#include "utils.h"
#include "hash.h"
#include "convert.h"
+#include "recur.h" /* declaration of get_urls_html */
#ifndef errno
extern int errno;
TAG_LAYER,
TAG_LINK,
TAG_META,
+ TAG_OBJECT,
TAG_OVERLAY,
TAG_SCRIPT,
TAG_TABLE,
{ 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_url_attributes documents which attributes of which tags contain
URLs to harvest. It is used by tag_find_urls. */
-/* Defines for the FLAGS field; currently only one flag is defined. */
+/* Defines for the FLAGS. */
-/* This tag points to an external document not necessary for rendering this
- document (i.e. it's not an inlined image, stylesheet, etc.). */
-#define TUA_EXTERNAL 1
+/* 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. */
const char *attr_name;
int flags;
} tag_url_attributes[] = {
- { TAG_A, "href", TUA_EXTERNAL },
- { TAG_APPLET, "code", 0 },
- { TAG_AREA, "href", TUA_EXTERNAL },
- { TAG_BGSOUND, "src", 0 },
- { TAG_BODY, "background", 0 },
- { TAG_EMBED, "href", TUA_EXTERNAL },
- { TAG_EMBED, "src", 0 },
- { TAG_FIG, "src", 0 },
- { TAG_FRAME, "src", 0 },
- { TAG_IFRAME, "src", 0 },
- { TAG_IMG, "href", 0 },
- { TAG_IMG, "lowsrc", 0 },
- { TAG_IMG, "src", 0 },
- { TAG_INPUT, "src", 0 },
- { TAG_LAYER, "src", 0 },
- { TAG_OVERLAY, "src", 0 },
- { TAG_SCRIPT, "src", 0 },
- { TAG_TABLE, "background", 0 },
- { TAG_TD, "background", 0 },
- { TAG_TH, "background", 0 }
+ { 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,
/* If --follow-tags is specified, use only those tags. */
if (opt.follow_tags)
{
- /* Create a new hash table with the intersection of tags in
- --follow-tags and known_tags, and use that as
- interesting_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 tags in --follow-tags. */
+ continue; /* ignore unknown --follow-tags entries. */
hash_table_put (intersect, *followed, t);
}
hash_table_destroy (interesting_tags);
}
/* Add the attributes we care about. */
- interesting_attributes = make_nocase_string_hash_table (17);
+ interesting_attributes = make_nocase_string_hash_table (10);
for (i = 0; i < countof (additional_attributes); i++)
- string_set_add (interesting_attributes, additional_attributes[i]);
+ hash_table_put (interesting_attributes, additional_attributes[i], "1");
for (i = 0; i < countof (tag_url_attributes); i++)
- string_set_add (interesting_attributes, tag_url_attributes[i].attr_name);
+ 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
size. */
static struct urlpos *
-append_one_url (const char *link_uri, int inlinep,
- struct taginfo *tag, int attrind, struct map_context *ctx)
+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;
DEBUGP (("appending \"%s\" to urlpos.\n", url->url));
- newel = (struct urlpos *)xmalloc (sizeof (struct urlpos));
- memset (newel, 0, sizeof (*newel));
-
- newel->next = NULL;
+ 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;
- newel->link_inline_p = inlinep;
/* A URL is relative if the host is not named, and the name does not
start with `/'. */
if (0 == strcasecmp (tag->attrs[attrind].name,
tag_url_attributes[i].attr_name))
{
- int flags = tag_url_attributes[i].flags;
- append_one_url (link, !(flags & TUA_EXTERNAL), tag, attrind, ctx);
+ struct urlpos *up = append_url (link, tag, attrind, ctx);
+ if (up)
+ {
+ int flags = tag_url_attributes[i].flags;
+ if (flags & ATTR_INLINE)
+ up->link_inline_p = 1;
+ if (flags & ATTR_HTML)
+ up->link_expect_html = 1;
+ }
}
}
}
if (!newbase)
return;
- base_urlpos = append_one_url (newbase, 0, tag, attrind, ctx);
+ base_urlpos = append_url (newbase, tag, attrind, ctx);
if (!base_urlpos)
return;
base_urlpos->ignore_when_downloading = 1;
char *action = find_attr (tag, "action", &attrind);
if (action)
{
- struct urlpos *action_urlpos = append_one_url (action, 0, tag,
- attrind, ctx);
- if (action_urlpos)
- action_urlpos->ignore_when_downloading = 1;
+ struct urlpos *up = append_url (action, tag, attrind, ctx);
+ if (up)
+ up->ignore_when_downloading = 1;
}
}
*/
if (href)
{
- char *rel = find_attr (tag, "rel", NULL);
- int inlinep = (rel
- && (0 == strcasecmp (rel, "stylesheet")
- || 0 == strcasecmp (rel, "shortcut icon")));
- append_one_url (href, inlinep, tag, attrind, ctx);
+ struct urlpos *up = append_url (href, tag, attrind, ctx);
+ if (up)
+ {
+ char *rel = find_attr (tag, "rel", NULL);
+ if (rel
+ && (0 == strcasecmp (rel, "stylesheet")
+ || 0 == strcasecmp (rel, "shortcut icon")))
+ up->link_inline_p = 1;
+ else
+ /* The external ones usually point to HTML pages, such as
+ <link rel="next" href="..."> */
+ up->link_expect_html = 1;
+ }
}
}
while (ISSPACE (*p))
++p;
- entry = append_one_url (p, 0, tag, attrind, ctx);
+ entry = append_url (p, tag, attrind, ctx);
if (entry)
{
entry->link_refresh_p = 1;
entry->refresh_timeout = timeout;
+ entry->link_expect_html = 1;
}
}
else if (name && 0 == strcasecmp (name, "robots"))
logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
return NULL;
}
- DEBUGP (("Loaded %s (size %ld).\n", file, fm->length));
+ DEBUGP (("Loaded %s (size %s).\n", file, number_to_static_string (fm->length)));
ctx.text = fm->content;
ctx.head = ctx.tail = NULL;
init_interesting ();
/* Specify MHT_TRIM_VALUES because of buggy HTML generators that
- generate <a href=" foo"> instead of <a href="foo"> (Netscape
- ignores spaces as well.) If you really mean space, use &32; or
- %20. */
+ generate <a href=" foo"> instead of <a href="foo"> (browsers
+ ignore spaces as well.) If you really mean space, use &32; or
+ %20. MHT_TRIM_VALUES also causes squashing of embedded newlines,
+ e.g. in <img src="foo.[newline]html">. Such newlines are also
+ ignored by IE and Mozilla and are presumably introduced by
+ writing HTML with editors that force word wrap. */
flags = MHT_TRIM_VALUES;
if (opt.strict_comments)
flags |= MHT_STRICT_COMMENTS;
if (meta_disallow_follow)
*meta_disallow_follow = ctx.nofollow;
- FREE_MAYBE (ctx.base);
+ xfree_null (ctx.base);
read_file_free (fm);
return ctx.head;
}
logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
return NULL;
}
- DEBUGP (("Loaded %s (size %ld).\n", file, fm->length));
+ DEBUGP (("Loaded %s (size %s).\n", file, number_to_static_string (fm->length)));
head = tail = NULL;
text = fm->content;
}
xfree (url_text);
- entry = (struct urlpos *)xmalloc (sizeof (struct urlpos));
- memset (entry, 0, sizeof (*entry));
+ entry = xnew0 (struct urlpos);
entry->next = NULL;
entry->url = url;
void
cleanup_html_url (void)
{
- FREE_MAYBE (interesting_tags);
- FREE_MAYBE (interesting_attributes);
+ /* Destroy the hash tables. The hash table keys and values are not
+ allocated by this code, so we don't need to free them here. */
+ if (interesting_tags)
+ hash_table_destroy (interesting_tags);
+ if (interesting_attributes)
+ hash_table_destroy (interesting_attributes);
}