From a3f276f502522cf3b81f4d10c1504d372bcfad5f Mon Sep 17 00:00:00 2001
From: Yuval Adam <_@yuv.al>
Date: Mon, 24 Oct 2022 09:06:30 +0300
Subject: Add initial scan script

---
 scan.py | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
 create mode 100644 scan.py

(limited to 'scan.py')

diff --git a/scan.py b/scan.py
new file mode 100644
index 0000000..34c65d8
--- /dev/null
+++ b/scan.py
@@ -0,0 +1,14 @@
+import requests
+
+from pathlib import Path
+
+ROOT_PATH = Path(__file__).parent
+
+DATA_DIR = ROOT_PATH / "data"
+
+TLDS_URL = "https://data.iana.org/TLD/tlds-alpha-by-domain.txt"
+
+with open(DATA_DIR / "tlds.txt", "w") as f:
+    res = requests.get(TLDS_URL)
+    tlds = [tld.lower() for tld in res.text.split("\n")[1:] if len(tld) < 4]
+    f.write("\n".join(tlds).strip())
\ No newline at end of file
-- 
cgit v1.3.1