From 167ef0c6817658c1a089c75c462482209e207db4 Mon Sep 17 00:00:00 2001
From: Carlos Garcia Campos <cgarcia@igalia.com>
Date: Thu, 22 Jan 2026 15:26:18 +0100
Subject: [PATCH] uri-utils: do host validation when checking if a GUri is
 valid

Currently we only check if the host is not NULL and not empty, but it
might contain invalid characters not allowed for a host name in a URL.
This patch replaces the SOUP_URI_IS_VALID internal macro by a function
that in addition to the existing checks, it also validates the host.

Closes #488

Conflict:Adapted libsoup2 based on libsoup3 and remove tests/uri-parsing-test.c
Reference: https://github.com/GNOME/libsoup/commit/167ef0c6817658c1a089c75c462482209e207db4
---
 libsoup/soup-auth.c    |  2 +-
 libsoup/soup-message.c |  6 ++++-
 libsoup/soup-uri.c     | 60 ++++++++++++++++++++++++++++++++++++++++++
 libsoup/soup-uri.h     |  3 +++
 4 files changed, 69 insertions(+), 2 deletions(-)

diff --git a/libsoup/soup-auth.c b/libsoup/soup-auth.c
index 1896aab..e205fe3 100644
--- a/libsoup/soup-auth.c
+++ b/libsoup/soup-auth.c
@@ -535,7 +535,7 @@ GSList *
 soup_auth_get_protection_space (SoupAuth *auth, SoupURI *source_uri)
 {
 	g_return_val_if_fail (SOUP_IS_AUTH (auth), NULL);
-	g_return_val_if_fail (source_uri != NULL, NULL);
+	g_return_val_if_fail (soup_uri_is_valid (source_uri), NULL);
 
 	return SOUP_AUTH_GET_CLASS (auth)->get_protection_space (auth, source_uri);
 }
diff --git a/libsoup/soup-message.c b/libsoup/soup-message.c
index da32b42..d528c37 100644
--- a/libsoup/soup-message.c
+++ b/libsoup/soup-message.c
@@ -1044,7 +1044,7 @@ soup_message_new (const char *method, const char *uri_string)
 	uri = soup_uri_new (uri_string);
 	if (!uri)
 		return NULL;
-	if (!uri->host) {
+	if (!soup_uri_is_valid (uri)) {
 		soup_uri_free (uri);
 		return NULL;
 	}
@@ -1066,6 +1066,8 @@ soup_message_new (const char *method, const char *uri_string)
 SoupMessage *
 soup_message_new_from_uri (const char *method, SoupURI *uri)
 {
+	g_return_val_if_fail (soup_uri_is_valid (uri), NULL);
+
 	return g_object_new (SOUP_TYPE_MESSAGE,
 			     SOUP_MESSAGE_METHOD, method,
 			     SOUP_MESSAGE_URI, uri,
@@ -1676,6 +1678,8 @@ soup_message_set_uri (SoupMessage *msg, SoupURI *uri)
 	SoupMessagePrivate *priv;
 
 	g_return_if_fail (SOUP_IS_MESSAGE (msg));
+	g_return_if_fail (soup_uri_is_valid (uri));
+
 	priv = soup_message_get_instance_private (msg);
 
 	if (priv->uri)
diff --git a/libsoup/soup-uri.c b/libsoup/soup-uri.c
index bdb7a17..17eb1eb 100644
--- a/libsoup/soup-uri.c
+++ b/libsoup/soup-uri.c
@@ -1342,6 +1342,66 @@ soup_uri_host_equal (gconstpointer v1, gconstpointer v2)
 	return g_ascii_strcasecmp (one->host, two->host) == 0;
 }
 
+static gboolean
+is_valid_character_for_host (char c)
+{
+        static const char forbidden_chars[] = { '\t', '\n', '\r', ' ', '#', '/', ':', '<', '>', '?', '@', '[', '\\', ']', '^', '|' };
+        int i;
+
+        for (i = 0; i < G_N_ELEMENTS (forbidden_chars); ++i) {
+                if (c == forbidden_chars[i])
+                        return FALSE;
+        }
+
+        return TRUE;
+}
+
+static gboolean
+is_host_valid (const char* host)
+{
+        int i;
+        gboolean is_valid;
+        char *ascii_host = NULL;
+
+        if (!host || !host[0])
+                return FALSE;
+
+        if (g_hostname_is_non_ascii (host)) {
+                ascii_host = g_hostname_to_ascii (host);
+                if (!ascii_host)
+                  return FALSE;
+
+                host = ascii_host;
+        }
+
+        if ((g_ascii_isdigit (host[0]) || strchr (host, ':')) && g_hostname_is_ip_address (host)) {
+                g_free (ascii_host);
+                return TRUE;
+        }
+
+        is_valid = TRUE;
+        for (i = 0; host[i] && is_valid; i++)
+                is_valid = is_valid_character_for_host (host[i]);
+
+        g_free (ascii_host);
+
+        return is_valid;
+}
+
+gboolean
+soup_uri_is_valid (SoupURI *uri)
+{
+        if (!uri)
+                return FALSE;
+
+        if (!is_host_valid (soup_uri_get_host (uri)))
+                return FALSE;
+
+        /* FIXME: validate other URI components? */
+
+        return TRUE;
+}
+
 gboolean
 soup_uri_is_http (SoupURI *uri, char **aliases)
 {
diff --git a/libsoup/soup-uri.h b/libsoup/soup-uri.h
index 8015e4f..46baea4 100644
--- a/libsoup/soup-uri.h
+++ b/libsoup/soup-uri.h
@@ -46,6 +46,9 @@ SOUP_VAR gpointer _SOUP_URI_SCHEME_FTP;
 SOUP_VAR gpointer _SOUP_URI_SCHEME_FILE, _SOUP_URI_SCHEME_DATA, _SOUP_URI_SCHEME_RESOURCE;
 SOUP_VAR gpointer _SOUP_URI_SCHEME_WS, _SOUP_URI_SCHEME_WSS;
 
+SOUP_AVAILABLE_IN_2_4
+gboolean     soup_uri_is_valid             (SoupURI    *uri);
+
 SOUP_AVAILABLE_IN_2_4
 SoupURI	   *soup_uri_new_with_base         (SoupURI    *base,
 					    const char *uri_string);
-- 
2.33.0