Automated merge.

[wget] / src / url.c

diff --git a/src/url.c b/src/url.c
index f5d621f9064ca4f4d68b8f3117361db9936e1d08..c937d05695e03ede8d38dc79dd2bcb01632427d9 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)
+url_parse (const char *url, int *error, struct iri *iri)
 {
   struct url *u;
   const char *p;
@@ -659,7 +659,7 @@ url_parse (const char *url, int *error)
   int port;
   char *user = NULL, *passwd = NULL;
 
-  char *url_encoded = NULL;
+  char *url_encoded = NULL, *new_url = NULL;
 
   int error_code;
 
@@ -670,9 +670,20 @@ url_parse (const char *url, int *error)
       goto error;
     }
 
-  url_encoded = reencode_escapes (url);
+  if (iri && iri->utf8_encode)
+    {
+      url_unescape ((char *) url);
+      iri->utf8_encode = remote_to_utf8 (iri, url, (const char **) &new_url);
+      if (!iri->utf8_encode)
+        new_url = NULL;
+    }
+
+  url_encoded = reencode_escapes (new_url ? new_url : url);
   p = url_encoded;
 
+  if (new_url && url_encoded != new_url)
+    xfree (new_url);
+
   p += strlen (supported_schemes[scheme].leading_string);
   uname_b = p;
   p = url_skip_credentials (p);
@@ -842,6 +853,18 @@ url_parse (const char *url, int *error)
     {
       url_unescape (u->host);
       host_modified = true;
+
+      /* Apply IDNA regardless of iri->utf8_encode status */
+      if (opt.enable_iri && iri)
+        {
+          char *new = idn_encode (iri, u->host);
+          if (new)
+            {
+              xfree (u->host);
+              u->host = new;
+              host_modified = true;
+            }
+        }
     }
 
   if (params_b)
@@ -851,7 +874,7 @@ url_parse (const char *url, int *error)
   if (fragment_b)
     u->fragment = strdupdelim (fragment_b, fragment_e);
 
-  if (path_modified || u->fragment || host_modified || path_b == path_e)
+  if (opt.enable_iri || path_modified || u->fragment || host_modified || path_b == path_e)
     {
       /* If we suspect that a transformation has rendered what
          url_string might return different from URL_ENCODED, rebuild
@@ -886,11 +909,29 @@ url_parse (const char *url, int *error)
 /* 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
@@ -1430,11 +1471,17 @@ 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);
@@ -1476,7 +1523,7 @@ 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. */