mirror of
https://github.com/moparisthebest/wget
synced 2024-07-03 16:38:41 -04:00
[svn] Fix downloading of duplicate URLs.
Published in <sxsvgfmu2bj.fsf@florida.arsdigita.de>.
This commit is contained in:
parent
abf215c1e3
commit
0fdc1bd8c0
@ -1,3 +1,22 @@
|
|||||||
|
2001-12-04 Hrvoje Niksic <hniksic@arsdigita.com>
|
||||||
|
|
||||||
|
* recur.c (retrieve_tree): Check whether the URL was already
|
||||||
|
downloaded before downloading it again.
|
||||||
|
(descend_child_p): Renamed to download_child_p.
|
||||||
|
(register_download): When one URL is downloaded to a file already
|
||||||
|
"owned" by another URL, delete all references that map any URL to
|
||||||
|
that file.
|
||||||
|
(register_delete_file): New function.
|
||||||
|
(retrieve_tree): Use it after deleting a file.
|
||||||
|
|
||||||
|
* url.c (url_parse): Re-canonicalize the URL also if the path is
|
||||||
|
empty, so that e.g. "http://www.server.com" ->
|
||||||
|
"http://www.server.com/".
|
||||||
|
(lowercase_str): Use ISUPPER instead of !ISLOWER.
|
||||||
|
|
||||||
|
* retr.c (retrieve_url): Use the canonical URL form when calling
|
||||||
|
register_download().
|
||||||
|
|
||||||
2001-12-04 Ian Abbott <abbotti@mev.co.uk>
|
2001-12-04 Ian Abbott <abbotti@mev.co.uk>
|
||||||
|
|
||||||
* snprintf.c (dopr): Use `unsigned int' as the second argument to
|
* snprintf.c (dopr): Use `unsigned int' as the second argument to
|
||||||
|
165
src/recur.c
165
src/recur.c
@ -58,6 +58,8 @@ static struct hash_table *dl_url_file_map;
|
|||||||
in order. If you need to check whether a file has been downloaded,
|
in order. If you need to check whether a file has been downloaded,
|
||||||
use a hash table, e.g. dl_file_url_map. */
|
use a hash table, e.g. dl_file_url_map. */
|
||||||
static slist *downloaded_html_files;
|
static slist *downloaded_html_files;
|
||||||
|
|
||||||
|
static void register_delete_file PARAMS ((const char *));
|
||||||
|
|
||||||
/* Functions for maintaining the URL queue. */
|
/* Functions for maintaining the URL queue. */
|
||||||
|
|
||||||
@ -150,7 +152,7 @@ url_dequeue (struct url_queue *queue,
|
|||||||
return 1;
|
return 1;
|
||||||
}
|
}
|
||||||
|
|
||||||
static int descend_url_p PARAMS ((const struct urlpos *, struct url *, int,
|
static int download_child_p PARAMS ((const struct urlpos *, struct url *, int,
|
||||||
struct url *, struct hash_table *));
|
struct url *, struct hash_table *));
|
||||||
static int descend_redirect_p PARAMS ((const char *, const char *, int,
|
static int descend_redirect_p PARAMS ((const char *, const char *, int,
|
||||||
struct url *, struct hash_table *));
|
struct url *, struct hash_table *));
|
||||||
@ -205,19 +207,30 @@ retrieve_tree (const char *start_url)
|
|||||||
|
|
||||||
if (downloaded_exceeds_quota ())
|
if (downloaded_exceeds_quota ())
|
||||||
break;
|
break;
|
||||||
|
|
||||||
if (status == FWRITEERR)
|
if (status == FWRITEERR)
|
||||||
break;
|
break;
|
||||||
|
|
||||||
/* Get the next URL from the queue. */
|
/* Get the next URL from the queue... */
|
||||||
|
|
||||||
if (!url_dequeue (queue,
|
if (!url_dequeue (queue,
|
||||||
(const char **)&url, (const char **)&referer,
|
(const char **)&url, (const char **)&referer,
|
||||||
&depth))
|
&depth))
|
||||||
break;
|
break;
|
||||||
|
|
||||||
/* And download it. */
|
/* ...and download it. Note that this download is in most cases
|
||||||
|
unconditional, as download_child_p already makes sure a file
|
||||||
|
doesn't get enqueued twice -- and yet this check is here, and
|
||||||
|
not in download_child_p. This is so that if you run `wget -r
|
||||||
|
URL1 URL2', and a random URL is encountered once under URL1
|
||||||
|
and again under URL2, but at a different (possibly smaller)
|
||||||
|
depth, we want the URL's children to be taken into account
|
||||||
|
the second time. */
|
||||||
|
if (dl_url_file_map && hash_table_contains (dl_url_file_map, url))
|
||||||
|
{
|
||||||
|
DEBUGP (("Already downloaded \"%s\", reusing it from \"%s\".\n",
|
||||||
|
url, (char *)hash_table_get (dl_url_file_map, url)));
|
||||||
|
}
|
||||||
|
else
|
||||||
{
|
{
|
||||||
int dt = 0;
|
int dt = 0;
|
||||||
char *redirected = NULL;
|
char *redirected = NULL;
|
||||||
@ -305,7 +318,7 @@ retrieve_tree (const char *start_url)
|
|||||||
continue;
|
continue;
|
||||||
if (dash_p_leaf_HTML && !child->link_inline_p)
|
if (dash_p_leaf_HTML && !child->link_inline_p)
|
||||||
continue;
|
continue;
|
||||||
if (descend_url_p (child, url_parsed, depth, start_url_parsed,
|
if (download_child_p (child, url_parsed, depth, start_url_parsed,
|
||||||
blacklist))
|
blacklist))
|
||||||
{
|
{
|
||||||
url_enqueue (queue, xstrdup (child->url->url),
|
url_enqueue (queue, xstrdup (child->url->url),
|
||||||
@ -338,6 +351,7 @@ retrieve_tree (const char *start_url)
|
|||||||
file);
|
file);
|
||||||
if (unlink (file))
|
if (unlink (file))
|
||||||
logprintf (LOG_NOTQUIET, "unlink: %s\n", strerror (errno));
|
logprintf (LOG_NOTQUIET, "unlink: %s\n", strerror (errno));
|
||||||
|
register_delete_file (file);
|
||||||
}
|
}
|
||||||
|
|
||||||
xfree (url);
|
xfree (url);
|
||||||
@ -379,7 +393,7 @@ retrieve_tree (const char *start_url)
|
|||||||
will help if those URLs are encountered many times. */
|
will help if those URLs are encountered many times. */
|
||||||
|
|
||||||
static int
|
static int
|
||||||
descend_url_p (const struct urlpos *upos, struct url *parent, int depth,
|
download_child_p (const struct urlpos *upos, struct url *parent, int depth,
|
||||||
struct url *start_url_parsed, struct hash_table *blacklist)
|
struct url *start_url_parsed, struct hash_table *blacklist)
|
||||||
{
|
{
|
||||||
struct url *u = upos->url;
|
struct url *u = upos->url;
|
||||||
@ -555,10 +569,10 @@ descend_url_p (const struct urlpos *upos, struct url *parent, int depth,
|
|||||||
return 0;
|
return 0;
|
||||||
}
|
}
|
||||||
|
|
||||||
/* This function determines whether we should descend the children of
|
/* This function determines whether we will consider downloading the
|
||||||
the URL whose download resulted in a redirection, possibly to
|
children of a URL whose download resulted in a redirection,
|
||||||
another host, etc. It is needed very rarely, and thus it is merely
|
possibly to another host, etc. It is needed very rarely, and thus
|
||||||
a simple-minded wrapper around descend_url_p. */
|
it is merely a simple-minded wrapper around download_child_p. */
|
||||||
|
|
||||||
static int
|
static int
|
||||||
descend_redirect_p (const char *redirected, const char *original, int depth,
|
descend_redirect_p (const char *redirected, const char *original, int depth,
|
||||||
@ -578,7 +592,7 @@ descend_redirect_p (const char *redirected, const char *original, int depth,
|
|||||||
memset (upos, 0, sizeof (*upos));
|
memset (upos, 0, sizeof (*upos));
|
||||||
upos->url = new_parsed;
|
upos->url = new_parsed;
|
||||||
|
|
||||||
success = descend_url_p (upos, orig_parsed, depth,
|
success = download_child_p (upos, orig_parsed, depth,
|
||||||
start_url_parsed, blacklist);
|
start_url_parsed, blacklist);
|
||||||
|
|
||||||
url_free (orig_parsed);
|
url_free (orig_parsed);
|
||||||
@ -592,21 +606,94 @@ descend_redirect_p (const char *redirected, const char *original, int depth,
|
|||||||
}
|
}
|
||||||
|
|
||||||
|
|
||||||
/* Register that URL has been successfully downloaded to FILE. */
|
#define ENSURE_TABLES_EXIST do { \
|
||||||
|
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); \
|
||||||
|
} while (0)
|
||||||
|
|
||||||
|
static int
|
||||||
|
dissociate_urls_from_file_mapper (void *key, void *value, void *arg)
|
||||||
|
{
|
||||||
|
char *mapping_url = (char *)key;
|
||||||
|
char *mapping_file = (char *)value;
|
||||||
|
char *file = (char *)arg;
|
||||||
|
|
||||||
|
if (0 == strcmp (mapping_file, file))
|
||||||
|
{
|
||||||
|
hash_table_remove (dl_url_file_map, mapping_url);
|
||||||
|
xfree (mapping_url);
|
||||||
|
xfree (mapping_file);
|
||||||
|
}
|
||||||
|
|
||||||
|
/* Continue mapping. */
|
||||||
|
return 0;
|
||||||
|
}
|
||||||
|
|
||||||
|
/* Remove all associations from various URLs to FILE from dl_url_file_map. */
|
||||||
|
|
||||||
|
static void
|
||||||
|
dissociate_urls_from_file (const char *file)
|
||||||
|
{
|
||||||
|
hash_table_map (dl_url_file_map, dissociate_urls_from_file_mapper,
|
||||||
|
(char *)file);
|
||||||
|
}
|
||||||
|
|
||||||
|
/* Register that URL has been successfully downloaded to FILE. This
|
||||||
|
is used by the link conversion code to convert references to URLs
|
||||||
|
to references to local files. It is also being used to check if a
|
||||||
|
URL has already been downloaded. */
|
||||||
|
|
||||||
void
|
void
|
||||||
register_download (const char *url, const char *file)
|
register_download (const char *url, const char *file)
|
||||||
{
|
{
|
||||||
if (!opt.convert_links)
|
char *old_file, *old_url;
|
||||||
return;
|
|
||||||
if (!dl_file_url_map)
|
ENSURE_TABLES_EXIST;
|
||||||
dl_file_url_map = make_string_hash_table (0);
|
|
||||||
if (!dl_url_file_map)
|
/* With some forms of retrieval, it is possible, although not
|
||||||
dl_url_file_map = make_string_hash_table (0);
|
likely, for different URLs to resolve to the same file name. For
|
||||||
|
example, "http://www.server.com/" and
|
||||||
|
"http://www.server.com/index.html" will both resolve to the same
|
||||||
|
file, "index.html". If both are downloaded, the second download
|
||||||
|
will override the first one.
|
||||||
|
|
||||||
|
If that happens, dissociate the old file name from the URL. */
|
||||||
|
|
||||||
|
if (hash_table_get_pair (dl_file_url_map, file, &old_file, &old_url))
|
||||||
|
{
|
||||||
|
if (0 == strcmp (url, old_url))
|
||||||
|
/* We have somehow managed to download the same URL twice.
|
||||||
|
Nothing to do. */
|
||||||
|
return;
|
||||||
|
|
||||||
|
hash_table_remove (dl_file_url_map, file);
|
||||||
|
xfree (old_file);
|
||||||
|
xfree (old_url);
|
||||||
|
|
||||||
|
/* Remove all the URLs that point to this file. Yes, there can
|
||||||
|
be more than one such URL, because we store redirections as
|
||||||
|
multiple entries in dl_url_file_map. For example, if URL1
|
||||||
|
redirects to URL2 which gets downloaded to FILE, we map both
|
||||||
|
URL1 and URL2 to FILE in dl_url_file_map. (dl_file_url_map
|
||||||
|
only points to URL2.) When another URL gets loaded to FILE,
|
||||||
|
we want both URL1 and URL2 dissociated from it.
|
||||||
|
|
||||||
|
This is a relatively expensive operation because it performs
|
||||||
|
a linear search of the whole hash table, but it should be
|
||||||
|
called very rarely, only when two URLs resolve to the same
|
||||||
|
file name, *and* the "<file>.1" extensions are turned off.
|
||||||
|
In other words, almost never. */
|
||||||
|
dissociate_urls_from_file (file);
|
||||||
|
}
|
||||||
|
|
||||||
|
/* A URL->FILE mapping is not possible without a FILE->URL mapping.
|
||||||
|
If the latter were present, it should have been removed by the
|
||||||
|
above `if'. */
|
||||||
|
assert (!hash_table_contains (dl_url_file_map, url));
|
||||||
|
|
||||||
if (!hash_table_contains (dl_file_url_map, file))
|
|
||||||
hash_table_put (dl_file_url_map, xstrdup (file), xstrdup (url));
|
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));
|
hash_table_put (dl_url_file_map, xstrdup (url), xstrdup (file));
|
||||||
}
|
}
|
||||||
|
|
||||||
@ -619,8 +706,7 @@ register_redirection (const char *from, const char *to)
|
|||||||
{
|
{
|
||||||
char *file;
|
char *file;
|
||||||
|
|
||||||
if (!opt.convert_links)
|
ENSURE_TABLES_EXIST;
|
||||||
return;
|
|
||||||
|
|
||||||
file = hash_table_get (dl_url_file_map, to);
|
file = hash_table_get (dl_url_file_map, to);
|
||||||
assert (file != NULL);
|
assert (file != NULL);
|
||||||
@ -628,7 +714,25 @@ register_redirection (const char *from, const char *to)
|
|||||||
hash_table_put (dl_url_file_map, xstrdup (from), xstrdup (file));
|
hash_table_put (dl_url_file_map, xstrdup (from), xstrdup (file));
|
||||||
}
|
}
|
||||||
|
|
||||||
/* Register that URL corresponds to the HTML file FILE. */
|
/* Register that the file has been deleted. */
|
||||||
|
|
||||||
|
static void
|
||||||
|
register_delete_file (const char *file)
|
||||||
|
{
|
||||||
|
char *old_url, *old_file;
|
||||||
|
|
||||||
|
ENSURE_TABLES_EXIST;
|
||||||
|
|
||||||
|
if (!hash_table_get_pair (dl_file_url_map, file, &old_file, &old_url))
|
||||||
|
return;
|
||||||
|
|
||||||
|
hash_table_remove (dl_file_url_map, file);
|
||||||
|
xfree (old_file);
|
||||||
|
xfree (old_url);
|
||||||
|
dissociate_urls_from_file (file);
|
||||||
|
}
|
||||||
|
|
||||||
|
/* Register that FILE is an HTML file that has been downloaded. */
|
||||||
|
|
||||||
void
|
void
|
||||||
register_html (const char *url, const char *file)
|
register_html (const char *url, const char *file)
|
||||||
@ -672,15 +776,16 @@ convert_all_links (void)
|
|||||||
struct urlpos *urls, *cur_url;
|
struct urlpos *urls, *cur_url;
|
||||||
char *url;
|
char *url;
|
||||||
|
|
||||||
DEBUGP (("Rescanning %s\n", html->string));
|
|
||||||
|
|
||||||
/* Determine the URL of the HTML file. get_urls_html will need
|
/* Determine the URL of the HTML file. get_urls_html will need
|
||||||
it. */
|
it. */
|
||||||
url = hash_table_get (dl_file_url_map, html->string);
|
url = hash_table_get (dl_file_url_map, html->string);
|
||||||
if (url)
|
if (!url)
|
||||||
DEBUGP (("It should correspond to %s.\n", url));
|
{
|
||||||
else
|
DEBUGP (("Apparently %s has been removed.\n", html->string));
|
||||||
DEBUGP (("I cannot find the corresponding URL.\n"));
|
continue;
|
||||||
|
}
|
||||||
|
|
||||||
|
DEBUGP (("Rescanning %s (from %s)\n", html->string, url));
|
||||||
|
|
||||||
/* Parse the HTML file... */
|
/* Parse the HTML file... */
|
||||||
urls = get_urls_html (html->string, url, NULL);
|
urls = get_urls_html (html->string, url, NULL);
|
||||||
|
@ -511,11 +511,11 @@ retrieve_url (const char *origurl, char **file, char **newloc,
|
|||||||
{
|
{
|
||||||
if (*dt & RETROKF)
|
if (*dt & RETROKF)
|
||||||
{
|
{
|
||||||
register_download (url, local_file);
|
register_download (u->url, local_file);
|
||||||
if (redirections)
|
if (redirections)
|
||||||
register_all_redirections (redirections, url);
|
register_all_redirections (redirections, u->url);
|
||||||
if (*dt & TEXTHTML)
|
if (*dt & TEXTHTML)
|
||||||
register_html (url, local_file);
|
register_html (u->url, local_file);
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
12
src/url.c
12
src/url.c
@ -595,7 +595,7 @@ lowercase_str (char *str)
|
|||||||
{
|
{
|
||||||
int change = 0;
|
int change = 0;
|
||||||
for (; *str; str++)
|
for (; *str; str++)
|
||||||
if (!ISLOWER (*str))
|
if (ISUPPER (*str))
|
||||||
{
|
{
|
||||||
change = 1;
|
change = 1;
|
||||||
*str = TOLOWER (*str);
|
*str = TOLOWER (*str);
|
||||||
@ -787,13 +787,11 @@ url_parse (const char *url, int *error)
|
|||||||
if (fragment_b)
|
if (fragment_b)
|
||||||
u->fragment = strdupdelim (fragment_b, fragment_e);
|
u->fragment = strdupdelim (fragment_b, fragment_e);
|
||||||
|
|
||||||
|
if (path_modified || u->fragment || host_modified || path_b == path_e)
|
||||||
if (path_modified || u->fragment || host_modified)
|
|
||||||
{
|
{
|
||||||
/* If path_simplify modified the path, or if a fragment is
|
/* If we suspect that a transformation has rendered what
|
||||||
present, or if the original host name had caps in it, make
|
url_string might return different from URL_ENCODED, rebuild
|
||||||
sure that u->url is equivalent to what would be printed by
|
u->url using url_string. */
|
||||||
url_string. */
|
|
||||||
u->url = url_string (u, 0);
|
u->url = url_string (u, 0);
|
||||||
|
|
||||||
if (url_encoded != url)
|
if (url_encoded != url)
|
||||||
|
Loading…
Reference in New Issue
Block a user