Sadly, this breaks path cookies on HTTPS sites. The correct
fix is to implement RFC6265 in full (probably replacing
urldb with something less complex, too).
This reverts commit 924f8844d4.
+ urldb API now takes URLs as nsurl, rather than string.
+ urldb internally stores full URLs with nsurl ref.
+ urldb internally stores schemes as lwc_string.
+ Load and save of cookies and URL file may be slower since
we now need to create a nsurl.
+ Everything else should be faster, and there should be much
less allocating/freeing and much less parsing of the same
url over and over again.
+ Updated urldbtest for new urldb API.
+ urldbtest now cleans up at the end
+ Added lwc_string itterator to end of urldbtest
+ Adding some broken URLs (such as http:domain/) will now
work, since nsurl fixes (http://domain/) them.
(prot_space_data): Added, it lives linked in the leaf host_part
struct and together with its scheme and port (which defins canonical root
url) and realm this defines a protection space.
(path_data): Removed auth_data field and replaced by a prot_space_data
pointer.
(host_part::prot_space): Added linked list of protection space data
structs.
(urldb_get_auth_details): Given an URL fetch fetches its auth.
(urldb_set_auth_details): Creates or updates the contents of a
protection space to which given URL belongs.
(urldb_destroy_host_tree): Delete protection data space structures
using urldb_destroy_prot_space.
(urldb_destroy_prot_space): Added.
- content/urldb.h(urldb_get_auth_details): Added realm parameter.
- content/fetchers/fetch_curl.c(fetch_curl_set_options): Update
urldb_get_auth_details call (we don't know realm at this point).
- content/fetchcache.c(fetchcache_callback, fetchcache_auth): At FETCH_AUTH,
use realm to determine if we really don't know auth data and if so,
refetch content.
- content/content.h(struct content): Add content::tried_with_auth.
- content/content.c(content_create): Initialize content::tried_with_auth.
- riscos/401login.c(ro_gui_401login_open): Show known authentication
data in dialogue so user can see what was wrong with it and correct it.
Solves bug #2830829.
svn path=/trunk/netsurf/; revision=9045
Lose unnecessary increment of rptr in cookie domain prefix matching.
Record that, in the long term, we need some kind of TLD service to consult so that (e.g.) .co.uk is not considered a valid domain prefix.
svn path=/trunk/netsurf/; revision=7600
John-Mark Bell