[Ver3.0 Supported GUI] FRESH HTTP(s)Proxy collection program by Python

# This program does not support languages other than Japanese. Please do not be offended.
# You can run this program to get a list of up to about 300,000 or more proxies.

import tkinter as tk
from tkinter import ttk, messagebox
from datetime import datetime, timedelta
import requests
import threading
import re
import json
import os

ALL_PROXIES = set()

def get_past_dates(days=31):
    today = datetime.utcnow()
    return [(today - timedelta(days=i)).strftime("%Y-%m-%d") for i in range(days)]

def fetch_checkerproxy(date):
    try:
        url = f"https://api.checkerproxy.net/v1/landing/archive/{date}"
        res = requests.get(url, timeout=10)
        if res.status_code == 200:
            data = res.json()
            proxies = data['data'].get('proxyList', [])
            return proxies, True
        else:
            return [], False
    except:
        return [], False

def extract_from_text_url(url):
    try:
        res = requests.get(url, timeout=10)
        return re.findall(r'\b(?:\d{1,3}\.){3}\d{1,3}:\d+\b', res.text)
    except:
        return []

def extract_from_json_url(url, ip_key="IP", port_key="PORT"):
    try:
        res = requests.get(url, timeout=10)
        data = res.json()
        proxies = []
        for entry in data.get("LISTA", []):
            ip = entry.get(ip_key)
            port = entry.get(port_key)
            if ip and port:
                proxies.append(f"{ip}:{port}")
        return proxies
    except:
        return []

def collect_proxies(log_widget, progress_bar, progress_label):
    global ALL_PROXIES
    ALL_PROXIES = set()
    now = datetime.now().strftime("%Y-%m-%d_%H-%M")
    folder = "proxies_tool_result"
    if not os.path.exists(folder):
        os.makedirs(folder)

    filename_txt = os.path.join(folder, f"httpProxies_{now}.txt")
    filename_json = os.path.join(folder, f"httpProxies_{now}.json")

    dates = get_past_dates(31)
    skipped_dates = []

    log_widget.insert(tk.END, "=== CheckerProxyの取得を開始します。 ===\n")
    for i, date in enumerate(dates):
        proxies, success = fetch_checkerproxy(date)
        if success and proxies:
            ALL_PROXIES.update(proxies)
            log_widget.insert(tk.END, f"{date} の取得完了。 ({len(proxies)}件)\n")
        else:
            skipped_dates.append(date)
            log_widget.insert(tk.END, f"{date} は取得できませんでした。\n")
        progress_bar["value"] = (i + 1) / len(dates) * 50
        progress_label.config(text=f"{int(progress_bar['value'])}%")
        log_widget.yview_moveto(1)
        log_widget.update()

    log_widget.insert(tk.END, "\n=== 他ソースからの取得を開始します。 ===\n")

    proxy_sources = [
        "https://raw.githubusercontent.com/roosterkid/openproxylist/main/HTTPS.txt",
        "https://raw.githubusercontent.com/TheSpeedX/PROXY-List/master/http.txt",
        "https://raw.githubusercontent.com/mmpx12/proxy-list/master/http.txt",
        "https://api.openproxylist.xyz/http.txt",
        "https://raw.githubusercontent.com/sunny9577/proxy-scraper/master/proxies.txt",
        "https://raw.githubusercontent.com/jetkai/proxy-list/main/online-proxies/txt/proxies-http.txt",
        "https://raw.githubusercontent.com/shiftytr/proxy-list/master/proxy.txt",
        "https://raw.githubusercontent.com/ShiftyTR/Proxy-List/master/http.txt",
        "https://raw.githubusercontent.com/monosans/proxy-list/main/proxies_anonymous/http.txt",
        "https://rootjazz.com/proxies/proxies.txt",
        "https://proxyspace.pro/http.txt",
        "https://alexa.lr2b.com/proxylist.txt",
        "https://raw.githubusercontent.com/clarketm/proxy-list/master/proxy-list-raw.txt"

    ]

    offset = 50 / len(proxy_sources)
    for idx, url in enumerate(proxy_sources):
        log_widget.insert(tk.END, f"{url} から取得中...\n")
        proxies = extract_from_text_url(url)
        ALL_PROXIES.update(proxies)
        progress_bar["value"] += offset
        progress_label.config(text=f"{int(progress_bar['value'])}%")
        log_widget.yview_moveto(1)
        log_widget.update()

    log_widget.insert(tk.END, f"\n=== 合計取得件数: {len(ALL_PROXIES)} 件 ===\n")

    with open(filename_txt, "w") as f:
        f.write("\n".join(sorted(ALL_PROXIES)))

    with open(filename_json, "w") as f:
        json.dump(sorted(ALL_PROXIES), f, indent=2)

    log_widget.insert(tk.END, f"\n保存完了: {filename_txt}, {filename_json}\n")
    log_widget.yview_moveto(1)
    log_widget.update()

    stop_btn.pack(pady=5)
    stop_btn.config(state="normal")

def stop_application():
    root.quit()

def start_collection():
    threading.Thread(target=collect_proxies, args=(log_text, progress, progress_label), daemon=True).start()

root = tk.Tk()
root.title("proxies_collect_tool")
root.geometry("640x480")

label = tk.Label(root, text="HTTP(s) Proxy収集ツール", font=("Meiryo", 14, "bold"))
label.pack(pady=10)

progress = ttk.Progressbar(root, length=500, mode='determinate')
progress.pack(pady=5)

progress_label = tk.Label(root, text="0%", font=("Meiryo", 10))
progress_label.place(relx=0.5, rely=0.3, anchor="center")

btn = tk.Button(root, text="Go!", command=start_collection, font=("Meiryo", 12))
btn.pack(pady=5)

log_text = tk.Text(root, height=20, wrap=tk.WORD)
log_text.pack(fill=tk.BOTH, padx=10, pady=10, expand=True)

stop_btn = tk.Button(root, text="終了", command=stop_application, font=("Meiryo", 12), state="disabled")

root.mainloop()