The repro is only my learners repro - I try hard to get a grip on the features.
The repro contains some stupid stuff about graph coloring. This is to try the basics.
Furthermore, there is a tiny calulator using parallel processes and messages.
$ rebar3 compile
$ rebar3 eunit
It implements parallel processing of terms expressed as reverse Łukasiewicz notation (RPN).
Each expression is notated as erlang list, e.g.: [1,2,add,2,mul].
On the rebar3 shell it works like this:
1> calculator:do([[1,2,add,2,mul]]).
....
[[6]]
2>calculator:do([[30,50,10,10,200,sum],[1,4,7,add,mul,17,add]]).
...
[[300],[28]
3>calculator:do([[30,50,10,10,200,sum],[1,4,7,add,mul,17,add],[0],[17,3,mul,fact]]).
...
...
[[0],
[1551118753287382280224243016469303211063259720016986112000000000000],
[300],
[28]]
4>
Currently only add with two arguments, mul with two arguments, sum with all available arguments on the stack, finally, fact for factorial.
This is how it works:
- with
calculator:do([Terms])all required processes are startet:- the operator units (add, mul, sum, fact),
- a supervisor, which collects the results,
- for each term a single calculation units, which handles the rpn, and calls the operator units accordingly
- if a calculation unit is finished, informs the supervisor and ends by itself.
- when all results are pathed to the supervisor, it tears down the opertor units.
The tiny experiments implements an algorithmus to color some graph. The world example is only an abstract Europe.
On the shell it looks like:
4> colorworld:color_world().
{[{"de",1},
{"fr",2},
{"oe",2},
{"it",1},
{"bw",1},
{"hu",3},
{"sl",4},
{"bl",3},
{"po",2},
{"ju",2},
{"v0",1},
{"ch",3},
{"tc",3},
{"lx",4},
{"sp",3},
{"pr",2},
{"nl",2},
{"gr",3},
{"dk",2}],
19,4}
5>
The result is a tuple containing:
{ color map, number of countries and numbers of colors },