IRI requirement: do not percent-encode already percent-encoded values (try1)

[wget] / src / url.c

diff --git a/src/url.c b/src/url.c
index e79cf8a2e0736688c0ce1495adc7d7b48286946c..8f0672503f07fdfef8cde9b8b4093b4b63841c71 100644 (file)
--- a/src/url.c
+++ b/src/url.c
@@ -619,7 +619,7 @@ static const char *parse_errors[] = {
 #define PE_NO_ERROR                     0
   N_("No error"),
 #define PE_UNSUPPORTED_SCHEME           1
-  N_("Unsupported scheme"),
+  N_("Unsupported scheme %s"),
 #define PE_INVALID_HOST_NAME            2
   N_("Invalid host name"),
 #define PE_BAD_PORT_NUMBER              3
@@ -640,7 +640,7 @@ static const char *parse_errors[] = {
    error, and if ERROR is not NULL, also set *ERROR to the appropriate
    error code. */
 struct url *
-url_parse (const char *url, int *error, struct iri *iri)
+url_parse (const char *url, int *error, struct iri *iri, bool percent_encode)
 {
   struct url *u;
   const char *p;
@@ -672,13 +672,19 @@ url_parse (const char *url, int *error, struct iri *iri)
 
   if (iri && iri->utf8_encode)
     {
-      url_unescape ((char *) url);
-      iri->utf8_encode = remote_to_utf8 (iri, url, (const char **) &new_url);
+      iri->utf8_encode = remote_to_utf8 (iri, iri->orig_url ? iri->orig_url : url, (const char **) &new_url);
       if (!iri->utf8_encode)
         new_url = NULL;
+      else
+        iri->orig_url = xstrdup (url);
     }
 
-  url_encoded = reencode_escapes (new_url ? new_url : url);
+  /* XXX XXX Could that change introduce (security) bugs ???  XXX XXX*/
+  if (percent_encode)
+    url_encoded = reencode_escapes (new_url ? new_url : url);
+  else
+     url_encoded = new_url ? new_url : url;
+
   p = url_encoded;
 
   if (new_url && url_encoded != new_url)
@@ -909,11 +915,29 @@ url_parse (const char *url, int *error, struct iri *iri)
 /* Return the error message string from ERROR_CODE, which should have
    been retrieved from url_parse.  The error message is translated.  */
 
-const char *
-url_error (int error_code)
+char *
+url_error (const char *url, int error_code)
 {
   assert (error_code >= 0 && ((size_t) error_code) < countof (parse_errors));
-  return _(parse_errors[error_code]);
+
+  if (error_code == PE_UNSUPPORTED_SCHEME)
+    {
+      char *error, *p;
+      char *scheme = xstrdup (url);
+      assert (url_has_scheme (url));
+
+      if ((p = strchr (scheme, ':')))
+        *p = '\0';
+      if (!strcasecmp (scheme, "https"))
+        asprintf (&error, _("HTTPS support not compiled in"));
+      else
+        asprintf (&error, _(parse_errors[error_code]), quote (scheme));
+      xfree (scheme);
+
+      return error;
+    }
+  else
+    return xstrdup (_(parse_errors[error_code]));
 }
 
 /* Split PATH into DIR and FILE.  PATH comes from the URL and is
@@ -1453,11 +1477,17 @@ url_file_name (const struct url *u)
 
   const char *u_file, *u_query;
   char *fname, *unique;
+  char *index_filename = "index.html"; /* The default index file is index.html */
 
   fnres.base = NULL;
   fnres.size = 0;
   fnres.tail = 0;
 
+  /* If an alternative index file was defined, change index_filename */
+  if (opt.default_page)
+    index_filename = opt.default_page;
+     
+
   /* Start with the directory prefix, if specified. */
   if (opt.dir_prefix)
     append_string (opt.dir_prefix, &fnres);
@@ -1499,7 +1529,7 @@ url_file_name (const struct url *u)
   /* Add the file name. */
   if (fnres.tail)
     append_char ('/', &fnres);
-  u_file = *u->file ? u->file : "index.html";
+  u_file = *u->file ? u->file : index_filename;
   append_uri_pathel (u_file, u_file + strlen (u_file), false, &fnres);
 
   /* Append "?query" to the file name. */
@@ -1968,12 +1998,12 @@ schemes_are_similar_p (enum url_scheme a, enum url_scheme b)
 \f
 static int
 getchar_from_escaped_string (const char *str, char *c)
-{  
+{
   const char *p = str;
 
   assert (str && *str);
   assert (c);
-  
+
   if (p[0] == '%')
     {
       if (!c_isxdigit(p[1]) || !c_isxdigit(p[2]))
@@ -2023,7 +2053,7 @@ are_urls_equal (const char *u1, const char *u2)
       p += pp;
       q += qq;
     }
-  
+
   return (*p == 0 && *q == 0 ? true : false);
 }
 \f
@@ -2132,7 +2162,7 @@ test_append_uri_pathel()
   } test_array[] = {
     { "http://www.yoyodyne.com/path/", "somepage.html", false, "http://www.yoyodyne.com/path/somepage.html" },
   };
-  
+
   for (i = 0; i < sizeof(test_array)/sizeof(test_array[0]); ++i) 
     {
       struct growable dest;