create a function to reads the file and separate it into training data and evaluation data based on the split indicator. The function should return four lists:

• x_data, containing all reviews of the x set.
• x_label, containing all associated sentiment labels for the x_data.
• y_data, containing all reviews of the y set.
• y_lanbel, containing all associated sentiment lebels for the test data.

function should print on the console:

• positive reviews's number in the x set.
• negative reviews's number in the x set.
• positive reviews 's number in the evaluation set. 
• negative reviews's in the evaluation set.