gnu/wget2
1
0
Fork 0
mirror of https://gitlab.com/gnuwget/wget2.git synced 2025-07-24 10:01:36 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix the rel attribute handling for link tags

Browse Source
This commit is contained in:
Tim Rühsen
2021-08-07 20:34:13 +02:00
parent d6e5641980
commit 10c5057a8a
1 changed files with 24 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
libwget/html_url.c
View File

@ -222,18 +222,36 @@ static void html_get_url(void *context, int flags, const char *tag, const char *
if ((*tag|0x20) == 'l' && !wget_strcasecmp_ascii(tag, "link")) { if ((*tag|0x20) == 'l' && !wget_strcasecmp_ascii(tag, "link")) {
if (!wget_strcasecmp_ascii(attr, "rel")) { if (!wget_strcasecmp_ascii(attr, "rel")) {
if (!wget_strncasecmp_ascii(val, "shortcut icon", len) ctx->link_inline = 0;
|| !wget_strncasecmp_ascii(val, "stylesheet", len)
|| !wget_strncasecmp_ascii(val, "preload", len)) // "rel" contains a space separated list of items.
ctx->link_inline = 1; // see https://html.spec.whatwg.org/multipage/semantics.html#attr-link-rel
else // see https://html.spec.whatwg.org/multipage/links.html#linkTypes
ctx->link_inline = 0; while (len) {
const char *p;
for (p = val;len && !c_isspace(*val); val++, len--); // find end of item
if (p == val) { val++; len--; continue; } // found a delimiter
// Check for items that may be important to display the page.
if (!wget_strncasecmp_ascii(p, "icon", val - p)
|| !wget_strncasecmp_ascii(p, "manifest", val - p)
|| !wget_strncasecmp_ascii(p, "modulepreload", val - p)
|| !wget_strncasecmp_ascii(p, "stylesheet", val - p)
|| !wget_strncasecmp_ascii(p, "prefetch", val - p)
|| !wget_strncasecmp_ascii(p, "preload", val - p))
{
ctx->link_inline = 1;
break;
}
}
if (ctx->uri_index >= 0) { if (ctx->uri_index >= 0) {
// href= came before rel= // href= came before rel=
wget_html_parsed_url *url = wget_vector_get(res->uris, ctx->uri_index); wget_html_parsed_url *url = wget_vector_get(res->uris, ctx->uri_index);
url->link_inline = ctx->link_inline; url->link_inline = ctx->link_inline;
} }
return;
} }
} }