Detect HTTP Content-Type server encoding

[wget] / src / http.c

diff --git a/src/http.c b/src/http.c
index 0252d3421f07f491a59c443f50ab08aa6480d2af..682258c073c3d6bf5472f99c7962b7ac5f5e1f96 100644 (file)
--- a/src/http.c
+++ b/src/http.c
@@ -1823,10 +1823,11 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy)
   /* TODO: perform this check only once. */
   if (!hs->existence_checked && file_exists_p (hs->local_file))
     {
-      if (opt.noclobber)
+      if (opt.noclobber && !opt.output_document)
         {
           /* If opt.noclobber is turned on and file already exists, do not
-             retrieve the file */
+             retrieve the file. But if the output_document was given, then this
+             test was already done and the file didn't exist. Hence the !opt.output_document */
           logprintf (LOG_VERBOSE, _("\
 File %s already there; not retrieving.\n\n"), quote (hs->local_file));
           /* If the file is there, we suppose it's retrieved OK.  */
@@ -2039,9 +2040,32 @@ File %s already there; not retrieving.\n\n"), quote (hs->local_file));
       char *tmp = strchr (type, ';');
       if (tmp)
         {
+#ifdef ENABLE_IRI
+          char *tmp2 = tmp + 1;
+#endif
+
           while (tmp > type && c_isspace (tmp[-1]))
             --tmp;
           *tmp = '\0';
+
+#ifdef ENABLE_IRI
+          if (opt.enable_iri && *tmp2 != '\0' &&
+              (tmp = strstr (tmp2, "charset=")) != NULL)
+            {
+              tmp += 8;
+              tmp2 = tmp;
+
+              while (*tmp2 && !c_isspace (*tmp2))
+                tmp2++;
+
+              if (tmp2 > tmp)
+                {
+                  *tmp2 = '\0';
+                  /* sXXXav : check given charset */
+                  logprintf (LOG_VERBOSE, "HTTP charset: `%s'\n", tmp);
+                }
+            }
+#endif
         }
     }
   hs->newloc = resp_header_strdup (resp, "Location");
@@ -2376,10 +2400,11 @@ http_loop (struct url *u, char **newloc, char **local_file, const char *referer,
 
   /* TODO: Ick! This code is now in both gethttp and http_loop, and is
    * screaming for some refactoring. */
-  if (got_name && file_exists_p (hstat.local_file) && opt.noclobber)
+  if (got_name && file_exists_p (hstat.local_file) && opt.noclobber && !opt.output_document)
     {
       /* If opt.noclobber is turned on and file already exists, do not
-         retrieve the file */
+         retrieve the file. But if the output_document was given, then this
+         test was already done and the file didn't exist. Hence the !opt.output_document */
       logprintf (LOG_VERBOSE, _("\
 File %s already there; not retrieving.\n\n"), 
                  quote (hstat.local_file));