res = -2;
goto out;
}
- if (opt.verbose)
+ if (progress)
progress_update (progress, sz, 0);
}
last_dltime = dltime;
}
- if (opt.verbose)
+ if (progress)
progress_update (progress, res, dltime);
*len += res;
}
res = -1;
out:
- if (opt.verbose)
+ if (progress)
progress_finish (progress, dltime);
if (elapsed)
*elapsed = dltime;
return dlrate;
}
\f
-static int
-register_redirections_mapper (void *key, void *value, void *arg)
-{
- const char *redirected_from = (const char *)key;
- const char *redirected_to = (const char *)arg;
- if (0 != strcmp (redirected_from, redirected_to))
- register_redirection (redirected_from, redirected_to);
- return 0;
-}
-
-/* Register the redirections that lead to the successful download of
- this URL. This is necessary so that the link converter can convert
- redirected URLs to the local file. */
-
-static void
-register_all_redirections (struct hash_table *redirections, const char *final)
-{
- hash_table_map (redirections, register_redirections_mapper, (void *)final);
-}
-
-#define USE_PROXY_P(u) (opt.use_proxy && getproxy((u)->scheme) \
- && no_proxy_match((u)->host, \
- (const char **)opt.no_proxy))
-
/* Maximum number of allowed redirections. 20 was chosen as a
"reasonable" value, which is low enough to not cause havoc, yet
high enough to guarantee that normal retrievals will not be hurt by
#define MAX_REDIRECTIONS 20
+#define SUSPEND_POST_DATA do { \
+ post_data_suspended = 1; \
+ saved_post_data = opt.post_data; \
+ saved_post_file_name = opt.post_file_name; \
+ opt.post_data = NULL; \
+ opt.post_file_name = NULL; \
+} while (0)
+
+#define RESTORE_POST_DATA do { \
+ if (post_data_suspended) \
+ { \
+ opt.post_data = saved_post_data; \
+ opt.post_file_name = saved_post_file_name; \
+ post_data_suspended = 0; \
+ } \
+} while (0)
+
/* Retrieve the given URL. Decides which loop to call -- HTTP, FTP,
FTP, proxy, etc. */
+/* #### This function should be rewritten so it doesn't return from
+ multiple points. */
+
uerr_t
retrieve_url (const char *origurl, char **file, char **newloc,
const char *refurl, int *dt)
uerr_t result;
char *url;
int location_changed, dummy;
- int use_proxy;
char *mynewloc, *proxy;
- struct url *u;
+ struct url *u, *proxy_url;
int up_error_code; /* url parse error code */
char *local_file;
- struct hash_table *redirections = NULL;
int redirection_count = 0;
+ int post_data_suspended = 0;
+ char *saved_post_data;
+ char *saved_post_file_name;
+
/* If dt is NULL, just ignore it. */
if (!dt)
dt = &dummy;
if (!u)
{
logprintf (LOG_NOTQUIET, "%s: %s.\n", url, url_error (up_error_code));
- if (redirections)
- string_set_free (redirections);
xfree (url);
return URLERROR;
}
result = NOCONERROR;
mynewloc = NULL;
local_file = NULL;
+ proxy_url = NULL;
- use_proxy = USE_PROXY_P (u);
- if (use_proxy)
+ proxy = getproxy (u);
+ if (proxy)
{
- struct url *proxy_url;
-
- /* Get the proxy server for the current scheme. */
- proxy = getproxy (u->scheme);
- if (!proxy)
- {
- logputs (LOG_NOTQUIET, _("Could not find proxy host.\n"));
- url_free (u);
- if (redirections)
- string_set_free (redirections);
- xfree (url);
- return PROXERR;
- }
-
/* Parse the proxy URL. */
proxy_url = url_parse (proxy, &up_error_code);
if (!proxy_url)
{
logprintf (LOG_NOTQUIET, _("Error parsing proxy URL %s: %s.\n"),
proxy, url_error (up_error_code));
- if (redirections)
- string_set_free (redirections);
xfree (url);
+ RESTORE_POST_DATA;
return PROXERR;
}
- if (proxy_url->scheme != SCHEME_HTTP)
+ if (proxy_url->scheme != SCHEME_HTTP && proxy_url->scheme != u->scheme)
{
logprintf (LOG_NOTQUIET, _("Error in proxy URL %s: Must be HTTP.\n"), proxy);
url_free (proxy_url);
- if (redirections)
- string_set_free (redirections);
xfree (url);
+ RESTORE_POST_DATA;
return PROXERR;
}
-
- result = http_loop (u, &mynewloc, &local_file, refurl, dt, proxy_url);
- url_free (proxy_url);
}
- else if (u->scheme == SCHEME_HTTP
+
+ if (u->scheme == SCHEME_HTTP
#ifdef HAVE_SSL
|| u->scheme == SCHEME_HTTPS
#endif
- )
+ || (proxy_url && proxy_url->scheme == SCHEME_HTTP))
{
- result = http_loop (u, &mynewloc, &local_file, refurl, dt, NULL);
+ result = http_loop (u, &mynewloc, &local_file, refurl, dt, proxy_url);
}
else if (u->scheme == SCHEME_FTP)
{
retrieval, so we save recursion to oldrec, and restore it
later. */
int oldrec = opt.recursive;
- if (redirections)
+ if (redirection_count)
opt.recursive = 0;
- result = ftp_loop (u, dt);
+ result = ftp_loop (u, dt, proxy_url);
opt.recursive = oldrec;
/* There is a possibility of having HTTP being redirected to
FTP. In these cases we must decide whether the text is HTML
- according to the suffix. The HTML suffixes are `.html' and
- `.htm', case-insensitive. */
- if (redirections && local_file && u->scheme == SCHEME_FTP)
+ according to the suffix. The HTML suffixes are `.html',
+ `.htm' and a few others, case-insensitive. */
+ if (redirection_count && local_file && u->scheme == SCHEME_FTP)
{
- char *suf = suffix (local_file);
- if (suf && (!strcasecmp (suf, "html") || !strcasecmp (suf, "htm")))
+ if (has_html_suffix_p (local_file))
*dt |= TEXTHTML;
}
}
+
+ if (proxy_url)
+ {
+ url_free (proxy_url);
+ proxy_url = NULL;
+ }
+
location_changed = (result == NEWLOCATION);
if (location_changed)
{
logprintf (LOG_NOTQUIET, "%s: %s.\n", mynewloc,
url_error (up_error_code));
url_free (u);
- if (redirections)
- string_set_free (redirections);
xfree (url);
xfree (mynewloc);
+ RESTORE_POST_DATA;
return result;
}
xfree (mynewloc);
mynewloc = xstrdup (newloc_parsed->url);
- if (!redirections)
- {
- redirections = make_string_hash_table (0);
- /* Add current URL immediately so we can detect it as soon
- as possible in case of a cycle. */
- string_set_add (redirections, u->url);
- }
-
- /* The new location is OK. Check for max. number of
- redirections. */
+ /* Check for max. number of redirections. */
if (++redirection_count > MAX_REDIRECTIONS)
{
logprintf (LOG_NOTQUIET, _("%d redirections exceeded.\n"),
MAX_REDIRECTIONS);
url_free (newloc_parsed);
url_free (u);
- if (redirections)
- string_set_free (redirections);
xfree (url);
xfree (mynewloc);
+ RESTORE_POST_DATA;
return WRONGCODE;
}
- /*Check for redirection cycle by
- peeking through the history of redirections. */
- if (string_set_contains (redirections, newloc_parsed->url))
- {
- logprintf (LOG_NOTQUIET, _("%s: Redirection cycle detected.\n"),
- mynewloc);
- url_free (newloc_parsed);
- url_free (u);
- if (redirections)
- string_set_free (redirections);
- xfree (url);
- xfree (mynewloc);
- return WRONGCODE;
- }
- string_set_add (redirections, newloc_parsed->url);
-
xfree (url);
url = mynewloc;
url_free (u);
u = newloc_parsed;
+
+ /* If we're being redirected from POST, we don't want to POST
+ again. Many requests answer POST with a redirection to an
+ index page; that redirection is clearly a GET. We "suspend"
+ POST data for the duration of the redirections, and restore
+ it when we're done. */
+ if (!post_data_suspended)
+ SUSPEND_POST_DATA;
+
goto redirected;
}
if (*dt & RETROKF)
{
register_download (u->url, local_file);
- if (redirections)
- register_all_redirections (redirections, u->url);
+ if (redirection_count && 0 != strcmp (origurl, u->url))
+ register_redirection (origurl, u->url);
if (*dt & TEXTHTML)
register_html (u->url, local_file);
}
url_free (u);
- if (redirections)
+ if (redirection_count)
{
- string_set_free (redirections);
if (newloc)
*newloc = url;
else
}
++global_download_count;
+ RESTORE_POST_DATA;
return result;
}
projects / wget / blobdiff