-\f
-/* Register that URL has been successfully downloaded to FILE. */
-
-void
-register_download (const char *url, const char *file)
-{
- if (!opt.convert_links)
- return;
- if (!dl_file_url_map)
- dl_file_url_map = make_string_hash_table (0);
- if (!dl_url_file_map)
- dl_url_file_map = make_string_hash_table (0);
-
- if (!hash_table_contains (dl_file_url_map, file))
- hash_table_put (dl_file_url_map, xstrdup (file), xstrdup (url));
- if (!hash_table_contains (dl_url_file_map, url))
- hash_table_put (dl_url_file_map, xstrdup (url), xstrdup (file));
-}
-
-/* Register that FROM has been redirected to TO. This assumes that TO
- is successfully downloaded and already registered using
- register_download() above. */
-
-void
-register_redirection (const char *from, const char *to)
-{
- char *file;
-
- if (!opt.convert_links)
- return;
-
- file = hash_table_get (dl_url_file_map, to);
- assert (file != NULL);
- if (!hash_table_contains (dl_url_file_map, from))
- hash_table_put (dl_url_file_map, xstrdup (from), xstrdup (file));
-}
-
-/* Register that URL corresponds to the HTML file FILE. */
-
-void
-register_html (const char *url, const char *file)
-{
- if (!opt.convert_links)
- return;
- downloaded_html_files = slist_prepend (downloaded_html_files, file);
-}
-
-/* This function is called when the retrieval is done to convert the
- links that have been downloaded. It has to be called at the end of
- the retrieval, because only then does Wget know conclusively which
- URLs have been downloaded, and which not, so it can tell which
- direction to convert to.
-
- The "direction" means that the URLs to the files that have been
- downloaded get converted to the relative URL which will point to
- that file. And the other URLs get converted to the remote URL on
- the server.
-
- All the downloaded HTMLs are kept in downloaded_html_files, and
- downloaded URLs in urls_downloaded. All the information is
- extracted from these two lists. */
-
-void
-convert_all_links (void)
-{
- slist *html;
- struct wget_timer *timer;
- long msecs;
- int file_count = 0;
-
- timer = wtimer_new ();
-
- /* Destructively reverse downloaded_html_files to get it in the right order.
- recursive_retrieve() used slist_prepend() consistently. */
- downloaded_html_files = slist_nreverse (downloaded_html_files);
-
- for (html = downloaded_html_files; html; html = html->next)
- {
- struct urlpos *urls, *cur_url;
- char *url;
-
- DEBUGP (("Rescanning %s\n", html->string));
-
- /* Determine the URL of the HTML file. get_urls_html will need
- it. */
- url = hash_table_get (dl_file_url_map, html->string);
- if (url)
- DEBUGP (("It should correspond to %s.\n", url));
- else
- DEBUGP (("I cannot find the corresponding URL.\n"));
-
- /* Parse the HTML file... */
- urls = get_urls_html (html->string, url, FALSE, NULL);
-
- /* We don't respect meta_disallow_follow here because, even if
- the file is not followed, we might still want to convert the
- links that have been followed from other files. */
-
- for (cur_url = urls; cur_url; cur_url = cur_url->next)
- {
- char *local_name;
- struct url *u = cur_url->url;
-
- if (cur_url->link_base_p)
- {
- /* Base references have been resolved by our parser, so
- we turn the base URL into an empty string. (Perhaps
- we should remove the tag entirely?) */
- cur_url->convert = CO_NULLIFY_BASE;
- continue;
- }
-
- /* We decide the direction of conversion according to whether
- a URL was downloaded. Downloaded URLs will be converted
- ABS2REL, whereas non-downloaded will be converted REL2ABS. */
- local_name = hash_table_get (dl_url_file_map, u->url);
- if (local_name)
- DEBUGP (("%s marked for conversion, local %s\n",
- u->url, local_name));
-
- /* Decide on the conversion type. */
- if (local_name)
- {
- /* We've downloaded this URL. Convert it to relative
- form. We do this even if the URL already is in
- relative form, because our directory structure may
- not be identical to that on the server (think `-nd',
- `--cut-dirs', etc.) */
- cur_url->convert = CO_CONVERT_TO_RELATIVE;
- cur_url->local_name = xstrdup (local_name);
- }
- else
- {
- /* We haven't downloaded this URL. If it's not already
- complete (including a full host name), convert it to
- that form, so it can be reached while browsing this
- HTML locally. */
- if (!cur_url->link_complete_p)
- cur_url->convert = CO_CONVERT_TO_COMPLETE;
- cur_url->local_name = NULL;
- }
- }
-
- /* Convert the links in the file. */
- convert_links (html->string, urls);
- ++file_count;
-
- /* Free the data. */
- free_urlpos (urls);
- }
-
- msecs = wtimer_elapsed (timer);
- wtimer_delete (timer);
- logprintf (LOG_VERBOSE, _("Converted %d files in %.2f seconds.\n"),
- file_count, (double)msecs / 1000);
-}
-
-/* Cleanup the data structures associated with recursive retrieving
- (the variables above). */
-void
-recursive_cleanup (void)
-{
- if (dl_file_url_map)
- {
- free_keys_and_values (dl_file_url_map);
- hash_table_destroy (dl_file_url_map);
- dl_file_url_map = NULL;
- }
- if (dl_url_file_map)
- {
- free_keys_and_values (dl_url_file_map);
- hash_table_destroy (dl_url_file_map);
- dl_url_file_map = NULL;
- }
- slist_free (downloaded_html_files);
- downloaded_html_files = NULL;
-}