feat: exclude IP addresses from link extraction

Author: reddevilmidzy

rook/src/git/link_extractor.rs

@@ -4,7 +4,8 @@ use std::collections::HashSet;
 
 use crate::RepoManager;
 
-const REGEX_URL: &str = r"https?://(www\.)?[-a-zA-Z0-9@:%._+~#=]{1,256}\.[a-zA-Z0-9()]{1,6}\b([-a-zA-Z0-9()@:%_+.~#?&/=]*)";
+const REGEX_DOMAIN: &str = r"https?://(www\.)?[-a-zA-Z0-9@:%._+~#=]{1,256}\.[a-zA-Z0-9()]{1,6}\b([-a-zA-Z0-9()@:%_+.~#?&/=]*)";
+const REGEX_IP_ADDRESS: &str = r"https?://(localhost|(?:\d{1,3}\.){3}\d{1,3})(?::\d+)?";
 
 #[derive(Debug, Clone, Serialize, Deserialize)]
 /// Represents a hyperlink found in a repository, along with its location.
@@ -65,11 +66,16 @@ pub fn extract_links_from_repo_url(
 }
 
 fn find_link_in_content(content: &str, file_path: String) -> HashSet<LinkInfo> {
-    let url_regex = Regex::new(REGEX_URL).unwrap();
+    let domain_regex = Regex::new(REGEX_DOMAIN).unwrap();
+    let ip_address_regex = Regex::new(REGEX_IP_ADDRESS).unwrap();
     let mut result = HashSet::new();
 
     for (line_num, line) in content.lines().enumerate() {
-        for mat in url_regex.find_iter(line) {
+        for mat in domain_regex.find_iter(line) {
+            if ip_address_regex.is_match(mat.as_str()) {
+                continue;
+            }
+
             let url = mat
                 .as_str()
                 .trim_end_matches(&[')', '>', '.', ',', ';'][..])
@@ -124,6 +130,22 @@ mod tests {
         }
     }
 
+    #[test]
+    fn test_skip_ip_addresses() {
+        let content = r#"
+        http://192.168.1.1
+        http://192.168.1.1/path
+        http://192.168.1.1/path?param=value
+        this is localhost ip address http://127.0.0.1
+        front server http://localhost:3000
+        backend server http://localhost:8080
+        "#;
+
+        let file_path = "test.txt".to_string();
+        let links = find_link_in_content(content, file_path);
+        assert!(links.is_empty(), "Expected no links");
+    }
+
     #[test]
     fn test_link_info_uniqueness() {
         let mut links = HashSet::new();
@@ -193,10 +215,10 @@ mod tests {
 
         assert!(!result.is_empty(), "No links found in the repository");
 
-        let url_regex = Regex::new(REGEX_URL).unwrap();
+        let domain_regex = Regex::new(REGEX_DOMAIN).unwrap();
         for link in &result {
             assert!(
-                url_regex.is_match(&link.url),
+                domain_regex.is_match(&link.url),
                 "Invalid URL found: {} at {}:{}",
                 link.url,
                 link.file_path,