follow the recommendation in the Makefile.
 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
-=-= compiling with Cygwin =-=
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 
-An effort is being made to allow windows users to compile calc using the
-Cygwin project (http://sources.redhat.com/cygwin/) with the GCC compiler
-and Un*x tools for Windows.
-
-The major porting work was performed by Thomas Jones-Low
-(tjoneslo at softstart dot com).  He said:
-
-    I had previous stated to this group that I have successfully managed
-    to port a version of Calc to Windows, and promised some point to
-    post what was required, so here it is.
-
-    One obvious manner of doing this port is to get the latest version
-    of the Cygwin project (http://sources.redhat.com/cygwin/) with the
-    GCC compiler and Un*x tools for Windows and recompile.
-
-    I built my working version using Calc ... I am using Visual C++
-    version 7.0, which is an older version of the Microsoft development
-    tools. The make file provided with Calc is not compatible with
-    NMAKE, so I used the Visual Studio tools to generate another one
-    (not included). Calc is built in two parts, calc.dll, which is the
-    library, and calc.exe which is the command line interface.
-
-He recommended that you generate by hand all of the header files that
-by the Makefile.  This has been done for you via the makefile rule:
-
-	make win32_hsrc
-
-which uses the Makefile variables in win32.mkdef to form these header
-files under win32 directory.
-
-You will find generated versions of these files located in the win32
-sub-directory.  These files may be appropriate for your Cygwin building
-needs.
-
-In particular:
-
-	Just copy the win32/*.[ch] files up into the top level calc
-	source directory, edit them (if needed) and build using the
-	Cygwin GCC compiler and Cygwin build environment.
-
-
-## Copyright (C) 2002-2007  Landon Curt Noll and Thomas Jones-Low
+## Copyright (C) 2002-2009  Landon Curt Noll and Thomas Jones-Low
 ##
 ## Calc is open software; you can redistribute it and/or modify it under
 ## the terms of the version 2.1 of the GNU Lesser General Public License
@@ -129,10 +157,6 @@ In particular:
 ## received a copy with calc; if not, write to Free Software Foundation, Inc.
 ## 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
 ##
-## @(#) $Revision: 30.1 $
-## @(#) $Id: README.WINDOWS,v 30.1 2007/03/16 11:09:46 chongo Exp $
-## @(#) $Source: /usr/local/src/cmd/calc/RCS/README.WINDOWS,v $
-##
 ## Under source code control:	2001/02/25 14:00:05
 ## File existed as early as:	2001
 ##

README.md

@@ -0,0 +1,263 @@
+# What is calc?
+
+Calc is an interactive calculator which provides for easy large
+numeric calculations, but which also can be easily programmed
+for difficult or long calculations.	 It can accept a command line
+argument, in which case it executes that single command and exits.
+Otherwise, it enters interactive mode.  In this mode, it accepts
+commands one at a time, processes them, and displays the answers.
+In the simplest case, commands are simply expressions which are
+evaluated.  For example, the following line can be input:
+
+```sh
+3 * (4 + 1)
+```
+
+and the calculator will print:
+
+```sh
+15
+```
+
+Calc has the usual collection of arithmetic operators +, -, /, * as
+well as ^ (exponentiation), % (modulus) and // (integer divide).
+For example:
+
+```sh
+3 * 19^43 - 1
+```
+
+will produce:
+
+```sh
+29075426613099201338473141505176993450849249622191102976
+```
+
+Notice that calc values can be very large.  For example:
+
+```sh
+2^23209-1
+```
+
+will print:
+
+```sh
+402874115778988778181873329071 ... many digits ... 3779264511
+```
+
+The special '.' symbol (called dot), represents the result of the
+last command expression, if any.  This is of great use when a series
+of partial results are calculated, or when the output mode is changed
+and the last result needs to be redisplayed.  For example, the above
+result can be modified by typing:
+
+```sh
+. % (2^127-1)
+```
+
+and the calculator will print:
+
+```sh
+47385033654019111249345128555354223304
+```
+
+For more complex calculations, variables can be used to save the
+intermediate results.  For example, the result of adding 7 to the
+previous result can be saved by typing:
+
+```sh
+curds = 15
+whey = 7 + 2*curds
+```
+
+Functions can be used in expressions.  There are a great number of
+pre-defined functions.  For example, the following will calculate
+the factorial of the value of 'whey':
+
+```sh
+fact(whey)
+```
+
+and the calculator prints:
+
+```sh
+13763753091226345046315979581580902400000000
+```
+
+The calculator also knows about complex numbers, so that typing:
+
+```sh
+(2+3i) * (4-3i)
+cos(.)
+```
+
+will print:
+
+```sh
+17+6i
+-55.50474777265624667147+193.9265235748927986537i
+```
+
+The calculator can calculate transcendental functions, and accept and
+display numbers in real or exponential format. For example, typing:
+
+```sh
+config("display", 70),
+epsilon(1e-70),
+sin(1)
+```
+
+prints:
+
+```sh
+0.8414709848078965066525023216302989996225630607983710656727517099919104
+```
+
+Calc can output values in terms of fractions, octal or hexadecimal.
+For example:
+
+```sh
+config("mode", "fraction"),
+(17/19)^23
+print
+base(16),
+(19/17)^29
+print
+log(79.3i)
+```
+
+ will print:
+
+```sh
+19967568900859523802559065713/257829627945307727248226067259
+
+0x9201e65bdbb801eaf403f657efcf863/0x5cd2e2a01291ffd73bee6aa7dcf7d1
+
+0x17b5164ac24ee836bf/0xc7b7a8e3ef5fcf752+0x883eaf5adadd26be3/0xc7b7a8e3ef5fcf752i
+```
+
+All numbers are represented as fractions with arbitrarily large
+numerators and denominators which are always reduced to lowest terms.
+Real or exponential format numbers can be input and are converted
+to the equivalent fraction.  Hex, binary, or octal numbers can be
+input by using numbers with leading '0x', '0b' or '0' characters.
+Complex numbers can be input using a trailing 'i', as in '2+3i'.
+Strings and characters are input by using single or double quotes.
+
+Commands are statements in a C-like language, where each input
+line is treated as the body of a procedure.  Thus the command
+line can contain variable declarations, expressions, labels,
+conditional tests, and loops.  Assignments to any variable name
+will automatically define that name as a global variable.  The
+other important thing to know is that all non-assignment expressions
+which are evaluated are automatically printed.  Thus, you can evaluate
+an expression's value by simply typing it in.
+
+Many useful built-in mathematical functions are available.  Use the:
+
+```sh
+help builtin
+```
+
+command to list them.
+
+You can also define your own functions by using the 'define' keyword,
+followed by a function declaration very similar to C.
+
+```sh
+define f2(n)
+{
+	local ans;
+
+	ans = 1;
+	while (n > 1)
+		ans *= (n -= 2);
+	return ans;
+}
+```
+
+Thus the input:
+
+```sh
+f2(79)
+```
+
+will produce:
+
+```sh
+1009847364737869270905302433221592504062302663202724609375
+```
+
+Functions which only need to return a simple expression can be defined
+using an equals sign, as in the example:
+
+```sh
+define sc(a,b) = a^3 + b^3
+```
+
+Thus the input:
+
+```sh
+sc(31, 61)
+```
+
+will produce:
+
+```sh
+256772
+```
+
+Variables in functions can be defined as either 'global', 'local',
+or 'static'.  Global variables are common to all functions and the
+command line, whereas local variables are unique to each function
+level, and are destroyed when the function returns.  Static variables
+are scoped within single input files, or within functions, and are
+never destroyed.  Variables are not typed at definition time, but
+dynamically change as they are used.
+
+For more information about the calc language and features, try:
+
+```sh
+help overview
+```
+
+Calc has a help command that will produce information about
+every builtin function, command as well as a number of other
+aspects of calc usage.  Try the command:
+
+```sh
+help help
+```
+
+for and overview of the help system.  The command:
+
+```sh
+help builtin
+```
+
+provides information on built-in mathematical functions, whereas:
+
+```sh
+help asinh
+```
+
+will provides information a specific function.  The following
+help files:
+
+```sh
+help command
+help define
+help operator
+help statement
+help variable
+```
+
+provide a good overview of the calc language.  If you are familiar
+with C, you should also try:
+
+```sh
+help unexpected
+```
+
+It contains information about differences between C and calc
+that may surprise C programmers.

addop.c

@@ -1,7 +1,7 @@
 /*
  * addop - add opcodes to a function being compiled
  *
- * Copyright (C) 1999-2007  David I. Bell and Ernest Bowen
+ * Copyright (C) 1999-2007,2021  David I. Bell and Ernest Bowen
  *
  * Primary author:  David I. Bell
  *
@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: addop.c,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/addop.c,v $
- *
  * Under source code control:	1990/02/15 01:48:10
  * File existed as early as:	before 1990
  *
@@ -32,6 +28,7 @@
 
 #include <stdio.h>
 #include "calc.h"
+#include "alloc.h"
 #include "opcodes.h"
 #include "str.h"
 #include "func.h"
@@ -40,6 +37,9 @@
 #include "symbol.h"
 
 
+#include "banned.h"	/* include after system header <> includes */
+
+
 #define FUNCALLOCSIZE	20	/* reallocate size for functions */
 #define OPCODEALLOCSIZE 100	/* reallocate size for opcodes in functions */
 

align32.c

@@ -1,7 +1,7 @@
 /*
  * align32 - determine if 32 bit accesses must be aligned
  *
- * Copyright (C) 1999  Landon Curt Noll
+ * Copyright (C) 1999,2021  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: align32.c,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/align32.c,v $
- *
  * Under source code control:	1995/11/23 05:18:06
  * File existed as early as:	1995
  *
@@ -38,6 +34,10 @@
 #include <unistd.h>
 #endif
 
+
+#include "banned.h"	/* include after system header <> includes */
+
+
 static void buserr(void);	/* catch alignment errors */
 
 

alloc.h

@@ -1,7 +1,7 @@
 /*
  * alloc - storage allocation and storage debug macros
  *
- * Copyright (C) 1999-2007  David I. Bell
+ * Copyright (C) 1999-2007,2014  David I. Bell
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: alloc.h,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/alloc.h,v $
- *
  * Under source code control:	1990/02/15 01:48:29
  * File existed as early as:	before 1990
  *
@@ -28,8 +24,8 @@
  */
 
 
-#if !defined(__ALLOC_H__)
-#define __ALLOC_H__
+#if !defined(INCLUDE_ALLOC_H)
+#define INCLUDE_ALLOC_H
 
 
 #if defined(CALC_SRC)	/* if we are building from the calc source tree */
@@ -46,11 +42,15 @@
 # include <string.h>
 
 #else
+#if defined(_WIN32) && defined(NOTCYGWIN)
+#include <stdio.h>
+#endif
 
 # if defined(HAVE_NEWSTR)
 E_FUNC void *memcpy();
 E_FUNC void *memset();
-#if defined(FORCE_STDC) || (defined(__STDC__) && __STDC__ != 0) || defined(__cplusplus)
+#if defined(FORCE_STDC) || \
+   (defined(__STDC__) && __STDC__ != 0) || defined(__cplusplus)
 E_FUNC size_t strlen();
 #  else
 E_FUNC long strlen();
@@ -79,7 +79,8 @@ E_FUNC int strcmp();
 
 #if !defined(HAVE_MEMMOVE)
 # undef MEMMOVE_SIZE_T
-#if defined(FORCE_STDC) || (defined(__STDC__) && __STDC__ != 0) || defined(__cplusplus)
+#if defined(FORCE_STDC) || \
+   (defined(__STDC__) && __STDC__ != 0) || defined(__cplusplus)
 #  define MEMMOVE_SIZE_T size_t
 # else
 #  define MEMMOVE_SIZE_T long
@@ -87,4 +88,4 @@ E_FUNC int strcmp();
 E_FUNC void *memmove(void *s1, CONST void *s2, MEMMOVE_SIZE_T n);
 #endif
 
-#endif /* !__ALLOC_H__ */
+#endif /* !INCLUDE_ALLOC_H */

assocfunc.c

@@ -1,7 +1,7 @@
 /*
  * assocfunc - association table routines
  *
- * Copyright (C) 1999-2007  David I. Bell
+ * Copyright (C) 1999-2007,2021  David I. Bell
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: assocfunc.c,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/assocfunc.c,v $
- *
  * Under source code control:	1993/07/20 23:04:27
  * File existed as early as:	1993
  *
@@ -40,6 +36,9 @@
 #include "value.h"
 
 
+#include "banned.h"	/* include after system header <> includes */
+
+
 #define MINHASHSIZE	31	/* minimum size of hash tables */
 #define GROWHASHSIZE	50	/* approximate growth for hash tables */
 #define CHAINLENGTH	10	/* desired number of elements on a hash chain */
@@ -83,7 +82,7 @@ associndex(ASSOC *ap, BOOL create, long dim, VALUE *indices)
 	 * so that we can first select the correct hash chain, and
 	 * also so we can quickly compare each element for a match.
 	 */
-	hash = FNV1_32_BASIS;
+	hash = QUICKHASH_BASIS;
 	for (i = 0; i < dim; i++)
 		hash = hashvalue(&indices[i], hash);
 
@@ -332,7 +331,8 @@ assoccopy(ASSOC *oldap)
 			oldep = oldep->e_next) {
 			ep = (ASSOCELEM *) malloc(ELEMSIZE(oldep->e_dim));
 			if (ep == NULL) {
-				math_error("Cannot allocate association element");
+				math_error("Cannot allocate "
+					   "association element");
 				/*NOTREACHED*/
 			}
 			ep->e_dim = oldep->e_dim;
@@ -340,7 +340,8 @@ assoccopy(ASSOC *oldap)
 			ep->e_value.v_type = V_NULL;
 			ep->e_value.v_subtype = V_NOSUBTYPE;
 			for (i = 0; i < ep->e_dim; i++)
-				copyvalue(&oldep->e_indices[i], &ep->e_indices[i]);
+				copyvalue(&oldep->e_indices[i],
+					  &ep->e_indices[i]);
 			copyvalue(&oldep->e_value, &ep->e_value);
 			listhead = &ap->a_table[ep->e_hash % ap->a_size];
 			ep->e_next = *listhead;

banned.h

@@ -0,0 +1,180 @@
+/*
+ * banned - optionally ban dqngerious functions
+ *
+ * Unless UNBAN is defined, this file will turn the use
+ * of certain dangerous functions into syntax errors.
+ *
+ * In the case of calc, we are motivated in part by the desire for
+ * calc to correctly calculate: even durings extremely long calculations.
+ *
+ * If UNBAN is NOT defined, then calling certain functions
+ * will result in a syntaxc error.
+ *
+ * If we define UNBAN, then the effect of this file is disabled.
+ *
+ * The banned.h attempts to ban the use of certain dangerous functions
+ * that, if improperly used, could compromise the computational integrity
+ * if calculations.
+ *
+ * In the case of calc, we are motivated in part by the desire for calc
+ * to correctly calculate: even durings extremely long calculations.
+ *
+ * If UNBAN is NOT defined, then calling certain functions
+ * will result in a call to a non-existent function (link error).
+ *
+ * While we do NOT encourage defining UNBAN, there may be
+ * a system / compiler environment where re-defining a
+ * function may lead to a fatal compiler complication.
+ * If that happens, consider compiling as:
+ *
+ *	make clobber all chk CCBAN=-DUNBAN
+ *
+ * as see if this is a work-a-round.
+ *
+ * If YOU discover a need for the -DUNBAN work-a-round, PLEASE tell us!
+ * Please send us a bug report.  See the file:
+ *
+ *	BUGS
+ *
+ * or the URL:
+ *
+ *	http://www.isthe.com/chongo/tech/comp/calc/calc-bugrept.html
+ *
+ * for how to send us such a bug report.
+ *
+ * Copyright (C) 2021  Landon Curt Noll
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.	 See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2021/03/06 21:07:31
+ * File existed as early as:	2021
+ *
+ * chongo <was here> /\oo/\	http://www.isthe.com/chongo/
+ * Share and enjoy!  :-)	http://www.isthe.com/chongo/tech/comp/calc/
+ */
+
+
+#if !defined(PRE_HAVE_BAN_PRAGMA_H)
+#include "have_ban_pragma.h"
+#endif /* ! PRE_HAVE_BAN_PRAGMA_H */
+
+
+#if !defined(INCLUDE_BANNED_H)
+#define INCLUDE_BANNED_H
+
+/*
+ * If we define UNBAN, then the effect of this file is disabled.
+ */
+#if !defined(UNBAN)
+
+/*
+ * In the spirit of:
+ *
+ *	https://github.com/git/git/blob/master/banned.h
+ *
+ * we will ban the use of certain unsafe functions by turning
+ * then into function calls that do not exist.
+ *
+ * In the case of calc, we are motivated in part by the desire
+ * for calc to correctly calculate: even durings extremely long
+ * calculations.
+ *
+ * If UNBAN is NOT defined, then calling certain functions
+ * will result in a syntaxc error.
+ *
+ * Unlike the above URL, we suggest an alternative function.
+ * In many cases, additional logic is required to use the
+ * alternative function, we cannot simply replace one function
+ * with another.
+ */
+
+/*
+ * If one is not careful, strcpy() can lead to buffer overflows.
+ * Use strlcpy() instead.
+ */
+#if defined(HAVE_PRAGMA_GCC_POSION)
+#undef strcpy
+#pragma GCC poison strcpy
+#endif /* HAVE_PRAGMA_GCC_POSION */
+
+/*
+ * If one is not careful, strcat() can lead to buffer overflows.
+ * Use strlcat() instead.
+ */
+#if defined(HAVE_PRAGMA_GCC_POSION)
+#undef strcat
+#pragma GCC poison strcat
+#endif /* HAVE_PRAGMA_GCC_POSION */
+
+/*
+ * If one is not careful, strncpy() can lead to buffer overflows.
+ * Use memccpy() instead.
+ */
+#if defined(HAVE_PRAGMA_GCC_POSION)
+#undef strncpy
+#pragma GCC poison strncpy
+#endif /* HAVE_PRAGMA_GCC_POSION */
+
+/*
+ * If one is not careful, strncat() can lead to buffer overflows.
+ * Use memccpy() instead.
+ */
+#if defined(HAVE_PRAGMA_GCC_POSION)
+#undef strncat
+#pragma GCC poison strncat
+#endif /* HAVE_PRAGMA_GCC_POSION */
+
+/*
+ * If one is not careful, sprintf() can lead to buffer overflows.
+ * Use snprintf() instead.
+ */
+#if defined(HAVE_PRAGMA_GCC_POSION)
+#undef sprintf
+#pragma GCC poison sprintf
+#endif /* HAVE_PRAGMA_GCC_POSION */
+
+/*
+ * If one is not careful, vsprintf() can lead to buffer overflows.
+ * Use vsnprintf() instead.
+ */
+#if defined(HAVE_PRAGMA_GCC_POSION)
+#undef vsprintf
+#pragma GCC poison vsprintf
+#endif /* HAVE_PRAGMA_GCC_POSION */
+
+/*
+ * XXX - As of 2021, functions such as:
+ *
+ *	gmtime_s
+ *	localtime_s
+ *	ctime_s
+ *	asctime_s
+ *
+ * are not universal.  We cannot yet ban the following
+ * functions because we do not have a portable AND
+ * widely available alternative.  Therefore we just
+ * have to be extra careful when using:
+ *
+ *	gmtime
+ *	localtime
+ *	ctime
+ *	ctime_r
+ *	asctime
+ *	asctime_r
+ */
+
+#endif /* !UNBAN */
+
+#endif /* !INCLUDE_BANNED_H */

blkcpy.c

@@ -1,7 +1,7 @@
 /*
  * blkcpy - general values and related routines used by the calculator
  *
- * Copyright (C) 1999-2007  Landon Curt Noll and Ernest Bowen
+ * Copyright (C) 1999-2007,2021  Landon Curt Noll and Ernest Bowen
  *
  * Primary author:  Landon Curt Noll
  *
@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: blkcpy.c,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/blkcpy.c,v $
- *
  * Under source code control:	1997/04/18 20:41:26
  * File existed as early as:	1997
  *
@@ -33,12 +29,16 @@
 #include <stdio.h>
 #include <sys/types.h>
 #include "calc.h"
+#include "alloc.h"
 #include "value.h"
 #include "file.h"
 #include "blkcpy.h"
 #include "str.h"
 
 
+#include "banned.h"	/* include after system header <> includes */
+
+
 /*
  * copystod - copy num indexed items from source value to destination value
  *
@@ -374,7 +374,8 @@ copyblk2mat(BLOCK *blk, long ssi, long num, MATRIX *dmat, long dsi)
  * copymat2blk - copy matrix to block
  */
 int
-copymat2blk(MATRIX *smat, long ssi, long num, BLOCK *dblk, long dsi, BOOL noreloc)
+copymat2blk(MATRIX *smat, long ssi, long num, BLOCK *dblk, long dsi,
+	    BOOL noreloc)
 {
 	long i;
 	long newlen;
@@ -720,7 +721,8 @@ copystr2file(STRING *str, long ssi, long num, FILEID id, long dsi)
  * copyblk2blk - copy block to block
  */
 int
-copyblk2blk(BLOCK *sblk, long ssi, long num, BLOCK *dblk, long dsi, BOOL noreloc)
+copyblk2blk(BLOCK *sblk, long ssi, long num, BLOCK *dblk, long dsi,
+	    BOOL noreloc)
 {
 	long	newlen;
 	long	newsize;
@@ -762,7 +764,8 @@ copyblk2blk(BLOCK *sblk, long ssi, long num, BLOCK *dblk, long dsi, BOOL noreloc
  * copystr2blk - copy string to block
  */
 int
-copystr2blk(STRING *str, long ssi, long num, BLOCK *dblk, long dsi, BOOL noreloc)
+copystr2blk(STRING *str, long ssi, long num, BLOCK *dblk, long dsi,
+	    BOOL noreloc)
 {
 	long	len;
 	long	newlen;
@@ -982,7 +985,8 @@ memmove(void *s1, CONST void *s2, MEMMOVE_SIZE_T n)
  * copynum2blk - copy number numerator to block
  */
 int
-copynum2blk(NUMBER *snum, long ssi, long num, BLOCK *dblk, long dsi, BOOL noreloc)
+copynum2blk(NUMBER *snum, long ssi, long num, BLOCK *dblk, long dsi,
+	    BOOL noreloc)
 {
 	size_t	newlen;
 	size_t	newsize;
@@ -1033,13 +1037,14 @@ copynum2blk(NUMBER *snum, long ssi, long num, BLOCK *dblk, long dsi, BOOL norelo
  * copyblk2num - copy block to number
  */
 int
-copyblk2num(BLOCK *sblk, long ssi, long num, NUMBER *dnum, long dsi, NUMBER **res)
+copyblk2num(BLOCK *sblk, long ssi, long num, NUMBER *dnum, long dsi,
+	    NUMBER **res)
 {
 	size_t	newlen;
 	NUMBER *ret;		/* cloned and modified numerator */
 #if CALC_BYTE_ORDER == BIG_ENDIAN
 	HALF *swapped;		/* byte swapped input data */
-	unsigned long halflen;	/* length of the input ounded up to HALFs */
+	unsigned long halflen;	/* length of the input rounded up to HALFs */
 	HALF *h;		/* copy byteswap pointer */
 	unsigned long i;
 #endif

blkcpy.h

@@ -1,7 +1,7 @@
 /*
  * blkcpy - general values and related routines used by the calculator
  *
- * Copyright (C) 1999-2007  Landon Curt Noll and Ernest Bowen
+ * Copyright (C) 1999-2007,2014  Landon Curt Noll and Ernest Bowen
  *
  * Primary author:  Landon Curt Noll
  *
@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: blkcpy.h,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/blkcpy.h,v $
- *
  * Under source code control:	1997/04/18 20:41:25
  * File existed as early as:	1997
  *
@@ -30,8 +26,8 @@
  */
 
 
-#if !defined(__BLKCPY_H__)
-#define __BLKCPY_H__
+#if !defined(INCLUDE_BLKCPY_H)
+#define INCLUDE_BLKCPY_H
 
 /*
  * the main copy gateway function
@@ -59,4 +55,4 @@ E_FUNC int copystr2blk(STRING *, long, long, BLOCK *, long, BOOL);
 E_FUNC int copystr2file(STRING *, long, long, FILEID, long);
 E_FUNC int copystr2str(STRING *, long, long, STRING *, long);
 
-#endif /* !__BLKCPY_H__ */
+#endif /* !INCLUDE_BLKCPY_H */

block.c

@@ -1,7 +1,7 @@
 /*
  * block - fixed, dynamic, fifo and circular memory blocks
  *
- * Copyright (C) 1999-2007  Landon Curt Noll and Ernest Bowen
+ * Copyright (C) 1999-2007,2021  Landon Curt Noll and Ernest Bowen
  *
  * Primary author:  Landon Curt Noll
  *
@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: block.c,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/block.c,v $
- *
  * Under source code control:	1997/02/27 00:29:40
  * File existed as early as:	1997
  *
@@ -32,6 +28,11 @@
 
 
 #include <stdio.h>
+#include "have_string.h"
+#ifdef HAVE_STRING_H
+# include <string.h>
+#endif
+#include "alloc.h"
 #include "value.h"
 #include "zmath.h"
 #include "config.h"
@@ -40,6 +41,10 @@
 #include "str.h"
 #include "calcerr.h"
 
+
+#include "banned.h"	/* include after system header <> includes */
+
+
 #define NBLOCKCHUNK 16
 
 STATIC long nblockcount = 0;
@@ -295,7 +300,7 @@ blkrealloc(BLOCK *blk, int newlen, int newchunk)
 	 * Set the data length
 	 *
 	 * We also know that the new block is not empty since we have
-	 * already dealth with that case above.
+	 * already dealt with that case above.
 	 *
 	 * After this section of code, limit and datalen will be
 	 * correct in terms of the new type.

block.h

@@ -1,7 +1,7 @@
 /*
  * block - fixed, dynamic, fifo and circular memory blocks
  *
- * Copyright (C) 1999-2007  Landon Curt Noll and Ernest Bowen
+ * Copyright (C) 1999-2007,2014,2021  Landon Curt Noll and Ernest Bowen
  *
  * Primary author:  Landon Curt Noll
  *
@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: block.h,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/block.h,v $
- *
  * Under source code control:	1997/02/21 05:03:39
  * File existed as early as:	1997
  *
@@ -31,8 +27,8 @@
  */
 
 
-#if !defined(__BLOCK_H__)
-#define __BLOCK_H__
+#if !defined(INCLUDE_BLOCK_H)
+#define INCLUDE_BLOCK_H
 
 
 /*
@@ -59,7 +55,7 @@
  *		blkchunk defaults to BLK_CHUNKSIZE
  *
  *	blkfree(x)
- *		Reduce storage down to 0 octetes.
+ *		Reduce storage down to 0 octets.
  *
  *	size(x)
  *		The length of data stored in the block.
@@ -92,7 +88,7 @@
  *	blkmove(dest, src, length [, dest_offset [, src_offset]])
  *		0 <= length <= blksize(x)
  *		offset's are restricted in value by block type
- *		overlapping moves are handeled correctly
+ *		overlapping moves are handled correctly
  *
  *	blkccpy(dest, src, stopval, length [, dest_offset [, src_offset]])
  *		0 <= length <= blksize(x)
@@ -108,7 +104,7 @@
  *		b = a*k for some integer k >= 1
  *
  *	scatter(src, dest1, dest2 [, dest3 ] ...)
- *		copy sucessive octets from src into dest1, dest2, ...
+ *		copy successive octets from src into dest1, dest2, ...
  *		    restarting with dest1 after end of list
  *		stops at end of src
  *
@@ -188,7 +184,7 @@ EXTERN int blk_debug;	/* 0 => debug off */
 /* length of data stored in a block */
 #define blklen(blk) ((blk)->datalen)
 
-/* block footpint in memory */
+/* block footprint in memory */
 #define blksizeof(blk) ((blk)->maxsize)
 
 /* block allocation chunk size */
@@ -222,4 +218,4 @@ E_FUNC int countnblocks(void);
 E_FUNC void shownblocks(void);
 
 
-#endif /* !__BLOCK_H__ */
+#endif /* !INCLUDE_BLOCK_H */

byteswap.c

@@ -1,7 +1,7 @@
 /*
  * byteswap - byte swapping routines
  *
- * Copyright (C) 1999  Landon Curt Noll
+ * Copyright (C) 1999,2021  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: byteswap.c,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/byteswap.c,v $
- *
  * Under source code control:	1995/10/11 04:44:01
  * File existed as early as:	1995
  *
@@ -33,6 +29,9 @@
 #include "byteswap.h"
 
 
+#include "banned.h"	/* include after system header <> includes */
+
+
 /*
  * swap_b8_in_HALFs - swap 8 and if needed, 16 bits in an array of HALFs
  *
@@ -81,7 +80,7 @@ swap_b8_in_HALFs(HALF *dest, HALF *src, LEN len)
  *		  NULL to allocate the storage
  *	src	- pointer to a ZVALUE to swap
  *	all	- TRUE => swap every element, FALSE => swap only the
- *		  multiprecision storage
+ *		  multi-precision storage
  *
  * returns:
  *	pointer to where the swapped src has been put
@@ -102,7 +101,8 @@ swap_b8_in_ZVALUE(ZVALUE *dest, ZVALUE *src, BOOL all)
 		 */
 		dest = malloc(sizeof(ZVALUE));
 		if (dest == NULL) {
-			math_error("swap_b8_in_ZVALUE: swap_b8_in_ZVALUE: Not enough memory");
+			math_error("swap_b8_in_ZVALUE: swap_b8_in_ZVALUE: "
+				    "Not enough memory");
 			/*NOTREACHED*/
 		}
 
@@ -126,8 +126,8 @@ swap_b8_in_ZVALUE(ZVALUE *dest, ZVALUE *src, BOOL all)
 	 * swap or copy the rest of the ZVALUE elements
 	 */
 	if (all) {
-		dest->len = (LEN)SWAP_B8_IN_LEN(&dest->len, &src->len);
-		dest->sign = (BOOL)SWAP_B8_IN_BOOL(&dest->sign, &src->sign);
+		SWAP_B8_IN_LEN(&dest->len, &src->len);
+		SWAP_B8_IN_BOOL(&dest->sign, &src->sign);
 	} else {
 		dest->len = src->len;
 		dest->sign = src->sign;
@@ -148,7 +148,7 @@ swap_b8_in_ZVALUE(ZVALUE *dest, ZVALUE *src, BOOL all)
  *		  NULL to allocate the storage
  *	src	- pointer to a NUMBER to swap
  *	all	- TRUE => swap every element, FALSE => swap only the
- *		  multiprecision storage
+ *		  multi-precision storage
  *
  * returns:
  *	pointer to where the swapped src has been put
@@ -192,7 +192,7 @@ swap_b8_in_NUMBER(NUMBER *dest, NUMBER *src, BOOL all)
 	 * swap or copy the rest of the NUMBER elements
 	 */
 	if (all) {
-		dest->links = (long)SWAP_B8_IN_LONG(&dest->links, &src->links);
+		SWAP_B8_IN_LONG(&dest->links, &src->links);
 	} else {
 		dest->links = src->links;
 	}
@@ -212,7 +212,7 @@ swap_b8_in_NUMBER(NUMBER *dest, NUMBER *src, BOOL all)
  *		  NULL to allocate the storage
  *	src	- pointer to a COMPLEX to swap
  *	all	- TRUE => swap every element, FALSE => swap only the
- *		  multiprecision storage
+ *		  multi-precision storage
  *
  * returns:
  *	pointer to where the swapped src has been put
@@ -256,7 +256,7 @@ swap_b8_in_COMPLEX(COMPLEX *dest, COMPLEX *src, BOOL all)
 	 * swap or copy the rest of the NUMBER elements
 	 */
 	if (all) {
-		dest->links = (long)SWAP_B8_IN_LONG(&dest->links, &src->links);
+		SWAP_B8_IN_LONG(&dest->links, &src->links);
 	} else {
 		dest->links = src->links;
 	}
@@ -316,7 +316,7 @@ swap_b16_in_HALFs(HALF *dest, HALF *src, LEN len)
  *		  NULL to allocate the storage
  *	src	- pointer to a ZVALUE to swap
  *	all	- TRUE => swap every element, FALSE => swap only the
- *		  multiprecision storage
+ *		  multi-precision storage
  *
  * returns:
  *	pointer to where the swapped src has been put
@@ -361,8 +361,8 @@ swap_b16_in_ZVALUE(ZVALUE *dest, ZVALUE *src, BOOL all)
 	 * swap or copy the rest of the ZVALUE elements
 	 */
 	if (all) {
-		dest->len = (LEN)SWAP_B16_IN_LEN(&dest->len, &src->len);
-		dest->sign = (BOOL)SWAP_B16_IN_BOOL(&dest->sign, &src->sign);
+		SWAP_B16_IN_LEN(&dest->len, &src->len);
+		SWAP_B16_IN_BOOL(&dest->sign, &src->sign);
 	} else {
 		dest->len = src->len;
 		dest->sign = src->sign;
@@ -383,7 +383,7 @@ swap_b16_in_ZVALUE(ZVALUE *dest, ZVALUE *src, BOOL all)
  *		  NULL to allocate the storage
  *	src	- pointer to a NUMBER to swap
  *	all	- TRUE => swap every element, FALSE => swap only the
- *		  multiprecision storage
+ *		  multi-precision storage
  *
  * returns:
  *	pointer to where the swapped src has been put
@@ -427,7 +427,7 @@ swap_b16_in_NUMBER(NUMBER *dest, NUMBER *src, BOOL all)
 	 * swap or copy the rest of the NUMBER elements
 	 */
 	if (all) {
-		dest->links = (long)SWAP_B16_IN_LONG(&dest->links, &src->links);
+		SWAP_B16_IN_LONG(&dest->links, &src->links);
 	} else {
 		dest->links = src->links;
 	}
@@ -447,7 +447,7 @@ swap_b16_in_NUMBER(NUMBER *dest, NUMBER *src, BOOL all)
  *		  NULL to allocate the storage
  *	src	- pointer to a COMPLEX to swap
  *	all	- TRUE => swap every element, FALSE => swap only the
- *		  multiprecision storage
+ *		  multi-precision storage
  *
  * returns:
  *	pointer to where the swapped src has been put
@@ -491,7 +491,7 @@ swap_b16_in_COMPLEX(COMPLEX *dest, COMPLEX *src, BOOL all)
 	 * swap or copy the rest of the NUMBER elements
 	 */
 	if (all) {
-		dest->links = (long)SWAP_B16_IN_LONG(&dest->links, &src->links);
+		SWAP_B16_IN_LONG(&dest->links, &src->links);
 	} else {
 		dest->links = src->links;
 	}
@@ -511,7 +511,7 @@ swap_b16_in_COMPLEX(COMPLEX *dest, COMPLEX *src, BOOL all)
  *		  NULL to allocate the storage
  *	src	- pointer to a ZVALUE to swap
  *	all	- TRUE => swap every element, FALSE => swap only the
- *		  multiprecision storage
+ *		  multi-precision storage
  *
  * returns:
  *	pointer to where the swapped src has been put
@@ -557,8 +557,8 @@ swap_HALF_in_ZVALUE(ZVALUE *dest, ZVALUE *src, BOOL all)
 	 * swap or copy the rest of the ZVALUE elements
 	 */
 	if (all) {
-		dest->len = (LEN)SWAP_HALF_IN_LEN(&dest->len, &src->len);
-		dest->sign = (BOOL)SWAP_HALF_IN_BOOL(&dest->sign, &src->sign);
+		SWAP_HALF_IN_LEN(&dest->len, &src->len);
+		SWAP_HALF_IN_BOOL(&dest->sign, &src->sign);
 	} else {
 		dest->len = src->len;
 		dest->sign = src->sign;
@@ -579,7 +579,7 @@ swap_HALF_in_ZVALUE(ZVALUE *dest, ZVALUE *src, BOOL all)
  *		  NULL to allocate the storage
  *	src	- pointer to a NUMBER to swap
  *	all	- TRUE => swap every element, FALSE => swap only the
- *		  multiprecision storage
+ *		  multi-precision storage
  *
  * returns:
  *	pointer to where the swapped src has been put
@@ -623,7 +623,7 @@ swap_HALF_in_NUMBER(NUMBER *dest, NUMBER *src, BOOL all)
 	 * swap or copy the rest of the NUMBER elements
 	 */
 	if (all) {
-		dest->links = (long)SWAP_HALF_IN_LONG(&dest->links,&src->links);
+		SWAP_HALF_IN_LONG(&dest->links, &src->links);
 	} else {
 		dest->links = src->links;
 	}
@@ -643,7 +643,7 @@ swap_HALF_in_NUMBER(NUMBER *dest, NUMBER *src, BOOL all)
  *		  NULL to allocate the storage
  *	src	- pointer to a COMPLEX to swap
  *	all	- TRUE => swap every element, FALSE => swap only the
- *		  multiprecision storage
+ *		  multi-precision storage
  *
  * returns:
  *	pointer to where the swapped src has been put
@@ -687,7 +687,7 @@ swap_HALF_in_COMPLEX(COMPLEX *dest, COMPLEX *src, BOOL all)
 	 * swap or copy the rest of the NUMBER elements
 	 */
 	if (all) {
-		dest->links = (long)SWAP_HALF_IN_LONG(&dest->links,&src->links);
+		SWAP_HALF_IN_LONG(&dest->links, &src->links);
 	} else {
 		dest->links = src->links;
 	}

byteswap.h

@@ -1,7 +1,7 @@
 /*
  * byteswap - byte swapping macros
  *
- * Copyright (C) 1999  Landon Curt Noll
+ * Copyright (C) 1999,2014  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: byteswap.h,v 30.1 2007/03/16 11:09:46 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/RCS/byteswap.h,v $
- *
  * Under source code control:	1995/10/11 04:44:01
  * File existed as early as:	1995
  *
@@ -29,8 +25,8 @@
  */
 
 
-#if !defined(__BYTESWAP_H__)
-#define __BYTESWAP_H__
+#if !defined(INCLUDE_BYTESWAP_H)
+#define INCLUDE_BYTESWAP_H
 
 
 #if defined(CALC_SRC)	/* if we are building from the calc source tree */
@@ -178,4 +174,4 @@
 #endif /* LONG_BITS == 64 */
 
 
-#endif /* !__BYTESWAP_H__ */
+#endif /* !INCLUDE_BYTESWAP_H */

cal/Makefile

@@ -2,7 +2,15 @@
 #
 # cal - makefile for calc standard resource files
 #
-# Copyright (C) 1999-2006  Landon Curt Noll
+# Copyright (C) 1999-2006,2017,2021  Landon Curt Noll
+#
+# SRC: cal/Makefile
+#
+#	The "# SRC: ... - ..." comment line above indicates
+#	the origin of this file.
+#
+# IMPORTANT: Please see the section on Makefiles near the
+#	     bottom of the HOWTO.INSTALL file.
 #
 # Calc is open software; you can redistribute it and/or modify it under
 # the terms of the version 2.1 of the GNU Lesser General Public License
@@ -18,10 +26,6 @@
 # received a copy with calc; if not, write to Free Software Foundation, Inc.
 # 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
 #
-# @(#) $Revision: 30.2 $
-# @(#) $Id: Makefile,v 30.2 2007/03/22 07:34:47 chongo Exp $
-# @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/Makefile,v $
-#
 # Under source code control:	1991/07/21 05:00:54
 # File existed as early as:	1991
 #
@@ -31,15 +35,64 @@
 # calculator by David I. Bell with help/mods from others
 # Makefile by Landon Curt Noll
 
-# required vars
+
+# The shell used by this Makefile
+#
+# On some systems, /bin/sh is a rather reduced shell with
+# deprecated behavior.
+#
+# If your system has a up to date, bash shell, then
+# you may wish to use:
+#
+#	SHELL= /bin/bash
+#
+# On some systems such as macOS, the bash shell is very
+# far behind to the point where is cannot be depended on.
+# On such systems, the zsh may be a much better alternative
+# shell for this Makefile to use:
+#
+#	SHELL= /bin/zsh
 #
 SHELL= /bin/sh
+#SHELL= /bin/bash
+#SHELL= /bin/zsh
+
 
 ####
 # Normally, the upper level makefile will set these values.  We provide
 # a default here just in case you want to build from this directory.
 ####
 
+# PREFIX - Top level location for calc
+#
+# The PREFIX is often prepended to paths within calc and calc Makefiles.
+#
+# Starting with calc v2.13.0.1, nearly all Makefile places that used
+# /usr/local now use ${PREFIX}.  An exception is the olduninstall rule
+# and, of course, this section. :-)
+#
+# NOTE: The ${PREFIX} is not the same as ${T}.  The ${T} specifies
+#	a top level directory under which calc installs things.
+#	While usually ${T} is empty, it can be specific path
+#	as if calc where "chrooted" during an install.
+#	The ${PREFIX} value, during install, is a path between
+#	the top level ${T} install directory and the object
+#	such as an include file.
+#
+# NOTE: See also, ${T}, below.
+#
+# There are some paths that do NOT call under ${PREFIX}, such as
+# ${CALCPATH}, that include paths not under ${PREFIX}, but those
+# too are exceptions to this general rule.
+#
+# When in doubt, try:
+#
+#	PREFIX= /usr/local
+#
+PREFIX= /usr/local
+#PREFIX= /usr
+#PREFIX= /usr/global
+
 # Normally certain files depend on the Makefile.  If the Makefile is
 # changed, then certain steps should be redone.  If MAKE_FILE is
 # set to Makefile, then these files will depend on Makefile.  If
@@ -69,11 +122,11 @@ TOP_MAKE_FILE= Makefile
 #	INCDIR= /usr/include
 #
 
-#INCDIR= /usr/local/include
+#INCDIR= ${PREFIX}/include
 #INCDIR= /dev/env/DJDIR/include
 INCDIR= /usr/include
 
-# where to install calc realted things
+# where to install calc related things
 #
 # ${BINDIR}		where to install calc binary files
 # ${LIBDIR}		where calc link library (*.a) files are installed
@@ -95,15 +148,15 @@ INCDIR= /usr/include
 #	LIBDIR= /usr/lib
 #	CALC_SHAREDIR= /usr/share/calc
 #
-#BINDIR= /usr/local/bin
+#BINDIR= ${PREFIX}/bin
 #BINDIR= /dev/env/DJDIR/bin
 BINDIR= /usr/bin
 
-#LIBDIR= /usr/local/lib
+#LIBDIR= ${PREFIX}/lib
 #LIBDIR= /dev/env/DJDIR/lib
 LIBDIR= /usr/lib
 
-#CALC_SHAREDIR= /usr/local/lib/calc
+#CALC_SHAREDIR= ${PREFIX}/lib/calc
 #CALC_SHAREDIR= /dev/env/DJDIR/share/calc
 CALC_SHAREDIR= /usr/share/calc
 
@@ -111,9 +164,6 @@ CALC_SHAREDIR= /usr/share/calc
 # ---------------------------------------------------------------
 # ${HELPDIR}		where the help directory is installed
 # ${CALC_INCDIR}	where the calc include files are installed
-# ${CUSTOMCALDIR}	where custom *.cal files are installed
-# ${CUSTOMHELPDIR}	where custom help files are installed
-# ${CUSTOMINCDIR}	where custom .h files are installed
 # ${SCRIPTDIR}		where calc shell scripts are installed
 #
 # NOTE: The install rule prepends installation paths with ${T}, which
@@ -124,16 +174,10 @@ CALC_SHAREDIR= /usr/share/calc
 #
 #	HELPDIR= ${CALC_SHAREDIR}/help
 #	CALC_INCDIR= ${INCDIR}/calc
-#	CUSTOMCALDIR= ${CALC_SHAREDIR}/custom
-#	CUSTOMHELPDIR= ${CALC_SHAREDIR}/custhelp
-#	CUSTOMINCDIR= ${CALC_INCDIR}/custom
 #	SCRIPTDIR= ${BINDIR}/cscript
 #
 HELPDIR= ${CALC_SHAREDIR}/help
 CALC_INCDIR= ${INCDIR}/calc
-CUSTOMCALDIR= ${CALC_SHAREDIR}/custom
-CUSTOMHELPDIR= ${CALC_SHAREDIR}/custhelp
-CUSTOMINCDIR= ${CALC_INCDIR}/custom
 SCRIPTDIR= ${BINDIR}/cscript
 
 # T - top level directory under which calc will be installed
@@ -170,39 +214,82 @@ T=
 
 # Makefile debug
 #
-# Q=@	do not echo internal makefile actions (quiet mode)
-# Q=	echo internal makefile actions (debug / verbose mode)
+# Q=@	do not echo internal Makefile actions (quiet mode)
+# Q=	echo internal Makefile actions (debug / verbose mode)
+#
+# H=@:	do not report hsrc file formation progress
+# H=@	do echo hsrc file formation progress
+#
+# S= >/dev/null 2>&1	slience ${CC} output during hsrc file formation
+# S=			full ${CC} output during hsrc file formation
+#
+# E= 2>/dev/null	slience command stderr during hsrc file formation
+# E=			full command stderr during hsrc file formation
+#
+# V=@:	do not echo debug statements (quiet mode)
+# V=@	do echo debug statements (debug / verbose mode)
 #
 #Q=
 Q=@
+#
+S= >/dev/null 2>&1
+#S=
+#
+E= 2>/dev/null
+#E=
+#
+#H=@:
+H=@
+#
+V=@:
+#V=@
 
 # standard tools
 #
 CHMOD= chmod
 CMP= cmp
-RM= rm
-MKDIR= mkdir
-RMDIR= rmdir
-CP= cp
-MV= mv
 CO= co
-TRUE= true
+CP= cp
+FMT= fmt
+MKDIR= mkdir
+MV= mv
+RM= rm
+RMDIR= rmdir
+SED= sed
+SORT= sort
 TOUCH= touch
+TRUE= true
 
 # The calc files to install
 #
-CALC_FILES= README bigprime.cal deg.cal ellip.cal lucas.cal lucas_chk.cal \
-	lucas_tbl.cal mersenne.cal mod.cal pell.cal pi.cal pix.cal \
-	pollard.cal poly.cal psqrt.cal quat.cal regress.cal solve.cal \
-	sumsq.cal surd.cal unitfrac.cal varargs.cal chrem.cal mfactor.cal \
-	bindings randmprime.cal test1700.cal randrun.cal linear.cal \
-	randbitrun.cal bernoulli.cal test2300.cal test2600.cal \
-	test2700.cal test3100.cal test3300.cal test3400.cal prompt.cal \
-	test3500.cal seedrandom.cal test4000.cal test4100.cal test4600.cal \
-	beer.cal hello.cal test5100.cal test5200.cal randombitrun.cal \
-	randomrun.cal repeat.cal xx_print.cal natnumset.cal qtime.cal \
-	test8400.cal test8500.cal test8600.cal chi.cal intfile.cal screen.cal \
-	dotest.cal set8700.cal set8700.line alg_config.cal sumtimes.cal
+# This list is produced by the detaillist rule when no WARNINGS are detected.
+#
+# Please use:
+#
+#	make calc_files_list
+#
+# to keep this list in nice sorted order.
+#
+CALC_FILES= README alg_config.cal beer.cal bernoulli.cal \
+	bernpoly.cal bigprime.cal bindings brentsolve.cal chi.cal chrem.cal \
+	constants.cal deg.cal dms.cal dotest.cal ellip.cal factorial.cal \
+	factorial2.cal gvec.cal hello.cal hms.cal infinities.cal intfile.cal \
+	intnum.cal lambertw.cal linear.cal lnseries.cal lucas.cal \
+	lucas_chk.cal mersenne.cal mfactor.cal mod.cal natnumset.cal pell.cal \
+	pi.cal pix.cal pollard.cal poly.cal prompt.cal psqrt.cal qtime.cal \
+	quat.cal randbitrun.cal randmprime.cal randombitrun.cal randomrun.cal \
+	randrun.cal regress.cal repeat.cal screen.cal seedrandom.cal \
+	set8700.cal set8700.line smallfactors.cal solve.cal \
+	specialfunctions.cal statistics.cal strings.cal sumsq.cal sumtimes.cal \
+	surd.cal test1700.cal test2300.cal test2600.cal test2700.cal \
+	test3100.cal test3300.cal test3400.cal test3500.cal test4000.cal \
+	test4100.cal test4600.cal test5100.cal test5200.cal test8400.cal \
+	test8500.cal test8600.cal test8900.cal toomcook.cal unitfrac.cal \
+	varargs.cal xx_print.cal zeta2.cal
+
+# These calc files are now obsolete and are removed by the install rule.
+#
+DEAD_CALC_FILES= lucas_tbl.cal
 
 # These files are found (but not built) in the distribution
 #
@@ -229,7 +316,7 @@ all: ${CALC_FILES} ${MAKE_FILE} .all
 # sub-directory called calc/cal.
 #
 # NOTE: Due to bogus shells found on one common system we must have
-#	an non-emoty else clause for every if condition.  *sigh*
+#	an non-empty else clause for every if condition.  *sigh*
 #
 ##
 
@@ -250,6 +337,21 @@ calcliblist:
 	    fi; \
 	done
 
+# These next rule help form the ${CALC_FILES} makefile variables above.
+#
+calc_files_list:
+	${Q} -(find .  -mindepth 1 -maxdepth 1 -type f -name '*.cal' -print | \
+	    while read i; do \
+		echo $$i; \
+	    done; \
+	    echo '--first_line--'; \
+	    echo README; \
+	    echo set8700.line; \
+	    echo bindings) | \
+	    ${SED} -e 's:^\./::' | LANG=C ${SORT} | ${FMT} -70 | \
+	    ${SED} -e '1s/--first_line--/CALC_FILES=/' -e '2,$$s/^/	/' \
+		   -e 's/$$/ \\/' -e '$$s/ \\$$//'
+
 ##
 #
 # rpm rules
@@ -273,6 +375,23 @@ clean:
 
 clobber: clean
 	${RM} -f .all
+	-${Q} if [ -e .DS_Store ]; then \
+	    echo ${RM} -rf .DS_Store; \
+	    ${RM} -rf .DS_Store; \
+	fi
+	-${Q} for i in ${DEAD_CALC_FILES} /dev/null; do \
+	    if [ "$$i" = "/dev/null" ]; then \
+		continue; \
+	    fi; \
+	    if [ -e "${T}${CALC_SHAREDIR}/$$i" ]; then \
+		echo "${RM} -f ${T}${CALC_SHAREDIR}/$$i"; \
+		${RM} -f ${T}${CALC_SHAREDIR}/$$i; \
+	    fi; \
+	    if [ -e "./$$i" ]; then \
+		echo "${RM} -f ./$$i"; \
+		${RM} -f ./$$i; \
+	    fi; \
+	done
 
 # install everything
 #
@@ -280,8 +399,8 @@ clobber: clean
 #
 install: all
 	-${Q} if [ ! -d ${T}${CALC_SHAREDIR} ]; then \
-	    echo ${MKDIR} ${T}${CALC_SHAREDIR}; \
-	    ${MKDIR} ${T}${CALC_SHAREDIR}; \
+	    echo ${MKDIR} -p ${T}${CALC_SHAREDIR}; \
+	    ${MKDIR} -p ${T}${CALC_SHAREDIR}; \
 	    if [ ! -d "${T}${CALC_SHAREDIR}" ]; then \
 		echo ${MKDIR} -p "${T}${CALC_SHAREDIR}"; \
 		${MKDIR} -p "${T}${CALC_SHAREDIR}"; \
@@ -301,16 +420,42 @@ install: all
 		${RM} -f ${T}${CALC_SHAREDIR}/$$i.new; \
 		${CP} -f $$i ${T}${CALC_SHAREDIR}/$$i.new; \
 		${CHMOD} 0444 ${T}${CALC_SHAREDIR}/$$i.new; \
-		${MV} -f ${T}${CALC_SHAREDIR}/$$i.new ${T}${CALC_SHAREDIR}/$$i; \
+		${MV} -f ${T}${CALC_SHAREDIR}/$$i.new ${T}${CALC_SHAREDIR}/$$i;\
 		echo "installed ${T}${CALC_SHAREDIR}/$$i"; \
 	    fi; \
 	done
+	${Q} for i in ${DEAD_CALC_FILES} /dev/null; do \
+	    if [ "$$i" = "/dev/null" ]; then \
+		continue; \
+	    fi; \
+	    if [ -e "${T}${CALC_SHAREDIR}/$$i" ]; then \
+		echo "${RM} -f ${T}${CALC_SHAREDIR}/$$i"; \
+		${RM} -f ${T}${CALC_SHAREDIR}/$$i; \
+	    fi; \
+	    if [ -e "./$$i" ]; then \
+		echo "${RM} -f ./$$i"; \
+		${RM} -f ./$$i; \
+	    fi; \
+	done
 
 # Try to remove everything that was installed
 #
 # NOTE: Keep the uninstall rule in reverse order to the install rule
 #
 uninstall:
+	- ${Q} for i in ${DEAD_CALC_FILES} /dev/null; do \
+	    if [ "$$i" = "/dev/null" ]; then \
+		continue; \
+	    fi; \
+	    if [ -e "${T}${CALC_SHAREDIR}/$$i" ]; then \
+		echo "${RM} -f ${T}${CALC_SHAREDIR}/$$i"; \
+		${RM} -f ${T}${CALC_SHAREDIR}/$$i; \
+	    fi; \
+	    if [ -e "./$$i" ]; then \
+		echo "${RM} -f ./$$i"; \
+		${RM} -f ./$$i; \
+	    fi; \
+	done
 	-${Q} for i in ${CALC_FILES} /dev/null; do \
 	    if [ "$$i" = "/dev/null" ]; then \
 		continue; \
@@ -320,7 +465,7 @@ uninstall:
 		if [ -f "${T}${CALC_SHAREDIR}/$$i" ]; then \
 		   echo "cannot uninstall ${T}${CALC_SHAREDIR}/$$i"; \
 		else \
-		   echo "uninstalled ${T}${CALC_SHAREDIR}/$$i"; \
+		   echo "un-installed ${T}${CALC_SHAREDIR}/$$i"; \
 		fi; \
 	    fi; \
 	done

cal/alg_config.cal

@@ -1,7 +1,7 @@
 /*
  * alg_config - help determine optimal values for algorithm levels
  *
- * Copyright (C) 2006  Landon Curt Noll
+ * Copyright (C) 2006,2014,2021  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: alg_config.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/alg_config.cal,v $
- *
  * Under source code control:	2006/06/07 14:10:11
  * File existed as early as:	2006
  *
@@ -28,8 +24,37 @@
  * Share and enjoy!  :-)	http://www.isthe.com/chongo/tech/comp/calc/
  */
 
+static test_time;	/* try for this many seconds in loop test */
 
-global test_time;	/* try for this many seconds in loop test */
+
+/*
+ * close_to_one - set to 1 if the ratio is close enough to 1
+ *
+ * given:
+ *	ratio	the ratio of time between two algorithms
+ *
+ * returns:
+ *	1	When ratio is near 1.0
+ *	0	otherwise
+ *
+ * We consider the range [0.995, 1.005] to be close enough to 1 to be call unity
+ * because of the precision of the CPU timing.
+ */
+define close_to_one(ratio)
+{
+    /* firewall */
+    if (!isreal(ratio)) {
+	quit "close: 1st arg: must be a real number";
+    }
+
+    /* check if the ratio is far from unity */
+    if ((ratio < 0.995) || (ratio > 1.005)) {
+	return 0;
+    }
+
+    /* we are close to unity */
+    return 1;
+}
 
 
 /*
@@ -80,7 +105,8 @@ define mul_loop(repeat, x)
     len = sizeof((*x)[0]) / baseb_bytes;
     for (i=1; i < 4; ++i) {
 	if ((sizeof((*x)[i]) / baseb_bytes) != len) {
-	    quit "mul_loop: 2nd arg matrix elements are not of equal BASEB-bit word length";
+	    quit "mul_loop: 2nd arg matrix elements are not of "
+	         "equal BASEB-bit word length";
 	}
     }
 
@@ -142,7 +168,7 @@ define mul_loop(repeat, x)
  *	len	length in BASEB-bit words to multiply
  *
  * return:
- *	ratio of (1st / 2nd) algorithm rate
+ *	ratio of (1st / 2nd) algorithm rate.
  *
  * When want to determine a rate to a precision of 1 part in 1000.
  * Most systems today return CPU time to at least 10 msec precision.
@@ -165,6 +191,7 @@ define mul_ratio(len)
     local tover;		/* est of time for loop overhead */
     local alg1_rate;		/* loop rate of 1st algorithm */
     local alg2_rate;		/* loop rate of 2nd algorithm */
+    local ret;			/* return ratio, or 1.0 */
     local i;
 
     /*
@@ -217,12 +244,12 @@ define mul_ratio(len)
      * determine the 1st algorithm rate
      */
     loops = max(1, ceil(loops * test_time / tlen));
-    if (loops < 8) {
+    if (loops < 16) {
 	if (config("user_debug") > 1) {
-	    printf("    we must expand loop test time to more than %d secs\n",
-		ceil(test_time * (8 / loops)));
+	    printf("    we must expand alg1 loop test time to about %d secs\n",
+		ceil(test_time * (16 / loops)));
 	}
-	loops = 8;
+	loops = 16;
     }
     if (config("user_debug") > 3) {
 	printf("\t    will try alg1 %d loops\n", loops);
@@ -263,12 +290,12 @@ define mul_ratio(len)
      * determine the 2nd algorithm rate
      */
     loops = max(1, ceil(loops * test_time / tlen));
-    if (loops < 8) {
+    if (loops < 16) {
 	if (config("user_debug") > 1) {
-	    printf("    we must expand loop test time to more than %d secs\n",
-		ceil(test_time * (8 / loops)));
+	    printf("    we must expand alg2 loop test time to about %d secs\n",
+		ceil(test_time * (16 / loops)));
 	}
-	loops = 8;
+	loops = 16;
     }
     tlen = mul_loop(loops, &x);
     if (config("user_debug") > 3) {
@@ -297,7 +324,12 @@ define mul_ratio(len)
     /*
      * return alg1 / alg2 rate ratio
      */
-    return (alg1_rate / alg2_rate);
+    ret = alg1_rate / alg2_rate;
+    if (config("user_debug") > 2) {
+	printf("\tprecise ratio is: %.f mul_ratio will return: %.3f\n",
+		alg1_rate / alg2_rate, ret);
+    }
+    return ret;
 }
 
 
@@ -319,30 +351,40 @@ define best_mul2()
 {
     local ratio;	/* previously calculated alg1/alg2 ratio */
     local low;		/* low loop value tested */
-    local high;		/* low loop value tested */
+    local high;		/* high loop value tested */
+    local mid;		/* between low and high */
+    local best_val;	/* value found with ratio closest to unity */
+    local best_ratio;	/* closest ratio found to unity */
     local expand;	/* how fast to expand the length */
 
     /*
      * setup
      */
-    test_time = 15.0;
-    if (config("user_debug") > 0) {
-	printf("will start with loop test time of %d secs\n", test_time);
+    printf("WARNING: This tool may not be computing the correct best value\n");
+    test_time = 5.0;
+    printf("The best_mul2() function will take a LONG time to run!\n");
+    printf("It is important that best_mul2() run on an otherwise idle host!\n");
+    if (config("user_debug") <= 0) {
+	printf("To monitor progress, set user_debug to 2: "
+	       "config(\"user_debug\", 2)\n");
     }
+    printf("Starting with loop test time of %d secs\n", test_time);
 
     /*
      * firewall - must have a >1 ratio for the initial length
      */
-    high = 16;
+    high = 8;
+    best_val = high;
     if (config("user_debug") > 0) {
 	printf("testing multiply alg1/alg2 ratio for len = %d\n", high);
     }
     ratio = mul_ratio(high);
+    best_ratio = ratio;
     if (config("user_debug") > 1) {
-	printf("    multiply alg1/alg2 ratio = %.3f\n", ratio);
+	printf("    multiply alg1/alg2 ratio = %.6f\n", ratio);
     }
-    if (ratio <= 1.0) {
-	quit "best_mul2: tests imply config(\"mul2\") should be < 4";
+    if (ratio < 1.0) {
+	quit "best_mul2: tests imply mul2 should be < 16, which seems bogus";
     }
 
     /*
@@ -350,12 +392,12 @@ define best_mul2()
      */
     do {
 	/*
-	 * determine the paramters of the next ratio test
+	 * determine the parameters of the next ratio test
 	 *
 	 * We will multiplicatively expand our test level until
 	 * the ratio drops below 1.0.
 	 */
-	expand = ((ratio >= 3.5) ? 16 : 2^round(ratio));
+	expand = 2;
 	low = high;
 	high *= expand;
 	if (config("user_debug") > 1) {
@@ -367,19 +409,64 @@ define best_mul2()
 	 * determine the alg1/alg2 test ratio for this new length
 	 */
 	if (high >= 2^31) {
-	    quit "best_mul2: tests imply config(\"mul2\") should be >= 2^31";
+	    quit "best_mul2: test implies mul2 >= 2^31, which seems bogus";
 	}
 	if (config("user_debug") > 0) {
 	    printf("testing multiply alg1/alg2 ratio for len = %d\n", high);
 	}
 	ratio = mul_ratio(high);
-	if (config("user_debug") > 1) {
-	    printf("    multiply alg1/alg2 ratio = %.3f\n", ratio);
+	if (abs(ratio - 1.0) < abs(best_ratio - 1.0)) {
+	    best_val = high;
+	    best_ratio = ratio;
+	    if (config("user_debug") > 1) {
+		printf("    len %d has a new closest ratio to unity: %.6f\n",
+		       best_val, best_ratio);
+	    }
 	}
-    } while (ratio >= 1.0);
+	if (config("user_debug") > 1) {
+	    printf("    multiply alg1/alg2 ratio = %.6f\n", ratio);
+	}
+    } while (ratio > 1.0);
+
+    /*
+     * If we previously expanded more than by a factor of 2, then
+     * we may have jumped over the crossover point.  So now
+     * drop down powers of two until the ratio is again >= 1.0
+     */
+    if (expand > 2) {
+	do {
+
+	    /*
+	     * contract by 2
+	     */
+	    high /= 2;
+	    low = high / 2;
+	    if (config("user_debug") > 0) {
+		printf("re-testing multiply alg1/alg2 ratio for len = %d\n",
+		       high);
+	    }
+	    ratio = mul_ratio(high);
+	    if (abs(ratio - 1.0) < abs(best_ratio - 1.0)) {
+		best_val = high;
+		best_ratio = ratio;
+		if (config("user_debug") > 1) {
+		    printf("    len %d has a new closest ratio "
+			   "to unity: %.6f\n",
+			   best_val, best_ratio);
+		}
+	    }
+	    if (config("user_debug") > 1) {
+		printf("    multiply alg1/alg2 ratio = %.6f\n", ratio);
+	    }
+
+	} while (ratio <= 1.0);
+
+	/* now that the ratio flipped again, use the previous range */
+	low = high;
+	high = high * 2;
+    }
     if (config("user_debug") > 0) {
-	printf("alg1/alg2 ratio now < 1.0, starting binary search between %d and %d\n",
-	    low, high);
+	printf("Starting binary search between %d and %d\n", low, high);
     }
 
     /*
@@ -388,30 +475,54 @@ define best_mul2()
     while (low+1 < high) {
 
     	/* try the mid-point */
+	mid = int((low+high)/2);
 	if (config("user_debug") > 0) {
-	    printf("testing multiply alg1/alg2 ratio for len = %d\n",
-	    	int((low+high)/2));
+	    printf("testing multiply alg1/alg2 ratio for len = %d\n", mid);
+	}
+	ratio = mul_ratio(mid);
+	if (abs(ratio - 1.0) < abs(best_ratio - 1.0)) {
+	    best_val = mid;
+	    best_ratio = ratio;
+	    if (config("user_debug") > 1) {
+		printf("    len %d has a new closest ratio to unity: %.6f\n",
+		       best_val, best_ratio);
+	    }
 	}
-	ratio = mul_ratio(int((low+high)/2));
 	if (config("user_debug") > 1) {
-	    printf("    multiply alg1/alg2 ratio = %.3f\n", ratio);
+	    printf("    len %d multiply alg1/alg2 ratio = %.6f\n", mid, ratio);
+	}
+
+	/* stop search if near unity */
+	if (close_to_one(ratio)) {
+	    low = mid;
+	    high = mid;
+	    if (config("user_debug") > 0) {
+	    	printf("\twe are close enough to unity ratio at: %d\n", mid);
+	    }
+	    break;
 	}
 
 	/* bump lower range up if we went over */
-	if (ratio >= 1.0) {
+	if (ratio > 1.0) {
 	    if (config("user_debug") > 2) {
 	    	printf("\tmove low from %d up to %d\n",
-		    low, int((low+high)/2));
+		    low, mid);
 	    }
-	    low = int((low+high)/2);
+	    low = mid;
 
 	/* drop higher range down if we went under */
 	} else {
 	    if (config("user_debug") > 2) {
 	    	printf("\tmove high from %d down to %d\n",
-		     high, int((low+high)/2));
+		     high, mid);
 	    }
-	    high = int((low+high)/2);
+	    high = mid;
+	}
+
+	/* report on test loop progress */
+	if (config("user_debug") > 1) {
+	    printf("\tsetting low: %d high: %d diff: %d\n",
+	    	   low, high, high-low);
 	}
     }
 
@@ -419,10 +530,15 @@ define best_mul2()
      * return on the suggested config("mul2") value
      */
     if (config("user_debug") > 0) {
-	printf("best value of config(\"mul2\") is %d\n",
-	    (ratio >= 1.0) ? high : low);
+	printf("Best value for multiply is near %d\n", best_val);
+	printf("Best multiply alg1/alg2 ratio is: %.6f\n", best_ratio);
+	printf("We suggest placing this line in your .calcrc:\n");
+	printf("config(\"mul2\", %d),;\n", best_val);
+	printf("WARNING: It is believed that the output "
+	       "of this resource file is bogus!\n");
+	printf("WARNING: You may NOT wish to follow the above suggestion.\n");
     }
-    return ((ratio >= 1.0) ? high : low);
+    return mid;
 }
 
 
@@ -472,7 +588,8 @@ define sq_loop(repeat, x)
     len = sizeof((*x)[0]) / baseb_bytes;
     for (i=1; i < 4; ++i) {
 	if ((sizeof((*x)[i]) / baseb_bytes) != len) {
-	    quit "sq_loop: 2nd arg matrix elements are not of equal BASEB-bit word length";
+	    quit "sq_loop: 2nd arg matrix elements are not of equal "
+	         "BASEB-bit word length";
 	}
     }
 
@@ -559,6 +676,7 @@ define sq_ratio(len)
     local tover;		/* est of time for loop overhead */
     local alg1_rate;		/* loop rate of 1st algorithm */
     local alg2_rate;		/* loop rate of 2nd algorithm */
+    local ret;			/* return ratio, or 1.0 */
     local i;
 
     /*
@@ -611,12 +729,12 @@ define sq_ratio(len)
      * determine the 1st algorithm rate
      */
     loops = max(1, ceil(loops * test_time / tlen));
-    if (loops < 8) {
+    if (loops < 16) {
 	if (config("user_debug") > 1) {
-	    printf("    we must expand loop test time to more than %d secs\n",
-		ceil(test_time * (8 / loops)));
+	    printf("    we must expand alg1 loop test time to about %d secs\n",
+		ceil(test_time * (16 / loops)));
 	}
-	loops = 8;
+	loops = 16;
     }
     tlen = sq_loop(loops, &x);
     if (config("user_debug") > 3) {
@@ -654,12 +772,12 @@ define sq_ratio(len)
      * determine the 2nd algorithm rate
      */
     loops = max(1, ceil(loops * test_time / tlen));
-    if (loops < 8) {
+    if (loops < 16) {
 	if (config("user_debug") > 1) {
-	    printf("    we must expand loop test time to more than %d secs\n",
-		ceil(test_time * (8 / loops)));
+	    printf("    we must expand alg2 loop test time to about %d secs\n",
+		ceil(test_time * (16 / loops)));
 	}
-	loops = 8;
+	loops = 16;
     }
     tlen = sq_loop(loops, &x);
     if (config("user_debug") > 3) {
@@ -688,7 +806,12 @@ define sq_ratio(len)
     /*
      * return alg1 / alg2 rate ratio
      */
-    return (alg1_rate / alg2_rate);
+    ret = alg1_rate / alg2_rate;
+    if (config("user_debug") > 2) {
+	printf("\tprecise ratio is: %.f sq_ratio will return: %.3f\n",
+		alg1_rate / alg2_rate, ret);
+    }
+    return ret;
 }
 
 
@@ -710,30 +833,40 @@ define best_sq2()
 {
     local ratio;	/* previously calculated alg1/alg2 ratio */
     local low;		/* low loop value tested */
-    local high;		/* low loop value tested */
+    local high;		/* high loop value tested */
+    local mid;		/* between low and high */
+    local best_val;	/* value found with ratio closest to unity */
+    local best_ratio;	/* closest ratio found to unity */
     local expand;	/* how fast to expand the length */
 
     /*
      * setup
      */
-    test_time = 15.0;
-    if (config("user_debug") > 0) {
-	printf("will start with loop test time of %d secs\n", test_time);
+    printf("WARNING: This tool may not be computing the correct best value\n");
+    test_time = 5.0;
+    printf("The best_sq2() function will take a LONG time to run!\n");
+    printf("It is important that best_sq2() run on an otherwise idle host!\n");
+    if (config("user_debug") <= 0) {
+	printf("To monitor progress, set user_debug to 2: "
+	       "config(\"user_debug\", 2)\n");
     }
+    printf("Starting with loop test time of %d secs\n", test_time);
 
     /*
      * firewall - must have a >1 ratio for the initial length
      */
-    high = 16;
+    high = 8;
+    best_val = high;
     if (config("user_debug") > 0) {
 	printf("testing square alg1/alg2 ratio for len = %d\n", high);
     }
     ratio = sq_ratio(high);
+    best_ratio = ratio;
     if (config("user_debug") > 1) {
 	printf("    square alg1/alg2 ratio = %.3f\n", ratio);
     }
-    if (ratio <= 1.0) {
-	quit "best_sq2: tests imply config(\"sq2\") should be < 4";
+    if (ratio < 1.0) {
+	quit "best_sq2: test implies sq2 < 16, which seems bogus";
     }
 
     /*
@@ -741,36 +874,81 @@ define best_sq2()
      */
     do {
 	/*
-	 * determine the paramters of the next ratio test
+	 * determine the parameters of the next ratio test
 	 *
 	 * We will multiplicatively expand our test level until
 	 * the ratio drops below 1.0.
 	 */
-	expand = ((ratio >= 3.5) ? 16 : 2^round(ratio));
+	expand = 2;
 	low = high;
 	high *= expand;
 	if (config("user_debug") > 1) {
 	    printf("    expand the next test range by a factor of %d\n",
-	           expand);
+	    	   expand);
 	}
 
 	/*
 	 * determine the alg1/alg2 test ratio for this new length
 	 */
 	if (high >= 2^31) {
-	    quit "best_sq2: tests imply config(\"sq2\") should be >= 2^31";
+	    quit "best_sq2: tests imply sq2 >= 2^31, which seems bogus";
 	}
 	if (config("user_debug") > 0) {
 	    printf("testing square alg1/alg2 ratio for len = %d\n", high);
 	}
 	ratio = sq_ratio(high);
+	if (abs(ratio - 1.0) < abs(best_ratio - 1.0)) {
+	    best_val = high;
+	    best_ratio = ratio;
+	    if (config("user_debug") > 1) {
+		printf("    len %d has a new closest ratio to unity: %.6f\n",
+		       best_val, best_ratio);
+	    }
+	}
 	if (config("user_debug") > 1) {
 	    printf("    square alg1/alg2 ratio = %.3f\n", ratio);
 	}
-    } while (ratio >= 1.0);
+    } while (ratio > 1.0);
+
+    /*
+     * If we previously expanded more than by a factor of 2, then
+     * we may have jumped over the crossover point.  So now
+     * drop down powers of two until the ratio is again >= 1.0
+     */
+    if (expand > 2) {
+	do {
+
+	    /*
+	     * contract by 2
+	     */
+	    high /= 2;
+	    low = high / 2;
+	    if (config("user_debug") > 0) {
+		printf("re-testing multiply alg1/alg2 ratio for len = %d\n",
+		       high);
+	    }
+	    ratio = mul_ratio(high);
+	    if (abs(ratio - 1.0) < abs(best_ratio - 1.0)) {
+		best_val = high;
+		best_ratio = ratio;
+		if (config("user_debug") > 1) {
+		    printf("    len %d has a new closest ratio "
+			   "to unity: %.6f\n",
+			   best_val, best_ratio);
+		}
+	    }
+	    if (config("user_debug") > 1) {
+		printf("    multiply alg1/alg2 ratio = %.6f\n", ratio);
+	    }
+
+	} while (ratio <= 1.0);
+
+	/* now that the ratio flipped again, use the previous range */
+	low = high;
+	high = high * 2;
+    }
     if (config("user_debug") > 0) {
-	printf("alg1/alg2 ratio now < 1.0, starting binary search between %d and %d\n",
-	    low, high);
+	printf("Starting binary search between %d and %d\n", low, high);
     }
 
     /*
@@ -779,41 +957,71 @@ define best_sq2()
     while (low+1 < high) {
 
     	/* try the mid-point */
+	mid = int((low+high)/2);
 	if (config("user_debug") > 0) {
-	    printf("testing square alg1/alg2 ratio for len = %d\n",
-	    	int((low+high)/2));
+	    printf("testing square alg1/alg2 ratio for len = %d\n", mid);
+	}
+	ratio = sq_ratio(mid);
+	if (abs(ratio - 1.0) < abs(best_ratio - 1.0)) {
+	    best_val = mid;
+	    best_ratio = ratio;
+	    if (config("user_debug") > 1) {
+		printf("    len %d has a new closest ratio to unity: %.6f\n",
+		       best_val, best_ratio);
+	    }
 	}
-	ratio = sq_ratio(int((low+high)/2));
 	if (config("user_debug") > 1) {
-	    printf("    square alg1/alg2 ratio = %.3f\n", ratio);
+	    printf("    len %d square alg1/alg2 ratio = %.6f\n", mid, ratio);
+	}
+
+	/* stop search if near unity */
+	if (close_to_one(ratio)) {
+	    low = mid;
+	    high = mid;
+	    if (config("user_debug") > 0) {
+	    	printf("\twe are close enough to unity ratio at: %d\n", mid);
+	    }
+	    break;
 	}
 
 	/* bump lower range up if we went over */
-	if (ratio >= 1.0) {
+	if (ratio > 1.0) {
 	    if (config("user_debug") > 2) {
 	    	printf("\tmove low from %d up to %d\n",
-		    low, int((low+high)/2));
+		    low, mid);
 	    }
-	    low = int((low+high)/2);
+	    low = mid;
 
 	/* drop higher range down if we went under */
 	} else {
 	    if (config("user_debug") > 2) {
 	    	printf("\tmove high from %d down to %d\n",
-		     high, int((low+high)/2));
+		     high, mid);
 	    }
-	    high = int((low+high)/2);
+	    high = mid;
+	}
+
+	/* report on test loop progress */
+	if (config("user_debug") > 1) {
+	    printf("\tsetting low: %d high: %d diff: %d\n",
+	    	   low, high, high-low);
 	}
     }
 
     /*
      * return on the suggested config("sq2") value
      */
+    mid = int((low+high)/2);
     if (config("user_debug") > 0) {
-	printf("best value of config(\"sq2\") is %d\n",
-	    (ratio >= 1.0) ? high : low);
+	printf("Best value for square is near %d\n", best_val);
+	printf("Best square alg1/alg2 ratio is: %.6f\n", best_ratio);
+	printf("We suggest placing this line in your .calcrc:\n");
+	printf("config(\"sq2\", %d),;\n", best_val);
+	printf("WARNING: It is believed that the output "
+	       "of this resource file is bogus!\n");
+	printf("WARNING: You may NOT wish to follow the above suggestion.\n");
     }
-    return ((ratio >= 1.0) ? high : low);
+    return mid;
 }
 
 
@@ -866,7 +1074,8 @@ define pow_loop(repeat, x, ex)
     len = sizeof((*x)[0]) / baseb_bytes;
     for (i=1; i < 4; ++i) {
 	if ((sizeof((*x)[i]) / baseb_bytes) != len) {
-	    quit "pow_loop: 2nd arg matrix elements are not of equal BASEB-bit word length";
+	    quit "pow_loop: 2nd arg matrix elements are not of "
+	         "equal BASEB-bit word length";
 	}
     }
     if (!isint(ex) || ex < 3) {
@@ -963,6 +1172,7 @@ define pow_ratio(len)
     local alg1_rate;		/* loop rate of 1st algorithm */
     local alg2_rate;		/* loop rate of 2nd algorithm */
     local ex;			/* exponent to use in pow_loop() */
+    local ret;			/* return ratio, or 1.0 */
     local i;
 
     /*
@@ -985,7 +1195,7 @@ define pow_ratio(len)
     /*
      * setup
      */
-    ex = 5;
+    ex = 7;
 
     /*
      * initialize x, the values we will pmod
@@ -1021,12 +1231,12 @@ define pow_ratio(len)
      * determine the 1st algorithm rate
      */
     loops = max(1, ceil(loops * test_time / tlen));
-    if (loops < 8) {
+    if (loops < 16) {
 	if (config("user_debug") > 1) {
-	    printf("    we must expand loop test time to more than %d secs\n",
-		ceil(test_time * (8 / loops)));
+	    printf("    we must expand alg1 loop test time to about %d secs\n",
+		ceil(test_time * (16 / loops)));
 	}
-	loops = 8;
+	loops = 16;
     }
     tlen = pow_loop(loops, &x, ex);
     if (config("user_debug") > 3) {
@@ -1065,12 +1275,12 @@ define pow_ratio(len)
      * determine the 2nd algorithm rate
      */
     loops = max(1, ceil(loops * test_time / tlen));
-    if (loops < 8) {
+    if (loops < 16) {
 	if (config("user_debug") > 1) {
-	    printf("    we must expand loop test time to more than %d secs\n",
-		ceil(test_time * (8 / loops)));
+	    printf("    we must expand alg2 loop test time to about %d secs\n",
+		ceil(test_time * (16 / loops)));
 	}
-	loops = 8;
+	loops = 16;
     }
     tlen = pow_loop(loops, &x, ex);
     if (config("user_debug") > 3) {
@@ -1099,7 +1309,12 @@ define pow_ratio(len)
     /*
      * return alg1 / alg2 rate ratio
      */
-    return (alg1_rate / alg2_rate);
+    ret = alg1_rate / alg2_rate;
+    if (config("user_debug") > 2) {
+	printf("\tprecise ratio is: %.f pow_ratio will return: %.3f\n",
+		alg1_rate / alg2_rate, ret);
+    }
+    return ret;
 }
 
 
@@ -1121,37 +1336,56 @@ define best_pow2()
 {
     local ratio;	/* previously calculated alg1/alg2 ratio */
     local low;		/* low loop value tested */
-    local high;		/* low loop value tested */
+    local high;		/* high loop value tested */
+    local mid;		/* between low and high */
+    local best_val;	/* value found with ratio closest to unity */
+    local best_ratio;	/* closest ratio found to unity */
     local expand;	/* how fast to expand the length */
     local looped;	/* 1 ==> we have expanded lengths before */
 
     /*
      * setup
      */
-    test_time = 15.0;
-    if (config("user_debug") > 0) {
-	printf("will start with loop test time of %d secs\n", test_time);
+    printf("WARNING: This tool may not be computing the correct best value\n");
+    test_time = 60.0;
+    printf("The best_pow2() function will take a LONG time to run!\n");
+    printf("It is important that best_pow2() run on an otherwise idle host!\n");
+    if (config("user_debug") <= 0) {
+	printf("To monitor progress, set user_debug to 2: "
+	       "config(\"user_debug\", 2)\n");
     }
+    printf("Starting with loop test time of %d secs\n", test_time);
 
     /*
      * firewall - must have a >1.02 ratio for the initial length
      *
      * We select 1.02 because of the precision of the CPU timing.  We
-     * want to firt move into an area where the 1st algoritm clearly
+     * want to first move into an area where the 1st algorithm clearly
      * dominates.
      */
     low = 4;
     high = 4;
+    best_val = high;
+    best_ratio = 1e10;	/* not a real value */
     do {
 	high *= 4;
 	if (config("user_debug") > 0) {
 	    printf("testing pmod alg1/alg2 ratio for len = %d\n", high);
 	}
 	ratio = pow_ratio(high);
+	if (abs(ratio - 1.0) < abs(best_ratio - 1.0)) {
+	    best_val = high;
+	    best_ratio = ratio;
+	    if (config("user_debug") > 1) {
+		printf("    len %d has a new closest ratio to unity: %.6f\n",
+		       best_val, best_ratio);
+	    }
+	}
 	if (config("user_debug") > 1) {
 	    printf("    pmod alg1/alg2 ratio = %.3f\n", ratio);
 	    if (ratio > 1.0 && ratio <= 1.02) {
-	    printf("    while alg1 is slightly better than alg2, it is not clearly better\n");
+	    printf("    while alg1 is slightly better than alg2, "
+	    	   "it is not clearly better\n");
 	    }
 	}
     } while (ratio <= 1.02);
@@ -1165,7 +1399,7 @@ define best_pow2()
     looped = 0;
     do {
 	/*
-	 * determine the paramters of the next ratio test
+	 * determine the parameters of the next ratio test
 	 *
 	 * We will multiplicatively expand our test level until
 	 * the ratio drops below 1.0.
@@ -1193,20 +1427,27 @@ define best_pow2()
 	 * determine the alg1/alg2 test ratio for this new length
 	 */
 	if (high >= 2^31) {
-	    quit "best_pow2: tests imply config(\"pow2\") should be >= 2^31";
+	    quit "best_pow2: test implies pow2 >= 2^31, which seems bogus";
 	}
 	if (config("user_debug") > 0) {
 	    printf("testing pmod alg1/alg2 ratio for len = %d\n", high);
 	}
 	ratio = pow_ratio(high);
+	if (abs(ratio - 1.0) < abs(best_ratio - 1.0)) {
+	    best_val = high;
+	    best_ratio = ratio;
+	    if (config("user_debug") > 1) {
+		printf("    len %d has a new closest ratio to unity: %.6f\n",
+		       best_val, best_ratio);
+	    }
+	}
 	if (config("user_debug") > 1) {
-	    printf("    pmod alg1/alg2 ratio = %.3f\n", ratio);
+	    printf("    pmod alg1/alg2 ratio = %.6f\n", ratio);
 	}
 	looped = 1;
-    } while (ratio >= 1.0);
+    } while (ratio > 1.0);
     if (config("user_debug") > 0) {
-	printf("alg1/alg2 ratio now < 1.0, starting binary search between %d and %d\n",
-	    low, high);
+	printf("Starting binary search between %d and %d\n", low, high);
     }
 
     /*
@@ -1215,39 +1456,69 @@ define best_pow2()
     while (low+1 < high) {
 
     	/* try the mid-point */
+	mid = int((low+high)/2);
 	if (config("user_debug") > 0) {
-	    printf("testing pmod alg1/alg2 ratio for len = %d\n",
-	    	int((low+high)/2));
+	    printf("testing pow2 alg1/alg2 ratio for len = %d\n", mid);
+	}
+	ratio = pow_ratio(mid);
+	if (abs(ratio - 1.0) < abs(best_ratio - 1.0)) {
+	    best_val = mid;
+	    best_ratio = ratio;
+	    if (config("user_debug") > 1) {
+		printf("    len %d has a new closest ratio to unity: %.6f\n",
+		       best_val, best_ratio);
+	    }
 	}
-	ratio = pow_ratio(int((low+high)/2));
 	if (config("user_debug") > 1) {
-	    printf("    pmod alg1/alg2 ratio = %.3f\n", ratio);
+	    printf("    len %d pmod alg1/alg2 ratio = %.6f\n", mid, ratio);
+	}
+
+	/* stop search if near unity */
+	if (close_to_one(ratio)) {
+	    low = mid;
+	    high = mid;
+	    if (config("user_debug") > 0) {
+	    	printf("\twe are close enough to unity ratio at: %d\n", mid);
+	    }
+	    break;
 	}
 
 	/* bump lower range up if we went over */
-	if (ratio >= 1.0) {
+	if (ratio > 1.0) {
 	    if (config("user_debug") > 2) {
 	    	printf("\tmove low from %d up to %d\n",
-		    low, int((low+high)/2));
+		    low, mid);
 	    }
-	    low = int((low+high)/2);
+	    low = mid;
 
 	/* drop higher range down if we went under */
 	} else {
 	    if (config("user_debug") > 2) {
 	    	printf("\tmove high from %d down to %d\n",
-		     high, int((low+high)/2));
+		     high, mid);
 	    }
-	    high = int((low+high)/2);
+	    high = mid;
+	}
+
+	/* report on test loop progress */
+	if (config("user_debug") > 1) {
+	    printf("\tsetting low: %d high: %d diff: %d\n",
+	    	   low, high, high-low);
 	}
     }
 
     /*
      * return on the suggested config("pow2") value
      */
+    mid = int((low+high)/2);
     if (config("user_debug") > 0) {
-	printf("best value of config(\"pow2\") is %d\n",
-	    (ratio >= 1.0) ? high : low);
+	printf("Best value for pmod is near %d\n", best_val);
+	printf("Best pmod alg1/alg2 ratio is: %.6f\n", best_ratio);
+	printf("We suggest placing this line in your .calcrc:\n");
+	printf("config(\"pow2\", %d),;\n", best_val);
+	printf("WARNING: It is believed that the output "
+	       "of this resource file is bogus!\n");
+	printf("WARNING: You may NOT wish to follow the above suggestion.\n");
     }
-    return ((ratio >= 1.0) ? high : low);
+    return mid;
 }

cal/beer.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: beer.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/beer.cal,v $
- *
  * Under source code control:	1996/11/13 13:21:05
  * File existed as early as:	1996
  *

cal/bernoulli.cal

@@ -1,7 +1,7 @@
 /*
- * bernoulli - clculate the Nth Bernoulli number B(n)
+ * bernoulli - calculate the Nth Bernoulli number B(n)
  *
- * Copyright (C) 2000  David I. Bell and Landon Curt Noll
+ * Copyright (C) 2000,2021  David I. Bell and Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: bernoulli.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/bernoulli.cal,v $
- *
  * Under source code control:	1991/09/30 11:18:41
  * File existed as early as:	1991
  *
@@ -30,9 +26,9 @@
 /*
  * Calculate the Nth Bernoulli number B(n).
  *
- * NOTE: This is now a bulitin function.
+ * NOTE: This is now a builtin function.
  *
- * The non-buildin code used the following symbolic formula to calculate B(n):
+ * The non-builtin code used the following symbolic formula to calculate B(n):
  *
  *	(b+1)^(n+1) - b^(n+1) = 0
  *
@@ -46,7 +42,7 @@
  *	B(3) = -(6*B(2) + 4*B(1) + 1) / 4
  *
  * The combinatorial factors in the expansion of the above formula are
- * calculated interatively, and we use the fact that B(2i+1) = 0 if i > 0.
+ * calculated interactively, and we use the fact that B(2i+1) = 0 if i > 0.
  * Since all previous B(n)'s are needed to calculate a particular B(n), all
  * values obtained are saved in an array for ease in repeated calculations.
  */

cal/bernpoly.cal

@@ -0,0 +1,55 @@
+/*
+ * bernpoly - Bernoulli polynomials B_n(z) for arbitrary n,z..
+ *
+ * Copyright (C) 2013,2021 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2013/08/11 01:31:28
+ * File existed as early as:	2013
+ */
+
+
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+read -once zeta2
+
+
+/* Idea by Don Zagier */
+define bernpoly(n,z){
+  local h s c k;
+  if(isint(n) && n>=0){
+    h=0;s=0;c=-1;
+    for(k=1;k<=n+1;k++){
+      c*=1-(n+2)/k;
+      s+=z^n;
+      z++;
+      h+=c*s/k;
+    }
+    return h;
+  }
+  else return -n*hurwitzzeta(1-n,z);
+}
+
+
+/*
+ * restore internal function from resource debugging
+ */
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "bernpoly(n,z)";
+}

cal/bigprime.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: bigprime.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/bigprime.cal,v $
- *
  * Under source code control:	1991/05/22 21:56:32
  * File existed as early as:	1991
  *

cal/bindings

@@ -16,10 +16,6 @@
 # received a copy with calc; if not, write to Free Software Foundation, Inc.
 # 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
 #
-# @(#) $Revision: 30.1 $
-# @(#) $Id: bindings,v 30.1 2007/03/16 11:09:54 chongo Exp $
-# @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/bindings,v $
-#
 # Under source code control:	1993/05/02 20:09:19
 # File existed as early as:	1993
 #

cal/brentsolve.cal

@@ -0,0 +1,254 @@
+/*
+ * brentsolve - Root finding with the Brent-Dekker trick
+ *
+ * Copyright (C) 2013,2021 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2013/08/11 01:31:28
+ * File existed as early as:	2013
+ */
+
+
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+/*
+  A short explanation is at http://en.wikipedia.org/wiki/Brent%27s_method
+  I tried to follow the description at wikipedia as much as possible to make
+  the slight changes I did more visible.
+  You may give http://people.sc.fsu.edu/~jburkardt/cpp_src/brent/brent.html a
+  short glimpse (Brent's originl Fortran77 versions and some translations of
+  it).
+*/
+
+static true = 1;
+static false = 0;
+define brentsolve(low, high,eps){
+  local a b c d fa fb fc fa2 fb2 fc2 s fs  tmp tmp2  mflag i places;
+  a = low;
+  b = high;
+  c = 0;
+
+  if(isnull(eps))
+    eps = epsilon(epsilon()*1e-3);
+  places = highbit(1 + int( 1/epsilon() ) ) + 1;
+
+  d = 1/eps;
+
+  fa = f(a);
+  fb = f(b);
+
+  fc = 0;
+  s = 0;
+  fs = 0;
+
+  if(fa * fb >= 0){
+    if(fa < fb){
+      epsilon(eps);
+      return a;
+    }
+    else{
+      epsilon(eps);
+      return b;
+    }
+  }
+
+  if(abs(fa) < abs(fb)){
+    tmp = a; a = b; b = tmp;
+    tmp = fa; fa = fb; fb = tmp;
+  }
+
+  c = a;
+  fc = fa;
+  mflag = 1;
+  i = 0;
+
+  while(!(fb==0) && (abs(a-b) > eps)){
+    if((fa != fc) && (fb != fc)){
+      /* Inverse quadratic interpolation*/
+      fc2 = fc^2;
+      fa2 = fa^2;
+      s = bround(((fb^2*((fc*a)-(c*fa)))+(fb*((c*fa2)-(fc2*a)))+(b*((fc2*fa)
+                    -(fc*fa2))))/((fc - fb)*(fa - fb)*(fc - fa)),places++);
+    }
+    else{
+      /* Secant Rule*/
+      s =bround( b - fb * (b - a) / (fb - fa),places++);
+    }
+    tmp2 = (3 * a + b) / 4;
+    if( (!( ((s > tmp2) && (s < b))||((s < tmp2) && (s > b))))
+        || (mflag && (abs(s - b) >= (abs(b - c) / 2)))
+        || (!mflag && (abs(s - b) >= (abs(c - d) / 2)))) {
+      s = (a + b) / 2;
+      mflag = true;
+    }
+    else{
+      if( (mflag && (abs(b - c) < eps))
+          || (!mflag && (abs(c - d) < eps))) {
+        s = (a + b) / 2;
+        mflag = true;
+      }
+      else
+        mflag = false;
+    }
+    fs = f(s);
+    c = b;
+    fc = fb;
+    if (fa * fs < 0){
+      b = s;
+      fb = fs;
+    }
+    else {
+      a = s;
+      fa = fs;
+    }
+
+    if (abs(fa) < abs(fb)){
+      tmp = a; a = b; b = tmp;
+      tmp = fa; fa = fb; fb = tmp;
+    }
+    i++;
+    if (i > 1000){
+      epsilon(eps);
+      return newerror("brentsolve: does not converge");
+    }
+  }
+  epsilon(eps);
+  return b;
+}
+
+/*
+  A variation of the solver to accept functions named differently from "f". The
+  code should explain it.
+*/
+define brentsolve2(low, high,which,eps){
+  local a b c d fa fb fc fa2 fb2 fc2 s fs  tmp tmp2  mflag i places;
+  a = low;
+  b = high;
+  c = 0;
+
+  switch(param(0)){
+    case 0:
+    case 1: return newerror("brentsolve2: not enough arguments");
+    case 2: eps = epsilon(epsilon()*1e-2);
+            which = 0;break;
+    case 3: eps = epsilon(epsilon()*1e-2);break;
+    default: break;
+  };
+  places = highbit(1 + int(1/epsilon())) + 1;
+
+  d = 1/eps;
+
+  switch(which){
+    case 1:  fa = __CZ__invbeta(a);
+             fb = __CZ__invbeta(b); break;
+    case 2:  fa = __CZ__invincgamma(a);
+             fb = __CZ__invincgamma(b); break;
+
+    default: fa = f(a);fb = f(b);   break;
+  };
+
+  fc = 0;
+  s = 0;
+  fs = 0;
+
+  if(fa * fb >= 0){
+    if(fa < fb)
+      return a;
+    else
+      return b;
+  }
+
+  if(abs(fa) < abs(fb)){
+    tmp = a; a = b; b = tmp;
+    tmp = fa; fa = fb; fb = tmp;
+  }
+
+  c = a;
+  fc = fa;
+  mflag = 1;
+  i = 0;
+
+  while(!(fb==0) && (abs(a-b) > eps)){
+
+    if((fa != fc) && (fb != fc)){
+      /* Inverse quadratic interpolation*/
+      fc2 = fc^2;
+      fa2 = fa^2;
+      s = bround(((fb^2*((fc*a)-(c*fa)))+(fb*((c*fa2)-(fc2*a)))+(b*((fc2*fa)
+                    -(fc*fa2))))/((fc - fb)*(fa - fb)*(fc - fa)),places);
+      places++;
+    }
+    else{
+      /* Secant Rule*/
+      s =bround( b - fb * (b - a) / (fb - fa),places);
+      places++;
+    }
+    tmp2 = (3 * a + b) / 4;
+    if( (!( ((s > tmp2) && (s < b))||((s < tmp2) && (s > b))))
+        || (mflag && (abs(s - b) >= (abs(b - c) / 2)))
+        || (!mflag && (abs(s - b) >= (abs(c - d) / 2)))) {
+      s = (a + b) / 2;
+      mflag = true;
+    }
+    else{
+      if( (mflag && (abs(b - c) < eps))
+          || (!mflag && (abs(c - d) < eps))) {
+        s = (a + b) / 2;
+        mflag = true;
+      }
+      else
+        mflag = false;
+    }
+    switch(which){
+      case 1:  fs = __CZ__invbeta(s); break;
+      case 2:  fs = __CZ__invincgamma(s); break;
+
+      default: fs = f(s);             break;
+    };
+    c = b;
+    fc = fb;
+    if (fa * fs < 0){
+      b = s;
+      fb = fs;
+    }
+    else {
+      a = s;
+      fa = fs;
+    }
+
+    if (abs(fa) < abs(fb)){
+      tmp = a; a = b; b = tmp;
+      tmp = fa; fa = fb; fb = tmp;
+    }
+    i++;
+    if (i > 1000){
+      return newerror("brentsolve2: does not converge");
+    }
+  }
+  return b;
+}
+
+
+/*
+ * restore internal function from resource debugging
+ */
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "brentsolve(low, high,eps)";
+    print "brentsolve2(low, high,which,eps)";
+}

cal/chi.cal

@@ -1,7 +1,7 @@
 /*
  * chi - chi^2 probabilities with degrees of freedom for null hypothesis
  *
- * Copyright (C) 2001  Landon Curt Noll
+ * Copyright (C) 2001,2021  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: chi.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/chi.cal,v $
- *
  * Under source code control:	2001/03/27 14:10:11
  * File existed as early as:	2001
  *
@@ -55,7 +51,7 @@ define Z(x, eps_term)
 
 
 /*
- * P(x[, eps]) asymtotic P(x) expansion for x>0 to an given epsilon error term
+ * P(x[, eps]) asymptotic P(x) expansion for x>0 to an given epsilon error term
  *
  * NOTE: If eps is omitted, the stored epsilon value is used.
  *
@@ -103,7 +99,7 @@ define P(x, eps_term)
     }
 
     /*
-     * aproximate sum(n=0; n < infinity){x^(2*n+1)/(1*3*5*...(2*n+1)}
+     * approximate sum(n=0; n < infinity){x^(2*n+1)/(1*3*5*...(2*n+1)}
      */
     x2 = x*x;
     x_term = x;
@@ -134,7 +130,7 @@ define P(x, eps_term)
  *
  *    The chi_prob() function does not work well with odd degrees of freedom.
  *    It is reasonable with even degrees of freedom, although one must give
- *    a sifficently small error term as the degress gets large (>100).
+ *    a sufficiently small error term as the degrees gets large (>100).
  *
  * NOTE: This function does not work well with odd degrees of freedom.
  *	 Can somebody help / find a bug / provide a better method of
@@ -190,7 +186,7 @@ define chi_prob(chi_sq, v, eps_term)
     local r;		/* index in finite sum */
     local r_lim;	/* limit value for r */
     local s;		/* sum */
-    local d;		/* demoninator (2*4*6*... or 1*3*5...) */
+    local d;		/* denominator (2*4*6*... or 1*3*5...) */
     local chi_term;	/* chi_sq^r */
     local ret;		/* return value */
 

cal/chrem.cal

@@ -1,7 +1,7 @@
 /*
- * chrem - chinese remainder theorem/problem solver
+ * chrem - Chinese remainder theorem/problem solver
  *
- * Copyright (C) 1999  Ernest Bowen and Landon Curt Noll
+ * Copyright (C) 1999,2021  Ernest Bowen and Landon Curt Noll
  *
  * Primary author:  Ernest Bowen
  *
@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: chrem.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/chrem.cal,v $
- *
  * Under source code control:	1992/09/26 01:00:47
  * File existed as early as:	1992
  *
@@ -30,7 +26,7 @@
  */
 
 /*
- * When possible, chrem finds solutions for x of a set of congruences
+ * When possible, chrem finds solutions for x of a set of congruence
  * of the form:
  *
  *			x = r1 (mod m1)
@@ -39,7 +35,7 @@
  *
  * where the residues r1, r2, ... and the moduli m1, m2, ... are
  * given integers.  The Chinese remainder theorem states that if
- * m1, m2, ... are relatively prime in pairs, the above congruences
+ * m1, m2, ... are relatively prime in pairs, the above congruence
  * have a unique solution modulo  m1 * m2 * ...	  If m1, m2, ...
  * are not relatively prime in pairs, it is possible that no solution
  * exists.  If solutions exist, the general solution is expressible as:

cal/constants.cal

@@ -0,0 +1,100 @@
+/*
+ * constants - implementation of different constants to arbitrary precision
+ *
+ * Copyright (C) 2013 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2013/08/11 01:31:28
+ * File existed as early as:	2013
+ */
+
+
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+static __CZ__euler_mascheroni = 0;
+static __CZ__euler_mascheroni_prec = 0;
+
+
+define e(){
+  local k temp1 temp2 ret eps factor upperlimit prec;
+
+  prec = digits(1/epsilon());
+  if(__CZ__euler_mascheroni != 0 &&  __CZ__euler_mascheroni_prec >= prec)
+    return __CZ__euler_mascheroni;
+  if(prec<=20) return 2.718281828459045235360287471;
+  if(prec<=1800){
+    __CZ__euler_mascheroni = exp(1);
+    __CZ__euler_mascheroni_prec = prec;
+  }
+
+  eps=epsilon(1e-20);
+  factor = 1;
+  k = 0;
+  upperlimit = prec * ln(10);
+  while(k<upperlimit){
+    k += ln(factor);
+    factor++;
+  }
+  epsilon(eps);
+  temp1 = 0;
+  ret   = 1;
+  for(k=3;k<=factor;k++){
+    temp2 = temp1;
+    temp1 = ret;
+    ret =  (k-1) *(temp1 + temp2);
+  }
+
+  ret = inverse( ret * inverse(factorial(factor) ) ) ;
+  __CZ__euler_mascheroni = ret;
+  __CZ__euler_mascheroni_prec = prec;
+  return ret;
+}
+
+
+/* Lupas' series */
+static __CZ__catalan = 0;
+static __CZ__catalan_prec = 0;
+define G(){
+  local eps a s t n;
+  eps = epsilon(epsilon()*1e-10);
+  if(__CZ__catalan != 0 &&  __CZ__catalan >= log(1/eps))
+    return __CZ__catalan;
+  a = 1;
+  s = 0;
+  t = 1;
+  n = 1;
+  while(abs(t)> eps){
+    a *= 32 * n^3 * (2*n-1);
+    a /=((3-16*n+16*n^2)^2);
+    t = a * (-1)^(n-1) * (40*n^2-24*n+3) / (n^3 * (2*n-1));
+    s += t;
+    n += 1;
+  }
+  s = s/64;
+  __CZ__catalan = s;
+  __CZ__catalan_prec = log(1/eps);
+  epsilon(eps);
+  return s;
+}
+
+
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "e()";
+    print "G()";
+}

cal/deg.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: deg.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/deg.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:33
  * File existed as early as:	before 1990
  *
@@ -28,9 +24,9 @@
  */
 
 
-obj dms {deg, min, sec};
+obj deg {deg, min, sec};
 
-define dms(deg, min, sec)
+define deg(deg, min, sec)
 {
 	local ans;
 
@@ -38,18 +34,18 @@ define dms(deg, min, sec)
 		sec = 0;
 	if (isnull(min))
 		min = 0;
-	obj dms ans;
+	obj deg ans;
 	ans.deg = deg;
 	ans.min = min;
 	ans.sec = sec;
-	fixdms(ans);
+	fixdeg(ans);
 	return ans;
 }
 
 
-define dms_add(a, b)
+define deg_add(a, b)
 {
-	local obj dms	ans;
+	local obj deg	ans;
 
 	ans.deg = 0;
 	ans.min = 0;
@@ -66,14 +62,14 @@ define dms_add(a, b)
 		ans.sec += b.sec;
 	} else
 		ans.deg += b;
-	fixdms(ans);
+	fixdeg(ans);
 	return ans;
 }
 
 
-define dms_neg(a)
+define deg_neg(a)
 {
-	local obj dms	ans;
+	local obj deg	ans;
 
 	ans.deg = -a.deg;
 	ans.min = -a.min;
@@ -82,15 +78,15 @@ define dms_neg(a)
 }
 
 
-define dms_sub(a, b)
+define deg_sub(a, b)
 {
 	return a - b;
 }
 
 
-define dms_mul(a, b)
+define deg_mul(a, b)
 {
-	local obj dms	ans;
+	local obj deg	ans;
 
 	if (istype(a, ans) && istype(b, ans))
 		quit "Cannot multiply degrees together";
@@ -103,24 +99,24 @@ define dms_mul(a, b)
 		ans.min = b.min * a;
 		ans.sec = b.sec * a;
 	}
-	fixdms(ans);
+	fixdeg(ans);
 	return ans;
 }
 
 
-define dms_print(a)
+define deg_print(a)
 {
 	print a.deg : 'd' : a.min : 'm' : a.sec : 's' :;
 }
 
 
-define dms_abs(a)
+define deg_abs(a)
 {
 	return a.deg + a.min / 60 + a.sec / 3600;
 }
 
 
-define fixdms(a)
+define fixdeg(a)
 {
 	a.min += frac(a.deg) * 60;
 	a.deg = int(a.deg);
@@ -134,5 +130,5 @@ define fixdms(a)
 }
 
 if (config("resource_debug") & 3) {
-    print "obj dms {deg, min, sec} defined";
+    print "obj deg {deg, min, sec} defined";
 }

cal/dms.cal

@@ -0,0 +1,364 @@
+/*
+ * dms - calculate in degrees, minutes, and seconds (based on deg)
+ *
+ * Copyright (C) 1999,2010,2021  David I. Bell and Landon Curt Noll
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.	 See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	1990/02/15 01:50:33
+ * File existed as early as:	before 1990
+ *
+ * Share and enjoy!  :-)	http://www.isthe.com/chongo/tech/comp/calc/
+ */
+
+
+obj dms {deg, min, sec};
+
+define dms(deg, min, sec)
+{
+    local obj dms ans;		/* return value */
+
+    /* default missing args to 0 */
+    if (isnull(sec)) {
+	sec = 0;
+    }
+    if (isnull(min)) {
+	min = 0;
+    }
+
+    /* load object */
+    ans.deg = deg;
+    ans.min = min;
+    ans.sec = sec;
+
+    /* return properly formed object */
+    ans = fixdms(ans);
+    return ans;
+}
+
+
+define dms_add(a, b)
+{
+    local obj dms ans;		/* return value */
+
+    /* initialize value to 1st arg */
+    if (istype(a, ans)) {
+	/* 1st arg is dms object, load it */
+	ans.deg = a.deg;
+	ans.min = a.min;
+	ans.sec = a.sec;
+    } else {
+	/* 1st arg is not dms, assume scalar degrees */
+	ans.deg = a;
+	ans.min = 0;
+	ans.sec = 0;
+    }
+
+    /* add value of 2nd arg */
+    if (istype(b, ans)) {
+	/* 2nd arg is dms object, add it */
+	ans.deg += b.deg;
+	ans.min += b.min;
+	ans.sec += b.sec;
+    } else {
+	/* 2nd arg is not dms, add scalar degrees */
+	ans.deg += b;
+    }
+
+    /* return normalized result */
+    ans = fixdms(ans);
+    return ans;
+}
+
+
+define dms_neg(a)
+{
+    local obj dms ans;		/* return value */
+
+    /* negate argument */
+    if (istype(a, ans)) {
+	/* 1st arg is dms object, load it */
+	ans.deg = -a.deg;
+	ans.min = -a.min;
+	ans.sec = -a.sec;
+    } else {
+	/* 2nd arg is not dms, negate scalar degrees */
+	ans.deg = -a;
+	ans.min = 0;
+	ans.sec = 0;
+    }
+
+    /* return normalized result */
+    ans = fixdms(ans);
+    return ans;
+}
+
+
+define dms_sub(a, b)
+{
+    local obj dms ans;		/* return value */
+
+    /* initialize value to 1st arg */
+    if (istype(a, ans)) {
+	/* 1st arg is dms object, load it */
+	ans.deg = a.deg;
+	ans.min = a.min;
+	ans.sec = a.sec;
+    } else {
+	/* 1st arg is not dms, assume scalar degrees */
+	ans.deg = a;
+	ans.min = 0;
+	ans.sec = 0;
+    }
+
+    /* subtract value of 2nd arg */
+    if (istype(b, ans)) {
+	/* 2nd arg is dms object, subtract it */
+	ans.deg -= b.deg;
+	ans.min -= b.min;
+	ans.sec -= b.sec;
+    } else {
+	/* 2nd arg is not dms, subtract scalar degrees */
+	ans.deg -= b;
+    }
+
+    /* return normalized result */
+    ans = fixdms(ans);
+    return ans;
+}
+
+
+define dms_mul(a, b)
+{
+    local obj dms ans;		/* return value */
+
+    /* dms object multiplication */
+    if (istype(a, ans) && istype(b, ans)) {
+	ans.deg = dms_abs(a) * dms_abs(b);
+	ans.min = 0;
+	ans.sec = 0;
+
+    /* scalar multiplication */
+    } else if (istype(a, ans)) {
+	ans.deg = a.deg * b;
+	ans.min = a.min * b;
+	ans.sec = a.sec * b;
+    } else {
+	ans.deg = b.deg * a;
+	ans.min = b.min * a;
+	ans.sec = b.sec * a;
+    }
+
+    /* return normalized result */
+    ans = fixdms(ans);
+    return ans;
+}
+
+
+define dms_print(a)
+{
+    local obj dms ans;		/* temp object for dms type testing */
+
+    /* firewall - arg must be a dms object */
+    if (! istype(a, ans)) {
+	quit "dms_print called with non dms object";
+    }
+
+    /* print in dms form */
+    print a.deg : 'd' : a.min : 'm' : a.sec : 's' :;
+}
+
+
+define dms_abs(a)
+{
+    local obj dms ans;		/* temp object for dms type testing */
+    local deg;			/* return scalar value */
+
+    /* firewall - just absolute value non dms objects */
+    if (! istype(a, ans)) {
+    	return abs(a);
+    }
+
+    /* compute degrees */
+    deg = a.deg + a.min / 60 + a.sec / 3600;
+
+    /* return degrees */
+    return deg;
+}
+
+
+define dms_norm(a)
+{
+    local obj dms ans;		/* temp object for dms type testing */
+    local deg;			/* degrees */
+
+    /* firewall - arg must be a dms object */
+    if (! istype(a, ans)) {
+	quit "dms_norm called with non dms object";
+    }
+
+    /* square degrees (norm is the square of absolute value */
+    deg = dms_abs(a);
+
+    /* return degrees */
+    return deg*deg;
+}
+
+
+define dms_test(a)
+{
+    local obj dms ans;		/* temp value */
+
+    /* firewall - arg must be a dms object */
+    if (! istype(a, ans)) {
+	quit "dms_test called with non dms object";
+    }
+
+    /* return false of non-zero */
+    ans = fixdms(a);
+    if (ans.deg == 0 && ans.min == 0 && ans.sec == 0) {
+	/* false */
+	return 0;
+    }
+    /* true */
+    return 1;
+}
+
+
+define dms_int(a)
+{
+    local obj dms ans;		/* return value */
+
+    /* firewall - arg must be a dms object */
+    if (! istype(a, ans)) {
+	quit "dms_int called with non dms object";
+    }
+
+    /* normalize the argument */
+    ans = fixdms(a);
+
+    /* truncate to the nearest second */
+    ans.sec = int(ans.sec);
+
+    /* return value to the nearest second */
+    return ans;
+}
+
+
+define dms_frac(a)
+{
+    local obj dms ans;		/* return value */
+
+    /* firewall - arg must be a dms object */
+    if (! istype(a, ans)) {
+	quit "dms_frac called with non dms object";
+    }
+
+    /* normalize the argument */
+    ans = fixdms(a);
+
+    /* remove all but fractional seconds */
+    ans.deg = 0;
+    ans.min = 0;
+    ans.sec = frac(ans.sec);
+
+    /* return value to the second fraction */
+    return ans;
+}
+
+
+define dms_rel(a,b)
+{
+    local abs_a, abs_b;		/* scalars of the arguments */
+
+    /* compute scalars of the arguments */
+    abs_a = dms_abs(a);
+    abs_b = dms_abs(b);
+
+    /* return the comparison */
+    return cmp(abs_a, abs_b);
+}
+
+
+define dms_cmp(a,b)
+{
+    local abs_a, abs_b;		/* scalars of the arguments */
+
+    /* compute scalars of the arguments */
+    abs_a = dms_abs(a);
+    abs_b = dms_abs(b);
+
+    /* return the equality comparison */
+    return (abs_a == abs_b);
+}
+
+
+define dms_inc(a)
+{
+    local obj dms ans;		/* return value */
+
+    /* increment a dms object */
+    if (istype(a, ans)) {
+	ans = a;
+	++ans.sec;
+
+	/* return normalized result */
+	ans = fixdms(ans);
+	return ans;
+    }
+
+    /* increment a scalar */
+    return a+1;
+}
+
+
+define dms_dec(a)
+{
+    local obj dms ans;		/* return value */
+
+    /* decrement a dms object */
+    if (istype(a, ans)) {
+	ans = a;
+	--ans.sec;
+
+	/* return normalized result */
+	ans = fixdms(ans);
+	return ans;
+    }
+
+    /* decrement a scalar */
+    return a-1;
+}
+
+
+define fixdms(a)
+{
+    local obj dms ans;		/* temp value */
+
+    /* firewall */
+    if (! istype(a, ans)) {
+	quit "attempt to fix a non dms object";
+    }
+
+    /* use builtin d2dms function */
+    d2dms(a.deg + a.min/60 + a.sec/3600, a.deg, a.min, a.sec),;
+
+    /* return normalized result */
+    return a;
+}
+
+if (config("resource_debug") & 3) {
+    print "obj dms {deg, min, sec} defined";
+}

cal/dotest.cal

@@ -8,18 +8,14 @@
  * copyright this dotest_code.
  *
  * ERNEST BOWEN AND LANDON CURT NOLL DISCLAIMS ALL WARRANTIES WITH REGARD TO
- * THIS	 SOFTWARE,  INCLUDING  ALL IMPLIED WARRANTIES OF MER-
- * CHANTABILITY AND FITNESS.  IN NO EVENT SHALL	 LANDON	 CURT
+ * THIS         SOFTWARE,  INCLUDING  ALL IMPLIED WARRANTIES OF MER-
+ * CHANTABILITY AND FITNESS.  IN NO EVENT SHALL         LANDON  CURT
  * NOLL	 BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL
  * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM  LOSS  OF
  * USE,	 DATA  OR  PROFITS, WHETHER IN AN ACTION OF CONTRACT,
  * NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR  IN
  * CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  *
- * @(#) $Revision: 30.2 $
- * @(#) $Id: dotest.cal,v 30.2 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/dotest.cal,v $
- *
  * This file is not covered under version 2.1 of the GNU LGPL.
  *
  * Under source dotest_code control:	2006/03/08 05:54:09
@@ -177,7 +173,7 @@ define dotest(dotest_file, dotest_code = 0, dotest_maxcond = -1)
     }
 
     /*
-     * preppare to return to the caller environment
+     * prepare to return to the caller environment
      *
      * We increase the caller's error count by the number
      * of line tests that failed, not the number of internal

cal/ellip.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: ellip.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/ellip.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:33
  * File existed as early as:	before 1990
  *

cal/factorial.cal

@@ -0,0 +1,200 @@
+/*
+ * factorial - implementation of different algorithms for the factorial
+ *
+ * Copyright (C) 2013 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2013/08/11 01:31:28
+ * File existed as early as:	2013
+ */
+
+
+/*
+ * hide internal function from resource debugging
+ */
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+/*
+  get dependencies
+*/
+read -once toomcook;
+
+
+/* A simple list to keep things...uhm...simple?*/
+static __CZ__primelist = list();
+
+/* Helper for primorial: fill list with primes in range a,b */
+define __CZ__fill_prime_list(a,b)
+{
+  local k;
+  k=a;
+  if(isprime(k))k--;
+  while(1){
+    k = nextprime(k);
+    if(k > b) break;
+    append(__CZ__primelist,k );
+  }
+}
+
+/* Helper for factorial: how often prime p divides the factorial of n */
+define __CZ__prime_divisors(n,p)
+{
+    local q,m;
+    q = n;
+    m = 0;
+    if (p > n) return 0;
+    if (p > n/2) return 1;
+    while (q >= p) {
+        q  = q//p;
+        m += q;
+    }
+    return m;
+}
+
+/*
+  Wrapper. Please set cut-offs to own taste and hardware.
+*/
+define factorial(n){
+  local prime result shift prime_list k k1 k2 expo_list pix cut primorial;
+
+  result = 1;
+  prime  = 2;
+
+  if(!isint(n))  {
+    return newerror("factorial(n): n is not an integer"); ## or gamma(n)?
+  }
+  if(n < 0)      return newerror("factorial(n): n < 0");
+  if(n < 9000 && !isdefined("test8900"))   {
+    ## builtin is implemented with splitting but only with
+    ## Toom-Cook 2 (by Karatsuba (the father))
+    return n!;
+  }
+
+  shift = __CZ__prime_divisors(n,prime);
+  prime = 3;
+  cut = n//2;
+  pix   = pix(cut);
+  prime_list = mat[pix];
+  expo_list  = mat[pix];
+
+  k = 0;
+/*
+  Peter Borwein's algorithm
+
+  @Article{journals/jal/Borwein85,
+  author = {Borwein, Peter B.},
+  title  = {On the Complexity of Calculating Factorials.},
+  journal = {J. Algorithms},
+  year   = {1985},
+  number = {3},
+  url    = {http://dblp.uni-trier.de/db/journals/jal/jal6.html#Borwein85}
+*/
+
+  do {
+    prime_list[k]  = prime;
+    expo_list[k++] = __CZ__prime_divisors(n,prime);
+    prime          = nextprime(prime);
+  }while(prime <= cut);
+
+  /* size of the largest exponent in bits  */
+  k1 = highbit(expo_list[0]);
+  k2 = size(prime_list)-1;
+
+  for(;k1>=0;k1--){
+    /*
+       the cut-off for T-C-4 ist still to low, using T-C-3 here
+       TODO: check cutoffs
+     */
+    result = toomcook3square(result);
+    /*
+      almost all time is spend in this loop, so cutting of the
+      upper half of the primes makes sense
+    */
+    for(k=0; k<=k2; k++) {
+      if((expo_list[k] & (1 << k1)) != 0) {
+        result *= prime_list[k];
+      }
+    }
+
+  }
+  primorial = primorial( cut, n);
+  result *= primorial;
+  result <<= shift;
+  return result;
+}
+
+/*
+  Helper for primorial: do the product with binary splitting
+  TODO: do it without the intermediate list
+*/
+define __CZ__primorial__lowlevel( a, b ,p)
+{
+  local  c;
+  if( b == a) return p ;
+  if( b-a > 1){
+    c= (b + a) >> 1;
+    return  __CZ__primorial__lowlevel( a   , c , __CZ__primelist[a] )
+         *  __CZ__primorial__lowlevel( c+1 , b , __CZ__primelist[b] ) ;
+  }
+  return  __CZ__primelist[a] * __CZ__primelist[b];
+}
+
+/*
+   Primorial, Product of consecutive primes in range a,b
+
+  Originally meant to do primorials with a start different from 2, but
+  found out that this is faster at about a=1,b>=10^5 than the builtin
+  function pfact().  With the moderately small list a=1,b=10^6 (78498
+  primes) it is 3 times faster.  A quick look-up showed what was already
+  guessed: pfact() does it linearly.  (BTW: what is the time complexity
+  of the primorial with the naive algorithm?)
+*/
+define primorial(a,b)
+{
+  local C1 C2;
+  if(!isint(a)) return newerror("primorial(a,b): a is not an integer");
+  else if(!isint(b)) return newerror("primorial(a,b): b is not an integer");
+  else if(a < 0) return newerror("primorial(a,b): a < 0");
+  else if( b < 2 ) return newerror("primorial(a,b): b < 2");
+  else if( b < a) return newerror("primorial(a,b): b < a");
+  else{
+    /* last prime < 2^32 is also  max. prime for nextprime()*/
+    if(b >= 4294967291) return newerror("primorial(a,b): max. prime exceeded");
+    if(b ==  2) return 2;
+    /*
+      Can be extended by way of pfact(b)/pfact(floor(a-1/2)) for small a
+     */
+    if(a<=2 && b < 10^5) return pfact(b);
+    /* TODO: use pix() and a simple array (mat[])instead*/
+     __CZ__primelist = list();
+     __CZ__fill_prime_list(a,b);
+    C1 = size(__CZ__primelist)-1;
+    return __CZ__primorial__lowlevel( 0, C1,1)
+  }
+}
+
+
+/*
+ * restore internal function from resource debugging
+ * report important interface functions
+ */
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "factorial(n)";
+    print "primorial(a, b)";
+}

cal/factorial2.cal

@@ -0,0 +1,719 @@
+/*
+ * factorial2 - implementation of different factorial related functions
+ *
+ * Copyright (C) 2013,2021 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2013/08/11 01:31:28
+ * File existed as early as:	2013
+ */
+
+
+/*
+ * hide internal function from resource debugging
+ */
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+/*
+  get dependencies
+*/
+read -once factorial toomcook specialfunctions;
+
+
+/*
+  Factorize a factorial and put the result in a 2-column matrix with pi(n) rows
+  mat[ primes , exponent ]
+  Result can be restricted to start at a prime different from 2 with the second
+  argument "start". That arguments gets taken at face value if it prime and
+  smaller than n, otherwise the next larger prime is taken if that prime is
+  smaller than n.
+*/
+
+define __CZ__factor_factorial(n,start){
+  local prime prime_list k pix stop;
+
+
+  if(!isint(n)) return
+    newerror("__CZ__factor_factorial(n,start): n is not integer");
+  if(n <  0)    return newerror("__CZ__factor_factorial(n,start): n < 0");
+  if(n == 1)    return newerror("__CZ__factor_factorial(n,start): n == 1");
+
+  if(start){
+    if(!isint(start) && start < 0 && start > n)
+      return newerror("__CZ__factor_factorial(n,start): value of "
+      		      "parameter 'start' out of range");
+    if(start == n && isprime(n)){
+      prime_list = mat[1 , 2];
+      prime_list[0,0] = n;
+      prime_list[0,1] = 1;
+    }
+    else if(!isprime(start) && nextprime(start) >n)
+      return newerror("__CZ__factor_factorial(n,start): value of parameter "
+      		      "'start' out of range");
+    else{
+      if(!isprime(start))  prime = nextprime(start);
+      else prime = start;
+    }
+  }
+  else
+    prime = 2;
+
+  pix   = pix(n);
+  if(start){
+    pix -= pix(prime) -1;
+  }
+  prime_list = mat[pix , 2];
+
+  k = 0;
+
+  do {
+    prime_list[k  ,0] = prime;
+    prime_list[k++,1] = __CZ__prime_divisors(n,prime);
+    prime          = nextprime(prime);
+  }while(prime <= n);
+
+  return prime_list;
+}
+
+/*
+
+  subtracts exponents of n_1! from exponents of n_2! with n_1<=n_2
+
+  Does not check for size or consecutiveness of the primes or a carry
+*/
+
+define __CZ__subtract_factored_factorials(matrix_2n,matrix_n){
+  local k ret len1,len2,tmp count p e;
+  len1 = size(matrix_n)/2;
+  len2 = size(matrix_2n)/2;
+  if(len2<len1){
+
+    swap(len1,len2);
+    tmp = matrix_n;
+    matrix_n = matrix_2n;
+    matrix_2n = tmp;
+  }
+  tmp = mat[len1,2];
+  k   = 0;
+
+  for(;k<len1;k++){
+     p = matrix_2n[k,0];
+     e = matrix_2n[k,1] - matrix_n[k,1];
+     if(e!=0){
+       tmp[count  ,0] = p;
+       tmp[count++,1] = e;
+     }
+  }
+  ret = mat[count + (len2-len1),2];
+  for(k=0;k<count;k++){
+     ret[k,0] = tmp[k,0];
+     ret[k,1] = tmp[k,1];
+  }
+
+  free(tmp);
+  for(k=len1;k<len2;k++){
+     ret[count,0] = matrix_2n[k,0];
+     ret[count++,1] = matrix_2n[k,1];
+  }
+  return ret;
+}
+
+/*
+
+  adds exponents of n_1! to exponents of n_2! with n_1<=n_2
+
+  Does not check for size or consecutiveness of the primes or a carry
+*/
+
+define __CZ__add_factored_factorials(matrix_2n,matrix_n){
+  local k ret len1,len2,tmp;
+  len1 = size(matrix_n)/2;
+  len2 = size(matrix_2n)/2;
+  if(len2<len1){
+    swap(len1,len2);
+    tmp = matrix_n;
+    matrix_n = matrix_2n;
+    matrix_2n = tmp;
+  }
+  ret = mat[len2,2];
+  k   = 0;
+  for(;k<len1;k++){
+     ret[k,0] = matrix_2n[k,0];
+     ret[k,1] = matrix_2n[k,1] + matrix_n[k,1];
+  }
+  for(;k<len2;k++){
+     ret[k,0] = matrix_2n[k,0];
+     ret[k,1] = matrix_2n[k,1];
+  }
+  return ret;
+}
+
+/*
+   Does not check if all exponents are positive
+
+
+                   timings
+                   this		  comb	    comb-this     rel. k/n
+;  benchmark_binomial(10,13)
+n=2^13 k=2^10 	 0.064004 	 0.016001	+ 	0.76923076923076923077
+n=2^13 k=2^11 	 0.064004 	 0.048003	+ 	0.84615384615384615385
+n=2^13 k=2^12 	 0.068004 	 0.124008	- 	0.92307692307692307692
+;  benchmark_binomial(10,15)
+n=2^15 k=2^10 	 0.216014 	 0.024001	+ 	0.66666666666666666667
+n=2^15 k=2^11 	 0.220014 	 0.064004	+ 	0.73333333333333333333
+n=2^15 k=2^12 	 0.228014 	 0.212014	+ 	0.8
+n=2^15 k=2^13 	 0.216013 	 0.664042	- 	0.86666666666666666667
+n=2^15 k=2^14 	 0.240015 	 1.868117	- 	0.93333333333333333333
+;  benchmark_binomial(11,15)
+n=2^15 k=2^11 	 0.216014 	 0.068004	+ 	0.73333333333333333333
+n=2^15 k=2^12 	 0.236015 	 0.212013	+ 	0.8
+n=2^15 k=2^13 	 0.216013 	 0.656041	- 	0.86666666666666666667
+n=2^15 k=2^14 	 0.244016 	 1.872117	- 	0.93333333333333333333
+;  benchmark_binomial(11,18)
+n=2^18 k=2^11 	 1.652103 	 0.100006	+ 	0.61111111111111111111
+n=2^18 k=2^12 	 1.608101 	 0.336021	+ 	0.66666666666666666667
+n=2^18 k=2^13 	 1.700106 	 1.140071	+ 	0.72222222222222222222
+n=2^18 k=2^14 	 1.756109 	 3.924245	- 	0.77777777777777777778
+n=2^18 k=2^15 	 2.036127 	 13.156822	- 	0.83333333333333333333
+n=2^18 k=2^16 	 2.172135 	 41.974624	- 	0.88888888888888888889
+n=2^18 k=2^17 	 2.528158 	 121.523594	- 	0.94444444444444444444
+;  benchmark_binomial(15,25)
+n=2^25 k=2^15 	 303.790985 	 38.266392	+ 	0.6
+;  benchmark_binomial(17,25)
+n=2^25 k=2^17 	 319.127944 	 529.025062	- 	0.68
+*/
+
+define benchmark_binomial(s,limit){
+  local ret k A B T1 T2 start end N K;
+    N = 2^(limit);
+  for(k=s;k<limit;k++){
+    K = 2^k;
+   start=usertime();A=binomial(N,K);end=usertime();
+   T1 = end-start;
+   start=usertime();B=comb(N,K);end=usertime();
+   T2 = end-start;
+   print "n=2^"limit,"k=2^"k,"	",T1,"	",T2,T1<T2?"-":"+","	"k/limit;
+   if(A!=B){
+     print "false";
+    break;
+   }
+  }
+}
+
+define __CZ__multiply_factored_factorial(matrix,stop){
+  local prime result shift prime_list k k1 k2 expo_list pix count start;
+  local hb flag;
+
+  result = 1;
+  shift  = 0;
+
+
+  if(!ismat(matrix))
+   return newerror("__CZ__multiply_factored_factorial(matrix): "
+   		   "argument matrix not a matrix ");
+  if(!matrix[0,0])
+    return
+      newerror("__CZ__multiply_factored_factorial(matrix): "
+      	       "matrix[0,0] is null/0");
+
+  if(!isnull(stop))
+    pix = stop;
+  else
+    pix   = size(matrix)/2-1;
+
+  if(matrix[0,0] == 2 && matrix[0,1] > 0){
+    shift = matrix[0,1];
+    if(pix-1 == 0)
+      return 2^matrix[0,1];
+  }
+
+  /*
+    This is a more general way to do the multiplication, so any optimization
+    must have been done by the caller.
+  */
+  k = 0;
+  /*
+     The size of the largest exponent in bits is calculated dynamically.
+     Can be done more elegantly and saves one run over the whole array if done
+     inside the main loop.
+   */
+  hb =0;
+  for(k=0;k<pix;k++){
+    k1=highbit(matrix[k,1]);
+    if(hb < k1)hb=k1;
+  }
+
+  k2 = pix;
+  start = 0;
+  if(shift) start++;
+
+  for(k1=hb;k1>=0;k1--){
+    /*
+       the cut-off for T-C-4 ist still too low, using T-C-3 here
+       TODO: check cutoffs
+     */
+    result = toomcook3square(result);
+
+    for(k=start; k<=k2; k++) {
+      if((matrix[k,1] & (1 << k1)) != 0) {
+        result *= matrix[k,0];
+      }
+    }
+  }
+
+  result <<= shift;
+  return result;
+}
+
+/*
+    Compute binomial coefficients n!/(k!(n-k)!)
+
+    One of the rare cases where a formula once meant to ease manual computation
+    is actually the (asymptotically) fastest way to do it (in July 2013) for
+    the extreme case binomial(2N,N) but for a high price, the memory
+    needed is pi(N)--theoretically.
+*/
+define binomial(n,k){
+  local ret factored_n factored_k factored_nk denom num quot K prime_list prime;
+  local pix diff;
+
+  if(!isint(n) || !isint(k))
+    return newerror("binomial(n,k): input is not integer");
+  if(n<0 || k<0)
+    return newerror("binomial(n,k): input is not >= 0"); ;
+  if(n<k )   return 0;
+  if(n==k)   return 1;
+  if(k==0)   return 1;
+  if(k==1)   return n;
+  if(n-k==1) return n;
+  /*
+      cut-off depends on real size of n,k and size of n/k
+      The current cut-off is to small for large n, e.g.:
+      for 2n=2^23, k=n-n/2 the quotient is q=2n/k=0.25. Empirical tests showed
+      that 2n=2^23 and k=2^16  with q=0.0078125 are still faster than the
+      builtin function.
+
+      The symmetry (n,k) = (n,n-k) is of not much advantage here. One way
+      might be to get closer to k=n/2 if k<n-k but only if the difference
+      is small and n very large.
+   */
+  if(n<2e4 && !isdefined("test8900"))  return comb(n,k);
+  if(n<2e4 && k< n-n/2 && !isdefined("test8900")) return comb(n,k);
+  /*
+    This should be done in parallel to save some memory, e.g. no temporary
+    arrays are needed, all can be done inline.
+    The theoretical memory needed is pi(k).
+    Which is still a lot.
+  */
+
+  prime = 2;
+  pix   = pix(n);
+  prime_list = mat[pix , 2];
+  K = 0;
+  do {
+    prime_list[K  ,0] = prime;
+    diff = __CZ__prime_divisors(n,prime)-
+	   ( __CZ__prime_divisors(n-k,prime)+__CZ__prime_divisors(k,prime));
+    if(diff != 0)
+      prime_list[K++,1] = diff;
+    prime          = nextprime(prime);
+  }while(prime <= k);
+
+  do {
+    prime_list[K  ,0] = prime;
+    diff = __CZ__prime_divisors(n,prime)-__CZ__prime_divisors(n-k,prime);
+    if(diff != 0)
+      prime_list[K++,1] = diff;
+    prime          = nextprime(prime);
+  }while(prime <= n-k);
+
+  do {
+    prime_list[K  ,0] = prime;
+    prime_list[K++,1] = __CZ__prime_divisors(n,prime);
+    prime          = nextprime(prime);
+  }while(prime <= n);
+  ##print K,pix(k),pix(n-k),pix(n);
+  ##factored_k  = __CZ__factor_factorial(k,1);
+  ##factored_nk = __CZ__factor_factorial(n-k,1);
+
+  ##denom = __CZ__add_factored_factorials(factored_k,factored_nk);
+    ##free(factored_k,factored_nk);
+  ##num  = __CZ__factor_factorial(n,1);
+  ##quot  = __CZ__subtract_factored_factorials( num , denom );
+    ##free(num,denom);
+
+  ret = __CZ__multiply_factored_factorial(`prime_list,K-1);
+
+  return ret;
+}
+
+/*
+    Compute large catalan numbers  C(n) = binomial(2n,n)/(n+1) with
+    cut-off: (n>5e4)
+    Needs a lot of memory.
+*/
+define bigcatalan(n){
+  if(!isint(n) )return newerror("bigcatalan(n): n is not integer");
+  if( n<0) return newerror("bigcatalan(n): n < 0");
+  if( n<5e4 && !isdefined("test8900") ) return catalan(n);
+  return binomial(2*n,n)/(n+1);
+}
+
+/*
+  df(-111) = -1/3472059605858239446587523014902616804783337112829102414124928
+  		7753332469144201839599609375
+
+ df(-3+1i) = 0.12532538977287649201-0.0502372106177184607i
+ df(2n + 1) = (2*n)!/(n!*2^n)
+*/
+define __CZ__double_factorial(n){
+  local n1 n2 diff prime pix K prime_list k;
+  prime = 3;
+  pix   = pix(2*n)+1;
+  prime_list = mat[pix , 2];
+  K = 0;
+  do {
+    prime_list[K  ,0] = prime;
+    diff = __CZ__prime_divisors(2*n,prime)-( __CZ__prime_divisors(n,prime));
+    if(diff != 0)
+      prime_list[K++,1] = diff;
+    prime          = nextprime(prime);
+  }while(prime <= n);
+  do {
+    prime_list[K  ,0] = prime;
+    prime_list[K++,1] = __CZ__prime_divisors(2*n,prime);
+    prime          = nextprime(prime);
+  }while(prime <= 2*n);
+  return __CZ__multiply_factored_factorial(prime_list,K);
+/*
+        n1=__CZ__factor_factorial(2*n,1);
+        n1[0,1] = n1[0,1]-n;
+        n2=__CZ__factor_factorial(n,1);
+        diff=__CZ__subtract_factored_factorials( n1 , n2 );
+        return __CZ__multiply_factored_factorial(diff);
+*/
+}
+
+##1, 1, 3, 15, 105, 945, 10395, 135135, 2027025, 34459425, 654729075,
+##13749310575, 316234143225, 7905853580625, 213458046676875,
+##6190283353629375, 191898783962510625, 6332659870762850625,
+##221643095476699771875, 8200794532637891559375
+
+## 1, 2, 8, 48, 384, 3840, 46080, 645120, 10321920, 185794560,
+##3715891200, 81749606400, 1961990553600, 51011754393600,
+##1428329123020800, 42849873690624000, 1371195958099968000,
+##46620662575398912000, 1678343852714360832000, 63777066403145711616000
+define doublefactorial(n){
+    local n1 n2 diff eps ret;
+    if(!isint(n) ){
+       /*
+           Probably one of the not-so-good ideas. See result of
+            http://www.wolframalpha.com/input/?i=doublefactorial%28a%2Bbi%29
+        */
+       eps=epsilon(epsilon()*1e-2);
+       ret =  2^(n/2-1/4 * cos(pi()* n)+1/4) * pi()^(1/4 *
+       	      cos(pi()* n)-1/4)* gamma(n/2+1);
+       epsilon(eps);
+       return ret;
+    }
+    if(n==2) return 2;
+    if(n==3) return 3;
+    switch(n){
+      case -1:
+      case 0 : return 1;break;
+      case 2 : return 2;break;
+      case 3 : return 3;break;
+      case 4 : return 8;break;
+      default: break;
+    }
+    if(isodd(n)){
+      /*
+        TODO: find reasonable cutoff
+        df(2n + 1) = (2*n)!/(n!*2^n)
+      */
+      if(n>0){
+         n = (n+1)//2;
+         return __CZ__double_factorial(n);
+      }
+      else{
+        if(n == -3 ) return -1;
+        n = ((-n)-1)/2;
+        return ((-1)^-n)/__CZ__double_factorial(n);
+       }
+    }
+    else{
+      /*
+         I'm undecided here. The formula for complex n is valid for the negative
+         integers, too.
+      */
+      n = n>>1;
+      if(n>0){
+        if(!isdefined("test8900"))
+          return factorial(n)<<n;
+        else
+          return n!<<n;
+      }
+      else
+        return newerror("doublefactorial(n): even(n) < 0");
+   }
+}
+
+/*
+    Algorithm 3.17,
+    Donald Kreher and Douglas Simpson,
+    Combinatorial Algorithms,
+    CRC Press, 1998, page 89.
+*/
+static __CZ__stirling1;
+static __CZ__stirling1_n = -1;
+static __CZ__stirling1_m = -1;
+
+define stirling1(n,m){
+  local i j k;
+  if(n<0)return newerror("stirling1(n,m): n <= 0");
+  if(m<0)return newerror("stirling1(n,m): m < 0");
+  if(n<m) return 0;
+  if(n==m) return 1;
+  if(m==0 || n==0) return 0;
+  /* We always use the list */
+  /*
+  if(m=1){
+    if(iseven(n)) return -factorial(n-1);
+    else return factorial(n-1);
+  }
+  if(m == n-1){
+    if(iseven(n)) return -binomial(n,2);
+    else return -binomial(n,2);
+  }
+  */
+  if(__CZ__stirling1_n >= n && __CZ__stirling1_m  >= m){
+    return __CZ__stirling1[n,m];
+  }
+  else{
+    __CZ__stirling1      = mat[n+1,m+1];
+    __CZ__stirling1[0,0] = 1;
+    for(i=1;i<=n;i++)
+      __CZ__stirling1[i,0] = 0;
+    for(i=1;i<=n;i++){
+      for(j=1;j<=m;j++){
+        if(j<=i){
+          __CZ__stirling1[i, j] =   __CZ__stirling1[i - 1, j - 1] - (i - 1)\
+                                  * __CZ__stirling1[i - 1, j];
+        }
+        else{
+         __CZ__stirling1[i, j] = 0;
+        }
+      }
+    }
+    __CZ__stirling1_n = n;
+    __CZ__stirling1_m = m;
+    return __CZ__stirling1[n,m];
+  }
+}
+
+define stirling2(n,m){
+  local k sum;
+  if(n<0)return newerror("stirling2(n,m): n < 0");
+  if(m<0)return newerror("stirling2(n,m): m < 0");
+  if(n<m) return 0;
+  if(n==0 && n!=m) return 0;
+  if(n==m) return 1;
+  if(m==0 )return 0;
+  if(m==1) return 1;
+  if(m==2) return 2^(n-1)-1;
+  /*
+    There are different methods to speed up alternating sums.
+    This one doesn't.
+   */
+  if(isdefined("test8900")){
+    for(k=0;k<=m;k++){
+      sum += (-1)^(m-k)*comb(m,k)*k^n;
+    }
+  return sum/(m!);
+  }
+  else{
+    for(k=0;k<=m;k++){
+      sum += (-1)^(m-k)*binomial(m,k)*k^n;
+    }
+  return sum/factorial(m);
+  }
+}
+
+static __CZ__stirling2;
+static __CZ__stirling2_n = -1;
+static __CZ__stirling2_m = -1;
+define stirling2caching(n,m){
+  local nm i j ;
+  if(n<0)return newerror("stirling2iter(n,m): n < 0");
+  if(m<0)return newerror("stirling2iter(n,m): m < 0");
+  /* no shortcuts here */
+
+  if(n<m) return 0;
+  if(n==0 && n!=m) return 0;
+  if(n==m) return 1;
+  if(m==0 )return 0;
+  if(m==1) return 1;
+  if(m==2) return 2^(n-1)-1;
+
+  nm = n-m;
+  if(__CZ__stirling2_n >= n && __CZ__stirling2_m >= m){
+    return __CZ__stirling2[n,m];
+  }
+  else{
+    __CZ__stirling2 = mat[n+1,m+1];
+    __CZ__stirling2[0,0] = 1;
+    for(i=1;i<=n;i++){
+      __CZ__stirling2[i,0] = 0;
+      for(j=1;j<=m;j++){
+        if(j<=i){
+          __CZ__stirling2[i, j] =   __CZ__stirling2[i -1, j -1] + (j )\
+                                  * __CZ__stirling2[i - 1, j];
+        }
+        else{
+         __CZ__stirling2[i, j] = 0;
+        }
+      }
+    }
+  }
+  __CZ__stirling2_n = (n);
+  __CZ__stirling2_m = (m);
+  return __CZ__stirling2[n,m];
+}
+
+define bell(n){
+  local sum s2list k A;
+
+  if(!isint(n)) return newerror("bell(n): n is not integer");
+  if(n < 0) return newerror("bell(n): n is not positive");
+  /* place some more shortcuts here?*/
+  if(n<=15){
+    mat A[16] = {
+                  1, 1, 2, 5, 15, 52, 203, 877, 4140, 21147, 115975, 678570,
+                  4213597, 27644437, 190899322, 1382958545
+                };
+    return A[n];
+  }
+  /* Start by generating the list of stirling numbers of the second kind */
+  s2list = stirling2caching(n,n//2);
+  if(iserror(s2list))
+    return newerror("bell(n): could not build stirling num. list");
+  sum = 0;
+  for(k=1;k<=n;k++){
+      sum += stirling2caching(n,k);
+  }
+  return sum;
+}
+
+define subfactorialrecursive(n){
+  if(n==0) return 1;
+  if(n==1) return 0;
+  if(n==2) return 1;
+  return n * subfactorialrecursive(n-1) + (-1)^n;
+}
+
+/* This is, quite amusingly, faster than the very same algorithm in
+   PARI/GP + GMP*/
+define subfactorialiterative(n){
+  local k temp1 temp2 ret;
+  if(n==0) return 1;
+  if(n==1) return 0;
+  if(n==2) return 1;
+  temp1 = 0;
+  ret   = 1;
+  for(k=3;k<=n;k++){
+    temp2 = temp1;
+    temp1 = ret;
+    ret =  (k-1) *(temp1 + temp2);
+  }
+  return ret;
+}
+
+define subfactorial(n){
+  local epsilon eps ret lnfact;
+  if(!isint(n))return  newerror("subfactorial(n): n is not integer.");
+  if(n < 0)return  newerror("subfactorial(n): n < 0");
+  return subfactorialiterative(n);
+}
+
+define risingfactorial(x,n){
+  local num denom quot ret;
+  if(n == 1) return x;
+  if(x==0) return newerror("risingfactorial(x,n): x == 0");
+  if(!isint(x) || !isint(n)){
+    return gamma(x+n)/gamma(x);
+  }
+  if(x<1)return newerror("risingfactorial(x,n): integer x and x < 1");
+  if(x+n < 1)return newerror("risingfactorial(x,n): integer x+n and x+n < 1");
+  if(x<9000&&n<9000){
+    return (x+n-1)!/(x-1)!;
+  }
+  else{
+    num   = __CZ__factor_factorial(x+n-1,1);
+    denom = __CZ__factor_factorial(x-1,1);
+    quot  = __CZ__subtract_factored_factorials( num , denom );
+      free(num,denom);
+    ret = __CZ__multiply_factored_factorial(quot);
+    return ret;
+  }
+}
+
+define fallingfactorial(x,n){
+  local num denom quot ret;
+  if(n == 0) return 1;
+
+  if(!isint(x) || !isint(n)){
+    if(x == n) return gamma(x+1);
+    return gamma(x+1)/gamma(x-n+1);
+  }
+  else{
+    if(x<0 || x-n < 0)
+     return newerror("fallingfactorial(x,n): integer x<0 or x-n < 0");
+    if(x == n) return factorial(x);
+    if(x<9000&&n<9000){
+      return (x)!/(x-n)!;
+    }
+    else{
+      num   = __CZ__factor_factorial(x,1);
+      denom = __CZ__factor_factorial(x-n,1);
+      quot  = __CZ__subtract_factored_factorials( num , denom );
+        free(num,denom);
+      ret = __CZ__multiply_factored_factorial(quot);
+      return ret;
+    }
+  }
+}
+
+
+/*
+ * restore internal function from resource debugging
+ * report important interface functions
+ */
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "binomial(n,k)";
+    print "bigcatalan(n)";
+    print "doublefactorial(n)";
+    print "subfactorial(n)";
+    print "stirling1(n,m)";
+    print "stirling2(n,m)";
+    print "stirling2caching(n,m)";
+    print "bell(n)";
+    print "subfactorial(n)";
+    print "risingfactorial(x,n)";
+    print "fallingfactorial(x,n)";
+}

cal/gvec.cal

@@ -0,0 +1,103 @@
+/*
+ * gvec - vectorize any single-input function or trailing operator
+ *
+ * This version accepts arbitrary number of arguments, but of course
+ * they must all be same length vectors.
+ *
+ * The gvec function is for use in either a two-arg function or a two-arg
+ * operation "function" must be first; calc doesn't care how many more
+ * arguments there actually are.
+ *
+ * Under source code control:   2011/03/31 17:54:55
+ * File existed as early as:    2010
+ *
+ * By Carl Witthoft carl at witthoft dot com
+ */
+
+define gvec(function, vector)
+{
+    local xlen,y,foo;
+    local precx = 1e-50;	/* default for now */
+    local argc = param(0)-1;
+    local old_tilde;		/* previous config("tilde") */
+
+    /*
+     * parse args
+     */
+    local plist = mat[argc];
+    if (config("resource_debug") & 8) {
+	print "plist=", plist;
+	print "argc=", argc;
+    }
+    for(local i = 0; i< argc; i++) {
+	local ii = i + 2;
+	if (config("resource_debug") & 8) {
+	    print "ii=", ii;
+	    print "param(" : ii : "}=", param(ii);
+	    print "size(param(" : ii : ")=", size(param(ii));
+	}
+	plist[i] = size(param(ii));
+    }
+    local slist=sort(plist);
+    if (config("resource_debug") & 8) {
+	print "plist=", plist;
+    }
+    local argm = argc-1;
+    if (config("resource_debug") & 8) {
+	print "argm=", argm;
+    }
+    if (slist[0] != slist[argm]) {
+	quit "lengths don't match";
+    }
+    xlen = size(vector);
+    y = mat[xlen];
+
+    /*
+     * We can't do str(vector[j]) outside loop, eval() petulantly refuses to
+     * look at local variables.
+     *
+     * Also we need to config("tilde",0) to turn off lead tilde
+     * (so str(vector[j]) looks like a number.
+     */
+    old_tilde = config("tilde",0);
+
+    /*
+     * Ok, now check to see if "function" is a function.  If not, it's an
+     * operation and it's up to user to make it valid
+     */
+    if (isdefined(function)) {
+
+	/* yep, it's a function, either builtin or user-defined */
+	for (local j=0; j<xlen; j++) {
+
+	    /* build the function call */
+	    foo = strcat(function, "(");
+	    for (local jj = 0; jj<argc; jj++) {
+		foo = strcat(foo , str(param(jj+2)[j]), ",");
+	    }
+	    foo = strcat(foo, str(precx), ")");
+	    if (config("resource_debug") & 8) {
+		print "foo=", foo;
+	    }
+	    y[j] = eval(foo);
+	}
+
+    /*
+     * it is an operator --  multi-argument operator makes no sense
+     */
+    } else {
+	if (argc > 1) {
+	    quit "Error: operator can accept only one argument";
+	}
+	for (j=0; j<xlen; j++) {
+	    foo = strcat(str(vector[j]), function);
+	    y[j] = eval(foo);
+	}
+    }
+
+    /* restore tilde mode if needed */
+    config("tilde", old_tilde);
+
+    /* return result */
+    return y;
+}

cal/hello.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: hello.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/hello.cal,v $
- *
  * Under source code control:	1996/11/13 13:25:43
  * File existed as early as:	1996
  *

cal/hms.cal

@@ -0,0 +1,364 @@
+/*
+ * hms - calculate in hours, minutes, and seconds
+ *
+ * Copyright (C) 2010,2021  Landon Curt Noll
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.	 See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2010/09/01 17:14:55
+ * File existed as early as:	2010
+ *
+ * Share and enjoy!  :-)	http://www.isthe.com/chongo/tech/comp/calc/
+ */
+
+
+obj hms {hour, min, sec};
+
+define hms(hour, min, sec)
+{
+    local obj hms ans;		/* return value */
+
+    /* default missing args to 0 */
+    if (isnull(sec)) {
+	sec = 0;
+    }
+    if (isnull(min)) {
+	min = 0;
+    }
+
+    /* load object */
+    ans.hour = hour;
+    ans.min = min;
+    ans.sec = sec;
+
+    /* return properly formed object */
+    ans = fixhms(ans);
+    return ans;
+}
+
+
+define hms_add(a, b)
+{
+    local obj hms ans;		/* return value */
+
+    /* initialize value to 1st arg */
+    if (istype(a, ans)) {
+	/* 1st arg is hms object, load it */
+	ans.hour = a.hour;
+	ans.min = a.min;
+	ans.sec = a.sec;
+    } else {
+	/* 1st arg is not hms, assume scalar hours */
+	ans.hour = a;
+	ans.min = 0;
+	ans.sec = 0;
+    }
+
+    /* add value of 2nd arg */
+    if (istype(b, ans)) {
+	/* 2nd arg is hms object, add it */
+	ans.hour += b.hour;
+	ans.min += b.min;
+	ans.sec += b.sec;
+    } else {
+	/* 2nd arg is not hms, add scalar hours */
+	ans.hour += b;
+    }
+
+    /* return normalized result */
+    ans = fixhms(ans);
+    return ans;
+}
+
+
+define hms_neg(a)
+{
+    local obj hms ans;		/* return value */
+
+    /* negate argument */
+    if (istype(a, ans)) {
+	/* 1st arg is hms object, load it */
+	ans.hour = -a.hour;
+	ans.min = -a.min;
+	ans.sec = -a.sec;
+    } else {
+	/* 2nd arg is not hms, negate scalar hours */
+	ans.hour = -a;
+	ans.min = 0;
+	ans.sec = 0;
+    }
+
+    /* return normalized result */
+    ans = fixhms(ans);
+    return ans;
+}
+
+
+define hms_sub(a, b)
+{
+    local obj hms ans;		/* return value */
+
+    /* initialize value to 1st arg */
+    if (istype(a, ans)) {
+	/* 1st arg is hms object, load it */
+	ans.hour = a.hour;
+	ans.min = a.min;
+	ans.sec = a.sec;
+    } else {
+	/* 1st arg is not hms, assume scalar hours */
+	ans.hour = a;
+	ans.min = 0;
+	ans.sec = 0;
+    }
+
+    /* subtract value of 2nd arg */
+    if (istype(b, ans)) {
+	/* 2nd arg is hms object, subtract it */
+	ans.hour -= b.hour;
+	ans.min -= b.min;
+	ans.sec -= b.sec;
+    } else {
+	/* 2nd arg is not hms, subtract scalar hours */
+	ans.hour -= b;
+    }
+
+    /* return normalized result */
+    ans = fixhms(ans);
+    return ans;
+}
+
+
+define hms_mul(a, b)
+{
+    local obj hms ans;		/* return value */
+
+    /* hms object multiplication */
+    if (istype(a, ans) && istype(b, ans)) {
+	ans.hour = hms_abs(a) * hms_abs(b);
+	ans.min = 0;
+	ans.sec = 0;
+
+    /* scalar multiplication */
+    } else if (istype(a, ans)) {
+	ans.hour = a.hour * b;
+	ans.min = a.min * b;
+	ans.sec = a.sec * b;
+    } else {
+	ans.hour = b.hour * a;
+	ans.min = b.min * a;
+	ans.sec = b.sec * a;
+    }
+
+    /* return normalized result */
+    ans = fixhms(ans);
+    return ans;
+}
+
+
+define hms_print(a)
+{
+    local obj hms ans;		/* temp object for hms type testing */
+
+    /* firewall - arg must be a hms object */
+    if (! istype(a, ans)) {
+	quit "hms_print called with non hms object";
+    }
+
+    /* print in hms form */
+    print a.hour : ':' : a.min : ':' : a.sec :;
+}
+
+
+define hms_abs(a)
+{
+    local obj hms ans;		/* temp object for hms type testing */
+    local hour;			/* return scalar value */
+
+    /* firewall - just absolute value non hms objects */
+    if (! istype(a, ans)) {
+    	return abs(a);
+    }
+
+    /* compute hours */
+    hour = a.hour + a.min / 60 + a.sec / 3600;
+
+    /* return hours */
+    return hour;
+}
+
+
+define hms_norm(a)
+{
+    local obj hms ans;		/* temp object for hms type testing */
+    local hour;			/* hours */
+
+    /* firewall - arg must be a hms object */
+    if (! istype(a, ans)) {
+	quit "hms_norm called with non hms object";
+    }
+
+    /* square hours (norm is the square of absolute value */
+    hour = hms_abs(a);
+
+    /* return hours */
+    return hour*hour;
+}
+
+
+define hms_test(a)
+{
+    local obj hms ans;		/* temp value */
+
+    /* firewall - arg must be a hms object */
+    if (! istype(a, ans)) {
+	quit "hms_test called with non hms object";
+    }
+
+    /* return false of non-zero */
+    ans = fixhms(a);
+    if (ans.hour == 0 && ans.min == 0 && ans.sec == 0) {
+	/* false */
+	return 0;
+    }
+    /* true */
+    return 1;
+}
+
+
+define hms_int(a)
+{
+    local obj hms ans;		/* return value */
+
+    /* firewall - arg must be a hms object */
+    if (! istype(a, ans)) {
+	quit "hms_int called with non hms object";
+    }
+
+    /* normalize the argument */
+    ans = fixhms(a);
+
+    /* truncate to the nearest second */
+    ans.sec = int(ans.sec);
+
+    /* return value to the nearest second */
+    return ans;
+}
+
+
+define hms_frac(a)
+{
+    local obj hms ans;		/* return value */
+
+    /* firewall - arg must be a hms object */
+    if (! istype(a, ans)) {
+	quit "hms_frac called with non hms object";
+    }
+
+    /* normalize the argument */
+    ans = fixhms(a);
+
+    /* remove all but fractional seconds */
+    ans.hour = 0;
+    ans.min = 0;
+    ans.sec = frac(ans.sec);
+
+    /* return value to the second fraction */
+    return ans;
+}
+
+
+define hms_rel(a,b)
+{
+    local abs_a, abs_b;		/* scalars of the arguments */
+
+    /* compute scalars of the arguments */
+    abs_a = hms_abs(a);
+    abs_b = hms_abs(b);
+
+    /* return the comparison */
+    return cmp(abs_a, abs_b);
+}
+
+
+define hms_cmp(a,b)
+{
+    local abs_a, abs_b;		/* scalars of the arguments */
+
+    /* compute scalars of the arguments */
+    abs_a = hms_abs(a);
+    abs_b = hms_abs(b);
+
+    /* return the equality comparison */
+    return (abs_a == abs_b);
+}
+
+
+define hms_inc(a)
+{
+    local obj hms ans;		/* return value */
+
+    /* increment a hms object */
+    if (istype(a, ans)) {
+	ans = a;
+	++ans.sec;
+
+	/* return normalized result */
+	ans = fixhms(ans);
+	return ans;
+    }
+
+    /* increment a scalar */
+    return a+1;
+}
+
+
+define hms_dec(a)
+{
+    local obj hms ans;		/* return value */
+
+    /* decrement a hms object */
+    if (istype(a, ans)) {
+	ans = a;
+	--ans.sec;
+
+	/* return normalized result */
+	ans = fixhms(ans);
+	return ans;
+    }
+
+    /* decrement a scalar */
+    return a-1;
+}
+
+
+define fixhms(a)
+{
+    local obj hms ans;		/* temp value */
+
+    /* firewall */
+    if (! istype(a, ans)) {
+	quit "attempt to fix a non hms object";
+    }
+
+    /* use builtin h2hms function */
+    h2hms(a.hour + a.min/60 + a.sec/3600, a.hour, a.min, a.sec),;
+
+    /* return normalized result */
+    return a;
+}
+
+if (config("resource_debug") & 3) {
+    print "obj hms {hour, min, sec} defined";
+}

cal/infinities.cal

@@ -0,0 +1,88 @@
+/*
+ * infinities - handle infinities symbolically, a little helper file
+ *
+ * Copyright (C) 2013 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ */
+
+
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+define isinfinite(x)
+{
+    if (isstr(x)) {
+	if (strncmp(x, "cinf", 4) == 0
+	    || strncmp(x, "pinf", 4) == 0 || strncmp(x, "ninf", 4) == 0)
+	    return 1;
+    }
+    return 0;
+}
+
+define iscinf(x)
+{
+    if (isstr(x)) {
+	if (strncmp(x, "cinf", 4) == 0)
+	    return 1;
+    }
+    return 0;
+}
+
+define ispinf(x)
+{
+    if (isstr(x)) {
+	if (strncmp(x, "pinf", 4) == 0)
+	    return 1;
+    }
+    return 0;
+}
+
+define isninf(x)
+{
+    if (isstr(x)) {
+	if (strncmp(x, "ninf", 4) == 0)
+	    return 1;
+    }
+    return 0;
+}
+
+define cinf()
+{
+    return "cinf";
+}
+
+define ninf()
+{
+    return "ninf";
+}
+
+define pinf()
+{
+    return "pinf";
+}
+
+
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "isinfinite(x)";
+    print "iscinf(x)";
+    print "ispinf(x)";
+    print "isninf(x)";
+    print "cinf()";
+    print "ninf()";
+    print "pinf()";
+}

cal/intfile.cal

@@ -1,7 +1,7 @@
 /*
  * intfile - integer to file and file to integer conversion
  *
- * Copyright (C) 2001  Landon Curt Noll
+ * Copyright (C) 2001,2021  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: intfile.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/intfile.cal,v $
- *
  * Under source code control:   2001/03/31 08:13:11
  * File existed as early as:    2001
  *
@@ -31,20 +27,20 @@
 
 /*
  * NOTE: Because leading HALF values are trimmed from integer, a file
- *	 that begins with lots of 0 bits (in the case of big endian)
- *	 or that ends with lots of 0 bits (in the case of little endian)
+ *	 that begins with lots of 0 bits (in the case of big Endian)
+ *	 or that ends with lots of 0 bits (in the case of little Endian)
  *	 will be changed when the subsequent integer is written back.
  */
 
 
 /*
- * file2be - convert a file into an big endian integer
+ * file2be - convert a file into an big Endian integer
  *
  * given:
  *	filename	filename to read
  *
  * returns:
- *	integer read from its contents on big endian order
+ *	integer read from its contents on big Endian order
  */
 define file2be(filename)
 {
@@ -79,13 +75,13 @@ define file2be(filename)
 
 
 /*
- * file2le - convert a file into an little endian integer
+ * file2le - convert a file into an little Endian integer
  *
  * given:
  *	filename	filename to read
  *
  * returns:
- *	integer read from its contents on little endian order
+ *	integer read from its contents on little Endian order
  */
 define file2le(filename)
 {
@@ -122,7 +118,7 @@ define file2le(filename)
 
 
 /*
- * be2file - convert a big endian integer into a file
+ * be2file - convert a big Endian integer into a file
  *
  * given:
  *	v		integer to write to the file
@@ -172,7 +168,7 @@ define be2file(v, filename)
 
 
 /*
- * le2file - convert a little endian integer into a file
+ * le2file - convert a little Endian integer into a file
  *
  * given:
  *	v		integer to write to the file

cal/intnum.cal

@@ -0,0 +1,728 @@
+/*
+ * intnum - implementation of tanhsinh- and Gauss-Legendre quadrature
+ *
+ * Copyright (C) 2013,2021 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ */
+
+
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+read -once infinities;
+
+static __CZ__tanhsinh_x;
+static __CZ__tanhsinh_w;
+static __CZ__tanhsinh_order;
+static __CZ__tanhsinh_prec;
+
+define quadtsdeletenodes()
+{
+    free(__CZ__tanhsinh_x);
+    free(__CZ__tanhsinh_w);
+    free(__CZ__tanhsinh_order);
+    free(__CZ__tanhsinh_prec);
+}
+
+define quadtscomputenodes(order, expo, eps)
+{
+    local t cht sht chp sum k PI places;
+    local h t0 x w;
+    if (__CZ__tanhsinh_order == order && __CZ__tanhsinh_prec == eps)
+	return 1;
+    __CZ__tanhsinh_order = order;
+    __CZ__tanhsinh_prec = eps;
+    __CZ__tanhsinh_x = list();
+    __CZ__tanhsinh_w = list();
+    /* The tanhsinh algorithm needs a slightly higher precision than G-L */
+    eps = epsilon(eps * 1e-2);
+    places = highbit(1 + int (1 / epsilon())) +1;
+    PI = pi();
+    sum = 0;
+    t0 = 2 ^ (-expo);
+    h = 2 * t0;
+    /*
+     * The author wanted to use the mpmath trick here which was
+     * advertised---and reasonably so!---to be faster. Didn't work out
+     * so well with calc.
+     * PI4 = PI/4;
+     * expt0 = bround(exp(t0),places);
+     * a = bround( PI4 * expt0,places);
+     * b = bround(PI4 / expt0,places);
+     * udelta = bround(exp(h),places);
+     * urdelta = bround(1/udelta,places);
+     */
+    /* make use of x(-t) = -x(t), w(-t) = w(t)  */
+    for (k = 0; k < 20 * order + 1; k++) {
+	/*
+	 * x = tanh(pi/2 * sinh(t))
+	 * w = pi/2 * cosh(t) / cosh(pi/2 * sinh(t))^2
+	 */
+	t = bround(t0 + k * h, places);
+
+	cht = bround(cosh(t), places);
+	sht = bround(sinh(t), places);
+	chp = bround(cosh(0.5 * PI * sht), places);
+	x = bround(tanh(0.5 * PI * sht), places);
+	w = bround((PI * h * cht) / (2 * chp ^ 2), places);
+	/*
+	 * c = bround(exp(a-b),places);
+	 * d = bround(1/c,places);
+	 * co =bround( (c+d)/2,places);
+	 * si =bround( (c-d)/2,places);
+	 * x = bround(si / co,places);
+	 * w = bround((a+b) / co^2,places);
+	 */
+	if (abs(x - 1) <= eps)
+	    break;
+
+	append(__CZ__tanhsinh_x, x);
+	append(__CZ__tanhsinh_w, w);
+	/*
+	 * a *= udelta;
+	 * b *= urdelta;
+	 */
+    }
+
+    /* Normalize the weights to make them add up to 2 (two) */
+    /*
+     * for(k=0;k < size(__CZ__tanhsinh_w);k++)
+     * sum = bround(sum + __CZ__tanhsinh_w[k],places);
+     * sum *= 2;
+     * for(k=0;k < size(__CZ__tanhsinh_w);k++)
+     * __CZ__tanhsinh_w[k] = bround(2.0 * __CZ__tanhsinh_w[k] / sum,places);
+     */
+
+    epsilon(eps);
+    return 1;
+}
+
+define quadtscore(a, b, n)
+{
+    local k c d order eps places sum ret x x1 x2 xm w w1 w2 m sizel;
+
+    eps = epsilon(epsilon() * 1e-2);
+    places = highbit(1 + int (1 / epsilon())) +1;
+    m = int (4 + max(0, ln(places / 30.0) / ln(2))) + 2;
+    if (!isnull(n)) {
+	order = n;
+	m = ilog(order / 3, 2) + 1;
+    } else
+	order = 3 * 2 ^ (m - 1);
+
+    quadtscomputenodes(order, m, epsilon());
+    sizel = size(__CZ__tanhsinh_w);
+
+    if (isinfinite(a) || isinfinite(b)) {
+	/*
+	 *           x
+	 * t =  ------------
+	 *                2
+	 *      sqrt(1 - y )
+	 */
+	if (isninf(a) && ispinf(b)) {
+	    for (k = 0; k < sizel; k++) {
+		x1 = __CZ__tanhsinh_x[k];
+		x2 = -__CZ__tanhsinh_x[k];
+		w1 = __CZ__tanhsinh_w[k];
+
+		x = bround(x1 * (1 - x1 ^ 2) ^ (-1 / 2), places);
+		xm = bround(x2 * (1 - x2 ^ 2) ^ (-1 / 2), places);
+		w = bround(w1 * (((1 - x1 ^ 2) ^ (-1 / 2)) / (1 - x1 ^ 2)),
+			   places);
+		w2 = bround(w1 * (((1 - x2 ^ 2) ^ (-1 / 2)) / (1 - x2 ^ 2)),
+			    places);
+		sum += bround(w * f(x), places);
+		sum += bround(w2 * f(xm), places);
+	    }
+	}
+	/*
+	 *        1
+	 * t =  - - + b + 1
+	 *        x
+	 */
+	else if (isninf(a) && !iscinf(b)) {
+	    for (k = 0; k < sizel; k++) {
+		x1 = __CZ__tanhsinh_x[k];
+		x2 = -__CZ__tanhsinh_x[k];
+		w1 = __CZ__tanhsinh_w[k];
+
+		x = bround((b + 1) - (2 / (x1 + 1)), places);
+		xm = bround((b + 1) - (2 / (x2 + 1)), places);
+		w = bround(w1 * (1 / 2 * (2 / (x1 + 1)) ^ 2), places);
+		w2 = bround(w1 * (1 / 2 * (2 / (x2 + 1)) ^ 2), places);
+		sum += bround(w * f(x), places);
+		sum += bround(w2 * f(xm), places);
+	    }
+	}
+	/*
+	 *      1
+	 * t =  - + a - 1
+	 *      x
+	 */
+	else if (!iscinf(a) && ispinf(b)) {
+	    for (k = 0; k < sizel; k++) {
+		x1 = __CZ__tanhsinh_x[k];
+		x2 = -__CZ__tanhsinh_x[k];
+		w1 = __CZ__tanhsinh_w[k];
+		x = bround((a - 1) + (2 / (x1 + 1)), places);
+		xm = bround((a - 1) + (2 / (x2 + 1)), places);
+		w = bround(w1 * (((1 / 2) * (2 / (x1 + 1)) ^ 2)), places);
+		w2 = bround(w1 * (((1 / 2) * (2 / (x2 + 1)) ^ 2)), places);
+		sum += bround(w * f(x), places);
+		sum += bround(w2 * f(xm), places);
+	    }
+	} else if (isninf(a) || isninf(b)) {
+	    /*TODO: swap(a,b) and negate(w)? Lookup! */
+	    return newerror("quadtscore: reverse limits?");
+	} else {
+	    return
+		newerror("quadtscore: complex infinity not yet implemented");
+        }
+	ret = sum;
+    } else {
+	/* Avoid rounding errors */
+	if (a == -1 && b == 1) {
+	    c = 1;
+	    d = 0;
+	} else {
+	    c = (b - a) / 2;
+	    d = (b + a) / 2;
+	}
+	sum = 0;
+	for (k = 0; k < sizel; k++) {
+	    sum +=
+		bround(__CZ__tanhsinh_w[k] * f(c * __CZ__tanhsinh_x[k] + d),
+		       places);
+	    sum +=
+		bround(__CZ__tanhsinh_w[k] * f(c * -__CZ__tanhsinh_x[k] + d),
+		       places);
+	}
+	ret = c * sum;
+    }
+    epsilon(eps);
+    return ret;
+}
+
+static __CZ__quadts_error;
+
+define quadts(a, b, points)
+{
+    local k sp results epsbits nsect interval length segment slope C ;
+    local x1 x2 y1 y2  sum D1 D2 D3 D4;
+    if (param(0) < 2)
+	return newerror("quadts: not enough arguments");
+    epsbits = highbit(1 + int (1 / epsilon())) +1;
+    if (param(0) < 3 || isnull(points)) {
+	/* return as given */
+	return quadtscore(a, b);
+    } else {
+	if ((isinfinite(a) || isinfinite(b))
+	    && (!ismat(points) && !islist(points)))
+	    return
+		newerror(strcat
+			 ("quadts: segments of infinite length ",
+			  "are not yet supported"));
+	if (ismat(points) || islist(points)) {
+	    sp = size(points);
+	    if (sp == 0)
+		return
+		    newerror(strcat
+			     ("quadts: variable 'points` must be a list or ",
+			      "1d-matrix of a length > 0"));
+	    /* check if all points are numbers */
+	    for (k = 0; k < sp; k++) {
+		if (!isnum(points[k]))
+		    return
+			newerror(strcat
+				 ("quadts: elements of 'points` must be",
+				  " numbers only"));
+	    }
+	    /* We have n-1 intervals and a and b, hence n-1 + 2 results */
+	    results = mat[sp + 1];
+            if (a != points[0]) {
+	        results[0] = quadtscore(a, points[0]);
+            } else {
+                results[0] = 0;
+            }
+	    if (sp == 1) {
+                if (b != points[0]) {
+		    results[1] = quadtscore(points[0], b);
+                } else {
+                    results[1] = 0;
+                }
+	    } else {
+		for (k = 1; k < sp; k++) {
+		    results[k] = quadtscore(points[k - 1], points[k]);
+		}
+                if (b != points[k - 1]) {
+		    results[k] = quadtscore(points[k - 1], b);
+                } else {
+                    results[k] = 0;
+                }
+	    }
+	} else {
+	    if (!isint(points) || points <= 0)
+		return newerror(strcat("quadts: variable 'points` must be a ",
+				       "list or a positive integer"));
+	    /* Taking "points" as the number of equally spaced intervals */
+	    results = mat[points + 1];
+	    /* It is easy if a,b lie on the real line */
+	    if (isreal(a) && isreal(b)) {
+		length = abs(a - b);
+		segment = length / points;
+
+		for (k = 1; k <= points; k++) {
+		    results[k - 1] =
+			quadtscore(a + (k - 1) * segment, a + k * segment);
+		}
+	    } else {
+		/* We have at least one complex limit but treat "points" still
+                 * as the number of equally spaced intervals on a straight line
+                 * connecting a and b. Computing the segments here is a bit
+                 * more complicated but not much, it should have been taught in
+                 * high school.
+		 * Other contours by way of a list of points */
+		slope = (im(b) - im(a)) / (re(b) - re(a));
+		C = (im(a) + slope) * re(a);
+		length = abs(re(a) - re(b));
+		segment = length / points;
+
+		/* y = mx+C where m is the slope, x is the real part and y the
+		 * imaginary part  */
+                if(re(a)>re(b))swap(a,b);
+		for (k = re(a); k <= (re(b)); k+=segment) {
+		    x1 = slope*(k) +  C;
+		    results[k] = quadtscore(k + x1 * 1i);
+		}
+	    }			/* else of isreal */
+	}			/* else of ismat|islist */
+    }				/* else of isnull(points) */
+    /* With a bit of undeserved luck we have a result by now. */
+    sp = size(results);
+    for (k = 0; k < sp; k++) {
+	sum += results[k];
+    }
+    return sum;
+}
+
+static __CZ__gl_x;
+static __CZ__gl_w;
+static __CZ__gl_order;
+static __CZ__gl_prec;
+
+define quadglcomputenodes(N)
+{
+    local places k l x w t1 t2 t3 t4 t5 r tmp;
+
+    if (__CZ__gl_order == N && __CZ__gl_prec == epsilon())
+	return;
+
+    __CZ__gl_x = mat[N];
+    __CZ__gl_w = mat[N];
+    __CZ__gl_order = N;
+    __CZ__gl_prec = epsilon();
+
+    places = highbit(1 + int (1 / epsilon())) +1;
+
+    /*
+     * Compute roots and weights (doing it inline seems to be fastest)
+     * Trick shamelessly stolen from D. Bailey et .al (program "arprec")
+     */
+    for (k = 1; k <= N//2; k++) {
+	r = bround(cos(pi() * (k - .25) / (N + .5)), places);
+	while (1) {
+	    t1 = 1, t2 = 0;
+	    for (l = 1; l <= N; l++) {
+		t3 = t2;
+		t2 = t1;
+		t1 = bround(((2 * l - 1) * r * t2 - (l - 1) * t3) / l, places);
+	    }
+	    t4 = bround(N * (r * t1 - t2) / ((r ^ 2) - 1), places);
+	    t5 = r;
+	    tmp = t1 / t4;
+	    r = r - tmp;
+	    if (abs(tmp) <= epsilon())
+		break;
+	}
+	x = r;
+	w = bround(2 / ((1 - r ^ 2) * t4 ^ 2), places);
+
+	__CZ__gl_x[k - 1] = x;
+	__CZ__gl_w[k - 1] = w;
+	__CZ__gl_x[N - k] = -__CZ__gl_x[k - 1];
+	__CZ__gl_w[N - k] = __CZ__gl_w[k - 1];
+    }
+    return;
+}
+
+define quadgldeletenodes()
+{
+    free(__CZ__gl_x);
+    free(__CZ__gl_w);
+    free(__CZ__gl_order);
+    free(__CZ__gl_prec);
+}
+
+define quadglcore(a, b, n)
+{
+    local k c d digs order eps places sum ret err x x1 w w1 m;
+    local phalf x2 px1 spx1 u b1 a1 half;
+
+    eps = epsilon(epsilon() * 1e-2);
+    places = highbit(1 + int (1 / epsilon())) +1;
+    if (!isnull(n))
+	order = n;
+    else {
+	m = int (4 + max(0, ln(places / 30.0) / ln(2))) + 2;
+	order = 3 * 2 ^ (m - 1);
+    }
+
+
+    quadglcomputenodes(order, 1);
+
+    if (isinfinite(a) || isinfinite(b)) {
+	if (isninf(a) && ispinf(b)) {
+	    for (k = 0; k < order; k++) {
+		x1 = __CZ__gl_x[k];
+		w1 = __CZ__gl_w[k];
+
+		x = bround(x1 * (1 - x1 ^ 2) ^ (-1 / 2), places);
+		w = bround(w1 * (((1 - x1 ^ 2) ^ (-1 / 2)) / (1 - x1 ^ 2)),
+			   places);
+		sum += bround(w * f(x), places);
+	    }
+	} else if (isninf(a) && !iscinf(b)) {
+	    for (k = 0; k < order; k++) {
+		x1 = __CZ__gl_x[k];
+		w1 = __CZ__gl_w[k];
+
+		x = bround((b + 1) - (2 / (x1 + 1)), places);
+		w = bround(w1 * (1 / 2 * (2 / (x1 + 1)) ^ 2), places);
+		sum += bround(w * f(x), places);
+	    }
+	} else if (!iscinf(a) && ispinf(b)) {
+	    for (k = 0; k < order; k++) {
+		x1 = __CZ__gl_x[k];
+		w1 = __CZ__gl_w[k];
+		x = bround((a - 1) + (2 / (x1 + 1)), places);
+		w = bround(w1 * (((1 / 2) * (2 / (x1 + 1)) ^ 2)), places);
+		sum += bround(w * f(x), places);
+	    }
+	} else if (isninf(a) || isninf(b)) {
+	    /*TODO: swap(a,b) and negate(w)? Lookup! */
+	    return newerror("quadglcore: reverse limits?");
+	} else
+	    return
+		newerror("quadglcore: complex infinity not yet implemented");
+	ret = sum;
+    } else {
+	/* Avoid rounding errors */
+	if (a == -1 && b == 1) {
+	    c = 1;
+	    d = 0;
+	} else {
+	    c = (b - a) / 2;
+	    d = (b + a) / 2;
+	}
+	sum = 0;
+	for (k = 0; k < order; k++) {
+	    sum += bround(__CZ__gl_w[k] * f(c * __CZ__gl_x[k] + d), places);
+	}
+	ret = c * sum;
+    }
+    epsilon(eps);
+    return ret;
+}
+
+define quadgl(a, b, points)
+{
+    local k sp results epsbits nsect interval length segment slope C x1 y1 x2
+	y2;
+    local sum D1 D2 D3 D4;
+    if (param(0) < 2)
+	return newerror("quadgl: not enough arguments");
+    epsbits = highbit(1 + int (1 / epsilon())) +1;
+    if (isnull(points)) {
+	/* return as given */
+	return quadglcore(a, b);
+    } else {
+	/* But if we could half the time needed to execute a single operation
+	 * we could do all of it in just twice that time. */
+	if (isinfinite(a) || isinfinite(b)
+	    && (!ismat(points) && !islist(points)))
+	    return
+		newerror(strcat
+			 ("quadgl: multiple segments of infinite length ",
+			  "are not yet supported"));
+	if (ismat(points) || islist(points)) {
+	    sp = size(points);
+	    if (sp == 0)
+		return
+		    newerror(strcat
+			     ("quadgl: variable 'points` must be a list or ",
+			      "1d-matrix of a length > 0"));
+	    /* check if all points are numbers */
+	    for (k = 0; k < sp; k++) {
+		if (!isnum(points[k]))
+		    return
+			newerror(strcat
+				 ("quadgl: elements of 'points` must be ",
+				  "numbers only"));
+	    }
+	    /* We have n-1 intervals and a and b, hence n-1 + 2 results */
+	    results = mat[sp + 1];
+            if (a != points[0]) {
+	        results[0] = quadglcore(a, points[0]);
+            } else {
+                results[0] = 0;
+            }
+	    if (sp == 1) {
+		if (b != points[0]) {
+		    results[1] = quadglcore(points[0], b);
+                } else {
+                    results[1] = 0;
+                }
+	    } else {
+		for (k = 1; k < sp; k++) {
+		    results[k] = quadglcore(points[k - 1], points[k]);
+		}
+                if (b != points[k - 1]) {
+		    results[k] = quadglcore(points[k - 1], b);
+                } else {
+                    results[k] = 0;
+                }
+	    }
+	} else {
+	    if (!isint(points) || points <= 0)
+		return newerror(strcat("quadgl: variable 'points` must be a ",
+				       "list or a positive integer"));
+	    /* Taking "points" as the number of equally spaced intervals */
+	    results = mat[points + 1];
+	    /* It is easy if a,b lie on the real line */
+	    if (isreal(a) && isreal(b)) {
+		length = abs(a - b);
+		segment = length / points;
+
+		for (k = 1; k <= points; k++) {
+		    results[k - 1] =
+			quadglcore(a + (k - 1) * segment, a + k * segment);
+		}
+	    } else {
+		/* Other contours by way of a list of points */
+		slope = (im(b) - im(a)) / (re(b) - re(a));
+		C = (im(a) + slope) * re(a);
+		length = abs(re(a) - re(b));
+		segment = length / points;
+
+		/* y = mx+C where m is the slope, x is the real part and y the
+		 * imaginary part  */
+                if(re(a)>re(b))swap(a,b);
+		for (k = re(a); k <= (re(b)); k+=segment) {
+		    x1 = slope*(k) +  C;
+		    results[k] = quadglcore(k + x1 * 1i);
+		}
+	    }			/* else of isreal */
+	}			/* else of ismat|islist */
+    }				/* else of isnull(points) */
+    /* With a bit of undeserved luck we have a result by now. */
+    sp = size(results);
+    for (k = 0; k < sp; k++) {
+	sum += results[k];
+    }
+    return sum;
+}
+
+define quad(a, b, points = -1, method = "tanhsinh")
+{
+    if (isnull(a) || isnull(b) || param(0) < 2)
+	return newerror("quad: both limits must be given");
+    if (isstr(a)) {
+	if (strncmp(a, "cinf", 1) == 0)
+	    return
+		newerror(strcat
+			 ("quad: complex infinity not yet supported, use",
+			  " 'pinf' or 'ninf' respectively"));
+    }
+    if (isstr(b)) {
+	if (strncmp(b, "cinf", 1) == 0)
+	    return
+		newerror(strcat
+			 ("quad: complex infinity not yet supported, use",
+			  " 'pinf' or 'ninf' respectively"));
+    }
+
+    if (param(0) == 3) {
+	if (isstr(points))
+	    method = points;
+    }
+
+    if (strncmp(method, "tanhsinh", 1) == 0) {
+	if (!isstr(points)) {
+	    if (points == -1) {
+		return quadts(a, b);
+	    } else {
+		return quadts(a, b, points);
+	    }
+	} else {
+	    return quadts(a, b);
+	}
+    }
+
+    if (strncmp(method, "gausslegendre", 1) == 0) {
+	if (!isstr(points)) {
+	    if (points == -1) {
+		return quadgl(a, b);
+	    } else {
+		return quadgl(a, b, points);
+	    }
+	} else {
+	    return quadgl(a, b);
+	}
+    }
+}
+
+define makerange(start, end, steps)
+{
+    local ret k l step C length slope x1 x2 y1 y2;
+    local segment;
+    steps = int (steps);
+    if (steps < 1) {
+	return newerror("makerange: number of steps must be > 0");
+    }
+    if (!isnum(start) || !isnum(end)) {
+	return newerror("makerange: only numbers are supported yet");
+    }
+    if (isreal(start) && isreal(end)) {
+	step = (end - start) / (steps);
+	print step;
+	ret = mat[steps + 1];
+	for (k = 0; k <= steps; k++) {
+	    ret[k] = k * step + start;
+	}
+    } else {
+	ret = mat[steps + 1];
+	if (re(start) > re(end)) {
+	    swap(start, end);
+	}
+
+	slope = (im(end) - im(start)) / (re(end) - re(start));
+	C = im(start) - slope * re(start);
+	length = abs(re(start) - re(end));
+	segment = length / (steps);
+
+	for (k = re(start), l = 0; k <= (re(end)); k += segment, l++) {
+	    x1 = slope * (k) + C;
+	    ret[l] = k + x1 * 1i;
+	}
+
+    }
+    return ret;
+}
+
+define makecircle(radius, center, points)
+{
+    local ret k a b twopi centerx centery;
+    if (!isint(points) || points < 2) {
+	return
+	    newerror("makecircle: number of points is not a positive integer");
+    }
+    if (!isnum(center)) {
+	return newerror("makecircle: center does not lie on the complex plane");
+    }
+    if (!isreal(radius) || radius <= 0) {
+	return newerror("makecircle: radius is not a real > 0");
+    }
+    ret = mat[points];
+    twopi = 2 * pi();
+    centerx = re(center);
+    centery = im(center);
+    for (k = 0; k < points; k++) {
+	a = centerx + radius * cos(twopi * k / points);
+	b = centery + radius * sin(twopi * k / points);
+	ret[k] = a + b * 1i;
+    }
+    return ret;
+}
+
+define makeellipse(angle, a, b, center, points)
+{
+    local ret k x y twopi centerx centery;
+    if (!isint(points) || points < 2) {
+	return
+	    newerror("makeellipse: number of points is not a positive integer");
+    }
+    if (!isnum(center)) {
+	return
+	    newerror("makeellipse: center does not lie on the complex plane");
+    }
+    if (!isreal(a) || a <= 0) {
+	return newerror("makecircle: a is not a real > 0");
+    }
+    if (!isreal(b) || b <= 0) {
+	return newerror("makecircle: b is not a real > 0");
+    }
+    if (!isreal(angle)) {
+	return newerror("makecircle: angle is not a real");
+    }
+    ret = mat[points];
+    twopi = 2 * pi();
+    centerx = re(center);
+    centery = im(center);
+    for (k = 0; k < points; k++) {
+	x = centerx + a * cos(twopi * k / points) * cos(angle)
+	    - b * sin(twopi * k / points) * sin(angle);
+	y = centerx + a * cos(twopi * k / points) * sin(angle)
+	    + b * sin(twopi * k / points) * cos(angle);
+	ret[k] = x + y * 1i;
+    }
+    return ret;
+}
+
+define makepoints()
+{
+    local ret k;
+    ret = mat[param(0)];
+    for (k = 0; k < param(0); k++) {
+	if (!isnum(param(k + 1))) {
+	    return
+		newerror(strcat
+			 ("makepoints: parameter number \"", str(k + 1),
+			  "\" is not a number"));
+	}
+	ret[k] = param(k + 1);
+    }
+    return ret;
+}
+
+
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "quadtsdeletenodes()";
+    print "quadtscomputenodes(order, expo, eps)";
+    print "quadtscore(a,b,n)";
+    print "quadts(a,b,points)";
+    print "quadglcomputenodes(N)";
+    print "quadgldeletenodes()";
+    print "quadglcore(a,b,n)";
+    print "quadgl(a,b,points)";
+    print "quad(a,b,points=-1,method=\"tanhsinh\")";
+    print "makerange(start, end, steps)";
+    print "makecircle(radius, center, points)";
+    print "makeellipse(angle, a, b, center, points)";
+    print "makepoints(a1,[...])";
+}

cal/lambertw.cal

@@ -0,0 +1,284 @@
+/*
+ * lambertw - Lambert's W-function
+ *
+ * Copyright (C) 2013,2021 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2013/08/11 01:31:28
+ * File existed as early as:	2013
+ */
+
+
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+/*
+
+     R. M. Corless and G. H. Gonnet and D. E. G. Hare and D. J. Jeffrey and
+     D. E. Knuth, "On the Lambert W Function", Advances n Computational
+     Mathematics, 329--359, (1996)
+     http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.112.6117
+
+     D. J. Jeffrey, D. E. G. Hare, R. M. Corless, "Unwinding the branches of the
+     Lambert W function", The Mathematical Scientist, 21, pp 1-7, (1996)
+     http://www.apmaths.uwo.ca/~djeffrey/Offprints/wbranch.pdf
+
+     Darko Verebic, "Having Fun with Lambert W(x) Function"
+     arXiv:1003.1628v1, March 2010, http://arxiv.org/abs/1003.1628
+
+     Winitzki, S. "Uniform Approximations for Transcendental Functions",
+     In Part 1 of Computational Science and its Applications - ICCSA 2003,
+     Lecture Notes in Computer Science, Vol. 2667, Springer-Verlag,
+     Berlin, 2003, 780-789. DOI 10.1007/3-540-44839-X_82
+     A copy may be found by Google.
+
+
+*/
+static true = 1;
+static false = 0;
+
+/* Branch 0, Winitzki (2003) , the well known Taylor series*/
+define __CZ__lambertw_0(z,eps){
+  local a=2.344e0, b=0.8842e0, c=0.9294e0, d=0.5106e0, e=-1.213e0;
+  local y=sqrt(2*exp(1)*z+2);
+  return (2*ln(1+b*y)-ln(1+c*ln(1+d*y))+e)/(1+1/(2*ln(1+b*y)+2*a));
+}
+/* branch -1 */
+define __CZ__lambertw_m1(z,eps){
+  local wn k;
+  /* Cut-off found in Maxima */
+  if(z < 0.3) return __CZ__lambertw_app(z,eps);
+  wn = z;
+  /*  Verebic (2010) eqs. 16-18*/
+  for(k=0;k<10;k++){
+    wn = ln(-z)-ln(-wn);
+  }
+  return wn;
+}
+
+/*
+  generic approximation
+
+  series for 1+W((z-2)/(2 e))
+
+  Corless et al (1996) (4.22)
+  Verebic (2010) eqs. 35-37; more coefficients given at the end of sect. 3.1
+  or online
+     http://www.wolframalpha.com/input/?
+       i=taylor+%28+1%2Bproductlog%28+%28z-2%29%2F%282*e%29+%29+%29
+  or by using the function lambertw_series_print() after running
+  lambertw_series(z,eps,branch,terms) at least once with the wanted number of
+  terms and z = 1 (which might throw an error because the series will not
+  converge in anybodies lifetime for something that far from the branch point).
+
+
+*/
+define __CZ__lambertw_app(z,eps){
+  local  b0=-1, b1=1, b2=-1/3, b3=11/72;
+  local  y=sqrt(2*exp(1)*z+2);
+  return b0 + ( y * (b1 + (y * (b2 + (b3 * y)))));
+}
+
+static __CZ__Ws_a;
+static __CZ__Ws_c;
+static __CZ__Ws_len=0;
+
+define lambertw_series_print(){
+  local k;
+  for(k=0;k<__CZ__Ws_len;k++){
+    print num(__CZ__Ws_c[k]):"/":den(__CZ__Ws_c[k]):"*p^":k;
+  }
+}
+
+/*
+   The series is fast but only if _very_ close to the branch point
+   The exact branch must be given explicitly, e.g.:
+
+; lambertw(-exp(-1)+.001)-lambertw_series(-exp(-1)+.001,epsilon()*1e-10,0)
+	-0.14758879113205794065490184399030194122136720202792-
+                  0.00000000000000000000000000000000000000000000000000i
+; lambertw(-exp(-1)+.001)-lambertw_series(-exp(-1)+.001,epsilon()*1e-10,1)
+	0.00000000000000000000000000000000000000000000000000-
+                  0.00000000000000000000000000000000000000000000000000i
+ */
+define lambertw_series(z,eps,branch,terms){
+  local k l limit tmp sum A C P PP epslocal;
+  if(!isnull(terms))
+    limit = terms;
+  else
+    limit = 100;
+
+  if(isnull(eps))
+    eps = epsilon(epsilon()*1e-10);
+  epslocal = epsilon(eps);
+
+  P = sqrt(2*(exp(1)*z+1));
+  if(branch != 0) P = -P;
+  tmp=0;sum=0;PP=P;
+
+    __CZ__Ws_a   = mat[limit+1];
+    __CZ__Ws_c   = mat[limit+1];
+    __CZ__Ws_len = limit;
+   /*
+      c0 = -1; c1 = 1
+      a0 =  2; a1 =-1
+    */
+    __CZ__Ws_c[0] = -1; __CZ__Ws_c[1] =  1;
+    __CZ__Ws_a[0] =  2; __CZ__Ws_a[1] = -1;
+    sum += __CZ__Ws_c[0];
+    sum += __CZ__Ws_c[1] * P;
+    PP *= P;
+    for(k=2;k<limit;k++){
+      for(l=2;l<k;l++){
+         __CZ__Ws_a[k] += __CZ__Ws_c[l]*__CZ__Ws_c[k+1-l];
+      }
+
+      __CZ__Ws_c[k] = (k-1) * (  __CZ__Ws_c[k-2]/2
+                            +__CZ__Ws_a[k-2]/4)/
+                      (k+1)-__CZ__Ws_a[k]/2-__CZ__Ws_c[k-1]/(k+1);
+      tmp = __CZ__Ws_c[k] * PP;
+      sum += tmp;
+      if(abs(tmp) <= eps){
+        epsilon(epslocal);
+        return sum;
+      }
+      PP *= P;
+    }
+    epsilon(epslocal);
+    return
+      newerror(strcat("lambertw_series: does not converge in ",
+                        str(limit)," terms" ));
+}
+
+/* */
+define lambertw(z,branch){
+  local eps epslarge ret branchpoint bparea w we ew w1e wn k places m1e;
+  local closeness;
+
+  eps = epsilon();
+  if(branch == 0){
+    if(!im(z)){
+      if(abs(z) <= eps) return 0;
+      if(abs(z-exp(1)) <= eps) return 1;
+      if(abs(z - (-ln(2)/2)) <= eps ) return -ln(2);
+      if(abs(z - (-pi()/2)) <= eps ) return 1i*pi()/2;
+    }
+  }
+
+  branchpoint = -exp(-1);
+  bparea = .2;
+  if(branch == 0){
+    if(!im(z) && abs(z-branchpoint) == 0) return -1;
+    ret = __CZ__lambertw_0(z,eps);
+   /* Yeah, C&P, I know, sorry */
+   ##ret = ln(z) + 2*pi()*1i*branch - ln(ln(z)+2*pi()*1i*branch);
+  }
+  else if(branch == 1){
+    if(im(z)<0 && abs(z-branchpoint) <= bparea)
+      ret = __CZ__lambertw_app(z,eps);
+   /* Does calc have a goto? Oh, it does!  */
+    ret =ln(z) + 2*pi()*1i*branch - ln(ln(z)+2*pi()*1i*branch);
+  }
+  else if(branch == -1){##print "-1";
+    if(!im(z) && abs(z-branchpoint) == 0) return -1;
+    if(!im(z) && z>branchpoint && z < 0){##print "0";
+      ret = __CZ__lambertw_m1(z,eps);}
+    if(im(z)>=0 && abs(z-branchpoint) <= bparea){##print "1";
+      ret = __CZ__lambertw_app(z,eps);}
+    ret =ln(z) + 2*pi()*1i*branch - ln(ln(z)+2*pi()*1i*branch);
+  }
+  else
+    ret = ln(z) + 2*pi()*1i*branch - ln(ln(z)+2*pi()*1i*branch);
+
+  /*
+     Such a high precision is only needed _very_ close to the branchpoint
+     and might even be insufficient if z has not been computed with
+     sufficient precision itself (M below was calculated by Mathematica and also
+     with the series above with epsilon(1e-200)):
+    ; epsilon(1e-50)
+	0.00000000000000000001
+    ; display(50)
+	20
+    ; M=-0.9999999999999999999999997668356018402875796636464119050387
+    ; lambertw(-exp(-1)+1e-50,0)-M
+	-0.00000000000000000000000002678416515423276355643684
+    ; epsilon(1e-60)
+	0.0000000000000000000000000000000000000000000000000
+    ; A=-exp(-1)+1e-50
+    ; epsilon(1e-50)
+	0.00000000000000000000000000000000000000000000000000
+    ; lambertw(A,0)-M
+	-0.00000000000000000000000000000000000231185460220585
+    ; lambertw_series(A,epsilon(),0)-M
+	-0.00000000000000000000000000000000000132145133161626
+    ; epsilon(1e-100)
+	0.00000000000000000000000000000000000000000000000001
+    ; A=-exp(-1)+1e-50
+    ; epsilon(1e-65)
+	0.00000000000000000000000000000000000000000000000000
+    ; lambertw_series(A,epsilon(),0)-M
+	0.00000000000000000000000000000000000000000000000000
+    ; lambertw_series(-exp(-1)+1e-50,epsilon(),0)-M
+	-0.00000000000000000000000000000000000000002959444084
+    ; epsilon(1e-74)
+	0.00000000000000000000000000000000000000000000000000
+    ; lambertw_series(-exp(-1)+1e-50,epsilon(),0)-M
+	-0.00000000000000000000000000000000000000000000000006
+   */
+  closeness = abs(z-branchpoint);
+  if( closeness< 1){
+    if(closeness != 0)
+      eps = epsilon(epsilon()*( closeness));
+    else
+      eps = epsilon(epsilon()^2);
+  }
+  else
+    eps = epsilon(epsilon()*1e-2);
+
+
+  epslarge =epsilon();
+
+  places = highbit(1 + int(1/epslarge)) + 1;
+  w = ret;
+  for(k=0;k<100;k++){
+    ew = exp(w);
+    we = w*ew;
+    if(abs(we-z)<= 4*epslarge*abs(z))break;
+    w1e = (1+w)*ew;
+    wn = bround(w- ((we - z) / ( w1e - ( (w+2)*(we-z) )/(2*w+2)  ) ),places++) ;
+    if( abs(wn - w) <= epslarge*abs(wn)) break;
+    else  w = wn;
+  }
+
+  if(k==100){
+    epsilon(eps);
+    return newerror("lambertw: Halley iteration does not converge");
+  }
+  /* The Maxima coders added a check if the iteration converged to the correct
+     branch. This coder deems it superfluous. */
+
+  epsilon(eps);
+  return wn;
+}
+
+
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "lambertw(z,branch)";
+    print "lambertw_series(z,eps,branch,terms)";
+    print "lambertw_series_print()";
+}

cal/linear.cal

@@ -1,7 +1,7 @@
 /*
  * linear - perform a simple two point 2D linear interpolation
  *
- * Copyright (C) 2005-2007  Landon Curt Noll
+ * Copyright (C) 2005-2007,2021  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.2 $
- * @(#) $Id: linear.cal,v 30.2 2007/03/17 05:57:42 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/linear.cal,v $
- *
  * Under source code control:	2005/12/12 06:41:50
  * File existed as early as:	2005
  *
@@ -33,7 +29,7 @@
  *
  * given:
  *	x0, y0		first known point on the line
- *	x1, y1		second knonw point on the line
+ *	x1, y1		second known point on the line
  *	x		a given point to interpolate on
  *
  * returns:

cal/lnseries.cal

@@ -0,0 +1,108 @@
+/*
+ * lnseries - special functions (e.g.: gamma, zeta, psi)
+ *
+ * Copyright (C) 2013 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2013/08/11 01:31:28
+ * File existed as early as:	2013
+ */
+
+
+/*
+ * hide internal function from resource debugging
+ */
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+static __CZ__int_logs;
+static __CZ__int_logs_limit;
+static __CZ__int_logs_prec;
+
+
+define deletelnseries(){
+  free(__CZ__int_logs,__CZ__int_logs_limit,__CZ__int_logs_prec);
+}
+
+define lnfromseries(n){
+  if(    isnull(__CZ__int_logs)
+      || __CZ__int_logs_limit < n
+      || __CZ__int_logs_prec < log(1/epsilon())){
+
+    lnseries(n+1);
+  }
+  return __CZ__int_logs[n,0];
+}
+
+define lnseries(limit){
+  local k j eps ;
+  if(    isnull(__CZ__int_logs)
+      || __CZ__int_logs_limit < limit
+      || __CZ__int_logs_prec < log(1/epsilon())){
+    __CZ__int_logs = mat[limit+1,2];
+    __CZ__int_logs_limit = limit;
+    __CZ__int_logs_prec = log(1/epsilon());
+
+    /* probably still too much */
+    eps = epsilon(epsilon()*10^(-(5+log(limit))));
+    k =2;
+    while(1){
+      /* the prime itself, compute logarithm */
+      __CZ__int_logs[k,0] = ln(k);
+      __CZ__int_logs[k,1] = k;
+
+      for(j = 2*k;j<=limit;j+=k){
+       /* multiples of prime k, add logarithm of k computed earlier */
+       __CZ__int_logs[j,0] += __CZ__int_logs[k,0];
+       /* First hit, set counter to number */
+       if(__CZ__int_logs[j,1] ==0)
+         __CZ__int_logs[j,1]=j;
+       /* reduce counter by prime added */
+       __CZ__int_logs[j,1] //= __CZ__int_logs[k,1];
+      }
+
+      k++;
+      if(k>=limit) break;
+      /* Erastothenes-sieve: look for next prime. */
+      while(__CZ__int_logs[k,0]!=0){
+        k++;
+        if(k>=limit) break;
+      }
+    }
+    /* Second run to include the last factor */
+    for(k=1;k<=limit;k++){
+      if(__CZ__int_logs[k,1] != k){
+        __CZ__int_logs[k,0] +=__CZ__int_logs[ __CZ__int_logs[k,1],0];
+        __CZ__int_logs[k,1] = 0;
+      }
+    }
+
+    epsilon(eps);
+  }
+  return 1;
+}
+
+
+/*
+ * restore internal function from resource debugging
+ */
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "lnseries(limit)";
+    print "lnfromseries(n)";
+    print "deletelnseries()";
+}

cal/lucas_chk.cal

@@ -1,7 +1,7 @@
 /*
  * lucas_chk - test all primes of the form h*2^n-1, 1<=h<200 and n <= high_n
  *
- * Copyright (C) 1999  Landon Curt Noll
+ * Copyright (C) 1999,2021  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: lucas_chk.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/lucas_chk.cal,v $
- *
  * Under source code control:	1991/01/11 05:41:43
  * File existed as early as:	1991
  *
@@ -307,7 +303,7 @@ read -once "lucas.cal";
  *	[quiet] if given and != 0, then do not print individual test results
  *
  * returns:
- *	1	all is ok
+ *	1	all is OK
  *	0	something went wrong
  */
 define

cal/lucas_tbl.cal

@@ -1,165 +0,0 @@
-/*
- * lucas_tbl - lucasian criteria for primality tables
- *
- * Copyright (C) 1999  Landon Curt Noll
- *
- * Calc is open software; you can redistribute it and/or modify it under
- * the terms of the version 2.1 of the GNU Lesser General Public License
- * as published by the Free Software Foundation.
- *
- * Calc is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
- * or FITNESS FOR A PARTICULAR PURPOSE.	 See the GNU Lesser General
- * Public License for more details.
- *
- * A copy of version 2.1 of the GNU Lesser General Public License is
- * distributed with calc under the filename COPYING-LGPL.  You should have
- * received a copy with calc; if not, write to Free Software Foundation, Inc.
- * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
- *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: lucas_tbl.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/lucas_tbl.cal,v $
- *
- * Under source code control:	1991/01/26 02:43:43
- * File existed as early as:	1991
- *
- * chongo <was here> /\oo/\	http://www.isthe.com/chongo/
- * Share and enjoy!  :-)	http://www.isthe.com/chongo/tech/comp/calc/
- */
-
-/*
- * Lucasian criteria for primality
- *
- * The following table is taken from:
- *
- *	"Lucasian Criteria for the Primality of N=h*2^n-1", by Hans Riesel,
- *	Mathematics of Computation, Vol 23 #108, p 872.
- *
- * The index of the *_val[] arrays correspond to the v(1) values found
- * in the table.  That is, for v(1) == x:
- *
- *	D == d_val[x]
- *	a == a_val[x]
- *	b == b_val[x]
- *	r == r_val[x]		(r == abs(a^2 - b^2*D))
- *
- *
- * Note that when *_val[i] is not a number, the related v(1) value
- * is not found in Table 1.
- */
-
-
-trymax = 100;
-mat d_val[trymax+1];
-mat a_val[trymax+1];
-mat b_val[trymax+1];
-mat r_val[trymax+1];
-/* v1= 0	    INVALID */
-/* v1= 1	    INVALID */
-/* v1= 2	    INVALID */
-d_val[ 3]=   5;	 a_val[ 3]= 1;	b_val[ 3]=1;  r_val[ 3]=4;
-d_val[ 4]=   3;	 a_val[ 4]= 1;	b_val[ 4]=1;  r_val[ 4]=2;
-d_val[ 5]=  21;	 a_val[ 5]= 3;	b_val[ 5]=1;  r_val[ 5]=12;
-d_val[ 6]=   2;	 a_val[ 6]= 1;	b_val[ 6]=1;  r_val[ 6]=1;
-/* v1= 7	    INVALID */
-d_val[ 8]=  15;	 a_val[ 8]= 3;	b_val[ 8]=1;  r_val[ 8]=6;
-d_val[ 9]=  77;	 a_val[ 9]= 7;	b_val[ 9]=1;  r_val[ 9]=28;
-d_val[10]=   6;	 a_val[10]= 2;	b_val[10]=1;  r_val[10]=2;
-d_val[11]=  13;	 a_val[11]= 3;	b_val[11]=1;  r_val[11]=4;
-d_val[12]=  35;	 a_val[12]= 5;	b_val[12]=1;  r_val[12]=10;
-d_val[13]= 165;	 a_val[13]=11;	b_val[13]=1;  r_val[13]=44;
-/* v1=14	    INVALID */
-d_val[15]= 221;	 a_val[15]=13;	b_val[15]=1;  r_val[15]=52;
-d_val[16]=   7;	 a_val[16]= 3;	b_val[16]=1;  r_val[16]=2;
-d_val[17]= 285;	 a_val[17]=15;	b_val[17]=1;  r_val[17]=60;
-/* v1=18	    INVALID */
-d_val[19]= 357;	 a_val[19]=17;	b_val[19]=1;  r_val[19]=68;
-d_val[20]=  11;	 a_val[20]= 3;	b_val[20]=1;  r_val[20]=2;
-d_val[21]= 437;	 a_val[21]=19;	b_val[21]=1;  r_val[21]=76;
-d_val[22]=  30;	 a_val[22]= 5;	b_val[22]=1;  r_val[22]=5;
-/* v1=23	    INVALID */
-d_val[24]= 143;	 a_val[24]=11;	b_val[24]=1;  r_val[24]=22;
-d_val[25]=  69;	 a_val[25]= 9;	b_val[25]=1;  r_val[25]=12;
-d_val[26]=  42;	 a_val[26]= 6;	b_val[26]=1;  r_val[26]=6;
-d_val[27]=  29;	 a_val[27]= 5;	b_val[27]=1;  r_val[27]=4;
-d_val[28]= 195;	 a_val[28]=13;	b_val[28]=1;  r_val[28]=26;
-d_val[29]=  93;	 a_val[29]= 9;	b_val[29]=1;  r_val[29]=12;
-d_val[30]=  14;	 a_val[30]= 4;	b_val[30]=1;  r_val[30]=2;
-d_val[31]= 957;	 a_val[31]=29;	b_val[31]=1;  r_val[31]=116;
-d_val[32]= 255;	 a_val[32]=15;	b_val[32]=1;  r_val[32]=30;
-d_val[33]=1085;	 a_val[33]=31;	b_val[33]=1;  r_val[33]=124;
-/* v1=34	    INVALID */
-d_val[35]=1221;	 a_val[35]=33;	b_val[35]=1;  r_val[35]=132;
-d_val[36]= 323;	 a_val[36]=17;	b_val[36]=1;  r_val[36]=34;
-d_val[37]=1365;	 a_val[37]=35;	b_val[37]=1;  r_val[37]=140;
-d_val[38]=  10;	 a_val[38]= 3;	b_val[38]=1;  r_val[38]=1;
-d_val[39]=1517;	 a_val[39]=37;	b_val[39]=1;  r_val[39]=148;
-d_val[40]= 399;	 a_val[40]=19;	b_val[40]=1;  r_val[40]=38;
-d_val[41]=1677;	 a_val[41]=39;	b_val[41]=1;  r_val[41]=156;
-d_val[42]= 110;	 a_val[42]=10;	b_val[42]=1;  r_val[42]=10;
-d_val[43]= 205;	 a_val[43]=15;	b_val[43]=1;  r_val[43]=20;
-d_val[44]= 483;	 a_val[44]=21;	b_val[44]=1;  r_val[44]=42;
-d_val[45]=2021;	 a_val[45]=43;	b_val[45]=1;  r_val[45]=172;
-d_val[46]=  33;	 a_val[46]= 6;	b_val[46]=1;  r_val[46]=3;
-/* v1=47	    INVALID */
-d_val[48]=  23;	 a_val[48]= 5;	b_val[48]=1;  r_val[48]=2;
-d_val[49]=2397;	 a_val[49]=47;	b_val[49]=1;  r_val[49]=188;
-d_val[50]=  39;	 a_val[50]= 6;	b_val[50]=1;  r_val[50]=3;
-d_val[51]=  53;	 a_val[51]= 7;	b_val[51]=1;  r_val[51]=4;
-/* v1=52	    INVALID */
-d_val[53]=2805;	 a_val[53]=51;	b_val[53]=1;  r_val[53]=204;
-d_val[54]= 182;	 a_val[54]=13;	b_val[54]=1;  r_val[54]=13;
-d_val[55]=3021;	 a_val[55]=53;	b_val[55]=1;  r_val[55]=212;
-d_val[56]=  87;	 a_val[56]= 9;	b_val[56]=1;  r_val[56]=6;
-d_val[57]=3245;	 a_val[57]=55;	b_val[57]=1;  r_val[57]=220;
-d_val[58]= 210;	 a_val[58]=14;	b_val[58]=1;  r_val[58]=14;
-d_val[59]=3477;	 a_val[59]=57;	b_val[59]=1;  r_val[59]=228;
-d_val[60]= 899;	 a_val[60]=29;	b_val[60]=1;  r_val[60]=58;
-d_val[61]= 413;	 a_val[61]=21;	b_val[61]=1;  r_val[61]=28;
-/* v1=62	    INVALID */
-d_val[63]=3965;	 a_val[63]=61;	b_val[63]=1;  r_val[63]=244;
-d_val[64]=1023;	 a_val[64]=31;	b_val[64]=1;  r_val[64]=62;
-d_val[65]= 469;	 a_val[65]=21;	b_val[65]=1;  r_val[65]=28;
-d_val[66]=  17;	 a_val[66]= 4;	b_val[66]=1;  r_val[66]=1;
-d_val[67]=4485;	 a_val[67]=65;	b_val[67]=1;  r_val[67]=260;
-d_val[68]=1155;	 a_val[68]=33;	b_val[68]=1;  r_val[68]=66;
-d_val[69]=4757;	 a_val[69]=67;	b_val[69]=1;  r_val[69]=268;
-d_val[70]=  34;	 a_val[70]= 6;	b_val[70]=1;  r_val[70]=2;
-d_val[71]=5037;	 a_val[71]=69;	b_val[71]=1;  r_val[71]=276;
-d_val[72]=1295;	 a_val[72]=35;	b_val[72]=1;  r_val[72]=70;
-d_val[73]= 213;	 a_val[73]=15;	b_val[73]=1;  r_val[73]=12;
-d_val[74]=  38;	 a_val[74]= 6;	b_val[74]=1;  r_val[74]=2;
-d_val[75]=5621;	 a_val[75]=73;	b_val[75]=1;  r_val[75]=292;
-d_val[76]=1443;	 a_val[76]=37;	b_val[76]=1;  r_val[76]=74;
-d_val[77]= 237;	 a_val[77]=15;	b_val[77]=1;  r_val[77]=12;
-d_val[78]=  95;	 a_val[78]=10;	b_val[78]=1;  r_val[78]=5;
-/* v1=79	    INVALID */
-d_val[80]=1599;	 a_val[80]=39;	b_val[80]=1;  r_val[80]=78;
-d_val[81]=6557;	 a_val[81]=79;	b_val[81]=1;  r_val[81]=316;
-d_val[82]= 105;	 a_val[82]=10;	b_val[82]=1;  r_val[82]=5;
-d_val[83]=  85;	 a_val[83]= 9;	b_val[83]=1;  r_val[83]=4;
-d_val[84]=1763;	 a_val[84]=41;	b_val[84]=1;  r_val[84]=82;
-d_val[85]=7221;	 a_val[85]=83;	b_val[85]=1;  r_val[85]=332;
-d_val[86]= 462;	 a_val[86]=21;	b_val[86]=1;  r_val[86]=21;
-d_val[87]=7565;	 a_val[87]=85;	b_val[87]=1;  r_val[87]=340;
-d_val[88]= 215;	 a_val[88]=15;	b_val[88]=1;  r_val[88]=10;
-d_val[89]=7917;	 a_val[89]=87;	b_val[89]=1;  r_val[89]=348;
-d_val[90]= 506;	 a_val[90]=22;	b_val[90]=1;  r_val[90]=22;
-d_val[91]=8277;	 a_val[91]=89;	b_val[91]=1;  r_val[91]=356;
-d_val[92]= 235;	 a_val[92]=15;	b_val[92]=1;  r_val[92]=10;
-d_val[93]=8645;	 a_val[93]=91;	b_val[93]=1;  r_val[93]=364;
-d_val[94]= 138;	 a_val[94]=12;	b_val[94]=1;  r_val[94]=6;
-d_val[95]=9021;	 a_val[95]=93;	b_val[95]=1;  r_val[95]=372;
-d_val[96]=  47;	 a_val[96]= 7;	b_val[96]=1;  r_val[96]=2;
-d_val[97]=1045;	 a_val[97]=33;	b_val[97]=1;  r_val[97]=44;
-/* v1=98	    INVALID */
-d_val[99]=9797;	 a_val[99]=97;	b_val[99]=1;  r_val[99]=388;
-d_val[100]=  51; a_val[100]= 7; b_val[100]=1; r_val[100]=2;
-
-if (config("resource_debug") & 3) {
-    print "d_val[100] defined";
-    print "a_val[100] defined";
-    print "b_val[100] defined";
-    print "r_val[100] defined";
-}

cal/mersenne.cal

@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: mersenne.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/mersenne.cal,v $
- *
  * Under source code control:	1991/05/22 21:56:36
  * File existed as early as:	1991
  *

cal/mfactor.cal

@@ -1,7 +1,7 @@
 /*
  * mfactor - return the lowest factor of 2^n-1, for n > 0
  *
- * Copyright (C) 1999  Landon Curt Noll
+ * Copyright (C) 1999,2021  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: mfactor.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/mfactor.cal,v $
- *
  * Under source code control:	1996/07/06 06:09:40
  * File existed as early as:	1996
  *
@@ -83,7 +79,7 @@
  *		hindx = 0;
  *	} while (test_factor < some_limit);
  *
- * The test, mfactor(67, 1, 10000) took on an 200 Mhz r4k (user CPU seconds):
+ * The test, mfactor(67, 1, 10000) took on an 200 MHz r4k (user CPU seconds):
  *
  *	210.83	(prior to use of hset[])
  *	 78.35	(hset[] for p_elim = 7)
@@ -103,7 +99,7 @@
  *	 57.78	(hset[] for p_elim = 17)
  *	 p_elim == 19 rejected because of memory size
  *
- * The p_elim == 17 overhead takes ~3 minutes on an 200 Mhz r4k CPU and
+ * The p_elim == 17 overhead takes ~3 minutes on an 200 MHz r4k CPU and
  * requires about ~13 Megs of memory.  The p_elim == 13 overhead
  * takes about 3 seconds and requires ~1.5 Megs of memory.
  *
@@ -260,7 +256,7 @@ define mfactor(n, start_k, rept_loop, p_elim)
 		return q;
 	} else {
 		/* report this loop */
-		printf("at 2*%d*%d+1, cpu: %f\n",
+		printf("at 2*%d*%d+1, CPU: %f\n",
 			(q-1)/(2*n), n, usertime());
 		fflush(files(1));
 		loop = 0;
@@ -273,14 +269,14 @@ define mfactor(n, start_k, rept_loop, p_elim)
 		 */
 		if (rept_loop <= ++loop) {
 			/* report this loop */
-			printf("at 2*%d*%d+1, cpu: %f\n",
+			printf("at 2*%d*%d+1, CPU: %f\n",
 				(q-1)/(2*n), n, usertime());
 			fflush(files(1));
 			loop = 0;
 		}
 
 		/*
-		 * skip if divisable by a prime <= 449
+		 * skip if divisible by a prime <= 449
 		 *
 		 * The value 281 was determined by timing loops
 		 * which found that 281 was at or near the
@@ -289,7 +285,7 @@ define mfactor(n, start_k, rept_loop, p_elim)
 		 * The addition of the do { ... } while (factor(q, 449)>1);
 		 * loop reduced the factoring loop time (36504 k values with
 		 * the hset[] initialization time removed) from 25.69 sec to
-		 * 15.62 sec of CPU time on a 200Mhz r4k.
+		 * 15.62 sec of CPU time on a 200MHz r4k.
 		 */
 		do {
 			/*

cal/mod.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: mod.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/mod.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:34
  * File existed as early as:	before 1990
  *

cal/natnumset.cal

@@ -1,7 +1,7 @@
 /*
  * natnumset - functions for sets of natural numbers not exceeding a fixed bound
  *
- * Copyright (C) 1999  Ernest Bowen
+ * Copyright (C) 1999,2021  Ernest Bowen
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: natnumset.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/natnumset.cal,v $
- *
  * Under source code control:	1997/09/07 23:53:51
  * File existed as early as:	1997
  *
@@ -80,7 +76,7 @@
  *	A \ B = set difference, integers in A but not in B
  *
  *	~A	= complement of A, integers not in A
- *	#A	= number ofintegers in A
+ *	#A	= number of integers in A
  *	!A	= 1 or 0 according as A is empty or not empty
  *	+A	= sum of the members of A
  *
@@ -104,7 +100,7 @@
  *	A >= B = (B <= A)
  *	A > B = (B < A)
  *
- * Expresssions may be formed from the above "arithmetic" operations in
+ * Expressions may be formed from the above "arithmetic" operations in
  * the usual way, with parentheses for variations from the usual precedence
  * rules.  For example
  *

cal/pell.cal

@@ -1,7 +1,7 @@
 /*
  * pell - solve Pell's equation
  *
- * Copyright (C) 1999  David I. Bell
+ * Copyright (C) 1999,2021  David I. Bell
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: pell.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/pell.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:34
  * File existed as early as:	before 1990
  *
@@ -29,7 +25,7 @@
 
 /*
  * Solve Pell's equation; Returns the solution X to: X^2 - D * Y^2 = 1.
- * Type the solution to pells equation for a particular D.
+ * Type the solution to Pell's equation for a particular D.
  */
 
 

cal/pi.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: pi.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/pi.cal,v $
- *
  * Under source code control:	1991/05/22 21:56:37
  * File existed as early as:	1991
  *

cal/pix.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: pix.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/pix.cal,v $
- *
  * Under source code control:	1996/07/09 03:14:14
  * File existed as early as:	1996
  *

cal/pollard.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: pollard.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/pollard.cal,v $
- *
  * Under source code control:	1991/05/22 21:56:37
  * File existed as early as:	1991
  *

cal/poly.cal

@@ -1,7 +1,7 @@
 /*
  * poly - calculate with polynomials of one variable
  *
- * Copyright (C) 1999  Ernest Bowen
+ * Copyright (C) 1999,2021  Ernest Bowen
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: poly.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/poly.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:35
  * File existed as early as:	before 1990
  *
@@ -39,7 +35,7 @@
  *	variable has only one name.  For some purposes, a name like
  *	"sin(t)" or "(a + b)" or "\lambda" might be useful;
  *	names like "*" or "-27" are legal but might give expressions
- *	that are difficult to intepret.
+ *	that are difficult to interpret.
  *
  * Polynomial expressions may be constructed from numbers and the
  *	independent variable and other polynomials by the algebraic
@@ -47,7 +43,7 @@
  *	The operations // and % are defined to have the quotient and
  *	remainder meanings as usually defined for polynomials.
  *
- * When polynomials are assigned to idenfifiers, it is convenient to
+ * When polynomials are assigned to identifiers, it is convenient to
  *	think of the polynomials as values.  For example, p = (x - 1)^2
  *	assigns to p a polynomial value in the same way as q = (7 - 1)^2
  *	would assign to q a number value.  As with number expressions

cal/prompt.cal

@@ -1,7 +1,7 @@
 /*
- * prompt - eemonstration of some uses of prompt() and eval()
+ * prompt - demonstration of some uses of prompt() and eval()
  *
- * Copyright (C) 1999  Ernest Bowen
+ * Copyright (C) 1999,2021  Ernest Bowen
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: prompt.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/prompt.cal,v $
- *
  * Under source code control:	1995/12/18 04:43:25
  * File existed as early as:	1995
  *
@@ -65,9 +61,9 @@
  * nothing to sum.  The last line returns the value 3, i.e. the last
  * non-null value found for the expressions separated by semicolons,
  * so sum will be increased by 3 after the "print sum^2;" command
- * is executed.	 xxx The terminating semicolon is essential in the
+ * is executed.	 XXX The terminating semicolon is essential in the
  * last two lines.  A command like eval("print 7;") is acceptable to
- * calc but eval("print 7") causes an exit from calc. xxx)
+ * calc but eval("print 7") causes an exit from calc. XXX)
  *
  * If the value returned is not a number (e.g. the name of a list or matrix,
  * or if the string has syntax errors as in "2 + ", in which case the
@@ -79,7 +75,7 @@
  *	"sin(x)", "x^2 + 3*x", "exp(x, 1e-5)".
  *
  * Values of the function so defined are returned for values of x
- * entered in reponse to the ? prompt.	Operation is terminated by
+ * entered in response to the ? prompt.	Operation is terminated by
  * entering "end", "exit" or "quit".
  */
 

cal/psqrt.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: psqrt.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/psqrt.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:35
  * File existed as early as:	before 1990
  *

cal/qtime.cal

@@ -20,10 +20,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: qtime.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/qtime.cal,v $
- *
  * Under source code control:	1999/10/13 04:10:33
  * File existed as early as:	1999
  *

cal/quat.cal

@@ -1,7 +1,7 @@
 /*
- * quat - alculate using quaternions of the form: a + bi + cj + dk
+ * quat - calculate using quaternions of the form: a + bi + cj + dk
  *
- * Copyright (C) 1999  David I. Bell
+ * Copyright (C) 1999,2021  David I. Bell
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: quat.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/quat.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:35
  * File existed as early as:	before 1990
  *
@@ -55,7 +51,8 @@ define quat(a,b,c,d)
 
 define quat_print(a)
 {
-	print "quat(" : a.s : ", " : a.v[0] : ", " : a.v[1] : ", " : a.v[2] : ")" :;
+	print "quat(" : a.s : ", " : a.v[0] : ", " :
+	      a.v[1] : ", " : a.v[2] : ")" :;
 }
 
 

cal/randbitrun.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: randbitrun.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/randbitrun.cal,v $
- *
  * Under source code control:	1995/02/13 03:43:11
  * File existed as early as:	1995
  *

cal/randmprime.cal

@@ -1,7 +1,7 @@
 /*
  * randmprime - generate a random prime of the form h*2^n-1
  *
- * Copyright (C) 1999  Landon Curt Noll
+ * Copyright (C) 1999,2021  Landon Curt Noll
  *
  * Calc is open software; you can redistribute it and/or modify it under
  * the terms of the version 2.1 of the GNU Lesser General Public License
@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: randmprime.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/randmprime.cal,v $
- *
  * Under source code control:	1994/03/14 23:11:21
  * File existed as early as:	1994
  *
@@ -49,9 +45,9 @@ randmprime(bits, seed, dbg)
     local n;		/* n as in h*2^n-1 */
     local h;		/* h as in h*2^n-1 */
     local plush;	/* value added to h since the beginning */
-    local init;		/* initial cpu time */
-    local start;	/* cpu time before last test */
-    local stop;		/* cpu time afte last test */
+    local init;		/* initial CPU time */
+    local start;	/* CPU time before last test */
+    local stop;		/* CPU time after last test */
     local tmp;		/* just a tmp place holder value */
     local ret;		/* h*2^n-1 that is prime */
 

cal/randombitrun.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: randombitrun.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/randombitrun.cal,v $
- *
  * Under source code control:	1995/02/13 03:43:11
  * File existed as early as:	1995
  *

cal/randomrun.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: randomrun.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/randomrun.cal,v $
- *
  * Under source code control:	1997/02/19 03:35:59
  * File existed as early as:	1997
  *

cal/randrun.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: randrun.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/randrun.cal,v $
- *
  * Under source code control:	1995/02/12 20:00:06
  * File existed as early as:	1995
  *

cal/repeat.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: repeat.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/repeat.cal,v $
- *
  * Under source code control:	2003/01/05 00:00:01
  * File existed as early as:	2003
  *

cal/screen.cal

@@ -15,10 +15,6 @@
  * NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR  IN
  * CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  *
- * @(#) $Revision: 30.2 $
- * @(#) $Id: screen.cal,v 30.2 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/screen.cal,v $
- *
  * This file is not covered under version 2.1 of the GNU LGPL.
  *
  * Under source code control:	2006/03/08 05:54:09

cal/seedrandom.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: seedrandom.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/seedrandom.cal,v $
- *
  * Under source code control:	1996/01/01 08:21:00
  * File existed as early as:	1996
  *

cal/set8700.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: set8700.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/set8700.cal,v $
- *
  * Under source code control:	2006/05/20 14:10:11
  * File existed as early as:	2006
  *

cal/set8700.line

@@ -17,10 +17,6 @@
 ## received a copy with calc; if not, write to Free Software Foundation, Inc.
 ## 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
 ##
-## @(#) $Revision: 30.1 $
-## @(#) $Id: set8700.line,v 30.1 2007/03/16 11:09:54 chongo Exp $
-## @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/set8700.line,v $
-##
 ## Under source code control:	2006/05/20 14:10:11
 ## File existed as early as:	2006
 ##

cal/smallfactors.cal

@@ -0,0 +1,71 @@
+/*
+ * smallfactors - find the factors of a number < 2^32
+ *
+ * Copyright (C) 2013 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ */
+
+
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+define smallfactors(x0)
+{
+    local d q x flist tuple w;
+
+    if (x >= (2 ^ 32) - 1)
+	return newerror("smallfactors: number must be < 2^32 -1");
+
+    tuple = mat[2];
+    flist = list();
+    x = x0;
+    d = 2;
+    q = 0;
+    tuple[0] = d;
+    if (x < 2)
+	return 0;
+    do {
+	q = x // d;
+	    while (x == (q * d)) {
+	    tuple[0] = d;
+	    tuple[1]++;
+	    x = floor(q);
+	    q = x // d;
+	}
+	d = nextprime(d);
+	if (tuple[1] > 0)
+	    append(flist, tuple);
+	tuple = mat[2];
+    } while (d <= x);
+    return flist;
+}
+
+define printsmallfactors(flist)
+{
+    local k;
+    for (k = 0; k < size(flist); k++) {
+	print flist[k][0]:"^":flist[k][1];
+    }
+}
+
+
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "smallfactors(x0)";
+    print "printsmallfactors(flist)";
+
+}

cal/solve.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: solve.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/solve.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:37
  * File existed as early as:	before 1990
  *
@@ -47,12 +43,13 @@ define solve(low, high, epsilon)
 	if (abs(flow) < epsilon)
 		return low;
 	fhigh = f(high);
-	if (abs(flow) < epsilon)
+	if (abs(fhigh) < epsilon)
 		return high;
 	if (sgn(flow) == sgn(fhigh))
 		quit "Non-opposite signs";
 	while (1) {
-		mid = bround(high - fhigh * (high - low) / (fhigh - flow), places);
+		mid = bround(high - fhigh * (high - low) / (fhigh - flow),
+			     places);
 		if ((mid == low) || (mid == high))
 			places++;
 		fmid = f(mid);

cal/statistics.cal

@@ -0,0 +1,498 @@
+/*
+ * statistics - Some assorted statistics functions.
+ *
+ * Copyright (C) 2013,2021 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ *
+ * Under source code control:	2013/08/11 01:31:28
+ * File existed as early as:	2013
+ */
+
+
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+
+/*
+ *  get dependencies
+ */
+read -once factorial2 brentsolve
+
+
+/*******************************************************************************
+ *
+ *
+ *                Continuous distributions
+ *
+ *
+ ******************************************************************************/
+
+/* regularized incomplete gamma function like in Octave, hence the name */
+define gammaincoctave(z,a){
+  local tmp;
+  tmp = gamma(z);
+  return (tmp-gammainc(a,z))/tmp;
+}
+
+/* Inverse incomplete beta function. Old and slow. */
+static __CZ__invbeta_a;
+static __CZ__invbeta_b;
+static __CZ__invbeta_x;
+define __CZ__invbeta(x){
+   return __CZ__invbeta_x-__CZ__ibetaas63(x,__CZ__invbeta_a,__CZ__invbeta_b);
+}
+
+define invbetainc_slow(x,a,b){
+  local flag ret eps;
+  /* place checks and balances here */
+  eps = epsilon();
+  if(.5 < x){
+    __CZ__invbeta_x = 1 - x;
+    __CZ__invbeta_a = b;
+    __CZ__invbeta_b = a;
+    flag = 1;
+  }
+  else{
+    __CZ__invbeta_x = x;
+    __CZ__invbeta_a = a;
+    __CZ__invbeta_b = b;
+    flag = 0;
+  }
+
+  ret =  brentsolve2(0,1,1);
+
+  if(flag == 1)
+    ret = 1-ret;
+  epsilon(eps);
+  return ret;
+}
+
+/* Inverse incomplete beta function. Still old but not as slow as the function
+   above. */
+/*
+  Purpose:
+
+    invbetainc computes inverse of the incomplete Beta function.
+
+  Licensing:
+
+    This code is distributed under the GNU LGPL license.
+
+  Modified:
+
+    10 August 2013
+
+  Author:
+
+    Original FORTRAN77 version by GW Cran, KJ Martin, GE Thomas.
+    C version by John Burkardt.
+    Calc version by Christoph Zurnieden
+
+  Reference:
+
+    GW Cran, KJ Martin, GE Thomas,
+    Remark AS R19 and Algorithm AS 109:
+    A Remark on Algorithms AS 63: The Incomplete Beta Integral
+    and AS 64: Inverse of the Incomplete Beta integral,
+    Applied Statistics,
+    Volume 26, Number 1, 1977, pages 111-114.
+
+  Parameters:
+
+    Input,  P, Q, the parameters of the incomplete
+    Beta function.
+
+    Input,  BETA, the logarithm of the value of
+    the complete Beta function.
+
+    Input,  ALPHA, the value of the incomplete Beta
+    function.  0 <= ALPHA <= 1.
+
+    Output,  the argument of the incomplete
+    Beta function which produces the value ALPHA.
+
+  Local Parameters:
+
+    Local, SAE, the most negative decimal exponent
+    which does not cause an underflow.
+*/
+define invbetainc(x,a,b){
+  return __CZ__invbetainc(a,b,lnbeta(a,b),x);
+}
+
+define __CZ__invbetainc(p,q,beta,alpha){
+  local a acu adj fpu g h iex indx pp prev qq r s sae sq t tx value;
+  local w xin y yprev places eps;
+
+  /* Dirty trick, don't try at home */
+  eps= epsilon(epsilon()^2);
+  sae = -((log(1/epsilon())/log(2))//2);
+  fpu = 10.0^sae;
+
+  places = highbit(1 + int(1/epsilon())) + 1;
+  value = alpha;
+  if( p <= 0.0 ){
+    epsilon(eps);
+    return newerror("invbeta: argument p <= 0");
+  }
+  if( q <= 0.0 ){
+    epsilon(eps);
+    return newerror("invbeta: argument q <= 0");
+  }
+
+  if( alpha < 0.0 || 1.0 < alpha ){
+    epsilon(eps);
+    return newerror("invbeta: argument alpha out of domain");
+  }
+  if( alpha == 0.0 ){
+    epsilon(eps);
+    return 0;
+  }
+  if( alpha == 1.0 ){
+    epsilon(eps);
+    return 1;
+  }
+  if ( 0.5 < alpha ){
+    a = 1.0 - alpha;
+    pp = q;
+    qq = p;
+    indx = 1;
+  }
+  else{
+    a = alpha;
+    pp = p;
+    qq = q;
+    indx = 0;
+  }
+  r = sqrt ( - ln ( a * a ) );
+
+  y = r-(2.30753+0.27061*r)/(1.0+(0.99229+0.04481*r)*r);
+
+  if ( 1.0 < pp && 1.0 < qq ){
+    r = ( y * y - 3.0 ) / 6.0;
+    s = 1.0 / ( pp + pp - 1.0 );
+    t = 1.0 / ( qq + qq - 1.0 );
+    h = 2.0 / ( s + t );
+    w = y*sqrt(h+r)/h-(t-s)*(r+5.0/6.0-2.0/(3.0*h));
+    value = pp / ( pp + qq * exp ( w + w ) );
+  }
+  else{
+    r = qq + qq;
+    t = 1.0 / ( 9.0 * qq );
+    t = r *  ( 1.0 - t + y * sqrt ( t )^ 3 );
+
+    if ( t <= 0.0 ){
+      value = 1.0 - exp ( ( ln ( ( 1.0 - a ) * qq ) + beta ) / qq );
+    }
+    else{
+      t = ( 4.0 * pp + r - 2.0 ) / t;
+
+      if ( t <= 1.0 ) {
+        value = exp ( ( ln ( a * pp ) + beta ) / pp );
+      }
+      else{
+        value = 1.0 - 2.0 / ( t + 1.0 );
+      }
+    }
+  }
+  r = 1.0 - pp;
+  t = 1.0 - qq;
+  yprev = 0.0;
+  sq = 1.0;
+  prev = 1.0;
+
+  if ( value < 0.0001 )
+    value = 0.0001;
+
+  if ( 0.9999 < value )
+    value = 0.9999;
+
+  acu = 10^sae;
+
+  for ( ; ; ){
+    y = bround(__CZ__ibetaas63( value, pp, qq, beta),places);
+    xin = value;
+    y = bround(exp(ln(y-a)+(beta+r*ln(xin)+t*ln(1.0- xin ) )),places);
+
+    if ( y * yprev <= 0.0 ) {
+      prev = max ( sq, fpu );
+    }
+
+    g = 1.0;
+
+    for ( ; ; ){
+      for ( ; ; ){
+        adj = g * y;
+        sq = adj * adj;
+        if ( sq < prev ){
+          tx = value - adj;
+          if ( 0.0 <= tx && tx <= 1.0 )  break;
+        }
+        g = g / 3.0;
+      }
+      if ( prev <= acu ){
+        if ( indx )
+          value = 1.0 - value;
+        epsilon(eps);
+        return value;
+      }
+      if ( y * y <= acu ){
+        if ( indx )
+          value = 1.0 - value;
+        epsilon(eps);
+        return value;
+      }
+      if ( tx != 0.0 && tx != 1.0 )
+        break;
+      g = g / 3.0;
+    }
+    if ( tx == value )  break;
+    value = tx;
+    yprev = y;
+  }
+  if ( indx )
+    value = 1.0 - value;
+
+  epsilon(eps);
+  return value;
+}
+
+/*******************************************************************************
+ *
+ *
+ *                Beta distribution
+ *
+ *
+ ******************************************************************************/
+
+define betapdf(x,a,b){
+  if(x<0 || x>1) return newerror("betapdf: parameter x out of domain");
+  if(a<=0) return newerror("betapdf: parameter a out of domain");
+  if(b<=0) return newerror("betapdf: parameter b out of domain");
+
+  return 1/beta(a,b) *x^(a-1)*(1-x)^(b-1);
+}
+
+define betacdf(x,a,b){
+  if(x<0 || x>1) return newerror("betacdf: parameter x out of domain");
+  if(a<=0) return newerror("betacdf: parameter a out of domain");
+  if(b<=0) return newerror("betacdf: parameter b out of domain");
+
+  return betainc(x,a,b);
+}
+
+define betacdfinv(x,a,b){
+  return invbetainc(x,a,b);
+}
+
+define betamedian(a,b){
+  local t106 t104 t103 t105 approx ret;
+  if(a == b) return 1/2;
+  if(a == 1 && b > 0) return 1-(1/2)^(1/b);
+  if(a > 0 && b == 1) return   (1/2)^(1/a);
+  if(a == 3 && b == 2){
+    /* Yes, the author is not ashamed to ask Maxima for the exact solution
+       of a quartic equation. */
+    t103 = ( (2^(3/2))/27 +4/27  )^(1/3);
+    t104 = sqrt( ( 9*t103^2 + 4*t103 + 2  )/(t103)  )/3;
+    t105 = -t103-2/(9*t103) +8/9;
+    t106 = sqrt( (27*t104*t105+16)/(t104)  )/(2*3^(3/2));
+    return -t106+t104/2+1/3;
+  }
+  if(a == 2 && b == 3){
+    t103 = ( (2^(3/2))/27 +4/27  )^(1/3);
+    t104 = sqrt( ( 9*t103^2 + 4*t103 + 2  )/(t103)  )/3;
+    t105 = -t103-2/(9*t103) +8/9;
+    t106 = sqrt( (27*t104*t105+16)/(t104)  )/(2*3^(3/2));
+    return 1-(-t106+t104/2+1/3);
+  }
+  return invbetainc(1/2,a,b);
+}
+
+define betamode(a,b){
+  if(a + b == 2) return newerror("betamod: a + b = 2 = division by zero");
+  return (a-1)/(a+b-2);
+}
+
+define betavariance(a,b){
+  return (a*b)/( (a+b)^2*(a+b+1) );
+}
+
+define betalnvariance(a,b){
+  return polygamma(1,a)-polygamma(a+b);
+}
+
+define betaskewness(a,b){
+  return (2*(b-a)*sqrt(a+b+1))/( (a+b+1)*sqrt(a*b) );
+}
+
+define betakurtosis(a,b){
+  local num denom;
+
+  num = 6*( (a-b)^2*(a+b+1)-a*b*(a+b+2));
+  denom = a*b*(a+b+2)*(a+b+3);
+  return num/denom;
+}
+
+define betaentropy(a,b){
+  return lnbeta(a,b)-(a-1)*psi(a)-(b-1)*psi(b)+(a+b+1)*psi(a+b);
+
+}
+
+/*******************************************************************************
+ *
+ *
+ *                Normal (Gaussian) distribution
+ *
+ *
+ ******************************************************************************/
+
+
+define normalpdf(x,mu,sigma){
+  return 1/(sqrt(2*pi()*sigma^2))*exp( ( (x-mu)^2 )/( 2*sigma^2 ) );
+}
+
+define normalcdf(x,mu,sigma){
+  return 1/2*(1+erf( ( x-mu )/( sqrt(2*sigma^2) )  )  );
+}
+
+define probit(p){
+  if(p<0 || p > 1) return newerror("probit: p out of domain 0<=p<=1");
+  return sqrt(2)*ervinv(2*p-1);
+}
+
+define normalcdfinv(p,mu,sigma){
+  if(p<0 || p > 1) return newerror("normalcdfinv: p out of domain 0<=p<=1");
+  return mu+ sigma*probit(p);
+}
+
+define normalmean(mu,sigma){return mu;}
+
+define normalmedian(mu,sigma){return mu;}
+
+define normalmode(mu,sigma){return mu;}
+
+define normalvariance(mu,sigma){return sigma^2;}
+
+define normalskewness(mu,sigma){return 0;}
+
+define normalkurtosis(mu,sigma){return 0;}
+
+define normalentropy(mu,sigma){
+  return 1/3*ln( 2*pi()*exp(1)*sigma^2 );
+}
+
+/* moment generating f. */
+define normalmgf(mu,sigma,t){
+  return exp(mu*t+1/2*sigma^2*t^2);
+}
+
+/* characteristic f. */
+define normalcf(mu,sigma,t){
+  return exp(mu*t-1/2*sigma^2*t^2);
+}
+
+
+/*******************************************************************************
+ *
+ *
+ *                Chi-squared distribution
+ *
+ *
+ ******************************************************************************/
+
+define chisquaredpdf(x,k){
+  if(!isint(k) || k<0) return newerror("chisquaredpdf: k not in N");
+  if(im(x) || x<0) return newerror("chisquaredpdf: x not in +R");
+  /* The gamma function does not check for half integers, do it here? */
+  return 1/(2^(k/2)*gamma(k/2))*x^((k/2)-1)*exp(-x/2);
+}
+
+define chisquaredpcdf(x,k){
+  if(!isint(k) || k<0) return newerror("chisquaredcdf: k not in N");
+  if(im(x) || x<0) return newerror("chisquaredcdf: x not in +R");
+
+  return 1/(gamma(k/2))*gammainc(k/2,x/2);
+}
+
+define chisquaredmean(x,k){return k;}
+
+define chisquaredmedian(x,k){
+  /* TODO: implement a FAST inverse incomplete gamma-{q,p} function */
+  return k*(1-2/(9*k))^3;
+}
+
+define chisquaredmode(x,k){return max(k-2,0);}
+define chisquaredvariance(x,k){return 2*k;}
+define chisquaredskewness(x,k){return sqrt(8/k);}
+define chisquaredkurtosis(x,k){return 12/k;}
+define chisquaredentropy(x,k){
+  return k/2+ln(2*gamma(k/2)) + (1-k/2)*psi(k/2);
+}
+
+define chisquaredmfg(k,t){
+  if(t>=1/2)return newerror("chisquaredmfg: t >= 1/2");
+  return (1-2*t)^(k/2);
+}
+
+define chisquaredcf(k,t){
+  return (1-2*1i*t)^(k/2);
+}
+
+
+/*
+ * restore internal function from resource debugging
+ */
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "gammaincoctave(z,a)";
+    print "invbetainc(x,a,b)";
+    print "betapdf(x,a,b)";
+    print "betacdf(x,a,b)";
+    print "betacdfinv(x,a,b)";
+    print "betamedian(a,b)";
+    print "betamode(a,b)";
+    print "betavariance(a,b)";
+    print "betalnvariance(a,b)";
+    print "betaskewness(a,b)";
+    print "betakurtosis(a,b)";
+    print "betaentropy(a,b)";
+    print "normalpdf(x,mu,sigma)";
+    print "normalcdf(x,mu,sigma)";
+    print "probit(p)";
+    print "normalcdfinv(p,mu,sigma)";
+    print "normalmean(mu,sigma)";
+    print "normalmedian(mu,sigma)";
+    print "normalmode(mu,sigma)";
+    print "normalvariance(mu,sigma)";
+    print "normalskewness(mu,sigma)";
+    print "normalkurtosis(mu,sigma)";
+    print "normalentropy(mu,sigma)";
+    print "normalmgf(mu,sigma,t)";
+    print "normalcf(mu,sigma,t)";
+    print "chisquaredpdf(x,k)";
+    print "chisquaredpcdf(x,k)";
+    print "chisquaredmean(x,k)";
+    print "chisquaredmedian(x,k)";
+    print "chisquaredmode(x,k)";
+    print "chisquaredvariance(x,k)";
+    print "chisquaredskewness(x,k)";
+    print "chisquaredkurtosis(x,k)";
+    print "chisquaredentropy(x,k)";
+    print "chisquaredmfg(k,t)";
+    print "chisquaredcf(k,t)";
+}
+

cal/strings.cal

@@ -0,0 +1,41 @@
+/*
+ * strings - implementation of some of the macros in ctype.h
+ *
+ * Copyright (C) 2013 Christoph Zurnieden
+ *
+ * Calc is open software; you can redistribute it and/or modify it under
+ * the terms of the version 2.1 of the GNU Lesser General Public License
+ * as published by the Free Software Foundation.
+ *
+ * Calc is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General
+ * Public License for more details.
+ *
+ * A copy of version 2.1 of the GNU Lesser General Public License is
+ * distributed with calc under the filename COPYING-LGPL.  You should have
+ * received a copy with calc; if not, write to Free Software Foundation, Inc.
+ * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
+ */
+
+
+static resource_debug_level;
+resource_debug_level = config("resource_debug", 0);
+
+define isascii(c){
+  c = ord(c);
+  return (c >= 0 && c< 128);
+}
+
+define isblank(c){
+  c = ord(c);
+  return ( c == 32 || c == 9 );
+}
+
+
+config("resource_debug", resource_debug_level),;
+if (config("resource_debug") & 3) {
+    print "isascii(c)";
+    print "isblank(c)";
+}
+

cal/sumsq.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: sumsq.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/sumsq.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:37
  * File existed as early as:	before 1990
  *

cal/sumtimes.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: sumtimes.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/sumtimes.cal,v $
- *
  * Under source code control:	2006/06/22 17:29
  * File existed as early as:	2006
  *

cal/surd.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: surd.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/surd.cal,v $
- *
  * Under source code control:	1990/02/15 01:50:38
  * File existed as early as:	before 1990
  *

cal/test1700.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test1700.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test1700.cal,v $
- *
  * Under source code control:	1994/03/14 23:12:51
  * File existed as early as:	1994
  *

cal/test2300.cal

@@ -17,10 +17,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test2300.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test2300.cal,v $
- *
  * Under source code control:	1995/07/09 06:12:13
  * File existed as early as:	1995
  *

cal/test2600.cal

@@ -1,7 +1,7 @@
 /*
  * test2600 - 2600 series of the regress.cal test suite
  *
- * Copyright (C) 1999  Ernest Bowen and Landon Curt Noll
+ * Copyright (C) 1999,2021  Ernest Bowen and Landon Curt Noll
  *
  * Primary author:  Ernest Bowen
  *
@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.2 $
- * @(#) $Id: test2600.cal,v 30.2 2007/07/11 22:57:23 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test2600.cal,v $
- *
  * Under source code control:	1995/10/13 00:13:14
  * File existed as early as:	1995
  *
@@ -54,7 +50,7 @@
  * of b in power(a, b, eps) is large, the computation required for
  * a test may be very heavy.
  *
- * Test funcations are called as:
+ * Test functions are called as:
  *
  *	testabc(str, ..., verbose)
  *
@@ -91,7 +87,8 @@ define testismult(str, n, verbose)
 		if (!ismult(c,a)) {
 			m++;
 			if (verbose > 1) {
-				printf("*** Failure with:\na = %d\nb = %d\n", a,b);
+				printf("*** Failure with:\na = %d\nb = %d\n",
+				       a,b);
 			}
 		}
 	}
@@ -133,7 +130,8 @@ define testsqrt(str, n, eps, verbose)
 		if (abs(c) > 1) {
 			m++;
 			if (verbose > 1) {
-				printf("*** Failure with:\na = %d\neps = %d\n", a,eps);
+				printf("*** Failure with:\na = %d\neps = %d\n",
+				       a,eps);
 			}
 		}
 	}
@@ -178,7 +176,8 @@ define testexp(str, n, eps, verbose)
 		if (abs(c) > 0.02) {
 			m++;
 			if (verbose > 1) {
-				printf("*** Failure with:\na = %d\neps = %d\n", a,eps);
+				printf("*** Failure with:\na = %d\neps = %d\n",
+				       a,eps);
 			}
 		}
 	}
@@ -235,7 +234,8 @@ define testln(str, n, eps, verbose)
 		if (abs(c) > 0.5) {
 			m++;
 			if (verbose > 1) {
-				printf("*** Failure with:\na = %d\neps = %d\n", a,eps);
+				printf("*** Failure with:\na = %d\neps = %d\n",
+				       a,eps);
 			}
 		}
 	}

cal/test2700.cal

@@ -1,7 +1,7 @@
 /*
  * test2700 - 2700 series of the regress.cal test suite
  *
- * Copyright (C) 1999  Ernest Bowen and Landon Curt Noll
+ * Copyright (C) 1999,2021  Ernest Bowen and Landon Curt Noll
  *
  * Primary author:  Ernest Bowen
  *
@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test2700.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test2700.cal,v $
- *
  * Under source code control:	1995/11/01 22:52:25
  * File existed as early as:	1995
  *
@@ -88,7 +84,7 @@ define mkfrac() = rand(2) ? mkposfrac() : -mkposfrac();
 define mksquarereal() = mknonnegreal()^2;
 
 /*
- * We might be able to do better than the following.  For nonsquare
+ * We might be able to do better than the following.  For non-square
  * positive integer less than 1e6, could use:
  *		 x = rand(1, 1000);
  *		 return rand(x^2 + 1, (x + 1)^2);
@@ -127,7 +123,8 @@ define testcsqrt(str, n, verbose)
 			if (p) {
 			if (verbose > 0)
 				printf(
-				 "*** Type %d failure for x = %r, y = %r, z = %d\n",
+				 "*** Type %d failure for x = %r, "
+				 "y = %r, z = %d\n",
 				    p, x, y, z);
 				m++;
 			}

cal/test3100.cal

@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test3100.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test3100.cal,v $
- *
  * Under source code control:	1995/11/28 11:56:57
  * File existed as early as:	1995
  *

cal/test3300.cal

@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test3300.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test3300.cal,v $
- *
  * Under source code control:	1995/12/02 04:27:41
  * File existed as early as:	1995
  *

cal/test3400.cal

@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test3400.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test3400.cal,v $
- *
  * Under source code control:	1995/12/02 05:20:11
  * File existed as early as:	1995
  *

cal/test3500.cal

@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test3500.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test3500.cal,v $
- *
  * Under source code control:	1995/12/18 22:50:46
  * File existed as early as:	1995
  *

cal/test4000.cal

@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test4000.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test4000.cal,v $
- *
  * Under source code control:	1996/03/13 02:38:45
  * File existed as early as:	1996
  *
@@ -199,7 +195,8 @@ define ctimes(str, N, n, count, skip, verbose)
 		p = ptest(A[i], count, skip);
 		if (p) {
 			if (verbose > 0) {
-				printf("*** Error, what should be rare has occurred for x = %d \n", A[i]);
+				printf("*** Error, what should be rare "
+				       "has occurred for x = %d \n", A[i]);
 				m++;
 			}
 		}
@@ -306,7 +303,8 @@ define ntimes(str, N, n, count, skip, residue, modulus, verbose)
 	}
 	tprev = round(usertime() - t, 4);
 	if (verbose > 0) {
-		printf("%d evaluations, nextcand: %d, prevcand: %d\n", n, tnext, tprev);
+		printf("%d evaluations, nextcand: %d, "
+		       "prevcand: %d\n", n, tnext, tprev);
 	}
 }
 

cal/test4100.cal

@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test4100.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test4100.cal,v $
- *
  * Under source code control:	1996/03/13 03:53:22
  * File existed as early as:	1996
  *

cal/test4600.cal

@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test4600.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test4600.cal,v $
- *
  * Under source code control:	1996/07/02 20:04:40
  * File existed as early as:	1996
  *

cal/test5100.cal

@@ -1,7 +1,7 @@
 /*
  * test5100 - 5100 series of the regress.cal test suite
  *
- * Copyright (C) 1999  Ernest Bowen and Landon Curt Noll
+ * Copyright (C) 1999,2021  Ernest Bowen and Landon Curt Noll
  *
  * Primary author:  Ernest Bowen
  *
@@ -19,10 +19,6 @@
  * received a copy with calc; if not, write to Free Software Foundation, Inc.
  * 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
  *
- * @(#) $Revision: 30.1 $
- * @(#) $Id: test5100.cal,v 30.1 2007/03/16 11:09:54 chongo Exp $
- * @(#) $Source: /usr/local/src/cmd/calc/cal/RCS/test5100.cal,v $
- *
  * Under source code control:	1996/12/02 23:57:10
  * File existed as early as:	1996
  *
@@ -39,7 +35,7 @@ defaultverbose = 1;	/* default verbose value */
  * with zero value, when the definition is read.
  *
  * The variable a5100 is initialized with the value x if and when this
- * function is first called with a positive even x. The varable b5100
+ * function is first called with a positive even x. The variable b5100
  * is similarly initialized if and when this function is first called positive
  * odd x.
  *