/* HTTP support.
- Copyright (C) 1996-2005 Free Software Foundation, Inc.
+ Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
+ 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012 Free Software Foundation,
+ Inc.
This file is part of GNU Wget.
GNU Wget is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
-the Free Software Foundation; either version 2 of the License, or
+the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
GNU Wget is distributed in the hope that it will be useful,
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
-along with Wget; if not, write to the Free Software Foundation, Inc.,
-51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+along with Wget. If not, see <http://www.gnu.org/licenses/>.
-In addition, as a special exception, the Free Software Foundation
-gives permission to link the code of its release of Wget with the
-OpenSSL project's "OpenSSL" library (or with modified versions of it
-that use the same license as the "OpenSSL" library), and distribute
-the linked executables. You must obey the GNU General Public License
-in all respects for all of the code used other than "OpenSSL". If you
-modify this file, you may extend this exception to your version of the
-file, but you are not obligated to do so. If you do not wish to do
-so, delete this exception statement from your version. */
+Additional permission under GNU GPL version 3 section 7
-#include <config.h>
+If you modify this program, or any covered work, by linking or
+combining it with the OpenSSL project's OpenSSL library (or a
+modified version of that library), containing parts covered by the
+terms of the OpenSSL or SSLeay licenses, the Free Software Foundation
+grants you additional permission to convey the resulting work.
+Corresponding Source for a non-source form of such a combination
+shall include the source code for the parts of OpenSSL used as well
+as that of the covered work. */
+
+#include "wget.h"
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
-#ifdef HAVE_UNISTD_H
-# include <unistd.h>
-#endif
+#include <unistd.h>
#include <assert.h>
#include <errno.h>
#include <time.h>
#include <locale.h>
-#include "wget.h"
+#include "hash.h"
#include "http.h"
#include "utils.h"
#include "url.h"
# include "http-ntlm.h"
#endif
#include "cookies.h"
-#ifdef ENABLE_DIGEST
-# include "gen-md5.h"
-#endif
+#include "md5.h"
#include "convert.h"
+#include "spider.h"
+#include "warc.h"
#ifdef TESTING
#include "test.h"
#endif
+#ifdef __VMS
+# include "vms.h"
+#endif /* def __VMS */
+
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 *, uerr_t *);
+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 MIN(x, y) ((x) > (y) ? (y) : (x))
#endif
#define TEXTHTML_S "text/html"
#define TEXTXHTML_S "application/xhtml+xml"
+#define TEXTCSS_S "text/css"
/* Some status code validation macros: */
+#define H_10X(x) (((x) >= 100) && ((x) < 200))
#define H_20X(x) (((x) >= 200) && ((x) < 300))
#define H_PARTIAL(x) ((x) == HTTP_STATUS_PARTIAL_CONTENTS)
#define H_REDIRECTED(x) ((x) == HTTP_STATUS_MOVED_PERMANENTLY \
int hcount, hcapacity;
};
-/* Create a new, empty request. At least request_set_method must be
- called before the request can be used. */
+extern int numurls;
+
+/* Create a new, empty request. Set the request's method and its
+ arguments. METHOD should be a literal string (or it should outlive
+ the request) because it will not be freed. ARG will be freed by
+ request_free. */
static struct request *
-request_new (void)
+request_new (const char *method, char *arg)
{
struct request *req = xnew0 (struct request);
req->hcapacity = 8;
req->headers = xnew_array (struct request_header, req->hcapacity);
- return req;
-}
-
-/* Set the request's method and its arguments. METH should be a
- literal string (or it should outlive the request) because it will
- not be freed. ARG will be freed by request_free. */
-
-static void
-request_set_method (struct request *req, const char *meth, char *arg)
-{
- req->method = meth;
+ req->method = method;
req->arg = arg;
+ return req;
}
/* Return the method string passed with the last call to
*/
static void
-request_set_header (struct request *req, char *name, char *value,
+request_set_header (struct request *req, const char *name, const char *value,
enum rp release_policy)
{
struct request_header *hdr;
/* A NULL value is a no-op; if freeing the name is requested,
free it now to avoid leaks. */
if (release_policy == rel_name || release_policy == rel_both)
- xfree (name);
+ xfree ((void *)name);
return;
}
{
/* Replace existing header. */
release_header (hdr);
- hdr->name = name;
- hdr->value = value;
+ hdr->name = (void *)name;
+ hdr->value = (void *)value;
hdr->release_policy = release_policy;
return;
}
req->headers = xrealloc (req->headers, req->hcapacity * sizeof (*hdr));
}
hdr = &req->headers[req->hcount++];
- hdr->name = name;
- hdr->value = value;
+ hdr->name = (void *)name;
+ hdr->value = (void *)value;
hdr->release_policy = release_policy;
}
return;
BOUNDED_TO_ALLOCA (header, p, name);
++p;
- while (ISSPACE (*p))
+ while (c_isspace (*p))
++p;
request_set_header (req, xstrdup (name), (char *) p, rel_name);
}
the header was actually removed, false otherwise. */
static bool
-request_remove_header (struct request *req, char *name)
+request_remove_header (struct request *req, const char *name)
{
int i;
for (i = 0; i < req->hcount; i++)
p += A_len; \
} while (0)
-/* Construct the request and write it to FD using fd_write. */
+/* Construct the request and write it to FD using fd_write.
+ If warc_tmp is set to a file pointer, the request string will
+ also be written to that file. */
static int
-request_send (const struct request *req, int fd)
+request_send (const struct request *req, int fd, FILE *warc_tmp)
{
char *request_string, *p;
int i, size, write_error;
APPEND (p, req->method); *p++ = ' ';
APPEND (p, req->arg); *p++ = ' ';
- memcpy (p, "HTTP/1.0\r\n", 10); p += 10;
+ memcpy (p, "HTTP/1.1\r\n", 10); p += 10;
for (i = 0; i < req->hcount; i++)
{
if (write_error < 0)
logprintf (LOG_VERBOSE, _("Failed writing HTTP request: %s.\n"),
fd_errstr (fd));
+ else if (warc_tmp != NULL)
+ {
+ /* Write a copy of the data to the WARC record. */
+ int warc_tmp_written = fwrite (request_string, 1, size - 1, warc_tmp);
+ if (warc_tmp_written != size - 1)
+ return -2;
+ }
return write_error;
}
xfree (req);
}
+static struct hash_table *basic_authed_hosts;
+
+/* Find out if this host has issued a Basic challenge yet; if so, give
+ * it the username, password. A temporary measure until we can get
+ * proper authentication in place. */
+
+static bool
+maybe_send_basic_creds (const char *hostname, const char *user,
+ const char *passwd, struct request *req)
+{
+ bool do_challenge = false;
+
+ if (opt.auth_without_challenge)
+ {
+ DEBUGP (("Auth-without-challenge set, sending Basic credentials.\n"));
+ do_challenge = true;
+ }
+ else if (basic_authed_hosts
+ && hash_table_contains(basic_authed_hosts, hostname))
+ {
+ DEBUGP (("Found %s in basic_authed_hosts.\n", quote (hostname)));
+ do_challenge = true;
+ }
+ else
+ {
+ DEBUGP (("Host %s has not issued a general basic challenge.\n",
+ quote (hostname)));
+ }
+ if (do_challenge)
+ {
+ request_set_header (req, "Authorization",
+ basic_authentication_encode (user, passwd),
+ rel_value);
+ }
+ return do_challenge;
+}
+
+static void
+register_basic_auth_host (const char *hostname)
+{
+ if (!basic_authed_hosts)
+ {
+ basic_authed_hosts = make_nocase_string_hash_table (1);
+ }
+ if (!hash_table_contains(basic_authed_hosts, hostname))
+ {
+ hash_table_put (basic_authed_hosts, xstrdup(hostname), NULL);
+ DEBUGP (("Inserted %s into basic_authed_hosts\n", quote (hostname)));
+ }
+}
+
+
/* Send the contents of FILE_NAME to SOCK. Make sure that exactly
PROMISED_SIZE bytes are sent over the wire -- if the file is
- longer, read only that much; if the file is shorter, report an error. */
+ longer, read only that much; if the file is shorter, report an error.
+ If warc_tmp is set to a file pointer, the post data will
+ also be written to that file. */
static int
-post_file (int sock, const char *file_name, wgint promised_size)
+body_file_send (int sock, const char *file_name, wgint promised_size, FILE *warc_tmp)
{
static char chunk[8192];
wgint written = 0;
int write_error;
FILE *fp;
- DEBUGP (("[writing POST file %s ... ", file_name));
+ DEBUGP (("[writing BODY file %s ... ", file_name));
fp = fopen (file_name, "rb");
if (!fp)
fclose (fp);
return -1;
}
+ if (warc_tmp != NULL)
+ {
+ /* Write a copy of the data to the WARC record. */
+ int warc_tmp_written = fwrite (chunk, 1, towrite, warc_tmp);
+ if (warc_tmp_written != towrite)
+ {
+ fclose (fp);
+ return -2;
+ }
+ }
written += towrite;
}
fclose (fp);
&& 0 == strncasecmp (b, name, name_len))
{
b += name_len + 1;
- while (b < e && ISSPACE (*b))
+ while (b < e && c_isspace (*b))
++b;
- while (b < e && ISSPACE (e[-1]))
+ while (b < e && c_isspace (e[-1]))
--e;
*begptr = b;
*endptr = e;
if (p < end && *p == '/')
{
++p;
- while (p < end && ISDIGIT (*p))
+ while (p < end && c_isdigit (*p))
++p;
if (p < end && *p == '.')
- ++p;
- while (p < end && ISDIGIT (*p))
+ ++p;
+ while (p < end && c_isdigit (*p))
++p;
}
- while (p < end && ISSPACE (*p))
+ while (p < end && c_isspace (*p))
++p;
- if (end - p < 3 || !ISDIGIT (p[0]) || !ISDIGIT (p[1]) || !ISDIGIT (p[2]))
+ if (end - p < 3 || !c_isdigit (p[0]) || !c_isdigit (p[1]) || !c_isdigit (p[2]))
return -1;
status = 100 * (p[0] - '0') + 10 * (p[1] - '0') + (p[2] - '0');
if (message)
{
- while (p < end && ISSPACE (*p))
+ while (p < end && c_isspace (*p))
++p;
- while (p < end && ISSPACE (end[-1]))
+ while (p < end && c_isspace (end[-1]))
--end;
*message = strdupdelim (p, end);
}
xfree (resp);
}
+/* Print a single line of response, the characters [b, e). We tried
+ getting away with
+ logprintf (LOG_VERBOSE, "%s%.*s\n", prefix, (int) (e - b), b);
+ but that failed to escape the non-printable characters and, in fact,
+ caused crashes in UTF-8 locales. */
+
+static void
+print_response_line(const char *prefix, const char *b, const char *e)
+{
+ char *copy;
+ BOUNDED_TO_ALLOCA(b, e, copy);
+ logprintf (LOG_ALWAYS, "%s%s\n", prefix,
+ quotearg_style (escape_quoting_style, copy));
+}
+
/* Print the server response, line by line, omitting the trailing CRLF
from individual header lines, and prefixed with PREFIX. */
--e;
if (b < e && e[-1] == '\r')
--e;
- /* This is safe even on printfs with broken handling of "%.<n>s"
- because resp->headers ends with \0. */
- logprintf (LOG_VERBOSE, "%s%.*s\n", prefix, e - b, b);
+ print_response_line(prefix, b, e);
}
}
HTTP spec. */
if (*hdr == ':')
++hdr;
- while (ISSPACE (*hdr))
+ while (c_isspace (*hdr))
++hdr;
if (!*hdr)
return false;
}
- if (!ISDIGIT (*hdr))
+ if (!c_isdigit (*hdr))
return false;
- for (num = 0; ISDIGIT (*hdr); hdr++)
+ for (num = 0; c_isdigit (*hdr); hdr++)
num = 10 * num + (*hdr - '0');
- if (*hdr != '-' || !ISDIGIT (*(hdr + 1)))
+ if (*hdr != '-' || !c_isdigit (*(hdr + 1)))
return false;
*first_byte_ptr = num;
++hdr;
- for (num = 0; ISDIGIT (*hdr); hdr++)
+ for (num = 0; c_isdigit (*hdr); hdr++)
num = 10 * num + (*hdr - '0');
- if (*hdr != '/' || !ISDIGIT (*(hdr + 1)))
+ if (*hdr != '/' || !c_isdigit (*(hdr + 1)))
return false;
*last_byte_ptr = num;
++hdr;
- for (num = 0; ISDIGIT (*hdr); hdr++)
- num = 10 * num + (*hdr - '0');
+ if (*hdr == '*')
+ num = -1;
+ else
+ for (num = 0; c_isdigit (*hdr); hdr++)
+ num = 10 * num + (*hdr - '0');
*entity_length_ptr = num;
return true;
}
mode, the body is displayed for debugging purposes. */
static bool
-skip_short_body (int fd, wgint contlen)
+skip_short_body (int fd, wgint contlen, bool chunked)
{
enum {
SKIP_SIZE = 512, /* size of the download buffer */
SKIP_THRESHOLD = 4096 /* the largest size we read */
};
+ wgint remaining_chunk_size = 0;
char dlbuf[SKIP_SIZE + 1];
dlbuf[SKIP_SIZE] = '\0'; /* so DEBUGP can safely print it */
- /* We shouldn't get here with unknown contlen. (This will change
- with HTTP/1.1, which supports "chunked" transfer.) */
- assert (contlen != -1);
-
/* If the body is too large, it makes more sense to simply close the
connection than to try to read the body. */
if (contlen > SKIP_THRESHOLD)
return false;
- DEBUGP (("Skipping %s bytes of body: [", number_to_static_string (contlen)));
-
- while (contlen > 0)
+ while (contlen > 0 || chunked)
{
- int ret = fd_read (fd, dlbuf, MIN (contlen, SKIP_SIZE), -1);
+ int ret;
+ if (chunked)
+ {
+ if (remaining_chunk_size == 0)
+ {
+ char *line = fd_read_line (fd);
+ char *endl;
+ if (line == NULL)
+ break;
+
+ remaining_chunk_size = strtol (line, &endl, 16);
+ xfree (line);
+
+ if (remaining_chunk_size == 0)
+ {
+ line = fd_read_line (fd);
+ xfree_null (line);
+ break;
+ }
+ }
+
+ contlen = MIN (remaining_chunk_size, SKIP_SIZE);
+ }
+
+ DEBUGP (("Skipping %s bytes of body: [", number_to_static_string (contlen)));
+
+ ret = fd_read (fd, dlbuf, MIN (contlen, SKIP_SIZE), -1);
if (ret <= 0)
{
/* Don't normally report the error since this is an
return false;
}
contlen -= ret;
+
+ if (chunked)
+ {
+ remaining_chunk_size -= ret;
+ if (remaining_chunk_size == 0)
+ {
+ char *line = fd_read_line (fd);
+ if (line == NULL)
+ return false;
+ else
+ xfree (line);
+ }
+ }
+
/* Safe even if %.*s bogusly expects terminating \0 because
we've zero-terminated dlbuf above. */
DEBUGP (("%.*s", ret, dlbuf));
return true;
}
-typedef struct {
- /* A token consists of characters in the [b, e) range. */
- const char *b, *e;
-} param_token;
+#define NOT_RFC2231 0
+#define RFC2231_NOENCODING 1
+#define RFC2231_ENCODING 2
+
+/* extract_param extracts the parameter name into NAME.
+ However, if the parameter name is in RFC2231 format then
+ this function adjusts NAME by stripping of the trailing
+ characters that are not part of the name but are present to
+ indicate the presence of encoding information in the value
+ or a fragment of a long parameter value
+*/
+static int
+modify_param_name(param_token *name)
+{
+ const char *delim1 = memchr (name->b, '*', name->e - name->b);
+ const char *delim2 = memrchr (name->b, '*', name->e - name->b);
+
+ int result;
+
+ if(delim1 == NULL)
+ {
+ result = NOT_RFC2231;
+ }
+ else if(delim1 == delim2)
+ {
+ if ((name->e - 1) == delim1)
+ {
+ result = RFC2231_ENCODING;
+ }
+ else
+ {
+ result = RFC2231_NOENCODING;
+ }
+ name->e = delim1;
+ }
+ else
+ {
+ name->e = delim1;
+ result = RFC2231_ENCODING;
+ }
+ return result;
+}
+
+/* extract_param extract the paramater value into VALUE.
+ Like modify_param_name this function modifies VALUE by
+ stripping off the encoding information from the actual value
+*/
+static void
+modify_param_value (param_token *value, int encoding_type )
+{
+ if (encoding_type == RFC2231_ENCODING)
+ {
+ const char *delim = memrchr (value->b, '\'', value->e - value->b);
+ if ( delim != NULL )
+ {
+ value->b = (delim+1);
+ }
+ }
+}
-/* Extract a parameter from the HTTP header at *SOURCE and advance
- *SOURCE to the next parameter. Return false when there are no more
- parameters to extract. The name of the parameter is returned in
- NAME, and the value in VALUE. If the parameter has no value, the
- token's value is zeroed out.
+/* Extract a parameter from the string (typically an HTTP header) at
+ **SOURCE and advance SOURCE to the next parameter. Return false
+ when there are no more parameters to extract. The name of the
+ parameter is returned in NAME, and the value in VALUE. If the
+ parameter has no value, the token's value is zeroed out.
For example, if *SOURCE points to the string "attachment;
filename=\"foo bar\"", the first call to this function will return
the token named "attachment" and no value, and the second call will
return the token named "filename" and value "foo bar". The third
- call will return false, indicating no more valid tokens. */
+ call will return false, indicating no more valid tokens.
-static bool
-extract_param (const char **source, param_token *name, param_token *value)
+ is_url_encoded is an out parameter. If not NULL, a boolean value will be
+ stored into it, letting the caller know whether or not the extracted value is
+ URL-encoded. The caller can then decode it with url_unescape(), which however
+ performs decoding in-place. URL-encoding is used by RFC 2231 to support
+ non-US-ASCII characters in HTTP header values. */
+
+bool
+extract_param (const char **source, param_token *name, param_token *value,
+ char separator, bool *is_url_encoded)
{
const char *p = *source;
+ int param_type;
+ if (is_url_encoded)
+ *is_url_encoded = false; /* initializing the out parameter */
- while (ISSPACE (*p)) ++p;
+ while (c_isspace (*p)) ++p;
if (!*p)
- return false; /* nothing more to extract */
+ {
+ *source = p;
+ return false; /* no error; nothing more to extract */
+ }
/* Extract name. */
name->b = p;
- while (*p && !ISSPACE (*p) && *p != '=' && *p != ';') ++p;
+ while (*p && !c_isspace (*p) && *p != '=' && *p != separator) ++p;
name->e = p;
- while (ISSPACE (*p)) ++p;
- if (*p == ';' || !*p) /* no value */
+ if (name->b == name->e)
+ return false; /* empty name: error */
+ while (c_isspace (*p)) ++p;
+ if (*p == separator || !*p) /* no value */
{
xzero (*value);
- if (*p == ';') ++p;
+ if (*p == separator) ++p;
*source = p;
return true;
}
if (*p != '=')
- return false; /* error */
+ return false; /* error */
/* *p is '=', extract value */
++p;
- while (ISSPACE (*p)) ++p;
- if (*p == '"') /* quoted */
+ while (c_isspace (*p)) ++p;
+ if (*p == '"') /* quoted */
{
value->b = ++p;
while (*p && *p != '"') ++p;
return false;
value->e = p++;
/* Currently at closing quote; find the end of param. */
- while (ISSPACE (*p)) ++p;
- while (*p && *p != ';') ++p;
- if (*p == ';')
- ++p;
+ while (c_isspace (*p)) ++p;
+ while (*p && *p != separator) ++p;
+ if (*p == separator)
+ ++p;
else if (*p)
- /* garbage after closed quote, e.g. foo="bar"baz */
- return false;
+ /* garbage after closed quote, e.g. foo="bar"baz */
+ return false;
}
- else /* unquoted */
+ else /* unquoted */
{
value->b = p;
- while (*p && *p != ';') ++p;
+ while (*p && *p != separator) ++p;
value->e = p;
- while (value->e != value->b && ISSPACE (value->e[-1]))
+ while (value->e != value->b && c_isspace (value->e[-1]))
--value->e;
- if (*p == ';') ++p;
+ if (*p == separator) ++p;
}
*source = p;
+
+ param_type = modify_param_name(name);
+ if (param_type != NOT_RFC2231)
+ {
+ if (param_type == RFC2231_ENCODING && is_url_encoded)
+ *is_url_encoded = true;
+ modify_param_value(value, param_type);
+ }
return true;
}
+#undef NOT_RFC2231
+#undef RFC2231_NOENCODING
+#undef RFC2231_ENCODING
+
+/* Appends the string represented by VALUE to FILENAME */
+
+static void
+append_value_to_filename (char **filename, param_token const * const value,
+ bool is_url_encoded)
+{
+ int original_length = strlen(*filename);
+ int new_length = strlen(*filename) + (value->e - value->b);
+ *filename = xrealloc (*filename, new_length+1);
+ memcpy (*filename + original_length, value->b, (value->e - value->b));
+ (*filename)[new_length] = '\0';
+ if (is_url_encoded)
+ url_unescape (*filename + original_length);
+}
+
#undef MAX
#define MAX(p, q) ((p) > (q) ? (p) : (q))
+/* Parse the contents of the `Content-Disposition' header, extracting
+ the information useful to Wget. Content-Disposition is a header
+ borrowed from MIME; when used in HTTP, it typically serves for
+ specifying the desired file name of the resource. For example:
+
+ Content-Disposition: attachment; filename="flora.jpg"
+
+ Wget will skip the tokens it doesn't care about, such as
+ "attachment" in the previous example; it will also skip other
+ unrecognized params. If the header is syntactically correct and
+ contains a file name, a copy of the file name is stored in
+ *filename and true is returned. Otherwise, the function returns
+ false.
+
+ The file name is stripped of directory components and must not be
+ empty.
+
+ Historically, this function returned filename prefixed with opt.dir_prefix,
+ now that logic is handled by the caller, new code should pay attention,
+ changed by crq, Sep 2010.
+
+*/
static bool
parse_content_disposition (const char *hdr, char **filename)
{
param_token name, value;
- while (extract_param (&hdr, &name, &value))
- if (BOUNDED_EQUAL_NO_CASE (name.b, name.e, "filename") && value.b != NULL)
- {
- /* Make the file name begin at the last slash or backslash. */
- const char *last_slash = memrchr (value.b, '/', value.e - value.b);
- const char *last_bs = memrchr (value.b, '\\', value.e - value.b);
- if (last_slash && last_bs)
- value.b = 1 + MAX (last_slash, last_bs);
- else if (last_slash || last_bs)
- value.b = 1 + (last_slash ? last_slash : last_bs);
- if (value.b == value.e)
- continue;
- *filename = strdupdelim (value.b, value.e);
- return true;
- }
- return false;
+ *filename = NULL;
+ bool is_url_encoded = false;
+ for ( ; extract_param (&hdr, &name, &value, ';', &is_url_encoded);
+ is_url_encoded = false)
+ {
+ int isFilename = BOUNDED_EQUAL_NO_CASE ( name.b, name.e, "filename" );
+ if ( isFilename && value.b != NULL)
+ {
+ /* Make the file name begin at the last slash or backslash. */
+ const char *last_slash = memrchr (value.b, '/', value.e - value.b);
+ const char *last_bs = memrchr (value.b, '\\', value.e - value.b);
+ if (last_slash && last_bs)
+ value.b = 1 + MAX (last_slash, last_bs);
+ else if (last_slash || last_bs)
+ value.b = 1 + (last_slash ? last_slash : last_bs);
+ if (value.b == value.e)
+ continue;
+
+ if (*filename)
+ append_value_to_filename (filename, &value, is_url_encoded);
+ else
+ {
+ *filename = strdupdelim (value.b, value.e);
+ if (is_url_encoded)
+ url_unescape (*filename);
+ }
+ }
+ }
+
+ if (*filename)
+ return true;
+ else
+ return false;
}
+
\f
/* Persistent connections. Currently, we cache the most recently used
connection as persistent, provided that the HTTP server agrees to
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. */
char *local_file; /* local file name. */
- bool timestamp_checked; /* true if pre-download time-stamping checks
+ bool existence_checked; /* true if we already checked for a file's
+ existence after having begun to download
+ (needed in gethttp for when connection is
+ interrupted/restarted. */
+ bool timestamp_checked; /* true if pre-download time-stamping checks
* have already been performed */
char *orig_file_name; /* name of file to compare for time-stamping
* (might be != local_file if -K is set) */
wgint orig_file_size; /* size of file to compare for time-stamping */
- time_t orig_file_tstamp; /* time-stamp of file to compare for
+ time_t orig_file_tstamp; /* time-stamp of file to compare for
* time-stamping */
};
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;
hs->error = NULL;
}
-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 load_cookies (void);
+static void
+get_file_flags (const char *filename, int *dt)
+{
+ logprintf (LOG_VERBOSE, _("\
+File %s already there; not retrieving.\n\n"), quote (filename));
+ /* If the file is there, we suppose it's retrieved OK. */
+ *dt |= RETROKF;
+
+ /* #### Bogusness alert. */
+ /* If its suffix is "html" or "htm" or similar, assume text/html. */
+ if (has_html_suffix_p (filename))
+ *dt |= TEXTHTML;
+}
+
+/* Download the response body from the socket and writes it to
+ an output file. The headers have already been read from the
+ socket. If WARC is enabled, the response body will also be
+ written to a WARC response record.
+
+ hs, contlen, contrange, chunked_transfer_encoding and url are
+ parameters from the gethttp method. fp is a pointer to the
+ output file.
+
+ url, warc_timestamp_str, warc_request_uuid, warc_ip, type
+ and statcode will be saved in the headers of the WARC record.
+ The head parameter contains the HTTP headers of the response.
+
+ If fp is NULL and WARC is enabled, the response body will be
+ written only to the WARC file. If WARC is disabled and fp
+ is a file pointer, the data will be written to the file.
+ If fp is a file pointer and WARC is enabled, the body will
+ be written to both destinations.
+
+ Returns the error code. */
+static int
+read_response_body (struct http_stat *hs, int sock, FILE *fp, wgint contlen,
+ wgint contrange, bool chunked_transfer_encoding,
+ char *url, char *warc_timestamp_str, char *warc_request_uuid,
+ ip_address *warc_ip, char *type, int statcode, char *head)
+{
+ int warc_payload_offset = 0;
+ FILE *warc_tmp = NULL;
+ int warcerr = 0;
+
+ if (opt.warc_filename != NULL)
+ {
+ /* Open a temporary file where we can write the response before we
+ add it to the WARC record. */
+ warc_tmp = warc_tempfile ();
+ if (warc_tmp == NULL)
+ warcerr = WARC_TMP_FOPENERR;
+
+ if (warcerr == 0)
+ {
+ /* We should keep the response headers for the WARC record. */
+ int head_len = strlen (head);
+ int warc_tmp_written = fwrite (head, 1, head_len, warc_tmp);
+ if (warc_tmp_written != head_len)
+ warcerr = WARC_TMP_FWRITEERR;
+ warc_payload_offset = head_len;
+ }
+
+ if (warcerr != 0)
+ {
+ if (warc_tmp != NULL)
+ fclose (warc_tmp);
+ return warcerr;
+ }
+ }
+
+ if (fp != NULL)
+ {
+ /* This confuses the timestamping code that checks for file size.
+ #### The timestamping code should be smarter about file size. */
+ if (opt.save_headers && hs->restval == 0)
+ fwrite (head, 1, strlen (head), fp);
+ }
+
+ /* Read the response body. */
+ int flags = 0;
+ if (contlen != -1)
+ /* If content-length is present, read that much; otherwise, read
+ until EOF. The HTTP spec doesn't require the server to
+ actually close the connection when it's done sending data. */
+ flags |= rb_read_exactly;
+ if (fp != NULL && hs->restval > 0 && contrange == 0)
+ /* If the server ignored our range request, instruct fd_read_body
+ to skip the first RESTVAL bytes of body. */
+ flags |= rb_skip_startpos;
+ if (chunked_transfer_encoding)
+ flags |= rb_chunked_transfer_encoding;
+
+ hs->len = hs->restval;
+ hs->rd_size = 0;
+ /* Download the response body and write it to fp.
+ If we are working on a WARC file, we simultaneously write the
+ response body to warc_tmp. */
+ hs->res = fd_read_body (hs->local_file, sock, fp, contlen != -1 ? contlen : 0,
+ hs->restval, &hs->rd_size, &hs->len, &hs->dltime,
+ flags, warc_tmp);
+ if (hs->res >= 0)
+ {
+ if (warc_tmp != NULL)
+ {
+ /* Create a response record and write it to the WARC file.
+ Note: per the WARC standard, the request and response should share
+ the same date header. We re-use the timestamp of the request.
+ The response record should also refer to the uuid of the request. */
+ bool r = warc_write_response_record (url, warc_timestamp_str,
+ warc_request_uuid, warc_ip,
+ warc_tmp, warc_payload_offset,
+ type, statcode, hs->newloc);
+
+ /* warc_write_response_record has closed warc_tmp. */
+
+ if (! r)
+ return WARC_ERR;
+ }
+
+ return RETRFINISHED;
+ }
+
+ if (warc_tmp != NULL)
+ fclose (warc_tmp);
+
+ if (hs->res == -2)
+ {
+ /* Error while writing to fd. */
+ return FWRITEERR;
+ }
+ else if (hs->res == -3)
+ {
+ /* Error while writing to warc_tmp. */
+ return WARC_TMP_FWRITEERR;
+ }
+ else
+ {
+ /* A read error! */
+ hs->rderrmsg = xstrdup (fd_errstr (sock));
+ return RETRFINISHED;
+ }
+}
#define BEGINS_WITH(line, string_constant) \
(!strncasecmp (line, string_constant, sizeof (string_constant) - 1) \
- && (ISSPACE (line[sizeof (string_constant) - 1]) \
+ && (c_isspace (line[sizeof (string_constant) - 1]) \
|| !line[sizeof (string_constant) - 1]))
+#ifdef __VMS
#define SET_USER_AGENT(req) do { \
if (!opt.useragent) \
request_set_header (req, "User-Agent", \
- aprintf ("Wget/%s", version_string), rel_value); \
+ aprintf ("Wget/%s (VMS %s %s)", \
+ version_string, vms_arch(), vms_vers()), \
+ rel_value); \
else if (*opt.useragent) \
request_set_header (req, "User-Agent", opt.useragent, rel_none); \
} while (0)
-
-/* The flags that allow clobbering the file (opening with "wb").
- Defined here to avoid repetition later. #### This will require
- rework. */
-#define ALLOW_CLOBBER (opt.noclobber || opt.always_rest || opt.timestamping \
- || opt.dirstruct || opt.output_document)
+#else /* def __VMS */
+#define SET_USER_AGENT(req) do { \
+ if (!opt.useragent) \
+ request_set_header (req, "User-Agent", \
+ aprintf ("Wget/%s (%s)", \
+ version_string, OS_TYPE), \
+ rel_value); \
+ else if (*opt.useragent) \
+ request_set_header (req, "User-Agent", opt.useragent, rel_none); \
+} while (0)
+#endif /* def __VMS [else] */
/* Retrieve a document through HTTP protocol. It recognizes status
code, and correctly handles redirections. It closes the network
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, int count)
{
struct request *req;
wgint contlen, contrange;
struct url *conn;
FILE *fp;
+ int err;
int sock = -1;
- int flags;
- /* Set to 1 when the authorization has failed permanently and should
+ /* Set to 1 when the authorization has already been sent and should
not be tried again. */
bool auth_finished = false;
+ /* Set to 1 when just globally-set Basic authorization has been sent;
+ * should prevent further Basic negotiations, but not other
+ * mechanisms. */
+ bool basic_auth_finished = false;
+
/* Whether NTLM authentication is used for this request. */
bool ntlm_seen = false;
char *head;
struct response *resp;
- char hdrval[256];
+ char hdrval[512];
char *message;
+ /* Declare WARC variables. */
+ bool warc_enabled = (opt.warc_filename != NULL);
+ FILE *warc_tmp = NULL;
+ char warc_timestamp_str [21];
+ char warc_request_uuid [48];
+ ip_address *warc_ip = NULL;
+ off_t warc_payload_offset = -1;
+
/* Whether this connection will be kept alive after the HTTP request
is done. */
bool keep_alive;
- /* Whether keep-alive should be inhibited.
+ /* Is the server using the chunked transfer encoding? */
+ bool chunked_transfer_encoding = false;
- RFC 2068 requests that 1.0 clients not send keep-alive requests
- to proxies. This is because many 1.0 proxies do not interpret
- the Connection header and transfer it to the remote server,
- causing it to not close the connection and leave both the proxy
- and the client hanging. */
+ /* Whether keep-alive should be inhibited. */
bool inhibit_keep_alive =
- !opt.http_keep_alive || opt.ignore_length || proxy != NULL;
+ !opt.http_keep_alive || opt.ignore_length;
/* Headers sent when using POST. */
- wgint post_data_size = 0;
+ wgint body_data_size = 0;
bool host_lookup_failed = false;
hs->newloc = NULL;
hs->remote_time = NULL;
hs->error = NULL;
+ hs->message = NULL;
conn = u;
/* Prepare the request to send. */
-
- req = request_new ();
{
char *meth_arg;
const char *meth = "GET";
if (head_only)
meth = "HEAD";
- else if (opt.post_file_name || opt.post_data)
- meth = "POST";
+ else if (opt.method)
+ meth = opt.method;
/* Use the full path, i.e. one that includes the leading slash and
the query string. E.g. if u->path is "foo/bar" and u->query is
"param=value", full_path will be "/foo/bar?param=value". */
meth_arg = xstrdup (u->url);
else
meth_arg = url_full_path (u);
- request_set_method (req, meth, meth_arg);
+ req = request_new (meth, meth_arg);
}
request_set_header (req, "Referer", (char *) hs->referer, rel_none);
if (*dt & SEND_NOCACHE)
- request_set_header (req, "Pragma", "no-cache", rel_none);
+ {
+ /* Cache-Control MUST be obeyed by all HTTP/1.1 caching mechanisms... */
+ request_set_header (req, "Cache-Control", "no-cache, must-revalidate", rel_none);
+
+ /* ... but some HTTP/1.0 caches doesn't implement Cache-Control. */
+ request_set_header (req, "Pragma", "no-cache", rel_none);
+ }
if (hs->restval)
request_set_header (req, "Range",
aprintf ("bytes=%s-",
user = user ? user : (opt.http_user ? opt.http_user : opt.user);
passwd = passwd ? passwd : (opt.http_passwd ? opt.http_passwd : opt.passwd);
- if (user && passwd)
- {
- /* We have the username and the password, but haven't tried
- any authorization yet. Let's see if the "Basic" method
- works. If not, we'll come back here and construct a
- proper authorization method with the right challenges.
-
- If we didn't employ this kind of logic, every URL that
- requires authorization would have to be processed twice,
- which is very suboptimal and generates a bunch of false
- "unauthorized" errors in the server log.
-
- #### But this logic also has a serious problem when used
- with stronger authentications: we *first* transmit the
- username and the password in clear text, and *then* attempt a
- stronger authentication scheme. That cannot be right! We
- are only fortunate that almost everyone still uses the
- `Basic' scheme anyway.
-
- There should be an option to prevent this from happening, for
- those who use strong authentication schemes and value their
- passwords. */
- request_set_header (req, "Authorization",
- basic_authentication_encode (user, passwd),
- rel_value);
- }
-
- proxyauth = NULL;
- if (proxy)
+ /* We only do "site-wide" authentication with "global" user/password
+ * values unless --auth-no-challange has been requested; URL user/password
+ * info overrides. */
+ if (user && passwd && (!u->user || opt.auth_without_challenge))
{
- char *proxy_user, *proxy_passwd;
- /* For normal username and password, URL components override
- command-line/wgetrc parameters. With proxy
- authentication, it's the reverse, because proxy URLs are
- normally the "permanent" ones, so command-line args
- should take precedence. */
- if (opt.proxy_user && opt.proxy_passwd)
- {
- proxy_user = opt.proxy_user;
- proxy_passwd = opt.proxy_passwd;
- }
- else
- {
- proxy_user = proxy->user;
- proxy_passwd = proxy->passwd;
- }
- /* #### This does not appear right. Can't the proxy request,
- say, `Digest' authentication? */
- if (proxy_user && proxy_passwd)
- proxyauth = basic_authentication_encode (proxy_user, proxy_passwd);
-
- /* If we're using a proxy, we will be connecting to the proxy
- server. */
- conn = proxy;
-
- /* Proxy authorization over SSL is handled below. */
-#ifdef HAVE_SSL
- if (u->scheme != SCHEME_HTTPS)
-#endif
- request_set_header (req, "Proxy-Authorization", proxyauth, rel_value);
+ /* If this is a host for which we've already received a Basic
+ * challenge, we'll go ahead and send Basic authentication creds. */
+ basic_auth_finished = maybe_send_basic_creds(u->host, user, passwd, req);
}
/* Generate the Host header, HOST:PORT. Take into account that:
rel_value);
}
- if (!inhibit_keep_alive)
- request_set_header (req, "Connection", "Keep-Alive", rel_none);
+ if (inhibit_keep_alive)
+ request_set_header (req, "Connection", "Close", rel_none);
+ else
+ {
+ if (proxy == NULL)
+ request_set_header (req, "Connection", "Keep-Alive", rel_none);
+ else
+ {
+ request_set_header (req, "Connection", "Close", rel_none);
+ request_set_header (req, "Proxy-Connection", "Keep-Alive", rel_none);
+ }
+ }
+
+ if (opt.method)
+ {
+
+ if (opt.body_data || opt.body_file)
+ {
+ request_set_header (req, "Content-Type",
+ "application/x-www-form-urlencoded", rel_none);
+
+ if (opt.body_data)
+ body_data_size = strlen (opt.body_data);
+ else
+ {
+ body_data_size = file_size (opt.body_file);
+ if (body_data_size == -1)
+ {
+ logprintf (LOG_NOTQUIET, _("BODY data file %s missing: %s\n"),
+ quote (opt.body_file), strerror (errno));
+ return FILEBADFILE;
+ }
+ }
+ request_set_header (req, "Content-Length",
+ xstrdup (number_to_static_string (body_data_size)),
+ rel_value);
+ }
+ }
+
+ retry_with_auth:
+ /* We need to come back here when the initial attempt to retrieve
+ without authorization header fails. (Expected to happen at least
+ for the Digest authorization scheme.) */
if (opt.cookies)
request_set_header (req, "Cookie",
),
rel_value);
- if (opt.post_data || opt.post_file_name)
- {
- request_set_header (req, "Content-Type",
- "application/x-www-form-urlencoded", rel_none);
- if (opt.post_data)
- post_data_size = strlen (opt.post_data);
- else
- {
- post_data_size = file_size (opt.post_file_name);
- if (post_data_size == -1)
- {
- logprintf (LOG_NOTQUIET, _("POST data file `%s' missing: %s\n"),
- opt.post_file_name, strerror (errno));
- post_data_size = 0;
- }
- }
- request_set_header (req, "Content-Length",
- xstrdup (number_to_static_string (post_data_size)),
- rel_value);
- }
-
/* Add the user headers. */
if (opt.user_headers)
{
request_set_user_header (req, opt.user_headers[i]);
}
- retry_with_auth:
- /* We need to come back here when the initial attempt to retrieve
- without authorization header fails. (Expected to happen at least
- for the Digest authorization scheme.) */
+ proxyauth = NULL;
+ if (proxy)
+ {
+ char *proxy_user, *proxy_passwd;
+ /* For normal username and password, URL components override
+ command-line/wgetrc parameters. With proxy
+ authentication, it's the reverse, because proxy URLs are
+ normally the "permanent" ones, so command-line args
+ should take precedence. */
+ if (opt.proxy_user && opt.proxy_passwd)
+ {
+ proxy_user = opt.proxy_user;
+ proxy_passwd = opt.proxy_passwd;
+ }
+ else
+ {
+ proxy_user = proxy->user;
+ proxy_passwd = proxy->passwd;
+ }
+ /* #### This does not appear right. Can't the proxy request,
+ say, `Digest' authentication? */
+ if (proxy_user && proxy_passwd)
+ proxyauth = basic_authentication_encode (proxy_user, proxy_passwd);
+
+ /* If we're using a proxy, we will be connecting to the proxy
+ server. */
+ conn = proxy;
- keep_alive = false;
+ /* Proxy authorization over SSL is handled below. */
+#ifdef HAVE_SSL
+ if (u->scheme != SCHEME_HTTPS)
+#endif
+ request_set_header (req, "Proxy-Authorization", proxyauth, rel_value);
+ }
+
+ keep_alive = true;
/* Establish the connection. */
- if (!inhibit_keep_alive)
+ if (inhibit_keep_alive)
+ keep_alive = false;
+ else
{
/* Look for a persistent connection to target host, unless a
proxy is used. The exception is when SSL is in use, in which
#endif
&host_lookup_failed))
{
+ int family = socket_family (pconn.socket, ENDPOINT_PEER);
sock = pconn.socket;
using_ssl = pconn.ssl;
- logprintf (LOG_VERBOSE, _("Reusing existing connection to %s:%d.\n"),
- escnonprint (pconn.host), pconn.port);
+#if ENABLE_IPV6
+ if (family == AF_INET6)
+ logprintf (LOG_VERBOSE, _("Reusing existing connection to [%s]:%d.\n"),
+ quotearg_style (escape_quoting_style, pconn.host),
+ pconn.port);
+ else
+#endif
+ logprintf (LOG_VERBOSE, _("Reusing existing connection to %s:%d.\n"),
+ quotearg_style (escape_quoting_style, pconn.host),
+ pconn.port);
DEBUGP (("Reusing fd %d.\n", sock));
if (pconn.authorized)
/* If the connection is already authorized, the "Basic"
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, quote (relevant->host));
return HOSTERR;
}
+ else if (sock != -1)
+ {
+ sock = -1;
+ }
+ }
+ if (sock < 0)
+ {
sock = connect_to_host (conn->host, conn->port);
if (sock == E_HOST)
{
{
/* When requesting SSL URLs through proxies, use the
CONNECT method to request passthrough. */
- struct request *connreq = request_new ();
- request_set_method (connreq, "CONNECT",
+ struct request *connreq = request_new ("CONNECT",
aprintf ("%s:%d", u->host, u->port));
SET_USER_AGENT (connreq);
if (proxyauth)
the regular request below. */
proxyauth = NULL;
}
- /* Examples in rfc2817 use the Host header in CONNECT
- requests. I don't see how that gains anything, given
- that the contents of Host would be exactly the same as
- the contents of CONNECT. */
+ request_set_header (connreq, "Host",
+ aprintf ("%s:%d", u->host, u->port),
+ rel_value);
- write_error = request_send (connreq, sock);
+ write_error = request_send (connreq, sock, 0);
request_free (connreq);
if (write_error < 0)
{
CLOSE_INVALIDATE (sock);
+ request_free (req);
return WRITEFAILED;
}
logprintf (LOG_VERBOSE, _("Failed reading proxy response: %s\n"),
fd_errstr (sock));
CLOSE_INVALIDATE (sock);
+ request_free (req);
return HERR;
}
message = NULL;
resp = resp_new (head);
statcode = resp_status (resp, &message);
+ if (statcode < 0)
+ {
+ char *tms = datetime_str (time (NULL));
+ logprintf (LOG_VERBOSE, "%d\n", statcode);
+ logprintf (LOG_NOTQUIET, _("%s ERROR %d: %s.\n"), tms, statcode,
+ quotearg_style (escape_quoting_style,
+ _("Malformed status line")));
+ xfree (head);
+ request_free (req);
+ return HERR;
+ }
+ hs->message = xstrdup (message);
resp_free (resp);
xfree (head);
if (statcode != 200)
{
failed_tunnel:
logprintf (LOG_NOTQUIET, _("Proxy tunneling failed: %s"),
- message ? escnonprint (message) : "?");
+ message ? quotearg_style (escape_quoting_style, message) : "?");
xfree_null (message);
+ request_free (req);
return CONSSLERR;
}
xfree_null (message);
if (conn->scheme == SCHEME_HTTPS)
{
- if (!ssl_connect (sock) || !ssl_check_certificate (sock, u->host))
+ if (!ssl_connect_wget (sock, u->host))
{
fd_close (sock);
+ request_free (req);
return CONSSLERR;
}
+ else if (!ssl_check_certificate (sock, u->host))
+ {
+ fd_close (sock);
+ request_free (req);
+ return VERIFCERTERR;
+ }
using_ssl = true;
}
#endif /* HAVE_SSL */
}
+ /* Open the temporary file where we will write the request. */
+ if (warc_enabled)
+ {
+ warc_tmp = warc_tempfile ();
+ if (warc_tmp == NULL)
+ {
+ CLOSE_INVALIDATE (sock);
+ request_free (req);
+ return WARC_TMP_FOPENERR;
+ }
+
+ if (! proxy)
+ {
+ warc_ip = (ip_address *) alloca (sizeof (ip_address));
+ socket_ip_address (sock, warc_ip, ENDPOINT_PEER);
+ }
+ }
+
/* Send the request to server. */
- write_error = request_send (req, sock);
+ write_error = request_send (req, sock, warc_tmp);
if (write_error >= 0)
{
- if (opt.post_data)
+ if (opt.body_data)
{
- DEBUGP (("[POST data: %s]\n", opt.post_data));
- write_error = fd_write (sock, opt.post_data, post_data_size, -1);
+ DEBUGP (("[BODY data: %s]\n", opt.body_data));
+ write_error = fd_write (sock, opt.body_data, body_data_size, -1);
+ if (write_error >= 0 && warc_tmp != NULL)
+ {
+ /* Remember end of headers / start of payload. */
+ warc_payload_offset = ftello (warc_tmp);
+
+ /* Write a copy of the data to the WARC record. */
+ int warc_tmp_written = fwrite (opt.body_data, 1, body_data_size, warc_tmp);
+ if (warc_tmp_written != body_data_size)
+ write_error = -2;
+ }
+ }
+ else if (opt.body_file && body_data_size != 0)
+ {
+ if (warc_tmp != NULL)
+ /* Remember end of headers / start of payload */
+ warc_payload_offset = ftello (warc_tmp);
+
+ write_error = body_file_send (sock, opt.body_file, body_data_size, warc_tmp);
}
- else if (opt.post_file_name && post_data_size != 0)
- write_error = post_file (sock, opt.post_file_name, post_data_size);
}
if (write_error < 0)
{
CLOSE_INVALIDATE (sock);
request_free (req);
- return WRITEFAILED;
+
+ if (warc_tmp != NULL)
+ fclose (warc_tmp);
+
+ if (write_error == -2)
+ return WARC_TMP_FWRITEERR;
+ else
+ return WRITEFAILED;
}
logprintf (LOG_VERBOSE, _("%s request sent, awaiting response... "),
proxy ? "Proxy" : "HTTP");
contrange = 0;
*dt &= ~RETROKF;
+
+ if (warc_enabled)
+ {
+ bool warc_result;
+ /* Generate a timestamp and uuid for this request. */
+ warc_timestamp (warc_timestamp_str);
+ warc_uuid_str (warc_request_uuid);
+
+ /* Create a request record and store it in the WARC file. */
+ warc_result = warc_write_request_record (u->url, warc_timestamp_str,
+ warc_request_uuid, warc_ip,
+ warc_tmp, warc_payload_offset);
+ if (! warc_result)
+ {
+ CLOSE_INVALIDATE (sock);
+ request_free (req);
+ return WARC_ERR;
+ }
+
+ /* warc_write_request_record has also closed warc_tmp. */
+ }
+
+
+read_header:
head = read_http_response_head (sock);
if (!head)
{
/* Check for status line. */
message = NULL;
statcode = resp_status (resp, &message);
+ if (statcode < 0)
+ {
+ char *tms = datetime_str (time (NULL));
+ logprintf (LOG_VERBOSE, "%d\n", statcode);
+ logprintf (LOG_NOTQUIET, _("%s ERROR %d: %s.\n"), tms, statcode,
+ quotearg_style (escape_quoting_style,
+ _("Malformed status line")));
+ CLOSE_INVALIDATE (sock);
+ resp_free (resp);
+ request_free (req);
+ xfree (head);
+ return HERR;
+ }
+
+ if (H_10X (statcode))
+ {
+ DEBUGP (("Ignoring response\n"));
+ resp_free (resp);
+ xfree (head);
+ goto read_header;
+ }
+
+ hs->message = xstrdup (message);
if (!opt.server_response)
logprintf (LOG_VERBOSE, "%2d %s\n", statcode,
- message ? escnonprint (message) : "");
+ message ? quotearg_style (escape_quoting_style, message) : "");
else
{
logprintf (LOG_VERBOSE, "\n");
print_server_response (resp, " ");
}
- /* Determine the local filename if needed. Notice that if -O is used
+ if (!opt.ignore_length
+ && resp_header_copy (resp, "Content-Length", hdrval, sizeof (hdrval)))
+ {
+ wgint parsed;
+ errno = 0;
+ parsed = str_to_wgint (hdrval, NULL, 10);
+ if (parsed == WGINT_MAX && errno == ERANGE)
+ {
+ /* Out of range.
+ #### If Content-Length is out of range, it most likely
+ means that the file is larger than 2G and that we're
+ compiled without LFS. In that case we should probably
+ refuse to even attempt to download the file. */
+ contlen = -1;
+ }
+ else if (parsed < 0)
+ {
+ /* Negative Content-Length; nonsensical, so we can't
+ assume any information about the content to receive. */
+ contlen = -1;
+ }
+ else
+ contlen = parsed;
+ }
+
+ /* Check for keep-alive related responses. */
+ if (!inhibit_keep_alive && contlen != -1)
+ {
+ if (resp_header_copy (resp, "Connection", hdrval, sizeof (hdrval)))
+ {
+ if (0 == strcasecmp (hdrval, "Close"))
+ keep_alive = false;
+ }
+ }
+
+ chunked_transfer_encoding = false;
+ if (resp_header_copy (resp, "Transfer-Encoding", hdrval, sizeof (hdrval))
+ && 0 == strcasecmp (hdrval, "chunked"))
+ chunked_transfer_encoding = true;
+
+ /* Handle (possibly multiple instances of) the Set-Cookie header. */
+ if (opt.cookies)
+ {
+ int scpos;
+ const char *scbeg, *scend;
+ /* The jar should have been created by now. */
+ assert (wget_cookie_jar != NULL);
+ for (scpos = 0;
+ (scpos = resp_header_locate (resp, "Set-Cookie", scpos,
+ &scbeg, &scend)) != -1;
+ ++scpos)
+ {
+ char *set_cookie; BOUNDED_TO_ALLOCA (scbeg, scend, set_cookie);
+ cookie_handle_set_cookie (wget_cookie_jar, u->host, u->port,
+ u->path, set_cookie);
+ }
+ }
+
+ if (keep_alive)
+ /* The server has promised that it will not close the connection
+ when we're done. This means that we can register it. */
+ register_persistent (conn->host, conn->port, sock, using_ssl);
+
+ if (statcode == HTTP_STATUS_UNAUTHORIZED)
+ {
+ /* Authorization is required. */
+
+ /* Normally we are not interested in the response body.
+ But if we are writing a WARC file we are: we like to keep everyting. */
+ if (warc_enabled)
+ {
+ int _err;
+ type = resp_header_strdup (resp, "Content-Type");
+ _err = read_response_body (hs, sock, NULL, contlen, 0,
+ chunked_transfer_encoding,
+ u->url, warc_timestamp_str,
+ warc_request_uuid, warc_ip, type,
+ statcode, head);
+ xfree_null (type);
+
+ if (_err != RETRFINISHED || hs->res < 0)
+ {
+ CLOSE_INVALIDATE (sock);
+ request_free (req);
+ xfree_null (message);
+ resp_free (resp);
+ xfree (head);
+ return _err;
+ }
+ else
+ CLOSE_FINISH (sock);
+ }
+ else
+ {
+ /* Since WARC is disabled, we are not interested in the response body. */
+ if (keep_alive && !head_only
+ && skip_short_body (sock, contlen, chunked_transfer_encoding))
+ CLOSE_FINISH (sock);
+ else
+ CLOSE_INVALIDATE (sock);
+ }
+
+ pconn.authorized = false;
+ uerr_t auth_err = RETROK;
+ if (!auth_finished && (user && passwd))
+ {
+ /* IIS sends multiple copies of WWW-Authenticate, one with
+ the value "negotiate", and other(s) with data. Loop over
+ all the occurrences and pick the one we recognize. */
+ int wapos;
+ const char *wabeg, *waend;
+ char *www_authenticate = NULL;
+ for (wapos = 0;
+ (wapos = resp_header_locate (resp, "WWW-Authenticate", wapos,
+ &wabeg, &waend)) != -1;
+ ++wapos)
+ if (known_authentication_scheme_p (wabeg, waend))
+ {
+ BOUNDED_TO_ALLOCA (wabeg, waend, www_authenticate);
+ break;
+ }
+
+ if (!www_authenticate)
+ {
+ /* If the authentication header is missing or
+ unrecognized, there's no sense in retrying. */
+ logputs (LOG_NOTQUIET, _("Unknown authentication scheme.\n"));
+ }
+ else if (!basic_auth_finished
+ || !BEGINS_WITH (www_authenticate, "Basic"))
+ {
+ char *pth = url_full_path (u);
+ const char *value;
+ uerr_t *auth_stat;
+ auth_stat = xmalloc (sizeof (uerr_t));
+ *auth_stat = RETROK;
+
+ value = create_authorization_line (www_authenticate,
+ user, passwd,
+ request_method (req),
+ pth,
+ &auth_finished,
+ auth_stat);
+
+ auth_err = *auth_stat;
+ if (auth_err == RETROK)
+ {
+ request_set_header (req, "Authorization", value, rel_value);
+
+ if (BEGINS_WITH (www_authenticate, "NTLM"))
+ ntlm_seen = true;
+ else if (!u->user && BEGINS_WITH (www_authenticate, "Basic"))
+ {
+ /* Need to register this host as using basic auth,
+ * so we automatically send creds next time. */
+ register_basic_auth_host (u->host);
+ }
+
+ xfree (pth);
+ xfree_null (message);
+ resp_free (resp);
+ xfree (head);
+ xfree (auth_stat);
+ goto retry_with_auth;
+ }
+ else
+ {
+ /* Creating the Authorization header went wrong */
+ }
+ }
+ else
+ {
+ /* We already did Basic auth, and it failed. Gotta
+ * give up. */
+ }
+ }
+ request_free (req);
+ xfree_null (message);
+ resp_free (resp);
+ xfree (head);
+ if (auth_err == RETROK)
+ return AUTHFAILED;
+ else
+ return auth_err;
+ }
+ else /* statcode != HTTP_STATUS_UNAUTHORIZED */
+ {
+ /* Kludge: if NTLM is used, mark the TCP connection as authorized. */
+ if (ntlm_seen)
+ pconn.authorized = true;
+ }
+
+ /* Determine the local filename if needed. Notice that if -O is used
* hstat.local_file is set by http_loop to the argument of -O. */
- if (!hs->local_file)
+ if (!hs->local_file)
{
+ char *local_file = NULL;
+
/* Honor Content-Disposition whether possible. */
- if (!resp_header_copy (resp, "Content-Disposition", hdrval, sizeof (hdrval))
- || !parse_content_disposition (hdrval, &hs->local_file))
+ if (!opt.content_disposition
+ || !resp_header_copy (resp, "Content-Disposition",
+ hdrval, sizeof (hdrval))
+ || !parse_content_disposition (hdrval, &local_file))
{
- /* Choose filename according to URL name. */
- hs->local_file = url_file_name (u);
+ /* The Content-Disposition header is missing or broken.
+ * Choose unique file name according to given URL. */
+ hs->local_file = url_file_name (u, NULL);
+ }
+ else
+ {
+ DEBUGP (("Parsed filename from Content-Disposition: %s\n",
+ local_file));
+ hs->local_file = url_file_name (u, local_file);
}
}
-
+
/* TODO: perform this check only once. */
- if (opt.noclobber && file_exists_p (hs->local_file))
+ if (!hs->existence_checked && file_exists_p (hs->local_file))
{
- /* If opt.noclobber is turned on and file already exists, do not
- retrieve the file */
- logprintf (LOG_VERBOSE, _("\
-File `%s' already there; not retrieving.\n\n"), hs->local_file);
- /* If the file is there, we suppose it's retrieved OK. */
- *dt |= RETROKF;
-
- /* #### Bogusness alert. */
- /* If its suffix is "html" or "htm" or similar, assume text/html. */
- if (has_html_suffix_p (hs->local_file))
- *dt |= TEXTHTML;
-
- return RETROK;
+ if (opt.noclobber && !opt.output_document)
+ {
+ /* If opt.noclobber is turned on and file already exists, do not
+ 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 */
+ get_file_flags (hs->local_file, dt);
+ request_free (req);
+ resp_free (resp);
+ xfree (head);
+ xfree_null (message);
+ return RETRUNNEEDED;
+ }
+ else if (!ALLOW_CLOBBER)
+ {
+ char *unique = unique_name (hs->local_file, true);
+ if (unique != hs->local_file)
+ xfree (hs->local_file);
+ hs->local_file = unique;
+ }
}
+ hs->existence_checked = true;
/* Support timestamping */
/* TODO: move this code out of gethttp. */
if (opt.timestamping && !hs->timestamp_checked)
{
size_t filename_len = strlen (hs->local_file);
- char *filename_plus_orig_suffix = alloca (filename_len + sizeof (".orig"));
+ char *filename_plus_orig_suffix = alloca (filename_len + sizeof (ORIG_SFX));
bool local_dot_orig_file_exists = false;
char *local_filename = NULL;
struct_stat st;
--hniksic */
memcpy (filename_plus_orig_suffix, hs->local_file, filename_len);
memcpy (filename_plus_orig_suffix + filename_len,
- ".orig", sizeof (".orig"));
+ ORIG_SFX, sizeof (ORIG_SFX));
/* Try to stat() the .orig file. */
if (stat (filename_plus_orig_suffix, &st) == 0)
{
- local_dot_orig_file_exists = 1;
+ 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>. */
}
}
- if (!opt.ignore_length
- && resp_header_copy (resp, "Content-Length", hdrval, sizeof (hdrval)))
- {
- wgint parsed;
- errno = 0;
- parsed = str_to_wgint (hdrval, NULL, 10);
- if (parsed == WGINT_MAX && errno == ERANGE)
- /* Out of range.
- #### If Content-Length is out of range, it most likely
- means that the file is larger than 2G and that we're
- compiled without LFS. In that case we should probably
- refuse to even attempt to download the file. */
- contlen = -1;
- else
- contlen = parsed;
- }
-
- /* Check for keep-alive related responses. */
- if (!inhibit_keep_alive && contlen != -1)
- {
- if (resp_header_copy (resp, "Keep-Alive", NULL, 0))
- keep_alive = true;
- else if (resp_header_copy (resp, "Connection", hdrval, sizeof (hdrval)))
- {
- if (0 == strcasecmp (hdrval, "Keep-Alive"))
- keep_alive = true;
- }
- }
- if (keep_alive)
- /* The server has promised that it will not close the connection
- when we're done. This means that we can register it. */
- register_persistent (conn->host, conn->port, sock, using_ssl);
-
- if (statcode == HTTP_STATUS_UNAUTHORIZED)
- {
- /* Authorization is required. */
- if (keep_alive && !head_only && skip_short_body (sock, contlen))
- CLOSE_FINISH (sock);
- else
- CLOSE_INVALIDATE (sock);
- pconn.authorized = false;
- if (!auth_finished && (user && passwd))
- {
- /* IIS sends multiple copies of WWW-Authenticate, one with
- the value "negotiate", and other(s) with data. Loop over
- all the occurrences and pick the one we recognize. */
- int wapos;
- const char *wabeg, *waend;
- char *www_authenticate = NULL;
- for (wapos = 0;
- (wapos = resp_header_locate (resp, "WWW-Authenticate", wapos,
- &wabeg, &waend)) != -1;
- ++wapos)
- if (known_authentication_scheme_p (wabeg, waend))
- {
- BOUNDED_TO_ALLOCA (wabeg, waend, www_authenticate);
- break;
- }
-
- if (!www_authenticate)
- /* If the authentication header is missing or
- unrecognized, there's no sense in retrying. */
- logputs (LOG_NOTQUIET, _("Unknown authentication scheme.\n"));
- else if (BEGINS_WITH (www_authenticate, "Basic"))
- /* If the authentication scheme is "Basic", which we send
- by default, there's no sense in retrying either. (This
- should be changed when we stop sending "Basic" data by
- default.) */
- ;
- else
- {
- char *pth;
- pth = url_full_path (u);
- request_set_header (req, "Authorization",
- create_authorization_line (www_authenticate,
- user, passwd,
- request_method (req),
- pth,
- &auth_finished),
- rel_value);
- if (BEGINS_WITH (www_authenticate, "NTLM"))
- ntlm_seen = true;
- xfree (pth);
- goto retry_with_auth;
- }
- }
- logputs (LOG_NOTQUIET, _("Authorization failed.\n"));
- request_free (req);
- return AUTHFAILED;
- }
- else /* statcode != HTTP_STATUS_UNAUTHORIZED */
- {
- /* Kludge: if NTLM is used, mark the TCP connection as authorized. */
- if (ntlm_seen)
- pconn.authorized = true;
- }
request_free (req);
hs->statcode = statcode;
char *tmp = strchr (type, ';');
if (tmp)
{
- while (tmp > type && ISSPACE (tmp[-1]))
+ /* sXXXav: only needed if IRI support is enabled */
+ char *tmp2 = tmp + 1;
+
+ while (tmp > type && c_isspace (tmp[-1]))
--tmp;
*tmp = '\0';
+
+ /* Try to get remote encoding if needed */
+ if (opt.enable_iri && !opt.encoding_remote)
+ {
+ tmp = parse_charset (tmp2);
+ if (tmp)
+ set_content_encoding (iri, tmp);
+ xfree_null(tmp);
+ }
}
}
hs->newloc = resp_header_strdup (resp, "Location");
hs->remote_time = resp_header_strdup (resp, "Last-Modified");
- /* Handle (possibly multiple instances of) the Set-Cookie header. */
- if (opt.cookies)
- {
- int scpos;
- const char *scbeg, *scend;
- /* The jar should have been created by now. */
- assert (wget_cookie_jar != NULL);
- for (scpos = 0;
- (scpos = resp_header_locate (resp, "Set-Cookie", scpos,
- &scbeg, &scend)) != -1;
- ++scpos)
- {
- char *set_cookie; BOUNDED_TO_ALLOCA (scbeg, scend, set_cookie);
- cookie_handle_set_cookie (wget_cookie_jar, u->host, u->port,
- u->path, set_cookie);
- }
- }
-
if (resp_header_copy (resp, "Content-Range", hdrval, sizeof (hdrval)))
{
wgint first_byte_pos, last_byte_pos, entity_length;
if (parse_content_range (hdrval, &first_byte_pos, &last_byte_pos,
&entity_length))
- contrange = first_byte_pos;
+ {
+ contrange = first_byte_pos;
+ contlen = last_byte_pos - first_byte_pos + 1;
+ }
}
resp_free (resp);
- /* 20x responses are counted among successful by default. */
- if (H_20X (statcode))
- *dt |= RETROKF;
+ /* 20x responses are counted among successful by default. */
+ if (H_20X (statcode))
+ *dt |= RETROKF;
+
+ if (statcode == HTTP_STATUS_NO_CONTENT)
+ {
+ /* 204 response has no body (RFC 2616, 4.3) */
+
+ /* In case the caller cares to look... */
+ hs->len = 0;
+ hs->res = 0;
+ hs->restval = 0;
+
+ CLOSE_FINISH (sock);
+ xfree_null (type);
+ xfree (head);
+
+ return RETRFINISHED;
+ }
/* Return if redirected. */
if (H_REDIRECTED (statcode) || statcode == HTTP_STATUS_MULTIPLE_CHOICES)
_("Location: %s%s\n"),
hs->newloc ? escnonprint_uri (hs->newloc) : _("unspecified"),
hs->newloc ? _(" [following]") : "");
- if (keep_alive && !head_only && skip_short_body (sock, contlen))
- CLOSE_FINISH (sock);
+
+ /* In case the caller cares to look... */
+ hs->len = 0;
+ hs->res = 0;
+ hs->restval = 0;
+
+ /* Normally we are not interested in the response body of a redirect.
+ But if we are writing a WARC file we are: we like to keep everyting. */
+ if (warc_enabled)
+ {
+ int _err = read_response_body (hs, sock, NULL, contlen, 0,
+ chunked_transfer_encoding,
+ u->url, warc_timestamp_str,
+ warc_request_uuid, warc_ip, type,
+ statcode, head);
+
+ if (_err != RETRFINISHED || hs->res < 0)
+ {
+ CLOSE_INVALIDATE (sock);
+ xfree_null (type);
+ xfree (head);
+ return _err;
+ }
+ else
+ CLOSE_FINISH (sock);
+ }
else
- CLOSE_INVALIDATE (sock);
+ {
+ /* Since WARC is disabled, we are not interested in the response body. */
+ if (keep_alive && !head_only
+ && skip_short_body (sock, contlen, chunked_transfer_encoding))
+ CLOSE_FINISH (sock);
+ else
+ CLOSE_INVALIDATE (sock);
+ }
+
xfree_null (type);
+ xfree (head);
+ /* From RFC2616: The status codes 303 and 307 have
+ been added for servers that wish to make unambiguously
+ clear which kind of reaction is expected of the client.
+
+ A 307 should be redirected using the same method,
+ in other words, a POST should be preserved and not
+ converted to a GET in that case.
+
+ With strict adherence to RFC2616, POST requests are not
+ converted to a GET request on 301 Permanent Redirect
+ or 302 Temporary Redirect.
+
+ A switch may be provided later based on the HTTPbis draft
+ that allows clients to convert POST requests to GET
+ requests on 301 and 302 response codes. */
+ switch (statcode)
+ {
+ case HTTP_STATUS_TEMPORARY_REDIRECT:
+ return NEWLOCATION_KEEP_POST;
+ case HTTP_STATUS_MOVED_PERMANENTLY:
+ if (opt.method && strcasecmp (opt.method, "post") != 0)
+ return NEWLOCATION_KEEP_POST;
+ break;
+ case HTTP_STATUS_MOVED_TEMPORARILY:
+ if (opt.method && strcasecmp (opt.method, "post") != 0)
+ return NEWLOCATION_KEEP_POST;
+ break;
+ default:
+ return NEWLOCATION;
+ }
return NEWLOCATION;
}
}
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 (last_period_in_local_filename == NULL
- || !(0 == strcasecmp (last_period_in_local_filename, ".htm")
- || 0 == strcasecmp (last_period_in_local_filename, ".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;
+ if (type &&
+ 0 == strncasecmp (type, TEXTCSS_S, strlen (TEXTCSS_S)))
+ *dt |= TEXTCSS;
+ else
+ *dt &= ~TEXTCSS;
+
+ if (opt.adjust_extension)
+ {
+ if (*dt & TEXTHTML)
+ /* -E / --adjust-extension / adjust_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". */
+ {
+ ensure_extension (hs, ".html", dt);
+ }
+ else if (*dt & TEXTCSS)
+ {
+ ensure_extension (hs, ".css", dt);
}
}
- if (statcode == HTTP_STATUS_RANGE_NOT_SATISFIABLE)
+ if (statcode == HTTP_STATUS_RANGE_NOT_SATISFIABLE
+ || (!opt.timestamping && hs->restval > 0 && statcode == HTTP_STATUS_OK
+ && contrange == 0 && contlen >= 0 && hs->restval >= contlen))
{
/* If `-c' is in use and the file has been fully downloaded (or
the remote file has shrunk), Wget effectively requests bytes
- after the end of file and the server response with 416. */
+ after the end of file and the server response with 416
+ (or 200 with a <= Content-Length. */
logputs (LOG_VERBOSE, _("\
\n The file is already fully retrieved; nothing to do.\n\n"));
/* In case the caller inspects. */
xfree_null (type);
CLOSE_INVALIDATE (sock); /* would be CLOSE_FINISH, but there
might be more bytes in the body. */
+ xfree (head);
return RETRUNNEEDED;
}
if ((contrange != 0 && contrange != hs->restval)
Bail out. */
xfree_null (type);
CLOSE_INVALIDATE (sock);
+ xfree (head);
return RANGEERR;
}
- hs->contlen = contlen + contrange;
+ if (contlen == -1)
+ hs->contlen = -1;
+ else
+ hs->contlen = contlen + contrange;
if (opt.verbose)
{
logputs (LOG_VERBOSE, number_to_static_string (contlen + contrange));
if (contlen + contrange >= 1024)
logprintf (LOG_VERBOSE, " (%s)",
- human_readable (contlen + contrange));
+ human_readable (contlen + contrange, 10, 1));
if (contrange)
{
if (contlen >= 1024)
logprintf (LOG_VERBOSE, _(", %s (%s) remaining"),
number_to_static_string (contlen),
- human_readable (contlen));
+ human_readable (contlen, 10, 1));
else
logprintf (LOG_VERBOSE, _(", %s remaining"),
number_to_static_string (contlen));
logputs (LOG_VERBOSE,
opt.ignore_length ? _("ignored") : _("unspecified"));
if (type)
- logprintf (LOG_VERBOSE, " [%s]\n", escnonprint (type));
+ logprintf (LOG_VERBOSE, " [%s]\n", quotearg_style (escape_quoting_style, type));
else
logputs (LOG_VERBOSE, "\n");
}
}
- xfree_null (type);
- type = NULL; /* We don't need it any more. */
/* Return if we have no intention of further downloading. */
- if (!(*dt & RETROKF) || head_only)
+ if ((!(*dt & RETROKF) && !opt.content_on_error) || head_only)
{
/* In case the caller cares to look... */
hs->len = 0;
hs->res = 0;
- xfree_null (type);
- if (head_only)
- /* Pre-1.10 Wget used CLOSE_INVALIDATE here. Now we trust the
- servers not to send body in response to a HEAD request, and
- those that do will likely be caught by test_socket_open.
- If not, they can be worked around using
- `--no-http-keep-alive'. */
- CLOSE_FINISH (sock);
- else if (keep_alive && skip_short_body (sock, contlen))
- /* Successfully skipped the body; also keep using the socket. */
- CLOSE_FINISH (sock);
+ hs->restval = 0;
+
+ /* Normally we are not interested in the response body of a error responses.
+ But if we are writing a WARC file we are: we like to keep everyting. */
+ if (warc_enabled)
+ {
+ int _err = read_response_body (hs, sock, NULL, contlen, 0,
+ chunked_transfer_encoding,
+ u->url, warc_timestamp_str,
+ warc_request_uuid, warc_ip, type,
+ statcode, head);
+
+ if (_err != RETRFINISHED || hs->res < 0)
+ {
+ CLOSE_INVALIDATE (sock);
+ xfree (head);
+ xfree_null (type);
+ return _err;
+ }
+ else
+ CLOSE_FINISH (sock);
+ }
else
- CLOSE_INVALIDATE (sock);
+ {
+ /* Since WARC is disabled, we are not interested in the response body. */
+ if (head_only)
+ /* Pre-1.10 Wget used CLOSE_INVALIDATE here. Now we trust the
+ servers not to send body in response to a HEAD request, and
+ those that do will likely be caught by test_socket_open.
+ If not, they can be worked around using
+ `--no-http-keep-alive'. */
+ CLOSE_FINISH (sock);
+ else if (keep_alive
+ && skip_short_body (sock, contlen, chunked_transfer_encoding))
+ /* Successfully skipped the body; also keep using the socket. */
+ CLOSE_FINISH (sock);
+ else
+ CLOSE_INVALIDATE (sock);
+ }
+
+ xfree (head);
+ xfree_null (type);
return RETRFINISHED;
}
- /* Print fetch message, if opt.verbose. */
- if (opt.verbose)
- {
- logprintf (LOG_NOTQUIET, _("Saving to: `%s'\n"),
- HYPHENP (hs->local_file) ? "STDOUT" : hs->local_file);
- }
-
+/* 2005-06-17 SMS.
+ For VMS, define common fopen() optional arguments.
+*/
+#ifdef __VMS
+# define FOPEN_OPT_ARGS "fop=sqo", "acc", acc_cb, &open_id
+# define FOPEN_BIN_FLAG 3
+#else /* def __VMS */
+# define FOPEN_BIN_FLAG true
+#endif /* def __VMS [else] */
+
/* Open the local file. */
if (!output_stream)
{
if (opt.backups)
rotate_backups (hs->local_file);
if (hs->restval)
- fp = fopen (hs->local_file, "ab");
- else if (ALLOW_CLOBBER)
- fp = fopen (hs->local_file, "wb");
+ {
+#ifdef __VMS
+ int open_id;
+
+ open_id = 21;
+ fp = fopen (hs->local_file, "ab", FOPEN_OPT_ARGS);
+#else /* def __VMS */
+ fp = fopen (hs->local_file, "ab");
+#endif /* def __VMS [else] */
+ }
+ else if (ALLOW_CLOBBER || count > 0)
+ {
+ if (opt.unlink && file_exists_p (hs->local_file))
+ {
+ int res = unlink (hs->local_file);
+ if (res < 0)
+ {
+ logprintf (LOG_NOTQUIET, "%s: %s\n", hs->local_file,
+ strerror (errno));
+ CLOSE_INVALIDATE (sock);
+ xfree (head);
+ xfree_null (type);
+ return UNLINKERR;
+ }
+ }
+
+#ifdef __VMS
+ int open_id;
+
+ open_id = 22;
+ fp = fopen (hs->local_file, "wb", FOPEN_OPT_ARGS);
+#else /* def __VMS */
+ fp = fopen (hs->local_file, "wb");
+#endif /* def __VMS [else] */
+ }
else
{
- fp = fopen_excl (hs->local_file, true);
+ fp = fopen_excl (hs->local_file, FOPEN_BIN_FLAG);
if (!fp && errno == EEXIST)
{
/* We cannot just invent a new name and use it (which is
_("%s has sprung into existence.\n"),
hs->local_file);
CLOSE_INVALIDATE (sock);
+ xfree (head);
+ xfree_null (type);
return FOPEN_EXCL_ERR;
}
}
{
logprintf (LOG_NOTQUIET, "%s: %s\n", hs->local_file, strerror (errno));
CLOSE_INVALIDATE (sock);
+ xfree (head);
+ xfree_null (type);
return FOPENERR;
}
}
else
fp = output_stream;
- /* This confuses the timestamping code that checks for file size.
- #### The timestamping code should be smarter about file size. */
- if (opt.save_headers && hs->restval == 0)
- fwrite (head, 1, strlen (head), fp);
+ /* Print fetch message, if opt.verbose. */
+ logprintf (LOG_VERBOSE, _("Saving to: %s\n"),
+ HYPHENP (hs->local_file) ? quote ("STDOUT") : quote (hs->local_file));
+
+
+ err = read_response_body (hs, sock, fp, contlen, contrange,
+ chunked_transfer_encoding,
+ u->url, warc_timestamp_str,
+ warc_request_uuid, warc_ip, type,
+ statcode, head);
/* Now we no longer need to store the response header. */
xfree (head);
-
- /* Download the request body. */
- flags = 0;
- if (contlen != -1)
- /* If content-length is present, read that much; otherwise, read
- until EOF. The HTTP spec doesn't require the server to
- actually close the connection when it's done sending data. */
- flags |= rb_read_exactly;
- if (hs->restval > 0 && contrange == 0)
- /* If the server ignored our range request, instruct fd_read_body
- to skip the first RESTVAL bytes of body. */
- flags |= rb_skip_startpos;
- hs->len = hs->restval;
- hs->rd_size = 0;
- hs->res = fd_read_body (sock, fp, contlen != -1 ? contlen : 0,
- hs->restval, &hs->rd_size, &hs->len, &hs->dltime,
- flags);
+ xfree_null (type);
if (hs->res >= 0)
CLOSE_FINISH (sock);
else
- {
- if (hs->res < 0)
- hs->rderrmsg = xstrdup (fd_errstr (sock));
- CLOSE_INVALIDATE (sock);
- }
+ CLOSE_INVALIDATE (sock);
if (!output_stream)
fclose (fp);
- if (hs->res == -2)
- return FWRITEERR;
- return RETRFINISHED;
+
+ return err;
}
/* The genuine HTTP loop! This is the part where the retrieval is
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)
+http_loop (struct url *u, struct url *original_url, char **newloc,
+ char **local_file, const char *referer, int *dt, struct url *proxy,
+ struct iri *iri)
{
int count;
- bool got_head = false; /* used for time-stamping */
+ bool got_head = false; /* used for time-stamping and filename detection */
+ bool time_came_from_head = false;
+ bool got_name = false;
char *tms;
const char *tmrate;
- uerr_t err;
+ 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;
+ char *file_name;
+ bool force_full_retrieve = false;
+
+
+ /* If we are writing to a WARC file: always retrieve the whole file. */
+ if (opt.warc_filename != NULL)
+ force_full_retrieve = 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;
here so that we don't go through the hoops if we're just using
FTP or whatever. */
if (opt.cookies)
- load_cookies();
+ load_cookies ();
/* Warn on (likely bogus) wildcard usage in HTTP. */
if (opt.ftp_glob && has_wildcards_p (u->path))
hstat.referer = referer;
if (opt.output_document)
- hstat.local_file = xstrdup (opt.output_document);
+ {
+ hstat.local_file = xstrdup (opt.output_document);
+ got_name = true;
+ }
+ else if (!opt.content_disposition)
+ {
+ hstat.local_file =
+ url_file_name (opt.trustservernames ? u : original_url, NULL);
+ got_name = true;
+ }
+
+ 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. But if the output_document was given, then this
+ test was already done and the file didn't exist. Hence the !opt.output_document */
+ get_file_flags (hstat.local_file, dt);
+ ret = RETROK;
+ goto exit;
+ }
/* Reset the counter. */
count = 0;
-
+
/* Reset the document type. */
*dt = 0;
-
+
+ /* Skip preliminary HEAD request if we're not in spider mode. */
+ if (!opt.spider)
+ send_head_first = false;
+
+ /* Send preliminary HEAD request if --content-disposition and -c are used
+ together. */
+ if (opt.content_disposition && opt.always_rest)
+ send_head_first = true;
+
+ /* Send preliminary HEAD request if -N is given and we have an existing
+ * destination file. */
+ if (!opt.output_document)
+ file_name = url_file_name (opt.trustservernames ? u : original_url, NULL);
+ else
+ file_name = xstrdup (opt.output_document);
+ if (opt.timestamping && (file_exists_p (file_name)
+ || opt.content_disposition))
+ send_head_first = true;
+ xfree (file_name);
+
/* THE loop */
do
{
/* Increment the pass counter. */
++count;
sleep_between_retrievals (count);
-
+
/* Get the current time string. */
- tms = time_str (NULL);
-
+ tms = datetime_str (time (NULL));
+
+ if (opt.spider && !got_head)
+ logprintf (LOG_VERBOSE, _("\
+Spider mode enabled. Check if remote file exists.\n"));
+
/* Print fetch message, if opt.verbose. */
if (opt.verbose)
{
- char *hurl = url_string (u, true);
-
- if (count > 1)
+ char *hurl = url_string (u, URL_AUTH_HIDE_PASSWD);
+
+ 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 (opt.spider || (opt.timestamping && !got_head))
+ if (send_head_first && !got_head)
*dt |= HEAD_ONLY;
else
*dt &= ~HEAD_ONLY;
/* Decide whether or not to restart. */
- if (opt.always_rest
+ if (force_full_retrieve)
+ hstat.restval = hstat.len;
+ else if (opt.start_pos >= 0)
+ hstat.restval = opt.start_pos;
+ else if (opt.always_rest
+ && got_name
&& stat (hstat.local_file, &st) == 0
&& S_ISREG (st.st_mode))
/* When -c is used, continue from on-disk size. (Can't use
we require a fresh get.
b) caching is explicitly inhibited. */
if ((proxy && count > 1) /* a */
- || !opt.allow_cache /* b */
- )
+ || !opt.allow_cache) /* b */
*dt |= SEND_NOCACHE;
else
*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, count);
/* Time? */
- tms = time_str (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:
/* Non-fatal errors continue executing the loop, which will
bring them to "while" statement at the end, to judge
whether the number of tries was exceeded. */
- /* free_hstat (&hstat); */
printwhat (count, opt.ntry);
continue;
- case HOSTERR: case CONIMPOSSIBLE: case PROXERR: case AUTHFAILED:
- case SSLINITFAILED: case CONTNOTSUPPORTED:
- /* Fatal errors just return from the function. */
- free_hstat (&hstat);
- return err;
case FWRITEERR: case FOPENERR:
/* Another fatal error. */
logputs (LOG_VERBOSE, "\n");
- logprintf (LOG_NOTQUIET, _("Cannot write to `%s' (%s).\n"),
- hstat.local_file, strerror (errno));
- free_hstat (&hstat);
- return err;
+ logprintf (LOG_NOTQUIET, _("Cannot write to %s (%s).\n"),
+ quote (hstat.local_file), strerror (errno));
+ case HOSTERR: case CONIMPOSSIBLE: case PROXERR: case SSLINITFAILED:
+ case CONTNOTSUPPORTED: case VERIFCERTERR: case FILEBADFILE:
+ case UNKNOWNATTR:
+ /* Fatal errors just return from the function. */
+ ret = err;
+ goto exit;
+ case ATTRMISSING:
+ /* A missing attribute in a Header is a fatal Protocol error. */
+ logputs (LOG_VERBOSE, "\n");
+ logprintf (LOG_NOTQUIET, _("Required attribute missing from Header received.\n"));
+ ret = err;
+ goto exit;
+ case AUTHFAILED:
+ logputs (LOG_VERBOSE, "\n");
+ logprintf (LOG_NOTQUIET, _("Username/Password Authentication Failed.\n"));
+ ret = err;
+ goto exit;
+ case WARC_ERR:
+ /* A fatal WARC error. */
+ logputs (LOG_VERBOSE, "\n");
+ logprintf (LOG_NOTQUIET, _("Cannot write to WARC file.\n"));
+ ret = err;
+ goto exit;
+ case WARC_TMP_FOPENERR: case WARC_TMP_FWRITEERR:
+ /* A fatal WARC error. */
+ logputs (LOG_VERBOSE, "\n");
+ logprintf (LOG_NOTQUIET, _("Cannot write to temporary WARC file.\n"));
+ ret = err;
+ goto exit;
case CONSSLERR:
/* Another fatal error. */
logprintf (LOG_NOTQUIET, _("Unable to establish SSL connection.\n"));
- free_hstat (&hstat);
- return err;
+ ret = err;
+ goto exit;
+ case UNLINKERR:
+ /* Another fatal error. */
+ logputs (LOG_VERBOSE, "\n");
+ logprintf (LOG_NOTQUIET, _("Cannot unlink %s (%s).\n"),
+ quote (hstat.local_file), strerror (errno));
+ ret = err;
+ goto exit;
case NEWLOCATION:
+ case NEWLOCATION_KEEP_POST:
/* Return the new location to the caller. */
if (!*newloc)
{
logprintf (LOG_NOTQUIET,
_("ERROR: Redirection (%d) without location.\n"),
hstat.statcode);
- free_hstat (&hstat);
- return WRONGCODE;
+ ret = WRONGCODE;
}
- free_hstat (&hstat);
- return NEWLOCATION;
+ else
+ {
+ ret = err;
+ }
+ goto exit;
case RETRUNNEEDED:
/* The file was already fully retrieved. */
- free_hstat (&hstat);
- return RETROK;
+ ret = RETROK;
+ goto exit;
case RETRFINISHED:
/* Deal with you later. */
break;
/* All possibilities should have been exhausted. */
abort ();
}
-
+
if (!(*dt & RETROKF))
{
+ char *hurl = NULL;
if (!opt.verbose)
{
/* #### Ugly ugly ugly! */
- char *hurl = url_string (u, true);
+ hurl = url_string (u, URL_AUTH_HIDE_PASSWD);
logprintf (LOG_NONVERBOSE, "%s:\n", hurl);
- xfree (hurl);
}
- logprintf (LOG_NOTQUIET, _("%s ERROR %d: %s.\n"),
- tms, hstat.statcode, escnonprint (hstat.error));
+
+ /* 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.
+ * 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)
+ hurl = url_string (u, URL_AUTH_HIDE_PASSWD);
+ nonexisting_url (hurl);
+ logprintf (LOG_NOTQUIET, _("\
+Remote file does not exist -- broken link!!!\n"));
+ }
+ else
+ {
+ logprintf (LOG_NOTQUIET, _("%s ERROR %d: %s.\n"),
+ tms, hstat.statcode,
+ quotearg_style (escape_quoting_style, hstat.error));
+ }
logputs (LOG_VERBOSE, "\n");
- free_hstat (&hstat);
- return WRONGCODE;
+ ret = WRONGCODE;
+ xfree_null (hurl);
+ goto exit;
}
/* Did we get the time-stamp? */
if (!got_head)
{
+ got_head = true; /* no more time-stamping */
+
if (opt.timestamping && !hstat.remote_time)
{
logputs (LOG_NOTQUIET, _("\
if (tmr == (time_t) (-1))
logputs (LOG_VERBOSE, _("\
Last-modified header invalid -- time-stamp ignored.\n"));
+ if (*dt & HEAD_ONLY)
+ time_came_from_head = true;
}
- }
- /* The time-stamping section. */
- if (opt.timestamping && !got_head)
- {
- got_head = true; /* no more time-stamping */
- *dt &= ~HEAD_ONLY;
- count = 0; /* the retrieve count for HEAD is reset */
-
- if (hstat.remote_time && tmr != (time_t) (-1))
+ if (send_head_first)
{
- /* Now time-stamping can be used validly. Time-stamping
- means that if the sizes of the local and remote file
- match, and local file is newer than the remote file,
- it will not be retrieved. Otherwise, the normal
- download procedure is resumed. */
- if (hstat.orig_file_tstamp >= tmr)
+ /* The time-stamping section. */
+ if (opt.timestamping)
{
- if (hstat.contlen == -1 || hstat.orig_file_size == hstat.contlen)
+ if (hstat.orig_file_name) /* Perform the following
+ checks only if the file
+ we're supposed to
+ download already exists. */
{
- logprintf (LOG_VERBOSE, _("\
-Server file no newer than local file `%s' -- not retrieving.\n\n"),
- hstat.orig_file_name);
- free_hstat (&hstat);
- return RETROK;
+ if (hstat.remote_time &&
+ tmr != (time_t) (-1))
+ {
+ /* Now time-stamping can be used validly.
+ Time-stamping means that if the sizes of
+ the local and remote file match, and local
+ file is newer than the remote file, it will
+ not be retrieved. Otherwise, the normal
+ download procedure is resumed. */
+ if (hstat.orig_file_tstamp >= tmr)
+ {
+ if (hstat.contlen == -1
+ || hstat.orig_file_size == hstat.contlen)
+ {
+ logprintf (LOG_VERBOSE, _("\
+Server file no newer than local file %s -- not retrieving.\n\n"),
+ quote (hstat.orig_file_name));
+ ret = RETROK;
+ goto exit;
+ }
+ else
+ {
+ logprintf (LOG_VERBOSE, _("\
+The sizes do not match (local %s) -- retrieving.\n"),
+ number_to_static_string (hstat.orig_file_size));
+ }
+ }
+ else
+ {
+ force_full_retrieve = true;
+ logputs (LOG_VERBOSE,
+ _("Remote file is newer, retrieving.\n"));
+ }
+
+ logputs (LOG_VERBOSE, "\n");
+ }
+ }
+
+ /* free_hstat (&hstat); */
+ hstat.timestamp_checked = true;
+ }
+
+ 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. */
+ }
}
else
{
- logprintf (LOG_VERBOSE, _("\
-The sizes do not match (local %s) -- retrieving.\n"),
- number_to_static_string (local_size));
+ if (*dt & TEXTHTML)
+ {
+ logprintf (LOG_VERBOSE, _("\
+Remote file exists and could contain further links,\n\
+but recursion is disabled -- not retrieving.\n\n"));
+ }
+ else
+ {
+ logprintf (LOG_VERBOSE, _("\
+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;
}
}
- else
- logputs (LOG_VERBOSE,
- _("Remote file is newer, retrieving.\n"));
- logputs (LOG_VERBOSE, "\n");
- }
-
- /* free_hstat (&hstat); */
- hstat.timestamp_checked = true;
- continue;
- }
-
- if ((tmr != (time_t) (-1))
- && !opt.spider
+ got_name = true;
+ *dt &= ~HEAD_ONLY;
+ count = 0; /* the retrieve count for HEAD is reset */
+ continue;
+ } /* send_head_first */
+ } /* !got_head */
+
+ if (opt.useservertimestamps
+ && (tmr != (time_t) (-1))
&& ((hstat.len == hstat.contlen) ||
((hstat.res == 0) && (hstat.contlen == -1))))
{
- /* #### This code repeats in http.c and ftp.c. Move it to a
- function! */
const char *fl = NULL;
- if (opt.output_document)
+ set_local_file (&fl, hstat.local_file);
+ if (fl)
{
- if (output_stream_regular)
- fl = opt.output_document;
+ time_t newtmr = -1;
+ /* Reparse time header, in case it's changed. */
+ if (time_came_from_head
+ && hstat.remote_time && hstat.remote_time[0])
+ {
+ newtmr = http_atotm (hstat.remote_time);
+ if (newtmr != (time_t)-1)
+ tmr = newtmr;
+ }
+ touch (fl, tmr);
}
- else
- fl = hstat.local_file;
- if (fl)
- touch (fl, tmr);
}
/* End of time-stamping section. */
- if (opt.spider)
- {
- logprintf (LOG_NOTQUIET, "%d %s\n\n", hstat.statcode,
- escnonprint (hstat.error));
- return RETROK;
- }
-
tmrate = retr_rate (hstat.rd_size, hstat.dltime);
total_download_time += hstat.dltime;
{
if (*dt & RETROKF)
{
+ bool write_to_stdout = (opt.output_document && HYPHENP (opt.output_document));
+
logprintf (LOG_VERBOSE,
- _("%s (%s) - `%s' saved [%s/%s]\n\n"),
- tms, tmrate, hstat.local_file,
+ write_to_stdout
+ ? _("%s (%s) - written to stdout %s[%s/%s]\n\n")
+ : _("%s (%s) - %s saved [%s/%s]\n\n"),
+ tms, tmrate,
+ write_to_stdout ? "" : quote (hstat.local_file),
number_to_static_string (hstat.len),
number_to_static_string (hstat.contlen));
logprintf (LOG_NONVERBOSE,
number_to_static_string (hstat.contlen),
hstat.local_file, count);
}
- ++opt.numurls;
- total_downloaded_bytes += hstat.len;
+ ++numurls;
+ total_downloaded_bytes += hstat.rd_size;
/* Remember that we downloaded the file for later ".orig" code. */
if (*dt & ADDED_HTML_EXTENSION)
- downloaded_file(FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED, hstat.local_file);
+ downloaded_file (FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED, hstat.local_file);
else
- downloaded_file(FILE_DOWNLOADED_NORMALLY, hstat.local_file);
+ downloaded_file (FILE_DOWNLOADED_NORMALLY, hstat.local_file);
- free_hstat (&hstat);
- return RETROK;
+ ret = RETROK;
+ goto exit;
}
else if (hstat.res == 0) /* No read error */
{
if (hstat.contlen == -1) /* We don't know how much we were supposed
- to get, so assume we succeeded. */
+ to get, so assume we succeeded. */
{
if (*dt & RETROKF)
{
+ bool write_to_stdout = (opt.output_document && HYPHENP (opt.output_document));
+
logprintf (LOG_VERBOSE,
- _("%s (%s) - `%s' saved [%s]\n\n"),
- tms, tmrate, hstat.local_file,
+ write_to_stdout
+ ? _("%s (%s) - written to stdout %s[%s]\n\n")
+ : _("%s (%s) - %s saved [%s]\n\n"),
+ tms, tmrate,
+ write_to_stdout ? "" : quote (hstat.local_file),
number_to_static_string (hstat.len));
logprintf (LOG_NONVERBOSE,
"%s URL:%s [%s] -> \"%s\" [%d]\n",
tms, u->url, number_to_static_string (hstat.len),
hstat.local_file, count);
}
- ++opt.numurls;
- total_downloaded_bytes += hstat.len;
+ ++numurls;
+ total_downloaded_bytes += hstat.rd_size;
/* Remember that we downloaded the file for later ".orig" code. */
if (*dt & ADDED_HTML_EXTENSION)
- downloaded_file(FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED, hstat.local_file);
+ downloaded_file (FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED, hstat.local_file);
else
- downloaded_file(FILE_DOWNLOADED_NORMALLY, hstat.local_file);
-
- free_hstat (&hstat);
- return RETROK;
+ downloaded_file (FILE_DOWNLOADED_NORMALLY, hstat.local_file);
+
+ ret = RETROK;
+ goto exit;
}
else if (hstat.len < hstat.contlen) /* meaning we lost the
connection too soon */
_("%s (%s) - Connection closed at byte %s. "),
tms, tmrate, number_to_static_string (hstat.len));
printwhat (count, opt.ntry);
- /* free_hstat (&hstat); */
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 */
{
tms, tmrate, number_to_static_string (hstat.len),
hstat.rderrmsg);
printwhat (count, opt.ntry);
- /* free_hstat (&hstat); */
continue;
}
else /* hstat.res == -1 and contlen is given */
number_to_static_string (hstat.contlen),
hstat.rderrmsg);
printwhat (count, opt.ntry);
- /* free_hstat (&hstat); */
continue;
}
}
/* not reached */
}
while (!opt.ntry || (count < opt.ntry));
-
- return TRYLIMEXC;
+
+exit:
+ if (ret == RETROK && local_file)
+ *local_file = xstrdup (hstat.local_file);
+ free_hstat (&hstat);
+
+ return ret;
}
\f
/* Check whether the result of strptime() indicates success.
{
if (!p)
return false;
- while (ISSPACE (*p))
+ while (c_isspace (*p))
++p;
if (!*p
|| (p[0] == 'G' && p[1] == 'M' && p[2] == 'T')
- || ((p[0] == '+' || p[0] == '-') && ISDIGIT (p[1])))
+ || ((p[0] == '+' || p[0] == '-') && c_isdigit (p[1])))
return true;
else
return false;
Netscape cookie specification.) */
};
const char *oldlocale;
- int i;
+ char savedlocale[256];
+ size_t i;
time_t ret = (time_t) -1;
/* Solaris strptime fails to recognize English month names in
non-English locales, which we work around by temporarily setting
locale to C before invoking strptime. */
oldlocale = setlocale (LC_TIME, NULL);
+ if (oldlocale)
+ {
+ size_t l = strlen (oldlocale) + 1;
+ if (l >= sizeof savedlocale)
+ savedlocale[0] = '\0';
+ else
+ memcpy (savedlocale, oldlocale, l);
+ }
+ else savedlocale[0] = '\0';
+
setlocale (LC_TIME, "C");
for (i = 0; i < countof (time_formats); i++)
}
/* Restore the previous locale. */
- setlocale (LC_TIME, oldlocale);
+ if (savedlocale[0])
+ setlocale (LC_TIME, savedlocale);
return ret;
}
}
#define SKIP_WS(x) do { \
- while (ISSPACE (*(x))) \
+ while (c_isspace (*(x))) \
++(x); \
} while (0)
#ifdef ENABLE_DIGEST
-/* Parse HTTP `WWW-Authenticate:' header. AU points to the beginning
- of a field in such a header. If the field is the one specified by
- ATTR_NAME ("realm", "opaque", and "nonce" are used by the current
- digest authorization code), extract its value in the (char*)
- variable pointed by RET. Returns negative on a malformed header,
- or number of bytes that have been parsed by this call. */
-static int
-extract_header_attr (const char *au, const char *attr_name, char **ret)
-{
- const char *ep;
- const char *cp = au;
-
- if (strncmp (cp, attr_name, strlen (attr_name)) == 0)
- {
- cp += strlen (attr_name);
- if (!*cp)
- return -1;
- SKIP_WS (cp);
- if (*cp != '=')
- return -1;
- if (!*++cp)
- return -1;
- SKIP_WS (cp);
- if (*cp != '\"')
- return -1;
- if (!*++cp)
- return -1;
- for (ep = cp; *ep && *ep != '\"'; ep++)
- ;
- if (!*ep)
- return -1;
- xfree_null (*ret);
- *ret = strdupdelim (cp, ep);
- return ep - au + 1;
- }
- else
- return 0;
-}
-
/* Dump the hexadecimal representation of HASH to BUF. HASH should be
an array of 16 bytes containing the hash keys, and BUF should be a
buffer of 33 writable characters (32 for hex digits plus one for
{
int i;
- for (i = 0; i < MD5_HASHLEN; i++, hash++)
+ for (i = 0; i < MD5_DIGEST_SIZE; i++, hash++)
{
*buf++ = XNUM_TO_digit (*hash >> 4);
*buf++ = XNUM_TO_digit (*hash & 0xf);
static char *
digest_authentication_encode (const char *au, const char *user,
const char *passwd, const char *method,
- const char *path)
+ const char *path, uerr_t *auth_err)
{
- static char *realm, *opaque, *nonce;
+ static char *realm, *opaque, *nonce, *qop, *algorithm;
static struct {
const char *name;
char **variable;
} options[] = {
{ "realm", &realm },
{ "opaque", &opaque },
- { "nonce", &nonce }
+ { "nonce", &nonce },
+ { "qop", &qop },
+ { "algorithm", &algorithm }
};
+ char cnonce[16] = "";
char *res;
+ int res_len;
+ size_t res_size;
+ param_token name, value;
+
- realm = opaque = nonce = NULL;
+ realm = opaque = nonce = algorithm = qop = NULL;
au += 6; /* skip over `Digest' */
- while (*au)
+ while (extract_param (&au, &name, &value, ',', NULL))
{
- int i;
-
- SKIP_WS (au);
+ size_t i;
+ size_t namelen = name.e - name.b;
for (i = 0; i < countof (options); i++)
- {
- int skip = extract_header_attr (au, options[i].name,
- options[i].variable);
- if (skip < 0)
- {
- xfree_null (realm);
- xfree_null (opaque);
- xfree_null (nonce);
- return NULL;
- }
- else if (skip)
- {
- au += skip;
- break;
- }
- }
- if (i == countof (options))
- {
- while (*au && *au != '=')
- au++;
- if (*au && *++au)
- {
- SKIP_WS (au);
- if (*au == '\"')
- {
- au++;
- while (*au && *au != '\"')
- au++;
- if (*au)
- au++;
- }
- }
- }
- while (*au && *au != ',')
- au++;
- if (*au)
- au++;
+ if (namelen == strlen (options[i].name)
+ && 0 == strncmp (name.b, options[i].name,
+ namelen))
+ {
+ *options[i].variable = strdupdelim (value.b, value.e);
+ break;
+ }
+ }
+
+ if (qop != NULL && strcmp(qop,"auth"))
+ {
+ logprintf (LOG_NOTQUIET, _("Unsupported quality of protection '%s'.\n"), qop);
+ xfree_null (qop); /* force freeing mem and return */
+ qop = NULL;
}
- if (!realm || !nonce || !user || !passwd || !path || !method)
+ else if (algorithm != NULL && strcmp (algorithm,"MD5") && strcmp (algorithm,"MD5-sess"))
+ {
+ logprintf (LOG_NOTQUIET, _("Unsupported algorithm '%s'.\n"), algorithm);
+ xfree_null (qop); /* force freeing mem and return */
+ qop = NULL;
+ }
+
+ if (!realm || !nonce || !user || !passwd || !path || !method || !qop)
{
xfree_null (realm);
xfree_null (opaque);
xfree_null (nonce);
+ xfree_null (qop);
+ xfree_null (algorithm);
+ if (!qop)
+ *auth_err = UNKNOWNATTR;
+ else
+ *auth_err = ATTRMISSING;
return NULL;
}
/* Calculate the digest value. */
{
- ALLOCA_MD5_CONTEXT (ctx);
- unsigned char hash[MD5_HASHLEN];
- char a1buf[MD5_HASHLEN * 2 + 1], a2buf[MD5_HASHLEN * 2 + 1];
- char response_digest[MD5_HASHLEN * 2 + 1];
+ struct md5_ctx ctx;
+ unsigned char hash[MD5_DIGEST_SIZE];
+ char a1buf[MD5_DIGEST_SIZE * 2 + 1], a2buf[MD5_DIGEST_SIZE * 2 + 1];
+ char response_digest[MD5_DIGEST_SIZE * 2 + 1];
/* A1BUF = H(user ":" realm ":" password) */
- gen_md5_init (ctx);
- gen_md5_update ((unsigned char *)user, strlen (user), ctx);
- gen_md5_update ((unsigned char *)":", 1, ctx);
- gen_md5_update ((unsigned char *)realm, strlen (realm), ctx);
- gen_md5_update ((unsigned char *)":", 1, ctx);
- gen_md5_update ((unsigned char *)passwd, strlen (passwd), ctx);
- gen_md5_finish (ctx, hash);
+ md5_init_ctx (&ctx);
+ md5_process_bytes ((unsigned char *)user, strlen (user), &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)realm, strlen (realm), &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)passwd, strlen (passwd), &ctx);
+ md5_finish_ctx (&ctx, hash);
+
dump_hash (a1buf, hash);
+ if (algorithm && !strcmp (algorithm, "MD5-sess"))
+ {
+ /* A1BUF = H( H(user ":" realm ":" password) ":" nonce ":" cnonce ) */
+ snprintf (cnonce, sizeof (cnonce), "%08x", random_number(INT_MAX));
+
+ md5_init_ctx (&ctx);
+ // md5_process_bytes (hash, MD5_DIGEST_SIZE, &ctx);
+ md5_process_bytes (a1buf, MD5_DIGEST_SIZE * 2, &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)nonce, strlen (nonce), &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)cnonce, strlen (cnonce), &ctx);
+ md5_finish_ctx (&ctx, hash);
+
+ dump_hash (a1buf, hash);
+ }
+
/* A2BUF = H(method ":" path) */
- gen_md5_init (ctx);
- gen_md5_update ((unsigned char *)method, strlen (method), ctx);
- gen_md5_update ((unsigned char *)":", 1, ctx);
- gen_md5_update ((unsigned char *)path, strlen (path), ctx);
- gen_md5_finish (ctx, hash);
+ md5_init_ctx (&ctx);
+ md5_process_bytes ((unsigned char *)method, strlen (method), &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)path, strlen (path), &ctx);
+ md5_finish_ctx (&ctx, hash);
dump_hash (a2buf, hash);
- /* RESPONSE_DIGEST = H(A1BUF ":" nonce ":" A2BUF) */
- gen_md5_init (ctx);
- gen_md5_update ((unsigned char *)a1buf, MD5_HASHLEN * 2, ctx);
- gen_md5_update ((unsigned char *)":", 1, ctx);
- gen_md5_update ((unsigned char *)nonce, strlen (nonce), ctx);
- gen_md5_update ((unsigned char *)":", 1, ctx);
- gen_md5_update ((unsigned char *)a2buf, MD5_HASHLEN * 2, ctx);
- gen_md5_finish (ctx, hash);
+ if (qop && (!strcmp(qop, "auth") || !strcmp (qop, "auth-int")))
+ {
+ /* RFC 2617 Digest Access Authentication */
+ /* generate random hex string */
+ if (!*cnonce)
+ snprintf(cnonce, sizeof(cnonce), "%08x", random_number(INT_MAX));
+
+ /* RESPONSE_DIGEST = H(A1BUF ":" nonce ":" noncecount ":" clientnonce ":" qop ": " A2BUF) */
+ md5_init_ctx (&ctx);
+ md5_process_bytes ((unsigned char *)a1buf, MD5_DIGEST_SIZE * 2, &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)nonce, strlen (nonce), &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)"00000001", 8, &ctx); /* TODO: keep track of server nonce values */
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)cnonce, strlen(cnonce), &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)qop, strlen(qop), &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)a2buf, MD5_DIGEST_SIZE * 2, &ctx);
+ md5_finish_ctx (&ctx, hash);
+ }
+ else
+ {
+ /* RFC 2069 Digest Access Authentication */
+ /* RESPONSE_DIGEST = H(A1BUF ":" nonce ":" A2BUF) */
+ md5_init_ctx (&ctx);
+ md5_process_bytes ((unsigned char *)a1buf, MD5_DIGEST_SIZE * 2, &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)nonce, strlen (nonce), &ctx);
+ md5_process_bytes ((unsigned char *)":", 1, &ctx);
+ md5_process_bytes ((unsigned char *)a2buf, MD5_DIGEST_SIZE * 2, &ctx);
+ md5_finish_ctx (&ctx, hash);
+ }
+
dump_hash (response_digest, hash);
- res = xmalloc (strlen (user)
- + strlen (user)
- + strlen (realm)
- + strlen (nonce)
- + strlen (path)
- + 2 * MD5_HASHLEN /*strlen (response_digest)*/
- + (opaque ? strlen (opaque) : 0)
- + 128);
- sprintf (res, "Digest \
-username=\"%s\", realm=\"%s\", nonce=\"%s\", uri=\"%s\", response=\"%s\"",
- user, realm, nonce, path, response_digest);
+ res_size = strlen (user)
+ + strlen (realm)
+ + strlen (nonce)
+ + strlen (path)
+ + 2 * MD5_DIGEST_SIZE /*strlen (response_digest)*/
+ + (opaque ? strlen (opaque) : 0)
+ + (algorithm ? strlen (algorithm) : 0)
+ + (qop ? 128: 0)
+ + strlen (cnonce)
+ + 128;
+
+ res = xmalloc (res_size);
+
+ if (qop && !strcmp (qop, "auth"))
+ {
+ res_len = snprintf (res, res_size, "Digest "\
+ "username=\"%s\", realm=\"%s\", nonce=\"%s\", uri=\"%s\", response=\"%s\""\
+ ", qop=auth, nc=00000001, cnonce=\"%s\"",
+ user, realm, nonce, path, response_digest, cnonce);
+
+ }
+ else
+ {
+ res_len = snprintf (res, res_size, "Digest "\
+ "username=\"%s\", realm=\"%s\", nonce=\"%s\", uri=\"%s\", response=\"%s\"",
+ user, realm, nonce, path, response_digest);
+ }
+
if (opaque)
{
- char *p = res + strlen (res);
- strcat (p, ", opaque=\"");
- strcat (p, opaque);
- strcat (p, "\"");
+ res_len += snprintf(res + res_len, res_size - res_len, ", opaque=\"%s\"", opaque);
+ }
+
+ if (algorithm)
+ {
+ snprintf(res + res_len, res_size - res_len, ", algorithm=\"%s\"", algorithm);
}
}
+
+ xfree_null (realm);
+ xfree_null (opaque);
+ xfree_null (nonce);
+ xfree_null (qop);
+ xfree_null (algorithm);
+
return res;
}
#endif /* ENABLE_DIGEST */
first argument and are followed by whitespace or terminating \0.
The comparison is case-insensitive. */
#define STARTS(literal, b, e) \
- ((e) - (b) >= STRSIZE (literal) \
+ ((e > b) \
+ && ((size_t) ((e) - (b))) >= STRSIZE (literal) \
&& 0 == strncasecmp (b, literal, STRSIZE (literal)) \
- && ((e) - (b) == STRSIZE (literal) \
- || ISSPACE (b[STRSIZE (literal)])))
+ && ((size_t) ((e) - (b)) == STRSIZE (literal) \
+ || c_isspace (b[STRSIZE (literal)])))
static bool
known_authentication_scheme_p (const char *hdrbeg, const char *hdrend)
static char *
create_authorization_line (const char *au, const char *user,
const char *passwd, const char *method,
- const char *path, bool *finished)
+ const char *path, bool *finished, uerr_t *auth_err)
{
/* We are called only with known schemes, so we can dispatch on the
first letter. */
- switch (TOUPPER (*au))
+ switch (c_toupper (*au))
{
case 'B': /* Basic */
*finished = true;
#ifdef ENABLE_DIGEST
case 'D': /* Digest */
*finished = true;
- return digest_authentication_encode (au, user, passwd, method, path);
+ return digest_authentication_encode (au, user, passwd, method, path, auth_err);
#endif
#ifdef ENABLE_NTLM
case 'N': /* NTLM */
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 - 1] = '\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
-char *
-test_parse_content_disposition()
+const char *
+test_parse_content_disposition(void)
{
- int i;
- struct {
- char *hdrval;
- char *filename;
+ unsigned i;
+ static const struct {
+ const char *hdrval;
+ const char *filename;
bool result;
} test_array[] = {
{ "filename=\"file.ext\"", "file.ext", true },
{ "attachment; filename=\"file.ext\"", "file.ext", true },
{ "attachment; filename=\"file.ext\"; dummy", "file.ext", true },
- { "attachment", NULL, false },
+ { "attachment", NULL, false },
+ { "attachement; filename*=UTF-8'en-US'hello.txt", "hello.txt", true },
+ { "attachement; filename*0=\"hello\"; filename*1=\"world.txt\"", "helloworld.txt", true },
};
-
- for (i = 0; i < sizeof(test_array)/sizeof(test_array[0]); ++i)
+
+ for (i = 0; i < countof(test_array); ++i)
{
char *filename;
- bool res = parse_content_disposition (test_array[i].hdrval, &filename);
+ bool res;
- mu_assert ("test_parse_content_disposition: wrong result",
+ res = parse_content_disposition (test_array[i].hdrval, &filename);
+
+ mu_assert ("test_parse_content_disposition: wrong result",
res == test_array[i].result
- && (res == false
+ && (res == false
|| 0 == strcmp (test_array[i].filename, filename)));
-
- /* printf ("test %d: %s\n", i, res == false ? "false" : filename); */
}
return NULL;
#endif /* TESTING */
/*
- * vim: et ts=2 sw=2
+ * vim: et sts=2 sw=2 cino+={s
*/
projects / wget / blobdiff