issue2085_pydecimal_wrong_distribution_for_min_max_values_of_dif_size

[AlexLitvino]

fixed issue 2085: if min and max values for pydecimal have different signs, faker returned only min_value for negatives

What does this change

• Changed calculation of left_number when '-' sign is selected and min value is specified

                left_number = str(self.random_int(int(max(max_value or 0, 0)), int(abs(min_value))))
->
                left_number = str(self.random_int(int(abs(min(max_value or 0, 0))), int(abs(min_value))))

• Added tests to check that all values could be selected from ranges: (-min ... -max), (-min ... +max), (+min ... +max)

What was wrong

When specified equal min and max values with different signs (-min and +max), then for selected '-' sign, left_number always calculated to -min value
left_number = str(self.random_int(int(max(max_value or 0, 0)), int(abs(min_value)))) # min_value is substituted with max_value as they equal by absolute value
->
left_number = str(self.random_int(int(max(max_value or 0, 0)), int(max_value)))
->
left_number = str(self.random_int(int(max_value), int(max_value)) -> max_value

How this fixes it

IF '-' sign is selected, AND min_value is specified, AND max_value is positive, max_value will be calculated as 0.

Fixes #2085

[fcurella]

Thank you!