from sklearn.metrics import precision_score, accuracy_score, recall_score, f1_score, classification_report
from sklearn.preprocessing import LabelEncoder

class TextClassifier:

    def __init__(self, train_features, train_targets, test_features, test_targets):
        self.train_features = train_features
        self.train_targets = train_targets
        self.test_features = test_features
        self.test_targets = test_targets

        self.model = None
        self.classification_report = None
        self.accuracy = None
        self.precision = None
        self.recall = None
        self.f1 = None

    def train(self) -> None:
        raise NotImplementedError

    def predict(self, text_samples:list, inverse_transform:bool=True) -> list:
        raise NotImplementedError

    def evaluate(self) -> dict:

        predictions = self.predict(self.test_features, inverse_transform=False)

        self.accuracy = accuracy_score(self.test_targets, predictions)
        self.precision = precision_score(self.test_targets, predictions, average='weighted')
        self.recall = recall_score(self.test_targets, predictions, average='weighted')
        self.f1 = f1_score(self.test_targets, predictions, average='weighted')
        self.classification_report = classification_report(self.test_targets, predictions)

        return {'accuracy' : self.accuracy,
                'precision' : self.precision,
                'recall' : self.recall}