Script en phyton para extraer archivos adjuntos de archivos mbox. Mbox attachment extractor.

Readme.txt

Idea original
https://gist.github.com/georgy7/3a80bce2cd8bf2f9985c

Descarga/Download Windows 10 . Portable.
https://mega.nz/file/40k0CJqT#xoZsG-D3fm0iDf0ZrsJBSJEGQOaYYkyVk0zRbsxR7sg

Ayuda con Grok.

El script genera una interface para seleccionar el archivo mbox y la ruta destino.
This script extract attachment of mbox archive. Easy graphical interface.

Se puede generar un archivo ejecutable con el siguiente comando.

pyinstaller --onefile --windowed mbox_extractor_gui.py

mbox_extractor_gui.py

import mailbox
import email
from email import policy
import os
import tkinter as tk
from tkinter import filedialog, messagebox, ttk
import shutil

class MboxExtractorApp:
    def __init__(self, root):
        self.root = root
        self.root.title("Extract Mbox Attachments")
        self.root.geometry("600x400")
        
        # Variables
        self.mbox_path = tk.StringVar()
        self.output_dir = tk.StringVar()
        
        # Interfaz
        tk.Label(root, text="Mbox Attachment Extractor", font=("Arial", 16)).pack(pady=10)
        
        # Selección de archivo mbox
        tk.Label(root, text="Mbox File:").pack()
        tk.Entry(root, textvariable=self.mbox_path, width=50).pack(pady=5)
        tk.Button(root, text="Browse Mbox", command=self.browse_mbox).pack(pady=5)
        
        # Selección de carpeta de salida
        tk.Label(root, text="Output Directory:").pack()
        tk.Entry(root, textvariable=self.output_dir, width=50).pack(pady=5)
        tk.Button(root, text="Browse Output Folder", command=self.browse_output).pack(pady=5)
        
        # Botón de extracción
        tk.Button(root, text="Extract Attachments", command=self.extract_attachments).pack(pady=20)
        
        # Barra de progreso
        self.progress = ttk.Progressbar(root, length=400, mode='determinate')
        self.progress.pack(pady=10)
        
        # Área de texto para resultados
        self.result_text = tk.Text(root, height=8, width=60)
        self.result_text.pack(pady=10)
        
    def browse_mbox(self):
        file_path = filedialog.askopenfilename(filetypes=[("Mbox files", "*.mbox"), ("All files", "*.*")])
        if file_path:
            self.mbox_path.set(file_path)
    
    def browse_output(self):
        folder_path = filedialog.askdirectory()
        if folder_path:
            self.output_dir.set(folder_path)
    
    def extract_attachments(self):
        mbox_file = self.mbox_path.get()
        output_dir = self.output_dir.get()
        
        if not mbox_file or not output_dir:
            messagebox.showerror("Error", "Please select both an mbox file and an output directory.")
            return
        
        if not os.path.exists(mbox_file):
            messagebox.showerror("Error", "The selected mbox file does not exist.")
            return
        
        try:
            os.makedirs(output_dir, exist_ok=True)
            mbox = mailbox.mbox(mbox_file)
            total_messages = len(mbox)
            self.progress['maximum'] = total_messages
            file_counter = 1
            
            self.result_text.delete(1.0, tk.END)
            self.result_text.insert(tk.END, "Starting extraction...\n")
            
            for i, message in enumerate(mbox):
                msg = email.message_from_bytes(message.as_bytes(), policy=policy.default)
                
                for part in msg.walk():
                    if part.get_content_disposition() == 'attachment':
                        filename = part.get_filename()
                        if filename:
                            base, ext = os.path.splitext(filename)
                            filepath = os.path.join(output_dir, f"{base}_{file_counter}{ext}")
                            file_counter += 1
                            try:
                                with open(filepath, 'wb') as f:
                                    f.write(part.get_payload(decode=True))
                                self.result_text.insert(tk.END, f"Saved: {filepath}\n")
                            except Exception as e:
                                self.result_text.insert(tk.END, f"Error saving {filename}: {str(e)}\n")
                
                self.progress['value'] = i + 1
                self.root.update()
            
            self.result_text.insert(tk.END, "Extraction completed.\n")
            messagebox.showinfo("Success", "Attachment extraction completed!")
        
        except Exception as e:
            messagebox.showerror("Error", f"An error occurred: {str(e)}")
        
        self.progress['value'] = 0

if __name__ == "__main__":
    root = tk.Tk()
    app = MboxExtractorApp(root)
    root.mainloop()

Idea original
https://gist.github.com/georgy7/3a80bce2cd8bf2f9985c

Descarga Win 10. Portable.
Download