http.c

一个从网络上自动下载文件的自由工具

/* HTTP support.
   Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003,
   2004, 2005, 2006, 2007, 2008 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 3 of the License, or
 (at your option) any later version.

GNU Wget is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with Wget.  If not, see <http://www.gnu.org/licenses/>.

Additional permission under GNU GPL version 3 section 7

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 <config.h>

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#ifdef HAVE_UNISTD_H
# include <unistd.h>
#endif
#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 "host.h"
#include "retr.h"
#include "connect.h"
#include "netrc.h"
#ifdef HAVE_SSL
# include "ssl.h"
#endif
#ifdef ENABLE_NTLM
# include "http-ntlm.h"
#endif
#include "cookies.h"
#ifdef ENABLE_DIGEST
# include "gen-md5.h"
#endif
#include "convert.h"
#include "spider.h"

#ifdef TESTING
#include "test.h"
#endif

extern char *version_string;

/* Forward decls. */
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);

#ifndef MIN
# define MIN(x, y) ((x) > (y) ? (y) : (x))
#endif


static bool cookies_loaded_p;
static struct cookie_jar *wget_cookie_jar;

#define TEXTHTML_S "text/html"
#define TEXTXHTML_S "application/xhtml+xml"

/* Some status code validation macros: */
#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          \
                         || (x) == HTTP_STATUS_MOVED_TEMPORARILY       \
                         || (x) == HTTP_STATUS_SEE_OTHER               \
                         || (x) == HTTP_STATUS_TEMPORARY_REDIRECT)

/* HTTP/1.0 status codes from RFC1945, provided for reference.  */
/* Successful 2xx.  */
#define HTTP_STATUS_OK                    200
#define HTTP_STATUS_CREATED               201
#define HTTP_STATUS_ACCEPTED              202
#define HTTP_STATUS_NO_CONTENT            204
#define HTTP_STATUS_PARTIAL_CONTENTS      206

/* Redirection 3xx.  */
#define HTTP_STATUS_MULTIPLE_CHOICES      300
#define HTTP_STATUS_MOVED_PERMANENTLY     301
#define HTTP_STATUS_MOVED_TEMPORARILY     302
#define HTTP_STATUS_SEE_OTHER             303 /* from HTTP/1.1 */
#define HTTP_STATUS_NOT_MODIFIED          304
#define HTTP_STATUS_TEMPORARY_REDIRECT    307 /* from HTTP/1.1 */

/* Client error 4xx.  */
#define HTTP_STATUS_BAD_REQUEST           400
#define HTTP_STATUS_UNAUTHORIZED          401
#define HTTP_STATUS_FORBIDDEN             403
#define HTTP_STATUS_NOT_FOUND             404
#define HTTP_STATUS_RANGE_NOT_SATISFIABLE 416

/* Server errors 5xx.  */
#define HTTP_STATUS_INTERNAL              500
#define HTTP_STATUS_NOT_IMPLEMENTED       501
#define HTTP_STATUS_BAD_GATEWAY           502
#define HTTP_STATUS_UNAVAILABLE           503

enum rp {
  rel_none, rel_name, rel_value, rel_both
};

struct request {
  const char *method;
  char *arg;

  struct request_header {
    char *name, *value;
    enum rp release_policy;
  } *headers;
  int hcount, hcapacity;
};

/* Create a new, empty request.  At least request_set_method must be
   called before the request can be used.  */

static struct request *
request_new (void)
{
  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->arg = arg;
}

/* Return the method string passed with the last call to
   request_set_method.  */

static const char *
request_method (const struct request *req)
{
  return req->method;
}

/* Free one header according to the release policy specified with
   request_set_header.  */

static void
release_header (struct request_header *hdr)
{
  switch (hdr->release_policy)
    {
    case rel_none:
      break;
    case rel_name:
      xfree (hdr->name);
      break;
    case rel_value:
      xfree (hdr->value);
      break;
    case rel_both:
      xfree (hdr->name);
      xfree (hdr->value);
      break;
    }
}

/* Set the request named NAME to VALUE.  Specifically, this means that
   a "NAME: VALUE\r\n" header line will be used in the request.  If a
   header with the same name previously existed in the request, its
   value will be replaced by this one.  A NULL value means do nothing.

   RELEASE_POLICY determines whether NAME and VALUE should be released
   (freed) with request_free.  Allowed values are:

    - rel_none     - don't free NAME or VALUE
    - rel_name     - free NAME when done
    - rel_value    - free VALUE when done
    - rel_both     - free both NAME and VALUE when done

   Setting release policy is useful when arguments come from different
   sources.  For example:

     // Don't free literal strings!
     request_set_header (req, "Pragma", "no-cache", rel_none);

     // Don't free a global variable, we'll need it later.
     request_set_header (req, "Referer", opt.referer, rel_none);

     // Value freshly allocated, free it when done.
     request_set_header (req, "Range",
                         aprintf ("bytes=%s-", number_to_static_string (hs->restval)),
                         rel_value);
   */

static void
request_set_header (struct request *req, char *name, char *value,
                    enum rp release_policy)
{
  struct request_header *hdr;
  int i;

  if (!value)
    {
      /* 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);
      return;
    }

  for (i = 0; i < req->hcount; i++)
    {
      hdr = &req->headers[i];
      if (0 == strcasecmp (name, hdr->name))
        {
          /* Replace existing header. */
          release_header (hdr);
          hdr->name = name;
          hdr->value = value;
          hdr->release_policy = release_policy;
          return;
        }
    }

  /* Install new header. */

  if (req->hcount >= req->hcapacity)
    {
      req->hcapacity <<= 1;
      req->headers = xrealloc (req->headers, req->hcapacity * sizeof (*hdr));
    }
  hdr = &req->headers[req->hcount++];
  hdr->name = name;
  hdr->value = value;
  hdr->release_policy = release_policy;
}

/* Like request_set_header, but sets the whole header line, as
   provided by the user using the `--header' option.  For example,
   request_set_user_header (req, "Foo: bar") works just like
   request_set_header (req, "Foo", "bar").  */

static void
request_set_user_header (struct request *req, const char *header)
{
  char *name;
  const char *p = strchr (header, ':');
  if (!p)
    return;
  BOUNDED_TO_ALLOCA (header, p, name);
  ++p;
  while (ISSPACE (*p))
    ++p;
  request_set_header (req, xstrdup (name), (char *) p, rel_name);
}

/* Remove the header with specified name from REQ.  Returns true if
   the header was actually removed, false otherwise.  */

static bool
request_remove_header (struct request *req, char *name)
{
  int i;
  for (i = 0; i < req->hcount; i++)
    {
      struct request_header *hdr = &req->headers[i];
      if (0 == strcasecmp (name, hdr->name))
        {
          release_header (hdr);
          /* Move the remaining headers by one. */
          if (i < req->hcount - 1)
            memmove (hdr, hdr + 1, (req->hcount - i - 1) * sizeof (*hdr));
          --req->hcount;
          return true;
        }
    }
  return false;
}

#define APPEND(p, str) do {                     \
  int A_len = strlen (str);                     \
  memcpy (p, str, A_len);                       \
  p += A_len;                                   \
} while (0)

/* Construct the request and write it to FD using fd_write.  */

static int
request_send (const struct request *req, int fd)
{
  char *request_string, *p;
  int i, size, write_error;

  /* Count the request size. */
  size = 0;

  /* METHOD " " ARG " " "HTTP/1.0" "\r\n" */
  size += strlen (req->method) + 1 + strlen (req->arg) + 1 + 8 + 2;

  for (i = 0; i < req->hcount; i++)
    {
      struct request_header *hdr = &req->headers[i];
      /* NAME ": " VALUE "\r\n" */
      size += strlen (hdr->name) + 2 + strlen (hdr->value) + 2;
    }

  /* "\r\n\0" */
  size += 3;

  p = request_string = alloca_array (char, size);

  /* Generate the request. */

  APPEND (p, req->method); *p++ = ' ';
  APPEND (p, req->arg);    *p++ = ' ';
  memcpy (p, "HTTP/1.0\r\n", 10); p += 10;

  for (i = 0; i < req->hcount; i++)
    {
      struct request_header *hdr = &req->headers[i];
      APPEND (p, hdr->name);
      *p++ = ':', *p++ = ' ';
      APPEND (p, hdr->value);
      *p++ = '\r', *p++ = '\n';
    }

  *p++ = '\r', *p++ = '\n', *p++ = '\0';
  assert (p - request_string == size);

#undef APPEND

  DEBUGP (("\n---request begin---\n%s---request end---\n", request_string));

  /* Send the request to the server. */

  write_error = fd_write (fd, request_string, size - 1, -1);
  if (write_error < 0)
    logprintf (LOG_VERBOSE, _("Failed writing HTTP request: %s.\n"),
               fd_errstr (fd));
  return write_error;
}

/* Release the resources used by REQ. */

static void
request_free (struct request *req)
{
  int i;
  xfree_null (req->arg);
  for (i = 0; i < req->hcount; i++)
    release_header (&req->headers[i]);
  xfree_null (req->headers);
  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 int
maybe_send_basic_creds (const char *hostname, const char *user,
                        const char *passwd, struct request *req)
{
  int did_challenge = 0;

  if (basic_authed_hosts
      && hash_table_contains(basic_authed_hosts, hostname))
    {
      DEBUGP(("Found `%s' in basic_authed_hosts.\n", hostname));
      request_set_header (req, "Authorization",
                          basic_authentication_encode (user, passwd),
                          rel_value);
      did_challenge = 1;
    }
  else
    {
      DEBUGP(("Host `%s' has not issued a general basic challenge.\n",
              hostname));