calc/help/script

Calc shell scripts
------------------

    There are several ways calc may be used in shell scripts.  The
    syntax for these varies widely for different shells and systems,
    but common to most are commands like echo, if, for, goto, shift,
    and exit, as well as the accessing of environment parameters, shell
    variables, and command-line arguments.

    As a simple example, assuming a C or Bourne shell, let add be a
    file containing just one line:

	    calc -q -- $1 + $2

    Then:

	    ./add 1.23 4.56

    should respond with the display of:

	    5.9

    The "-q" was included in the command to avoid reading of any
    start-up calc files which could contain commands not wanted
    here.  The "--" indicates that there are no more options;
    without it, if $1 began with '-', calc would interpret it as
    the first character of another option.  To execute the file,
    the strings "1.23" and "4.56" were assigned to $1 and $2, so
    calc was in effect asked to evaluate the string "1.23 + 4.56".

    By making add executable by a command like:

	    chmod u+x add

    the command used here may be simplified to:

	    ./add 1.23 4.56

    Here we shall assume that any script we refer to has been made
    executable in this way.

    Because $1 and $2, and instructions in the script, are to read
    by calc as expressions or commands, they may be much more
    complicated than in the above example, but if they involve
    characters with special interpretations by the shell (spaces
    for word separation, * or ? or [ ...] for file-name expansion,
    ! (without immediately following space) for history expansion,
    ( ... ) for shell-function arguments, { ... } for brace
    expansion, $ for parameter or variable expansion, <, <<, >, >>
    for redirection of input or output, etc.) it will usually be
    necessary to quote or escape th characters, or usually more
    conveniently, quote whole expressions with single or double
    quotes.

    For example, the add script should have no problem with
    commands like:

	    ./add "sqrt(2)" "3 * 4"

	    ./add "mat A[2,2] = {1,2,3,4}" "A^2"

	    ./add "2 + 3i" "(3 + 4i)^2"

    If the shell arguments are to be integers, one could use
    scripts like the following with arithmetic expansion
    for the bash and ksh:

	    declare -i a=$1
	    declare -i b=$2
	    calc -q -- $a + $b

    and for csh:

	    @ a = $1
	    @ b = $2
	    calc -q -- $a + $b

    Specifying the shell for a script may be done by including
    in the script a first line with the "magic number" "#!" and
    the full file path for the shell as in:

	    #!/bin/bash
	    declare -i a=$1
	    declare -i b=$2
	    calc -q -- $a + $b

    For a script to multiply rather than add two expressions, one
    could have a file mul with the one line:

	    calc -q -- $1 \* $2
    or:
	    calc -q -- "$1 * $2"

    which will work so long as $1 and $2 are literal numbers, but
    will not work for:

	    ./mul 2+3 4
    or:
	    ./mul "2 + 3" 4

    both of which calc interprets as evaluating 2 + 3 * 4.  What should
    work for most shells is:

	    calc -q -- "($1) * ($2)"

    For adding an arbitrary number of expressions that evaluate to
    rational numbers expressible with at most 20 decimal places,
    simple shell script could be used:

	    s=0
	    for i do
		    s=`calc -q -- $s + $i`
	    done
	    echo sum = $s

    This is not particularly efficient since it calls calc once for
    each argument.  Also, a more serious script would permit more
    general numbers.

    Another way of handling a sum of several expressions is with
    the script addall2 with a here document:

	    calc "-q -s" $* << +
	    global i, n, s;
	    n = argv();
	    for (i = 1; i < n; i++)
		    s += eval(argv(i));
	    print "sum =", s;
	    +

    In executing the command:

	    ./addall2 2 3 4

    the $* in this script expands to  2 3 4, and because of the "-s"
    in the options, calc starts with argv(1) = "2", argv(2) = "3",
    argv(3)= "4".  As there is only one calc process involved and
    the eval() function accepts as argument any string that
    represents the body of a calc function, the strings argv(1),
    argv(2), ... could evaluate to any value types for which the
    additions to be performed are defined, and variables defined in
    one argv() can be used in later arguments.

    In case you are wondering, argv(0) returns the program
    or calc script name.  In the case of the above example,
    argv(0) = "./addall2".

    For systems that support interpreter files, essentially the
    same thing may be done more efficiently by using calc as an
    interpreter.  Assuming the full path for calc is
    /usr/local/bin/calc, one could use the file addall3 with contents

	    #!/usr/bin/calc -q -f
	    global i, n, s;
	    n = argv();
	    for (i = 1; i < n; i++)
		    s += eval(argv(i));
	    print "sum =", s;

	IMPORTANT NOTE:

	    The -f flag must be at the very end of the #! line.
	    The #! line must be the first line of the executable file.
	    The path after the #! must be the full path to the calc executable.

    After the command:

	    ./addall3 2 3 4

    the arguments calc receives are argv(0) = "./addall3", argv(1) =
    "2", argv(3) = "3", argv(4) = "4".

    Another kind of script that can be useful is sqrts1:

	    calc -q 'global s; while (scanf("%s", s) == 1) print sqrt(eval(s));'

    or what is essentially an interpreter equivalent sqrts2:

	    #!/usr/local/bin/calc -q -f
	    global s;
	    while (scanf('%s', s) == 1)
		    print sqrt(eval(s));

    If sqrts is either of these scripts, the command:

	    echo 27 2+3i | sqrts

    or, if datafile contains the one line:

	    27 2+3i

    or the two lines:

	    27
	    2+3i

    either:

	    cat datafile | ./sqrts
    or:
	    ./sqrts < datafile

    should display the square-roots of 27 and 2+3i.  The output could
    be piped to another command by | or directed to a file by use of
    ; or >>.

    With no specified input, either sqrts1 or sqrts2 will wait
    without any prompt for input from the keyboard and as each line
    is completed display the square-roots of the expressions
    entered.  Exit can be achieved by entering exit or entering
    ctrl-D (interpreted as EOF) rather than a line of input.

    One advantage of an interpreter file like sqrts2 (which has only
    options, but neither "-s" nor "--" in its first line) is that it
    can be invoked with further options as in

	    echo 2 3 4 | ./sqrts2 -i -D 32

    An advantage of non-interpreter files is that they can use shell
    features.  For example, for unquoted arguments or arguments in
    double quotes parameter expansion (indicated by unquoted '$') and
    command substitution (using backquotes) occur before lines are
    compiled by calc.  For example, if doit is an executable
    script with contents

	    calc -q -- "$1($2)"

    it may be used as in:

	    ./doit sqrt 7
    and:
	    ./doit exp 7

    to display the values of sqrt(7) and exp(7).  The "--" prevents a
    leading '-' in the $1 argument as indicating one or more additional
    options. E.g., without the "--" in doit,

	    ./doit -sqrt 7

    would be interpreted as:

	    calc -q "-sqrt(7)"

    in which the dash in the quoted part would be taken as indicating a
    list of options -s, -q, -r, etc.; this would give an "illegal option"
    error as calc has no -r option.

    In invoking the doit script it is not necessary that $1 expand to a
    calc function name and $2 to an expression; all that is required is
    that:

	    $1($2)

    expands to a string that calc will recognize as a command.  E.g.:

	    ./doit "define f(x) = x^2; 2 + mod" "f(7), 6"

    does the same as:

	    calc -q -- "define f(x) = x^2; 2 + mod(f(7), 6)"

    Essentially the same is achieved by the contents of doit is changed to:

	    calc -q -p -- << +
	    $1($2)
	    +

    The "-p" stops calc going interactive; without it the effect would be
    be the same as that of a script with the one line:

	    calc -q -i -- "$1($2)"

For more information use the following calc commands:

    help usage
    help argv
    help config
    help cscript

## Copyright (C) 2000,2014,2021  Landon Curt Noll and 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
## 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:	1999/11/30 05:29:48
## File existed as early as:	1999
##
## chongo <was here> /\oo/\	http://www.isthe.com/chongo/
## Share and enjoy!  :-)	http://www.isthe.com/chongo/tech/comp/calc/