exp(1) a multiple expession calculator.


exp [-vn0..4] [-o output-file] [input-file]


Input is an ascii format file including numeric expessions with variables. Input file can includes other input files thanks to an inclusion directive. exp reads the input files to write it as is in the output file, but the numeric expessions are replaced by their value.


verbose mode
0 : quiet mode
1 : messages mess(...) are printed
2 : few statistics
3 : all expressions are decomposed
4 : yacc messages
-o output-file
where exp writes (stdout by default).
where exp reads (stdin by default).


Except the expessionsi and offline comments, the output file is identical to the flattened view of the input files whatever they are. The expession forms can be either [expessions] or {expessions}. The first form is for floating point expessions, the second one is for integer expessions. By default, the print format is respectively %7.3f and %4d (cf. sprintf(3)). Offline comments begins with // and ends at the beginning of the next line.


  • [ 3.0 + 12.1] print 15.100
  • { 3.0 + 12.1} print 15

It is possible to have more that one expession separated by ; (semi-column). All the expessions of a list are computed, but only the last one is printed. If the last expession is empty, exp do not print anything (see Examples section).


  • [ 6. *2.0;

      3.0 + 12.1 ] print 15.100 
  • [ 6. *2.0;

      3.0 + 12.1; ] print nothing 

It is possible to have comments in expessions. A comment begins with # (diese) and ends to the carriage return. Comment in expression is not copied to the output file.


  • [ 3.0 + 12.1 # comment
    ] print 15.100


Expession is a multi-level numeric expession using numbers, variables, arithmetic operators and numeric functions.


  • [ i=3.0; i*12.1] print 15.100

  • {i=3} print 3
    {i++} print 4

Arithmetic Operators

The operators, in order of increasing priority, are

+ -
Addition and subtraction.
* /
Multiplication and division.
++ --
post-incrementation of 1, post-decrementation of 1.
( )
Variable affectation. The return value is the one affected.

Boolean Operators

> <
respectively greater than and lower than
True whenever variable is defined yet
True whenever variable is not defined yet


Expession values may be stored in simple variables. There are three forms of variable name. First, they can begin with a letter followed by any number of letters, digits and underscores; second, they can begin with " (double quote) followed by any number of any char and ended by a "; Third they can begin with ' (quote) then a regular expression (see regex(7)) ended by '. exp is case sensitive.


  • [ VAR_1=3.0; # first form

      "variable numero 2"=12.0; # second form
    ] print nothing
  • [ VAR_1 * "variable numero 2" ] print 15.100

When a variable appears in the right member of an affectation, exp uses its value. If it has never been defined, this causes a fatal error. If it appears the left member, the first time it is automatically created and its value is set, or its value is changed.

Special variables

float_fmt, integer_fmt and string_fmt special variables to redefine print format of floats, integers and strings. Defaults are %7.3f, %-8s and %4d. Those variables are the only variables with a non numerical value.


  • [float_fmt = "%7.1f";] print nothing
  • [ 3.0 + 12.1] print 15.1

verbose is the variable passed in argument list, which can be modified by the program itself.

Examples get details of expression calculation

  • [tmp = verbose; verbose = 3; ..expressions..;
    verbose = tmp;]

Numeric functions

Few numeric functions are available. The form is fun(args). The arguments take the form of a list of expessions, separated with a , (comma). The number of arguments depends on functions. It is possible to make a list of arguments with a regular expession (see regex(7)). Then all matching variable names are part of the list.

min(args) max(args)
The minimum (resp. maximum) value of its arguments.


  • [min(3.0,12.1)] print 3.000
  • [min('RW_ALU.*')] print min value of all variables begining by RW_ALU

inf(step,val) sup(step,val)
Two arguments. inf (resp. sup) function rounds the second argument (val) downwards (resp. upwards) to an integer number of the first argument (step).


  • [step=0.3;value=1.6;inf(step,value)] print 1.500

Special functions and directives

if(condition, expr1, expr2, ...)
Calculates the condition if it is true (means greater than zero), the following expressions are all calculated.
sort(args) rsort(args)
sort (resp. reverse rsort) numerically all its arguments, each argument must be a variable, not directly a numeric expession. The return value is the sorted list of its arguments.
writes its arguments to stdout using float_fmt, one argument per line. The form is : variable_name = value;, value is omitted if the variable has never been defined.
  • [message('"'this is a message'"');] print
    this is a message
  • [a1b=0; a2b=1O; a3b=5; message(a*b);] print
    a1b = 0.000
    a2b = 1O.000
    a3b = 5.000
  • [string_fmt=%6s; message(sort(a*b));] print

       a1b =  0.000 

       a3b =  5.000

       a2b = 1O.000 

#include "filename"
Opens the file in argument then returns to the current file as soon as the new one is empty.


Input file
# this is a test file
[ # few variables
  WITDH = 2;
  LENGTH = 25 ;
this message is unchanged but all expresions are computed
  length_div_2 = [LENGTH/2]
  length_mul_2 = {LENGTH*2}
  result = [max ('leng.*')]
Output file
# this is a test file
this message is unchanged but all expresions are computed
  length_div_2 =   12.500
  length_mul_2 =   50
  result =   12.500


Written by Franck Wajsburt.