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
#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))
char *remote_time; /* remote time-stamp string */
char *error; /* textual HTTP error */
int statcode; /* status code */
+ char *message; /* status message */
wgint rd_size; /* amount of data read from socket */
double dltime; /* time it took to download the data */
const char *referer; /* value of the referer header. */
xfree_null (hs->rderrmsg);
xfree_null (hs->local_file);
xfree_null (hs->orig_file_name);
+ xfree_null (hs->message);
/* Guard against being called twice. */
hs->newloc = NULL;
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;
hs->newloc = NULL;
hs->remote_time = NULL;
hs->error = NULL;
+ hs->message = NULL;
conn = u;
resp = resp_new (head);
statcode = resp_status (resp, &message);
+ hs->message = xstrdup (message);
resp_free (resp);
xfree (head);
if (statcode != 200)
/* Check for status line. */
message = NULL;
statcode = resp_status (resp, &message);
+ hs->message = xstrdup (message);
if (!opt.server_response)
logprintf (LOG_VERBOSE, "%2d %s\n", statcode,
message ? quotearg_style (escape_quoting_style, message) : "");
hs->local_file = url_file_name (u);
}
}
-
+
/* TODO: perform this check only once. */
if (!hs->existence_checked && file_exists_p (hs->local_file))
{
local_dot_orig_file_exists = true;
local_filename = filename_plus_orig_suffix;
}
- }
+ }
if (!local_dot_orig_file_exists)
/* Couldn't stat() <file>.orig, so try to stat() <file>. */
char *tmp = strchr (type, ';');
if (tmp)
{
-#ifdef ENABLE_IRI
+ /* sXXXav: only needed if IRI support is enabled */
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)
+ /* Try to get remote encoding if needed */
+ if (opt.enable_iri && !opt.encoding_remote)
{
- 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);
- }
+ tmp = parse_charset (tmp2);
+ if (tmp)
+ set_content_encoding (iri, tmp);
}
-#endif
}
}
hs->newloc = resp_header_strdup (resp, "Location");
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);
}
}
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 */
uerr_t err, ret = TRYLIMEXC;
time_t tmr = -1; /* remote time-stamp */
struct http_stat hstat; /* HTTP status */
- struct_stat st;
+ struct_stat st;
bool send_head_first = true;
/* Assert that no value for *LOCAL_FILE was passed. */
assert (local_file == NULL || *local_file == NULL);
-
+
/* Set LOCAL_FILE parameter. */
if (local_file && opt.output_document)
*local_file = HYPHENP (opt.output_document) ? NULL : xstrdup (opt.output_document);
-
+
/* Reset NEWLOC parameter. */
*newloc = NULL;
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"),
+File %s already there; not retrieving.\n\n"),
quote (hstat.local_file));
/* If the file is there, we suppose it's retrieved OK. */
*dt |= RETROKF;
/* Reset the counter. */
count = 0;
-
+
/* 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
/* Send preliminary HEAD request if -N is given and we have an existing
* destination file. */
- if (opt.timestamping
+ if (opt.timestamping
&& !opt.content_disposition
&& file_exists_p (url_file_name (u)))
send_head_first = true;
-
+
/* THE loop */
do
{
/* Increment the pass counter. */
++count;
sleep_between_retrievals (count);
-
+
/* Get the current time string. */
tms = datetime_str (time (NULL));
-
+
if (opt.spider && !got_head)
logprintf (LOG_VERBOSE, _("\
Spider mode enabled. Check if remote file exists.\n"));
if (opt.verbose)
{
char *hurl = url_string (u, URL_AUTH_HIDE_PASSWD);
-
- if (count > 1)
+
+ if (count > 1)
{
char tmp[256];
sprintf (tmp, _("(try:%2d)"), count);
logprintf (LOG_NOTQUIET, "--%s-- %s %s\n",
tms, tmp, hurl);
}
- else
+ else
{
logprintf (LOG_NOTQUIET, "--%s-- %s\n",
tms, hurl);
}
-
+
#ifdef WINDOWS
ws_changetitle (hurl);
#endif
/* Default document type is empty. However, if spider mode is
on or time-stamping is employed, HEAD_ONLY commands is
encoded within *dt. */
- if (send_head_first && !got_head)
+ if (send_head_first && !got_head)
*dt |= HEAD_ONLY;
else
*dt &= ~HEAD_ONLY;
*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));
-
+
/* Get the new location (with or without the redirection). */
if (hstat.newloc)
*newloc = xstrdup (hstat.newloc);
hstat.statcode);
ret = WRONGCODE;
}
- else
+ else
{
ret = NEWLOCATION;
}
/* All possibilities should have been exhausted. */
abort ();
}
-
+
if (!(*dt & RETROKF))
{
char *hurl = NULL;
continue;
}
/* Maybe we should always keep track of broken links, not just in
- * spider mode. */
- else if (opt.spider)
+ * spider mode.
+ * 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)
+ if (!hurl)
hurl = url_string (u, URL_AUTH_HIDE_PASSWD);
nonexisting_url (hurl);
logprintf (LOG_NOTQUIET, _("\
else
{
logprintf (LOG_NOTQUIET, _("%s ERROR %d: %s.\n"),
- tms, hstat.statcode,
+ tms, hstat.statcode,
quotearg_style (escape_quoting_style, hstat.error));
}
logputs (LOG_VERBOSE, "\n");
if (opt.spider)
{
+ bool finished = true;
if (opt.recursive)
{
if (*dt & TEXTHTML)
{
logputs (LOG_VERBOSE, _("\
Remote file exists and could contain links to other resources -- retrieving.\n\n"));
+ finished = false;
}
else
{
logprintf (LOG_VERBOSE, _("\
Remote file exists but does not contain any link -- not retrieving.\n\n"));
ret = RETROK; /* RETRUNNEEDED is not for caller. */
- goto exit;
}
}
else
Remote file exists.\n\n"));
}
ret = RETROK; /* RETRUNNEEDED is not for caller. */
+ }
+
+ if (finished)
+ {
+ logprintf (LOG_NONVERBOSE,
+ _("%s URL:%s %2d %s\n"),
+ tms, u->url, hstat.statcode,
+ hstat.message ? quotearg_style (escape_quoting_style, hstat.message) : "");
goto exit;
}
}
printwhat (count, opt.ntry);
continue;
}
- else
+ else if (hstat.len != hstat.restval)
/* Getting here would mean reading more data than
requested with content-length, which we never do. */
abort ();
+ else
+ {
+ /* Getting here probably means that the content-length was
+ * _less_ than the original, local size. We should probably
+ * truncate or re-read, or something. FIXME */
+ ret = RETROK;
+ goto exit;
+ }
}
else /* from now on hstat.res can only be -1 */
{
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
projects / wget / blobdiff