From aa2cd37fb35c791bb9906d80b69206d311d78a11 Mon Sep 17 00:00:00 2001
From: qiushiyang <qiushiyang@corp.netease.com>
Date: Wed, 18 Jan 2023 06:08:46 +0000
Subject: [PATCH] use more accurate regex for {domain}:{port}

---
 libs/hbb_common/src/lib.rs | 40 ++++++++++++++++++++++++--------------
 src/client.rs              | 12 ++++--------
 2 files changed, 29 insertions(+), 23 deletions(-)

diff --git a/libs/hbb_common/src/lib.rs b/libs/hbb_common/src/lib.rs
index 29b066c66..e57994f34 100644
--- a/libs/hbb_common/src/lib.rs
+++ b/libs/hbb_common/src/lib.rs
@@ -322,10 +322,15 @@ pub fn is_ip_str(id: &str) -> bool {
 }
 
 #[inline]
-pub fn is_hostname_port_str(id: &str) -> bool {
-    regex::Regex::new(r"^[\-.0-9a-zA-Z]+:\d{1,5}$")
-        .unwrap()
-        .is_match(id)
+pub fn is_domain_port_str(id: &str) -> bool {
+    // modified regex for RFC1123 hostname. check https://stackoverflow.com/a/106223 for original version for hostname.
+    // according to [TLD List](https://data.iana.org/TLD/tlds-alpha-by-domain.txt) version 2023011700,
+    // there is no digits in TLD, and length is 2~63.
+    regex::Regex::new(
+        r"(?i)^([a-z0-9]([a-z0-9-]{0,61}[a-z0-9])?\.)+[a-z][a-z-]{0,61}[a-z]:\d{1,5}$",
+    )
+    .unwrap()
+    .is_match(id)
 }
 
 #[cfg(test)]
@@ -350,17 +355,22 @@ mod test_lib {
 
     #[test]
     fn test_hostname_port() {
-        assert_eq!(is_hostname_port_str("a:12"), true);
-        assert_eq!(is_hostname_port_str("a.b.c:12"), true);
-        assert_eq!(is_hostname_port_str("test.com:12"), true);
-        assert_eq!(is_hostname_port_str("1.2.3:12"), true);
-        assert_eq!(is_hostname_port_str("a.b.c:123456"), false);
-        // todo: should we also check for these edge case?
+        assert_eq!(is_domain_port_str("a:12"), false);
+        assert_eq!(is_domain_port_str("a.b.c:12"), false);
+        assert_eq!(is_domain_port_str("test.com:12"), true);
+        assert_eq!(is_domain_port_str("test-UPPER.com:12"), true);
+        assert_eq!(is_domain_port_str("some-other.domain.com:12"), true);
+        assert_eq!(is_domain_port_str("under_score:12"), false);
+        assert_eq!(is_domain_port_str("a@bc:12"), false);
+        assert_eq!(is_domain_port_str("1.1.1.1:12"), false);
+        assert_eq!(is_domain_port_str("1.2.3:12"), false);
+        assert_eq!(is_domain_port_str("1.2.3.45:12"), false);
+        assert_eq!(is_domain_port_str("a.b.c:123456"), false);
+        assert_eq!(is_domain_port_str("---:12"), false);
+        assert_eq!(is_domain_port_str(".:12"), false);
+        // todo: should we also check for these edge cases?
         // out-of-range port
-        assert_eq!(is_hostname_port_str("test.com:0"), true);
-        assert_eq!(is_hostname_port_str("test.com:98989"), true);
-        // invalid hostname
-        assert_eq!(is_hostname_port_str("---:12"), true);
-        assert_eq!(is_hostname_port_str(".:12"), true);
+        assert_eq!(is_domain_port_str("test.com:0"), true);
+        assert_eq!(is_domain_port_str("test.com:98989"), true);
     }
 }
diff --git a/src/client.rs b/src/client.rs
index 304314326..493448c3b 100644
--- a/src/client.rs
+++ b/src/client.rs
@@ -7,10 +7,10 @@ use cpal::{
 use magnum_opus::{Channels::*, Decoder as AudioDecoder};
 use sha2::{Digest, Sha256};
 use std::{
-    str::FromStr,
     collections::HashMap,
     net::SocketAddr,
     ops::{Deref, Not},
+    str::FromStr,
     sync::{atomic::AtomicBool, mpsc, Arc, Mutex, RwLock},
 };
 use uuid::Uuid;
@@ -181,14 +181,10 @@ impl Client {
                 true,
             ));
         }
-        // Allow connect to {hostname}:{port}
-        if hbb_common::is_hostname_port_str(peer) {
+        // Allow connect to {domain}:{port}
+        if hbb_common::is_domain_port_str(peer) {
             return Ok((
-                socket_client::connect_tcp(
-                    peer,
-                    RENDEZVOUS_TIMEOUT,
-                )
-                .await?,
+                socket_client::connect_tcp(peer, RENDEZVOUS_TIMEOUT).await?,
                 true,
             ));
         }