This is a fast calculator (faster than bc on some big operation!). It deal with arbitrary base, arbitrary symbol arithmetic and arbitrary number size.
It is design for 64 bits architecture yet run pretty fast on 32 bits too.
Run "make" in a terminal to compile.
/!\ Big endian robot or other computing stuff, pass your way, your are not welcome /!\
/!\ This is a Little endian bistro, you might get seriously drunk results here /!\
./calc <base_symbols> <operators> <operation_size> < <operation_file>
Note that every symbol in the operation must be defined in <base_symbols> or <operators>, even space. Otherwise calc throws Syntax error.
()+-*/% Symbols must be defined by typing them in the same order
echo '30-2*(4-1234%17)' | ./calc '0123456789' '()+-*/%' 16
output : 42
echo '10001x[~00011+00101]\01011' | ./calc '01' '[]+~x/\' 26
output : 1
-
./bc_vs_calc <operation>
compute <operation> with classical symbols, compare performance and result with the UNIX calculator bc -
./ultimate_bc_vs_calc :
benchmark test for both calc and bc using a giant random operation -
./simple_calc <operation> [<base_symbols> [<operators>]]:
simply compute <operation> without need to specify anything else if you dont want to.
This program is originaly an exercice for entrance in the school 42 (www.42.fr)
The original compilation flag have been turn off and replace with -O3 for performance.
The team epically failed to debug the program in time,
hopefully after changing two lines and 3 characters you can now enjoy it.
In addition of temporay tables to optimize mult, div and mod, the main trick is to add or substract 8 digits instantly over 64 bits integers (each digit stands in a byte, max base size is 128).
Some binary sweets propagate the carry instantly over the 8 digits.
If you are interested in more documentation over the code, send me your motivation at math.guin@gmail.com
PS : I will not help student to cheat, I am far too expensive