Fix silly scoping bug.

[wget] / src / http.c

diff --git a/src/http.c b/src/http.c
index 414e8e696d39aeb8e52ebf7c32ca2a170ef2c235..11af939a170fc20d073795988b13ef717d0a1b5b 100644 (file)
--- a/src/http.c
+++ b/src/http.c
@@ -1,5 +1,6 @@
 /* HTTP support.
-   Copyright (C) 1996-2006 Free Software Foundation, Inc.
+   Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003,
+   2004, 2005, 2006, 2007 Free Software Foundation, Inc.
 
 This file is part of GNU Wget.
 
@@ -1621,19 +1622,18 @@ gethttp (struct url *u, struct http_stat *hs, int *dt, struct url *proxy)
                only hurts us.  */
             request_remove_header (req, "Authorization");
         }
-    }
-
-  if (sock < 0)
-    {
-      /* In its current implementation, persistent_available_p will
-         look up conn->host in some cases.  If that lookup failed, we
-         don't need to bother with connect_to_host.  */
-      if (host_lookup_failed)
+      else if (host_lookup_failed)
         {
           request_free (req);
+          logprintf(LOG_NOTQUIET,
+                    _("%s: unable to resolve host address `%s'\n"),
+                    exec_name, relevant->host);
           return HOSTERR;
         }
+    }
 
+  if (sock < 0)
+    {
       sock = connect_to_host (conn->host, conn->port);
       if (sock == E_HOST)
         {
@@ -1820,7 +1820,7 @@ File `%s' already there; not retrieving.\n\n"), hs->local_file);
           if (has_html_suffix_p (hs->local_file))
             *dt |= TEXTHTML;
 
-          return RETROK;
+          return RETRUNNEEDED;
         }
       else if (!ALLOW_CLOBBER)
         {
@@ -2311,9 +2311,9 @@ http_loop (struct url *u, char **newloc, char **local_file, const char *referer,
   const char *tmrate;
   uerr_t err, ret = TRYLIMEXC;
   time_t tmr = -1;               /* remote time-stamp */
-  wgint local_size = 0;          /* the size of the local file */
   struct http_stat hstat;        /* HTTP status */
   struct_stat st;  
+  bool send_head_first = true;
 
   /* Assert that no value for *LOCAL_FILE was passed. */
   assert (local_file == NULL || *local_file == NULL);
@@ -2351,6 +2351,19 @@ http_loop (struct url *u, char **newloc, char **local_file, const char *referer,
   /* Reset the document type. */
   *dt = 0;
   
+  /* Skip preliminary HEAD request if we're not in spider mode AND
+   * if -O was given or HTTP Content-Disposition support is disabled. */
+  if (!opt.spider
+      && (got_name || !opt.content_disposition))
+    send_head_first = false;
+
+  /* Send preliminary HEAD request if -N is given and we have an existing 
+   * destination file. */
+  if (opt.timestamping 
+      && !opt.content_disposition
+      && file_exists_p (url_file_name (u)))
+    send_head_first = true;
+  
   /* THE loop */
   do
     {
@@ -2359,7 +2372,7 @@ http_loop (struct url *u, char **newloc, char **local_file, const char *referer,
       sleep_between_retrievals (count);
       
       /* Get the current time string.  */
-      tms = time_str (time (NULL));
+      tms = datetime_str (time (NULL));
       
       if (opt.spider && !got_head)
         logprintf (LOG_VERBOSE, _("\
@@ -2392,8 +2405,7 @@ Spider mode enabled. Check if remote file exists.\n"));
       /* Default document type is empty.  However, if spider mode is
          on or time-stamping is employed, HEAD_ONLY commands is
          encoded within *dt.  */
-      if (((opt.spider || opt.timestamping) && !got_head)
-          || (opt.always_rest && !got_name))
+      if (send_head_first && !got_head) 
         *dt |= HEAD_ONLY;
       else
         *dt &= ~HEAD_ONLY;
@@ -2429,12 +2441,12 @@ Spider mode enabled. Check if remote file exists.\n"));
       err = gethttp (u, &hstat, dt, proxy);
 
       /* Time?  */
-      tms = time_str (time (NULL));
+      tms = datetime_str (time (NULL));
       
       /* Get the new location (with or without the redirection).  */
       if (hstat.newloc)
         *newloc = xstrdup (hstat.newloc);
-      
+
       switch (err)
         {
         case HERR: case HEOF: case CONSOCKERR: case CONCLOSED:
@@ -2485,7 +2497,7 @@ Spider mode enabled. Check if remote file exists.\n"));
           /* All possibilities should have been exhausted.  */
           abort ();
         }
-     
+      
       if (!(*dt & RETROKF))
         {
           char *hurl = NULL;
@@ -2495,9 +2507,17 @@ Spider mode enabled. Check if remote file exists.\n"));
               hurl = url_string (u, URL_AUTH_HIDE_PASSWD);
               logprintf (LOG_NONVERBOSE, "%s:\n", hurl);
             }
+
+          /* Fall back to GET if HEAD fails with a 500 or 501 error code. */
+          if (*dt & HEAD_ONLY
+              && (hstat.statcode == 500 || hstat.statcode == 501))
+            {
+              got_head = true;
+              continue;
+            }
           /* Maybe we should always keep track of broken links, not just in
            * spider mode.  */
-          if (opt.spider)
+          else if (opt.spider)
             {
               /* #### Again: ugly ugly ugly! */
               if (!hurl) 
@@ -2518,7 +2538,7 @@ Remote file does not exist -- broken link!!!\n"));
         }
 
       /* Did we get the time-stamp? */
-      if (!got_head)
+      if (send_head_first && !got_head)
         {
           bool restart_loop = false;
 
@@ -2568,7 +2588,7 @@ Server file no newer than local file `%s' -- not retrieving.\n\n"),
                             {
                               logprintf (LOG_VERBOSE, _("\
 The sizes do not match (local %s) -- retrieving.\n"),
-                                         number_to_static_string (local_size));
+                                         number_to_static_string (hstat.orig_file_size));
                             }
                         }
                       else
@@ -2584,12 +2604,6 @@ The sizes do not match (local %s) -- retrieving.\n"),
               restart_loop = true;
             }
           
-          if (opt.always_rest)
-            {
-              got_name = true;
-              restart_loop = true;
-            }
-          
           if (opt.spider)
             {
               if (opt.recursive)
@@ -2604,7 +2618,7 @@ Remote file exists and could contain links to other resources -- retrieving.\n\n
                     {
                       logprintf (LOG_VERBOSE, _("\
 Remote file exists but does not contain any link -- not retrieving.\n\n"));
-                      ret = RETRUNNEEDED;
+                      ret = RETROK; /* RETRUNNEEDED is not for caller. */
                       goto exit;
                     }
                 }
@@ -2612,11 +2626,17 @@ Remote file exists but does not contain any link -- not retrieving.\n\n"));
                 {
                   logprintf (LOG_VERBOSE, _("\
 Remote file exists but recursion is disabled -- not retrieving.\n\n"));
-                  ret = RETRUNNEEDED;
+                  ret = RETROK; /* RETRUNNEEDED is not for caller. */
                   goto exit;
                 }
             }
 
+          if (send_head_first)
+            {
+              got_name = true;
+              restart_loop = true;
+            }
+          
           got_head = true;    /* no more time-stamping */
           *dt &= ~HEAD_ONLY;
           count = 0;          /* the retrieve count for HEAD is reset */