Script Linguistica Computazionale

# --------- pulizia_gui_candidati(v.2).py  --------

import re

import os

import tkinter as tk

from tkinter import messagebox, Listbox, Scrollbar

from tkinterdnd2 import DND_FILES, TkinterDnD

# ---------- PULIZIA ----------

def pulisci_testo_per_speaker(testo, speaker_target):

    righe = testo.split('\n')

    risultato = []

    in_blocco_giusto = False

    speaker_target = speaker_target.lower()

    for riga in righe:

        riga = riga.strip()

        if not riga:

            continue

        # Riconoscere intestazioni tipo "Donald Trump 00:00:00..."

        match_speaker = re.match(r'^([A-Z][a-zA-Z.]*\s+[A-Z][a-zA-Z.]+)\s\d{2}:\d{2}:\d{2}', riga)

        if match_speaker:

            speaker_name = match_speaker.group(1).lower()

            # Tolleranza: se contiene la parola "trump" o "harris"

            if speaker_target in speaker_name:

                in_blocco_giusto = True

            else:

                in_blocco_giusto = False

            continue

        # Salta metadati o rifiuti

        if re.match(r"^(No Signal|No StressLens|Weak|Note|Audience|Unidentified|[A-Z][a-z]+(?:\s[A-Z][a-z]+)*:)", riga):

            continue

        # Rimuove [tag tra parentesi quadre]

        riga = re.sub(r"\[.*?\]", "", riga).strip()

        if re.match(r"^(Strong|Medium|Weak)\s\(\d+\.\d+\)", riga):

            continue

        if re.match(r"^\d+\s+of\s+\d+\s+paragraphs\s+scored\.$", riga):

            continue

        if in_blocco_giusto and riga:

            risultato.append(riga)

    return "\n".join(risultato)

def pulisci_file(file_path, speaker_target, output_dir):

    try:

        with open(file_path, 'r', encoding='utf-8') as f:

            testo = f.read()

        pulito = pulisci_testo_per_speaker(testo, speaker_target)

        numero = re.search(r'\d+', os.path.basename(file_path))

        numero = numero.group() if numero else "000"

        output_path = os.path.join(output_dir, f"cleaned_speech{numero}.txt")

        with open(output_path, 'w', encoding='utf-8') as out:

            out.write(pulito)

        return output_path

    except Exception as e:

        return f"Errore con {os.path.basename(file_path)}: {e}"

# ---------- GUI APP ----------

class SpeechCleanerApp:

    def __init__(self, master):

        self.master = master

        master.title("Speech Cleaner - Trump / Harris")

        master.geometry("600x450")

        self.label = tk.Label(master, text="Trascina qui i file .txt da pulire", font=("Arial", 14))

        self.label.pack(pady=10)

        self.listbox = Listbox(master, selectmode=tk.MULTIPLE, width=80)

        self.listbox.pack(padx=10, pady=10, expand=True)

        scrollbar = Scrollbar(master, orient="vertical")

        scrollbar.config(command=self.listbox.yview)

        scrollbar.pack(side="right", fill="y")

        self.listbox.config(yscrollcommand=scrollbar.set)

        # Bottoni pulizia

        self.button_trump = tk.Button(master, text="Pulisci Trump", bg="lightblue", command=lambda: self.pulisci_tutti("Trump"))

        self.button_trump.pack(pady=5)

        self.button_harris = tk.Button(master, text="Pulisci Harris", bg="lightgreen", command=lambda: self.pulisci_tutti("Harris"))

        self.button_harris.pack(pady=5)

        master.drop_target_register(DND_FILES)

        master.dnd_bind('<<Drop>>', self.aggiungi_file)

        self.files = []

    def aggiungi_file(self, event):

        paths = master.tk.splitlist(event.data)

        for path in paths:

            if path.endswith('.txt') and path not in self.files:

                self.files.append(path)

                self.listbox.insert(tk.END, os.path.basename(path))

    def pulisci_tutti(self, speaker_target):

        if not self.files:

            messagebox.showwarning("Nessun file", "Trascina almeno un file .txt")

            return

        output_dir = os.path.join(os.getcwd(), speaker_target)

        os.makedirs(output_dir, exist_ok=True)

        successi = []

        errori = []

        for file_path in self.files:

            risultato = pulisci_file(file_path, speaker_target, output_dir)

            if risultato.startswith("Errore"):

                errori.append(risultato)

            else:

                successi.append(os.path.basename(risultato))

        msg = ""

        if successi:

            msg += f"\nPuliti: {len(successi)} file\n" + "\n".join(successi)

        if errori:

            msg += f"\nErrori:\n" + "\n".join(errori)

        messagebox.showinfo("Operazione completata", msg)

# ---------- AVVIO ----------

if __name__ == "__main__":

    master = TkinterDnD.Tk()

    app = SpeechCleanerApp(master)

    master.mainloop()