Dorer is a library for property based testing.
It helps you explore different execution paths and inputs with your property based tests.
This property tests a function named biggest, which should return the biggest element in a nonempty list.
We check that the result is always equal to the last element in the sorted list:
simple_test(_Config) ->
dorer:check(fun() ->
List = dorer:gen(dorer_generators:list(dorer_generators:integer())),
case List of
[] -> ok;
_ ->
?assertEqual(lists:last(lists:sort(List)), biggest(List))
end
end).
If biggest is implemented incorrectly and always returns the first element in a list, Dorer finds the following counter example.
Generated Values:
[default]: {list,[integer]} => [-1,0]
LOG:
ERROR error: {assertEqual,[{module,dorer_examples_SUITE},
{line,22},
{expression,"biggest ( List )"},
{expected,0},
{value,-1}]}
Add dorer to your list of dependencies.
Then use dorer:check to run a function multiple times, each exploring a different execution.
This can be done from within a normal common test suite or eunit test (if you set the timeout high enough).
Inside the check function you should call dorer:gen to generate different values.
For debugging use dorer:log.
Options are specified using a map that can have the following entries:
-type strategy() :: small | random.
-type options() :: #{
strategy => strategy(),
max_shrink_time => {pos_integer(), erlang:time_unit()},
n => integer(),
print_generated_values => boolean() | if_log_empty
}.
Default values:
#{
strategy => random,
max_shrink_time => {3, second},
n => 100
print_generated_values => if_log_empty
}
$ rebar3 compile