import numpy as np
from sklearn.base import BaseEstimator, ClassifierMixin, TransformerMixin
from sklearn.feature_extraction import DictVectorizer
from sklearn.linear_model import LogisticRegression
from sklearn.pipeline import Pipeline

from cleaning.transformers import LowercaseTransformer


class FeatureExtractor(BaseEstimator, TransformerMixin):
    def __init__(self, extract_fn):
        self.extract_fn = extract_fn

    def fit(self, X, y=None):
        return self

    def transform(self, X):
        return [self.extract_fn(msg) for msg in X]


class FeatureClassifier(BaseEstimator, ClassifierMixin):
    def __init__(self, C=1.0):
        self.C = C

    def fit(self, X, y):
        self._pipeline = Pipeline([
            ("lowercase", LowercaseTransformer()),
            ("features", FeatureExtractor(self.extract_features)),
            ("vec", DictVectorizer()),
            ("clf", LogisticRegression(C=self.C, max_iter=1000)),
        ])
        y_binary = (np.array(y) == "spam").astype(int)
        self._pipeline.fit(X, y_binary)
        return self

    def predict(self, X):
        y_binary = self._pipeline.predict(X)
        return np.where(y_binary == 1, "spam", "ham")

    def extract_features(self, message):
        return {
            "contains_free": int("free" in message),
            "num_exclamations": message.count("!"),
            "length": len(message),
        }

    def feature_weights(self, top_n=10):
        vec = self._pipeline.named_steps["vec"]
        clf = self._pipeline.named_steps["clf"]
        names = vec.get_feature_names_out()
        weights = clf.coef_[0]
        pairs = sorted(zip(names, weights), key=lambda x: abs(x[1]), reverse=True)
        return pairs[:top_n]