[epiphany/gnome-3-38] web-view: encode data in error pages

[Michael Catanzaro <mcatanzaro src gnome org>]

commit 77755d107e4d1cbe92bdefc6c394491e66a708e1
Author: Michael Catanzaro <mcatanzaro redhat com>
Date:   Tue Dec 14 17:37:05 2021 -0600

    web-view: encode data in error pages
    
    Page titles and URLs are untrusted and could be nasty, so we need to
    encode them appropriately when injecting them into HTML.
    
    Part-of: <https://gitlab.gnome.org/GNOME/epiphany/-/merge_requests/1045>

 embed/ephy-web-view.c | 37 +++++++++++++++++++++++++++----------
 1 file changed, 27 insertions(+), 10 deletions(-)
---
diff --git a/embed/ephy-web-view.c b/embed/ephy-web-view.c
index bba27ae45..e5206aec3 100644
--- a/embed/ephy-web-view.c
+++ b/embed/ephy-web-view.c
@@ -38,6 +38,7 @@
 #include "ephy-gsb-utils.h"
 #include "ephy-history-service.h"
 #include "ephy-lib-type-builtins.h"
+#include "ephy-output-encoding.h"
 #include "ephy-permissions-manager.h"
 #include "ephy-prefs.h"
 #include "ephy-reader-handler.h"
@@ -1772,6 +1773,8 @@ format_network_error_page (const char  *uri,
                            const char **icon_name,
                            const char **style)
 {
+  g_autofree char *encoded_uri = NULL;
+  g_autofree char *encoded_origin = NULL;
   g_autofree char *formatted_origin = NULL;
   g_autofree char *formatted_reason = NULL;
   g_autofree char *first_paragraph = NULL;
@@ -1783,7 +1786,8 @@ format_network_error_page (const char  *uri,
   /* Message title when a site cannot be loaded due to a network error. */
   *message_title = g_strdup (_("Unable to display this website"));
 
-  formatted_origin = g_strdup_printf ("<strong>%s</strong>", origin);
+  encoded_origin = ephy_encode_for_html_entity (origin);
+  formatted_origin = g_strdup_printf ("<strong>%s</strong>", encoded_origin);
   /* Error details when a site cannot be loaded due to a network error. */
   first_paragraph = g_strdup_printf (_("The site at %s seems to be "
                                        "unavailable."),
@@ -1805,7 +1809,8 @@ format_network_error_page (const char  *uri,
 
   /* The button on the network error page. DO NOT ADD MNEMONICS HERE. */
   *button_label = g_strdup (_("Reload"));
-  *button_action = g_strdup_printf ("window.location = '%s';", uri);
+  encoded_uri = ephy_encode_for_javascript (uri);
+  *button_action = g_strdup_printf ("window.location = '%s';", encoded_uri);
   /* Mnemonic for the Reload button on browser error pages. */
   *button_accesskey = C_("reload-access-key", "R");
 
@@ -1824,6 +1829,8 @@ format_crash_error_page (const char  *uri,
                          const char **icon_name,
                          const char **style)
 {
+  g_autofree char *html_encoded_uri = NULL;
+  g_autofree char *js_encoded_uri = NULL;
   g_autofree char *formatted_uri = NULL;
   g_autofree char *formatted_distributor = NULL;
   g_autofree char *first_paragraph = NULL;
@@ -1835,7 +1842,8 @@ format_crash_error_page (const char  *uri,
   /* Message title when a site cannot be loaded due to a page crash error. */
   *message_title = g_strdup (_("Oops! There may be a problem"));
 
-  formatted_uri = g_strdup_printf ("<strong>%s</strong>", uri);
+  html_encoded_uri = ephy_encode_for_html_entity (uri);
+  formatted_uri = g_strdup_printf ("<strong>%s</strong>", html_encoded_uri);
   /* Error details when a site cannot be loaded due to a page crash error. */
   first_paragraph = g_strdup_printf (_("The page %s may have caused Web to "
                                        "close unexpectedly."),
@@ -1854,7 +1862,8 @@ format_crash_error_page (const char  *uri,
 
   /* The button on the page crash error page. DO NOT ADD MNEMONICS HERE. */
   *button_label = g_strdup (_("Reload"));
-  *button_action = g_strdup_printf ("window.location = '%s';", uri);
+  js_encoded_uri = ephy_encode_for_javascript (uri);
+  *button_action = g_strdup_printf ("window.location = '%s';", js_encoded_uri);
   /* Mnemonic for the Reload button on browser error pages. */
   *button_accesskey = C_("reload-access-key", "R");
 
@@ -1873,6 +1882,7 @@ format_process_crash_error_page (const char  *uri,
                                  const char **icon_name,
                                  const char **style)
 {
+  g_autofree char *encoded_uri = NULL;
   const char *first_paragraph;
 
   /* Page title when a site cannot be loaded due to a process crash error. */
@@ -1888,7 +1898,8 @@ format_process_crash_error_page (const char  *uri,
 
   /* The button on the process crash error page. DO NOT ADD MNEMONICS HERE. */
   *button_label = g_strdup (_("Reload"));
-  *button_action = g_strdup_printf ("window.location = '%s';", uri);
+  encoded_uri = ephy_encode_for_javascript (uri);
+  *button_action = g_strdup_printf ("window.location = '%s';", encoded_uri);
   /* Mnemonic for the Reload button on browser error pages. */
   *button_accesskey = C_("reload-access-key", "R");
 
@@ -1912,6 +1923,7 @@ format_tls_error_page (EphyWebView  *view,
                        const char  **icon_name,
                        const char  **style)
 {
+  g_autofree char *encoded_origin = NULL;
   g_autofree char *formatted_origin = NULL;
   g_autofree char *first_paragraph = NULL;
 
@@ -1921,7 +1933,8 @@ format_tls_error_page (EphyWebView  *view,
   /* Message title when a site is not loaded due to an invalid TLS certificate. */
   *message_title = g_strdup (_("This Connection is Not Secure"));
 
-  formatted_origin = g_strdup_printf ("<strong>%s</strong>", origin);
+  encoded_origin = ephy_encode_for_html_entity (origin);
+  formatted_origin = g_strdup_printf ("<strong>%s</strong>", encoded_origin);
   /* Error details when a site is not loaded due to an invalid TLS certificate. */
   first_paragraph = g_strdup_printf (_("This does not look like the real %s. "
                                        "Attackers might be trying to steal or "
@@ -1966,6 +1979,7 @@ format_unsafe_browsing_error_page (EphyWebView  *view,
                                    const char  **icon_name,
                                    const char  **style)
 {
+  g_autofree char *encoded_origin = NULL;
   g_autofree char *formatted_origin = NULL;
   g_autofree char *first_paragraph = NULL;
 
@@ -1975,7 +1989,8 @@ format_unsafe_browsing_error_page (EphyWebView  *view,
   /* Message title on the unsafe browsing error page. */
   *message_title = g_strdup (_("Unsafe website detected!"));
 
-  formatted_origin = g_strdup_printf ("<strong>%s</strong>", origin);
+  encoded_origin = ephy_encode_for_html_entity (origin);
+  formatted_origin = g_strdup_printf ("<strong>%s</strong>", encoded_origin);
   /* Error details on the unsafe browsing error page.
    * https://developers.google.com/safe-browsing/v4/usage-limits#UserWarnings
    */
@@ -2046,7 +2061,8 @@ format_no_such_file_error_page (EphyWebView  *view,
                                 const char  **icon_name,
                                 const char  **style)
 {
-  g_autofree gchar *formatted_origin = NULL;
+  g_autofree gchar *encoded_address = NULL;
+  g_autofree gchar *formatted_address = NULL;
   g_autofree gchar *first_paragraph = NULL;
   g_autofree gchar *second_paragraph = NULL;
 
@@ -2056,10 +2072,11 @@ format_no_such_file_error_page (EphyWebView  *view,
   /* Message title on the no such file error page. */
   *message_title = g_strdup (_("File not found"));
 
-  formatted_origin = g_strdup_printf ("<strong>%s</strong>", view->address);
+  encoded_address = ephy_encode_for_html_entity (view->address);
+  formatted_address = g_strdup_printf ("<strong>%s</strong>", encoded_address);
 
   first_paragraph = g_strdup_printf (_("%s could not be found."),
-                                     formatted_origin);
+                                     formatted_address);
   second_paragraph = g_strdup_printf (_("Please check the file name for "
                                         "capitalization or other typing errors. Also check if "
                                         "it has been moved, renamed, or deleted."));