import streamlit as st
import pandas as pd
import xml.etree.ElementTree as ET
import csv
from fuzzywuzzy import fuzz
from fuzzywuzzy import process
from tqdm import tqdm

# Funzione per calcolare il punteggio di similarità tra due stringhe
def similarity(string1, string2):
    return fuzz.ratio(string1.lower(), string2.lower())


# Function to merge and enrich datasets
def merge_and_enrich_datasets(scholar_df, semantics_df, dblp_df):
    # Unisci i record in base al titolo con similarità superiore all'80%
    merged_records = []
    # Calcola il totale dei record da elaborare per la barra di avanzamento
    total_records = len(dblp_df)
    # with open('outputMerged.csv', mode='a', newline='', encoding='utf-8') as file:
        # Utilizza tqdm per mostrare la barra di avanzamento
    with tqdm(total=total_records, desc="Unione dei record") as pbar:
        for _, dblp_row in dblp_df.iterrows():
        
            dblp_title = dblp_row['Title']

            # Trova record simili nei DataFrames Scholar e Semantics
            scholar_df['Title'] = scholar_df['Title'].astype(str) 
            semantics_df['Title'] = semantics_df['Title'].astype(str) 

            scholar_matches = scholar_df[scholar_df['Title'].apply(lambda x: similarity(str(dblp_title), str(x))) > 80]
            semantics_matches = semantics_df[semantics_df['Title'].apply(lambda x: similarity(str(dblp_title), str(x))) > 80]
            try:
                merged_record = dblp_row.copy()
                if not scholar_matches.empty:
                    # Aggiungi solo le colonne 'doi' e 'citazioni' da scholar_matches
                    merged_record['doi_scholar'] = scholar_matches.iloc[0]['DOI']
                    merged_record['cites_scholar'] = scholar_matches.iloc[0]['Cites']
                    merged_record['abstract_scholar'] = scholar_matches.iloc[0]['Abstract']
                if not semantics_matches.empty:
                    # Aggiungi solo le colonne 'doi' e 'citazioni' da semantics_matches
                    merged_record['doi_semantic'] = semantics_matches.iloc[0]['DOI']
                    merged_record['cites_semantic'] = semantics_matches.iloc[0]['Cites']
                    merged_record['abstract_scholar'] = 'None'
                    
                merged_records.append(merged_record)
                # Scrivi il record nel file CSV
                # file.write(merged_record.to_csv(index=False, sep=';', encoding='utf-8') + '\n')
            except Exception as e:
                print(f"Errore durante l'unione dei record: {str(e)}")
                continue
            # Aggiorna la barra di avanzamento
            pbar.update(1)
            

    # Crea un DataFrame finale con i record uniti
    final_df = pd.DataFrame(merged_records)
   
    final_df.to_csv('output/scholar_dblp_semantics.csv', index=False, sep=';' , encoding='utf-8')

    return final_df
    
def main():
    # Interfaccia utente Streamlit
    st.title("Knowledge enrichment")

    if st.button("create enriched dataset"):
        # Carica i file CSV in DataFrame
        scholar_df = pd.read_csv("output/googleScholarcsv.csv", sep=";", encoding='utf-8')
        semantics_df = pd.read_csv("output/semanticscholarcsv.csv", sep=";", encoding='utf-8')
        dblp_df = pd.read_csv("output/crawler_doi_citation.csv", sep=";", encoding='utf-8')

        # Call the merge_and_enrich_datasets function
        enriched_dataset = merge_and_enrich_datasets(scholar_df, semantics_df, dblp_df)

        # Display the enriched dataset
        st.write("Enriched Dataset:")
        st.dataframe(enriched_dataset)