Age | Commit message (Collapse) | Author | Files | Lines |
|
This reverts commit 7bde7f66284b47a75bbceadc360e7f03550ace21.
because some older expat version do not support this function I don't want to
make it conditional for now.
Conflicts:
lib/discovery.cc
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
somehow I've managed to put too many tidying options into conditional.
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
now when parsing document that we expect might be html we also save first 16K
of the document to the buffer and if the parser choked we run the saved data
through htmltidy and feed the output to the parser again.
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
- added libtidy detection to configure
- added tidy flags to libopkele.pc.in
- added primitive wrapper classes
- added tidy exception class
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
It kept prepending a '/' to the trailing segment even if the segment was past
[?#]
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
The xrds_location should not be cleared when doing pure xrd discovery.
Otherwise it may get overwritten before curl actually uses it. Previously it
was cleared twice, which is a waste of cpu cycles, anyway.
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
1. removed checking that secret_t is exactly 20 bytes, because its size is not
so constant anymore.
2. added a query_string function that produces a query string suitable for use
in GET and POST requests. I'm making use of it when performing direct
request when establishing associations.
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
which was mistakenly added to the list of endpoint URIs instead of local ids.
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
|
|
The discovery, which does both XRDS-based (Yadis, XRI, for XRI, using proxy)
and HTML-based search, now returns results in opkele:idiscovery_t structure.
It uses expat-based parser idigger_t, which itself is not exposed via any
header files, but hidden in lib/discovery.cc, the discovery testing program is
renamed from openid_resolve to idiscover.
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
- almost irrelevant documentation updates
- introduced id_res_expired_on_delivery exception to be thrown if the
association has expired right before it could've been used for signature
verification.
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
if no service found in yadis document, but it is found in html, use
html-discovered service.
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
This commit adds openid service resolver that does discovery using XRI (proxy
only), Yadis protocol and html-based discovery. It uses expat as xml parsing
engine, which makes it a bit more strict about html it receives, but I think
failing to discover links in *severely* broken html is better than
misdetecting links, hidden in comments or such.
This is highly experimental code and needs more thoughts and testing.
Thanks everyone pushing me towards this development. Namely Joseph, John,
Gen.
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|
|
Signed-off-by: Michael Krelin <hacker@klever.net>
|