AVFP

This is an implementation of the FP programming language written by Andy
Valencia (vandys@lindy.stanford.edu or br.ajv@rlg.BITNET).  He wrote it in C
and the source code is public domain.  It is a full implementation of John
Backus' FP language, but only for numeric data.  For the sake of discussions 
below, we will refer to this implementation as AVFP (Andy Valencia's FP).

To use, just type

     % fp

as long as /usr/local/bin is in your path.

FP doesn't really give a prompt.  It just indents about 7 spaces and waits for
your input.  FP's response is printed in column 1.

To quit, do

     )quit

or just press CONTROL-D.

The rest of this document describes the differences between the classical
syntax and Andy Valencia's implementation.


1.  Simple selection and structure primitives:

    cons is not defined in AVFP, but cons is superfluous anyway since apndl
    (append on the left) is defined.

    -i works as a component selector of a list, taking the ith element from
    the end, rather than from the beginning:

        2:<5,6,8,4,7>    gives 6
        -2:<5,6,8,4,7>   gives 4

    Any erroneous use of a function returns ?

        5:<1,2,3>        (there is no fifth element to the list)    ?

    The commas between elements of a list are unnecessary but they do not 
    harm anything, either.  Thus <1 2 3> is exactly the same list as <1,2,3>.

    <> is the empty list.  Nil cannot be used to stand for it.

2.  Arithmetic

    / is division and mod is modulus.  If division would give a fractional
    amount then the result is a real number.


3.  Boolean

     =          equality
     eq         equality
     ~=         not equal
     <
     >
     >=
     <=

 
4.  Complex Selection and structure


4.5  Other built-ins

        tl
        div
        hd
        eq
        atom
        pick
        not
        nil
        reverse
        distl
        distr
        length
        trans
        apndl
        apndr
        tlr
        front
        rotl
        rotr
        iota     -- input is a single integer
                    output is a list of ints from 1 up to that integer
                    example:    iota:7   result=<1 2 3 4 5 6 7>
        pair
        split
        concat
        last
        first
        out
        sin
        cos
        tan
        asin
        acos
        atan
        log
        exp
        mod
        or
        and
        xor
        id

    

X.  Functional Forms

    & stands for "apply to all"

       &tl:<<5,3,2>,<8,9,10,11>,<6,1>>   gives   <<3,2>,<9,10,11>,<1>>

    % stands for "constant"

        %5:3   gives   5   (no matter what the argument is)

    @ stands for "function composition" (perhaps think of "AND")

        iota@+:<2,3>    gives    <1 2 3 4 5>

    [ ] stands for construct (as in the book)

        [id %0]:8       gives    <8 0>

    | stands for insertion (instead of /)
    ! can be used synonymously for |

        |+:<5,3,2>      gives    10, since 5+3+2=10

    ( p -> g ; h) stands for condition (as in the book)

        (< -> + ; -):<4,6> gives 10

    (while p f) stands for repetition (as in the book)

        (while >@[length %0] [tl |+]):<3,5,6>

    {name ....} stands for definition

        {junk id} defines a new function called "junk" whose action is "id":

        junk:5   gives  5

        {add1 +@[id,%1]}
        add1:7             gives 8


XX. External files and commands

     )quit          -- leave AVFP
     )help          -- print out a simple help menu
     )load filename -- get input from a file

=====

This file was found on the web at 
http://www-cs.canisius.edu/PL_TUTORIALS/FP/AVFP/howtouse
  This copy was made on 8 March 1999
  The original does not have "last modified" information, although
    it appears that it is dated 19 Feb 1995