CVE-2025-22952

@@ -1,16 +1,30 @@
 package httpgetter
 
 import (
-	"errors"
 	"io"
 	"net"
 	"net/http"
 	"net/url"
 
+	"github.com/pkg/errors"
 	"golang.org/x/net/html"
 	"golang.org/x/net/html/atom"
 )
 
+var ErrInternalIP = errors.New("internal IP addresses are not allowed")
+
+var httpClient = &http.Client{
+	CheckRedirect: func(req *http.Request, via []*http.Request) error {
+		if err := validateURL(req.URL.String()); err != nil {
+			return errors.Wrap(err, "redirect to internal IP")
+		}
+		if len(via) >= 10 {
+			return errors.New("too many redirects")
+		}
+		return nil
+	},
+}
+
 type HTMLMeta struct {
 	Title       string `json:"title"`
 	Description string `json:"description"`
@@ -22,7 +36,7 @@ func GetHTMLMeta(urlStr string) (*HTMLMeta, error) {
 		return nil, err
 	}
 
-	response, err := http.Get(urlStr)
+	response, err := httpClient.Get(urlStr)
 	if err != nil {
 		return nil, err
 	}
@@ -110,12 +124,28 @@ func validateURL(urlStr string) error {
 		return errors.New("only http/https protocols are allowed")
 	}
 
-	if host := u.Hostname(); host != "" {
-		ip := net.ParseIP(host)
-		if ip != nil {
-			if ip.IsLoopback() || ip.IsPrivate() || ip.IsLinkLocalUnicast() {
-				return errors.New("internal IP addresses are not allowed")
-			}
+	host := u.Hostname()
+	if host == "" {
+		return errors.New("empty hostname")
+	}
+
+	// check if the hostname is an IP
+	if ip := net.ParseIP(host); ip != nil {
+		if ip.IsLoopback() || ip.IsPrivate() || ip.IsLinkLocalUnicast() {
+			return errors.Wrap(ErrInternalIP, ip.String())
+		}
+		return nil
+	}
+
+	// check if it's a hostname, resolve it and check all returned IPs
+	ips, err := net.LookupIP(host)
+	if err != nil {
+		return errors.Errorf("failed to resolve hostname: %v", err)
+	}
+
+	for _, ip := range ips {
+		if ip.IsLoopback() || ip.IsPrivate() || ip.IsLinkLocalUnicast() {
+			return errors.Wrapf(ErrInternalIP, "host=%s, ip=%s", host, ip.String())
 		}
 	}
 

@@ -1,6 +1,8 @@
 package httpgetter
 
 import (
+	"errors"
+	"strings"
 	"testing"
 
 	"github.com/stretchr/testify/require"
@@ -17,3 +19,21 @@ func TestGetHTMLMeta(t *testing.T) {
 		require.Equal(t, test.htmlMeta, *metadata)
 	}
 }
+
+func TestGetHTMLMetaForInternal(t *testing.T) {
+	// test for internal IP
+	if _, err := GetHTMLMeta("http://192.168.0.1"); !errors.Is(err, ErrInternalIP) {
+		t.Errorf("Expected error for internal IP, got %v", err)
+	}
+
+	// test for resolved internal IP
+	if _, err := GetHTMLMeta("http://localhost"); !errors.Is(err, ErrInternalIP) {
+		t.Errorf("Expected error for resolved internal IP, got %v", err)
+	}
+
+	// test for redirected internal IP
+	// 49.232.126.226:1110 will redirects to 127.0.0.1
+	if _, err := GetHTMLMeta("http://49.232.126.226:1110"); !(errors.Is(err, ErrInternalIP) && strings.Contains(err.Error(), "redirect")) {
+		t.Errorf("Expected error for redirected internal IP, got %v", err)
+	}
+}

@@ -3,6 +3,7 @@ package httpgetter
 import (
 	"errors"
 	"io"
+	"net"
 	"net/http"
 	"net/url"
 
@@ -17,7 +18,7 @@ type HTMLMeta struct {
 }
 
 func GetHTMLMeta(urlStr string) (*HTMLMeta, error) {
-	if _, err := url.Parse(urlStr); err != nil {
+	if err := validateURL(urlStr); err != nil {
 		return nil, err
 	}
 
@@ -35,6 +36,8 @@ func GetHTMLMeta(urlStr string) (*HTMLMeta, error) {
 		return nil, errors.New("not a HTML page")
 	}
 
+	// TODO: limit the size of the response body
+
 	htmlMeta := extractHTMLMeta(response.Body)
 	return htmlMeta, nil
 }
@@ -96,3 +99,25 @@ func extractMetaProperty(token html.Token, prop string) (content string, ok bool
 	}
 	return content, ok
 }
+
+func validateURL(urlStr string) error {
+	u, err := url.Parse(urlStr)
+	if err != nil {
+		return errors.New("invalid URL format")
+	}
+
+	if u.Scheme != "http" && u.Scheme != "https" {
+		return errors.New("only http/https protocols are allowed")
+	}
+
+	if host := u.Hostname(); host != "" {
+		ip := net.ParseIP(host)
+		if ip != nil {
+			if ip.IsLoopback() || ip.IsPrivate() || ip.IsLinkLocalUnicast() {
+				return errors.New("internal IP addresses are not allowed")
+			}
+		}
+	}
+
+	return nil
+}