From 12304fe245f25cff9886eccbbfe7380ceb39bd50 Mon Sep 17 00:00:00 2001
From: iBug <git@ibugone.com>
Date: Mon, 27 Jul 2020 02:35:59 +0800
Subject: [PATCH] Add GFWList parser script

---
 .gitignore |  1 +
 gfwlist.py | 75 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 76 insertions(+)
 create mode 100755 gfwlist.py

diff --git a/.gitignore b/.gitignore
index 85d347b..941f743 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,4 +1,5 @@
 # Output
+gfwlist.txt
 release-info.md
 *.pac
 dist/
diff --git a/gfwlist.py b/gfwlist.py
new file mode 100755
index 0000000..e1743b3
--- /dev/null
+++ b/gfwlist.py
@@ -0,0 +1,75 @@
+#!/usr/bin/python3
+
+import base64
+import json
+import urllib.parse
+import requests
+
+
+GFWLIST_URL = 'https://raw.githubusercontent.com/gfwlist/gfwlist/master/gfwlist.txt'
+
+
+def get_gfwlist():
+    r = requests.get(GFWLIST_URL)
+    r.raise_for_status()
+    return base64.b64decode(r.text).decode("utf-8").rstrip("\n")
+
+
+def update_domains(domains, host, mode=0):
+    segments = host.strip(".").split(".")[::-1]
+
+    this = domains
+    for segment in segments:
+        if segment not in this:
+            this[segment] = {}
+        this = this[segment]
+    this["@"] = mode
+
+
+def parse_gfwlist(text):
+    domains = {}
+    blackpat = []  # blacklisted patterns
+    whitepat = []  # whitelisted patterns
+
+    for line in text.splitlines():
+        if not line.strip() or line.startswith("!"):
+            continue  # ignore comments and empty lines
+
+        mode = 0  # default to blacklist
+        if line.startswith("@@"):
+            mode = 1  # now it's whitelist
+            line = line[2:]
+
+        if line.startswith("||"):
+            # domain prefix
+            update_domains(domains, line[2:], mode)
+        else:
+            # Keyword pattern
+            # Single vertical line at either side means string boundary
+            if line.startswith("|"):
+                line = line[1:]
+            else:
+                line = "*" + line
+            if line.endswith("|"):
+                line = line[:-1]
+            else:
+                line = line + "*"
+            if mode == 0:
+                blackpat.append(line)
+            else:
+                whitepat.append(line)
+    return domains, blackpat, whitepat
+
+
+def generate_pac_partial():
+    gfwlist = get_gfwlist()
+    domains, blackpat, whitepat = parse_gfwlist(gfwlist)
+    return "var DOMAINS = {};\n\nvar BLACKPAT = {};\n\nvar WHITEPAT = {};\n".format(
+        json.dumps(domains, indent=2),
+        json.dumps(blackpat, indent=2),
+        json.dumps(whitepat, indent=2),
+    )
+
+
+if __name__ == '__main__':
+    print(generate_pac_partial())