Automated merge.

[wget] / src / http.c

diff --git a/src/http.c b/src/http.c
index f79327c32ebfd4b30d5deb700fe593eda753219c..589e18eebe8ab5627a15db7971757839caa48533 100644 (file)
--- a/src/http.c
+++ b/src/http.c
@@ -49,7 +49,6 @@ as that of the covered work.  */
 #include "retr.h"
 #include "connect.h"
 #include "netrc.h"
-#include "iri.h"
 #ifdef HAVE_SSL
 # include "ssl.h"
 #endif
@@ -70,11 +69,13 @@ as that of the covered work.  */
 extern char *version_string;
 
 /* Forward decls. */
+struct http_stat;
 static char *create_authorization_line (const char *, const char *,
                                         const char *, const char *,
                                         const char *, bool *);
 static char *basic_authentication_encode (const char *, const char *);
 static bool known_authentication_scheme_p (const char *, const char *);
+static void ensure_extension (struct http_stat *, const char *, int *);
 static void load_cookies (void);
 
 #ifndef MIN
@@ -87,6 +88,7 @@ static struct cookie_jar *wget_cookie_jar;
 
 #define TEXTHTML_S "text/html"
 #define TEXTXHTML_S "application/xhtml+xml"
+#define TEXTCSS_S "text/css"
 
 /* Some status code validation macros: */
 #define H_20X(x)        (((x) >= 200) && ((x) < 300))
@@ -1362,7 +1364,8 @@ free_hstat (struct http_stat *hs)
    If PROXY is non-NULL, the connection will be made to the proxy
    server, and u->url will be requested.  */
 static uerr_t
-gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy)
+gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy,
+         struct iri *iri)
 {
   struct request *req;
 
@@ -2055,7 +2058,11 @@ File %s already there; not retrieving.\n\n"), quote (hs->local_file));
 
           /* Try to get remote encoding if needed */
           if (opt.enable_iri && !opt.encoding_remote)
-            set_current_charset (parse_charset (tmp2));
+            {
+              tmp = parse_charset (tmp2);
+              if (tmp)
+                set_content_encoding (iri, tmp);
+            }
         }
     }
   hs->newloc = resp_header_strdup (resp, "Location");
@@ -2130,34 +2137,25 @@ File %s already there; not retrieving.\n\n"), quote (hs->local_file));
   else
     *dt &= ~TEXTHTML;
 
-  if (opt.html_extension && (*dt & TEXTHTML))
-    /* -E / --html-extension / html_extension = on was specified, and this is a
-       text/html file.  If some case-insensitive variation on ".htm[l]" isn't
-       already the file's suffix, tack on ".html". */
-    {
-      char *last_period_in_local_filename = strrchr (hs->local_file, '.');
+  if (type &&
+      0 == strncasecmp (type, TEXTCSS_S, strlen (TEXTCSS_S)))
+    *dt |= TEXTCSS;
+  else
+    *dt &= ~TEXTCSS;
 
-      if (last_period_in_local_filename == NULL
-          || !(0 == strcasecmp (last_period_in_local_filename, ".htm")
-               || 0 == strcasecmp (last_period_in_local_filename, ".html")))
+  if (opt.html_extension)
+    {
+      if (*dt & TEXTHTML)
+        /* -E / --html-extension / html_extension = on was specified,
+           and this is a text/html file.  If some case-insensitive
+           variation on ".htm[l]" isn't already the file's suffix,
+           tack on ".html". */
         {
-          int local_filename_len = strlen (hs->local_file);
-          /* Resize the local file, allowing for ".html" preceded by
-             optional ".NUMBER".  */
-          hs->local_file = xrealloc (hs->local_file,
-                                     local_filename_len + 24 + sizeof (".html"));
-          strcpy(hs->local_file + local_filename_len, ".html");
-          /* If clobbering is not allowed and the file, as named,
-             exists, tack on ".NUMBER.html" instead. */
-          if (!ALLOW_CLOBBER && file_exists_p (hs->local_file))
-            {
-              int ext_num = 1;
-              do
-                sprintf (hs->local_file + local_filename_len,
-                         ".%d.html", ext_num++);
-              while (file_exists_p (hs->local_file));
-            }
-          *dt |= ADDED_HTML_EXTENSION;
+          ensure_extension (hs, ".html", dt);
+        }
+      else if (*dt & TEXTCSS)
+        {
+          ensure_extension (hs, ".css", dt);
         }
     }
 
@@ -2339,7 +2337,7 @@ File %s already there; not retrieving.\n\n"), quote (hs->local_file));
    retried, and retried, and retried, and...  */
 uerr_t
 http_loop (struct url *u, char **newloc, char **local_file, const char *referer,
-           int *dt, struct url *proxy)
+           int *dt, struct url *proxy, struct iri *iri)
 {
   int count;
   bool got_head = false;         /* used for time-stamping and filename detection */
@@ -2503,7 +2501,7 @@ Spider mode enabled. Check if remote file exists.\n"));
         *dt &= ~SEND_NOCACHE;
 
       /* Try fetching the document, or at least its head.  */
-      err = gethttp (u, &hstat, dt, proxy);
+      err = gethttp (u, &hstat, dt, proxy, iri);
 
       /* Time?  */
       tms = datetime_str (time (NULL));
@@ -2582,9 +2580,9 @@ Spider mode enabled. Check if remote file exists.\n"));
             }
           /* Maybe we should always keep track of broken links, not just in
            * spider mode.
-           * Don't log error if it was utf8 encoded because we will try
-           * one unencoded. */
-          else if (opt.spider && !get_utf8_encode ())
+           * Don't log error if it was UTF-8 encoded because we will try
+           * once unencoded. */
+          else if (opt.spider && !iri->utf8_encode)
             {
               /* #### Again: ugly ugly ugly! */
               if (!hurl)
@@ -3222,6 +3220,42 @@ http_cleanup (void)
     cookie_jar_delete (wget_cookie_jar);
 }
 
+void
+ensure_extension (struct http_stat *hs, const char *ext, int *dt)
+{
+  char *last_period_in_local_filename = strrchr (hs->local_file, '.');
+  char shortext[8];
+  int len = strlen (ext);
+  if (len == 5)
+    {
+      strncpy (shortext, ext, len - 1);
+      shortext[len - 2] = '\0';
+    }
+
+  if (last_period_in_local_filename == NULL
+      || !(0 == strcasecmp (last_period_in_local_filename, shortext)
+           || 0 == strcasecmp (last_period_in_local_filename, ext)))
+    {
+      int local_filename_len = strlen (hs->local_file);
+      /* Resize the local file, allowing for ".html" preceded by
+         optional ".NUMBER".  */
+      hs->local_file = xrealloc (hs->local_file,
+                                 local_filename_len + 24 + len);
+      strcpy (hs->local_file + local_filename_len, ext);
+      /* If clobbering is not allowed and the file, as named,
+         exists, tack on ".NUMBER.html" instead. */
+      if (!ALLOW_CLOBBER && file_exists_p (hs->local_file))
+        {
+          int ext_num = 1;
+          do
+            sprintf (hs->local_file + local_filename_len,
+                     ".%d%s", ext_num++, ext);
+          while (file_exists_p (hs->local_file));
+        }
+      *dt |= ADDED_HTML_EXTENSION;
+    }
+}
+
 
 #ifdef TESTING