import pandas as pd
from pydriller import Repository
import plotly.graph_objects as go
from get_gender import GenderPredictor

class CommitInfo:
    def __init__(self, repo_url):
        self.repo_url = repo_url
        self.df = self.get_commit_info()

    def get_commit_info(self):
        commit_data = []

        for commit in Repository(self.repo_url).traverse_commits():
            commit_info = {
                'Author': commit.author.name,
                'Committer_Date': commit.committer_date,
                'Author_Timezone': commit.author_timezone,
            }
            commit_data.append(commit_info)

        df = pd.DataFrame(commit_data)

        # Assuming you have your DataFrame named df
        # Convert the "Committer_Date" column to pandas datetime with utc=True
        df["Committer_Date"] = pd.to_datetime(df["Committer_Date"], utc=True)

        # Extract the year from the "Committer_Date" column and create a new column "Year"
        df["Year"] = df["Committer_Date"].dt.year

        # Print the updated DataFrame
        return df

    def get_first_commit_dates(self):
        # Group the DataFrame by 'Author' and find the minimum commit date and timezone for each author
        first_commit_dates = self.df.groupby('Author').agg({
            'Committer_Date': 'min',
            'Author_Timezone': 'first'
        }).reset_index()

        # Rename the columns for clarity
        first_commit_dates.columns = ['Author', 'First_Commit_Date', 'Author_Timezone']

        return self.df,first_commit_dates

# Example usage:
if __name__ == "__main__":
    modelpath="saved_model/bestmodel.tf"
    
    gender_predictor = GenderPredictor(modelpath)
    
    repo_url = 'https://github.com/Amstf/DDoS-Attacks-Detection-Using-Adversarial-Neural-Network'
    commit_info = CommitInfo(repo_url)


    # Get and print the DataFrame with first commit dates for each author
    df,first_commit_dates = commit_info.get_first_commit_dates()
    first_commit_dates[['Predicted_Gender', 'Confidence']] = first_commit_dates['Author'].apply(lambda name: pd.Series(gender_predictor.predict_gender(name)))
    merged_df = df.merge(first_commit_dates[["Author","Predicted_Gender","Confidence"]], on=["Author"])
    # Group by Year and Predicted_Gender, then count the occurrences
    gender_counts = merged_df.groupby(['Year', 'Predicted_Gender']).size().reset_index(name='Count')

    counts = first_commit_dates['Predicted_Gender'].value_counts()
    labels = ["Male", "Female", "Unknown"]
    colors = ["blue", "pink", "red"]
    fig = go.Figure(data=[go.Pie(labels=first_commit_dates['Predicted_Gender'].unique(), values=counts, marker=dict(colors=colors))])
    # Convert the chart to HTML and return it
    chart_html = fig.to_html(full_html=False)

    print(first_commit_dates)