From be438420d59612e2501d8d0f9d6f139002417384 Mon Sep 17 00:00:00 2001
From: agatha <agatha@juggalol.com>
Date: Sun, 15 Mar 2026 16:22:42 -0400
Subject: [PATCH] feat: add protocol-prefix parsers for socks5://ip:port format

---
 .../builtin/parsers/protocol_prefix.py        | 54 +++++++++++++++++++
 1 file changed, 54 insertions(+)
 create mode 100644 src/proxy_pool/plugins/builtin/parsers/protocol_prefix.py

diff --git a/src/proxy_pool/plugins/builtin/parsers/protocol_prefix.py b/src/proxy_pool/plugins/builtin/parsers/protocol_prefix.py
new file mode 100644
index 0000000..563a293
--- /dev/null
+++ b/src/proxy_pool/plugins/builtin/parsers/protocol_prefix.py
@@ -0,0 +1,54 @@
+from __future__ import annotations
+
+import re
+from typing import Any
+
+from proxy_pool.config import Settings
+from proxy_pool.plugins.protocols import DiscoveredProxy
+
+_PROTOCOL_LINE_RE = re.compile(
+    r"^(https?|socks[45])://(\d{1,3}(?:\.\d{1,3}){3}):(\d{2,5})$"
+)
+
+SUPPORTED_PROTOCOLS = {"http", "https", "socks4", "socks5"}
+
+
+class ProtocolPrefixParser:
+    name = "protocol_prefix"
+
+    def supports(self, url: str) -> bool:
+        return False  # Must be explicitly configured
+
+    async def parse(
+        self,
+        raw: bytes,
+        source_url: str,
+        source_id: Any,
+        default_protocol: str,
+    ) -> list[DiscoveredProxy]:
+        results: list[DiscoveredProxy] = []
+        text = raw.decode("utf-8", errors="ignore")
+
+        for line in text.splitlines():
+            line = line.strip().lower()
+            match = _PROTOCOL_LINE_RE.match(line)
+            if match:
+                protocol = match.group(1)
+                if protocol in SUPPORTED_PROTOCOLS:
+                    results.append(
+                        DiscoveredProxy(
+                            ip=match.group(2),
+                            port=int(match.group(3)),
+                            protocol=protocol,
+                            source_id=source_id,
+                        )
+                    )
+
+        return results
+
+    def default_schedule(self) -> str | None:
+        return "*/30 * * * *"
+
+
+def create_plugin(settings: Settings) -> ProtocolPrefixParser:
+    return ProtocolPrefixParser()