html-url.c

一个从网络上自动下载文件的自由工具

/* Collect URLs from HTML source.
   Copyright (C) 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 <string.h>
#include <stdlib.h>
#include <errno.h>
#include <assert.h>

#include "wget.h"
#include "html-parse.h"
#include "url.h"
#include "utils.h"
#include "hash.h"
#include "convert.h"
#include "recur.h"              /* declaration of get_urls_html */

struct map_context;

typedef void (*tag_handler_t) (int, struct taginfo *, struct map_context *);

#define DECLARE_TAG_HANDLER(fun)                                \
  static void fun (int, struct taginfo *, struct map_context *)

DECLARE_TAG_HANDLER (tag_find_urls);
DECLARE_TAG_HANDLER (tag_handle_base);
DECLARE_TAG_HANDLER (tag_handle_form);
DECLARE_TAG_HANDLER (tag_handle_link);
DECLARE_TAG_HANDLER (tag_handle_meta);

enum {
  TAG_A,
  TAG_APPLET,
  TAG_AREA,
  TAG_BASE,
  TAG_BGSOUND,
  TAG_BODY,
  TAG_EMBED,
  TAG_FIG,
  TAG_FORM,
  TAG_FRAME,
  TAG_IFRAME,
  TAG_IMG,
  TAG_INPUT,
  TAG_LAYER,
  TAG_LINK,
  TAG_META,
  TAG_OBJECT,
  TAG_OVERLAY,
  TAG_SCRIPT,
  TAG_TABLE,
  TAG_TD,
  TAG_TH
};

/* The list of known tags and functions used for handling them.  Most
   tags are simply harvested for URLs. */
static struct known_tag {
  int tagid;
  const char *name;
  tag_handler_t handler;
} known_tags[] = {
  { TAG_A,       "a",           tag_find_urls },
  { TAG_APPLET,  "applet",      tag_find_urls },
  { TAG_AREA,    "area",        tag_find_urls },
  { TAG_BASE,    "base",        tag_handle_base },
  { TAG_BGSOUND, "bgsound",     tag_find_urls },
  { TAG_BODY,    "body",        tag_find_urls },
  { TAG_EMBED,   "embed",       tag_find_urls },
  { TAG_FIG,     "fig",         tag_find_urls },
  { TAG_FORM,    "form",        tag_handle_form },
  { TAG_FRAME,   "frame",       tag_find_urls },
  { TAG_IFRAME,  "iframe",      tag_find_urls },
  { TAG_IMG,     "img",         tag_find_urls },
  { TAG_INPUT,   "input",       tag_find_urls },
  { TAG_LAYER,   "layer",       tag_find_urls },
  { TAG_LINK,    "link",        tag_handle_link },
  { TAG_META,    "meta",        tag_handle_meta },
  { TAG_OBJECT,  "object",      tag_find_urls },
  { TAG_OVERLAY, "overlay",     tag_find_urls },
  { TAG_SCRIPT,  "script",      tag_find_urls },
  { TAG_TABLE,   "table",       tag_find_urls },
  { TAG_TD,      "td",          tag_find_urls },
  { TAG_TH,      "th",          tag_find_urls }
};

/* tag_url_attributes documents which attributes of which tags contain
   URLs to harvest.  It is used by tag_find_urls.  */

/* Defines for the FLAGS. */

/* The link is "inline", i.e. needs to be retrieved for this document
   to be correctly rendered.  Inline links include inlined images,
   stylesheets, children frames, etc.  */
#define ATTR_INLINE     1

/* The link is expected to yield HTML contents.  It's important not to
   try to follow HTML obtained by following e.g. <img src="...">
   regardless of content-type.  Doing this causes infinite loops for
   "images" that return non-404 error pages with links to the same
   image.  */
#define ATTR_HTML       2

/* For tags handled by tag_find_urls: attributes that contain URLs to
   download. */
static struct {
  int tagid;
  const char *attr_name;
  int flags;
} tag_url_attributes[] = {
  { TAG_A,              "href",         ATTR_HTML },
  { TAG_APPLET,         "code",         ATTR_INLINE },
  { TAG_AREA,           "href",         ATTR_HTML },
  { TAG_BGSOUND,        "src",          ATTR_INLINE },
  { TAG_BODY,           "background",   ATTR_INLINE },
  { TAG_EMBED,          "href",         ATTR_HTML },
  { TAG_EMBED,          "src",          ATTR_INLINE | ATTR_HTML },
  { TAG_FIG,            "src",          ATTR_INLINE },
  { TAG_FRAME,          "src",          ATTR_INLINE | ATTR_HTML },
  { TAG_IFRAME,         "src",          ATTR_INLINE | ATTR_HTML },
  { TAG_IMG,            "href",         ATTR_INLINE },
  { TAG_IMG,            "lowsrc",       ATTR_INLINE },
  { TAG_IMG,            "src",          ATTR_INLINE },
  { TAG_INPUT,          "src",          ATTR_INLINE },
  { TAG_LAYER,          "src",          ATTR_INLINE | ATTR_HTML },
  { TAG_OBJECT,         "data",         ATTR_INLINE },
  { TAG_OVERLAY,        "src",          ATTR_INLINE | ATTR_HTML },
  { TAG_SCRIPT,         "src",          ATTR_INLINE },
  { TAG_TABLE,          "background",   ATTR_INLINE },
  { TAG_TD,             "background",   ATTR_INLINE },
  { TAG_TH,             "background",   ATTR_INLINE }
};

/* The lists of interesting tags and attributes are built dynamically,
   from the information above.  However, some places in the code refer
   to the attributes not mentioned here.  We add them manually.  */
static const char *additional_attributes[] = {
  "rel",                        /* used by tag_handle_link */
  "http-equiv",                 /* used by tag_handle_meta */
  "name",                       /* used by tag_handle_meta */
  "content",                    /* used by tag_handle_meta */
  "action"                      /* used by tag_handle_form */
};

static struct hash_table *interesting_tags;
static struct hash_table *interesting_attributes;

static void
init_interesting (void)
{
  /* Init the variables interesting_tags and interesting_attributes
     that are used by the HTML parser to know which tags and
     attributes we're interested in.  We initialize this only once,
     for performance reasons.

     Here we also make sure that what we put in interesting_tags
     matches the user's preferences as specified through --ignore-tags
     and --follow-tags.  */

  int i;
  interesting_tags = make_nocase_string_hash_table (countof (known_tags));

  /* First, add all the tags we know hot to handle, mapped to their
     respective entries in known_tags.  */
  for (i = 0; i < countof (known_tags); i++)
    hash_table_put (interesting_tags, known_tags[i].name, known_tags + i);

  /* Then remove the tags ignored through --ignore-tags.  */
  if (opt.ignore_tags)
    {
      char **ignored;
      for (ignored = opt.ignore_tags; *ignored; ignored++)
        hash_table_remove (interesting_tags, *ignored);
    }

  /* If --follow-tags is specified, use only those tags.  */
  if (opt.follow_tags)
    {
      /* Create a new table intersecting --follow-tags and known_tags,
         and use it as interesting_tags.  */
      struct hash_table *intersect = make_nocase_string_hash_table (0);
      char **followed;
      for (followed = opt.follow_tags; *followed; followed++)
        {
          struct known_tag *t = hash_table_get (interesting_tags, *followed);
          if (!t)
            continue;           /* ignore unknown --follow-tags entries. */
          hash_table_put (intersect, *followed, t);
        }
      hash_table_destroy (interesting_tags);
      interesting_tags = intersect;
    }

  /* Add the attributes we care about. */
  interesting_attributes = make_nocase_string_hash_table (10);
  for (i = 0; i < countof (additional_attributes); i++)
    hash_table_put (interesting_attributes, additional_attributes[i], "1");
  for (i = 0; i < countof (tag_url_attributes); i++)
    hash_table_put (interesting_attributes,
                    tag_url_attributes[i].attr_name, "1");
}

/* Find the value of attribute named NAME in the taginfo TAG.  If the
   attribute is not present, return NULL.  If ATTRIND is non-NULL, the
   index of the attribute in TAG will be stored there.  */

static char *
find_attr (struct taginfo *tag, const char *name, int *attrind)
{
  int i;
  for (i = 0; i < tag->nattrs; i++)
    if (!strcasecmp (tag->attrs[i].name, name))
      {
        if (attrind)
          *attrind = i;
        return tag->attrs[i].value;
      }
  return NULL;
}

struct map_context {
  char *text;                   /* HTML text. */
  char *base;                   /* Base URI of the document, possibly
                                   changed through <base href=...>. */
  const char *parent_base;      /* Base of the current document. */
  const char *document_file;    /* File name of this document. */
  bool nofollow;                /* whether NOFOLLOW was specified in a
                                   <meta name=robots> tag. */

  struct urlpos *head, *tail;   /* List of URLs that is being
                                   built. */
};

/* Append LINK_URI to the urlpos structure that is being built.

   LINK_URI will be merged with the current document base.  TAG and
   ATTRIND are the necessary context to store the position and
   size.  */

static struct urlpos *
append_url (const char *link_uri,
            struct taginfo *tag, int attrind, struct map_context *ctx)
{
  int link_has_scheme = url_has_scheme (link_uri);
  struct urlpos *newel;
  const char *base = ctx->base ? ctx->base : ctx->parent_base;
  struct url *url;

  if (!base)
    {
      DEBUGP (("%s: no base, merge will use \"%s\".\n",
               ctx->document_file, link_uri));

      if (!link_has_scheme)
        {
          /* Base URL is unavailable, and the link does not have a
             location attached to it -- we have to give up.  Since
             this can only happen when using `--force-html -i', print
             a warning.  */
          logprintf (LOG_NOTQUIET,
                     _("%s: Cannot resolve incomplete link %s.\n"),
                     ctx->document_file, link_uri);
          return NULL;
        }

      url = url_parse (link_uri, NULL);
      if (!url)
        {
          DEBUGP (("%s: link \"%s\" doesn't parse.\n",
                   ctx->document_file, link_uri));
          return NULL;
        }
    }
  else
    {
      /* Merge BASE with LINK_URI, but also make sure the result is
         canonicalized, i.e. that "../" have been resolved.
         (parse_url will do that for us.) */

      char *complete_uri = uri_merge (base, link_uri);

      DEBUGP (("%s: merge(\"%s\", \"%s\") -> %s\n",
               ctx->document_file, base, link_uri, complete_uri));

      url = url_parse (complete_uri, NULL);
      if (!url)
        {
          DEBUGP (("%s: merged link \"%s\" doesn't parse.\n",
                   ctx->document_file, complete_uri));
          xfree (complete_uri);
          return NULL;
        }
      xfree (complete_uri);
    }

  DEBUGP (("appending \"%s\" to urlpos.\n", url->url));

  newel = xnew0 (struct urlpos);
  newel->url = url;
  newel->pos = tag->attrs[attrind].value_raw_beginning - ctx->text;
  newel->size = tag->attrs[attrind].value_raw_size;

  /* A URL is relative if the host is not named, and the name does not
     start with `/'.  */
  if (!link_has_scheme && *link_uri != '/')
    newel->link_relative_p = 1;
  else if (link_has_scheme)
    newel->link_complete_p = 1;

  if (ctx->tail)
    {
      ctx->tail->next = newel;
      ctx->tail = newel;
    }
  else
    ctx->tail = ctx->head = newel;

  return newel;
}

/* All the tag_* functions are called from collect_tags_mapper, as
   specified by KNOWN_TAGS.  */

/* Default tag handler: collect URLs from attributes specified for
   this tag by tag_url_attributes.  */

static void
tag_find_urls (int tagid, struct taginfo *tag, struct map_context *ctx)
{
  int i, attrind;
  int first = -1;

  for (i = 0; i < countof (tag_url_attributes); i++)
    if (tag_url_attributes[i].tagid == tagid)
      {