aboutsummaryrefslogtreecommitdiffstats
path: root/lib
Commit message (Collapse)AuthorAgeLines
* xt_from_string() will return NULL if the string wasn't terminated properly.Wilmer van der Gaast2012-09-24-3/+6
| | | | | | | Adding an extra layer of defense against truncated responses from Twitter. But as long as the response from xt_from_string() isn't NULL-checked this won't help much. So that's my next step. :-)
* Since I can't figure out where the stalls are coming from at this point, atWilmer van der Gaast2012-09-22-1/+1
| | | | | | | | least have a work-around for it. After hitting the Twitter timer a few times while a previous fetch still seems to be running, close the connection. Auto-reconnect will do the rest.
* Little cleanup. Use xt_from_string() where possible.Wilmer van der Gaast2012-09-22-3/+6
|
* Solve a whole bunch of Twitter module crashes: Twitter responses seem toWilmer van der Gaast2012-09-22-6/+19
| | | | | | be getting truncated sometimes. Treat this as an error in http_client already instead of returning partial data.
* Allow building position-independent executables. #981, patch from brainsmoke.Wilmer van der Gaast2012-09-15-1/+1
|
* NSS module fixes from mcepl in #714. This removes des.c since it's noWilmer van der Gaast2012-08-19-781/+249
| | | | | longer necessary.
* Shut up a flood of GLib-related compiler warnings.Wilmer van der Gaast2012-06-04-0/+3
|
* Work-around for what turned out to be a GnuTLS bug (#938). From 3.0.13Wilmer van der Gaast2012-04-10-0/+15
| | | | | | | until 3.0.18 gnutls_record_check_pending() returns non-0 even if the data read so far is an incomplete record and can not yet be read. This can get BitlBee's http_client stuck in a semi-infinite loop.
* Rename SRCDIR directory to _SRCDIR_. I guess #907 was caused by SRCDIRWilmer van der Gaast2012-02-11-3/+3
| | | | | unexpectedly being set to something already.
* Added sha1_random_uuid function, which I will use later to generate randomWilmer van der Gaast2012-02-10-0/+32
| | | | | Jabber roomnames.
* Fixing one compiler warning that shouldn't indicate any real problem (onlyWilmer van der Gaast2012-01-04-1/+1
| | | | | causes unpredictable behaviour on wrongly formatted JSON input).
* A few more SSL fixes merged from AopicieR. This also fixes OpenSSL compileWilmer van der Gaast2012-01-03-10/+5
| | | | | issues (bug #881).
* Keep only one xcred object globally instead of one per connection. WithWilmer van der Gaast2011-12-29-16/+20
| | | | | | verification, this object gets pretty huge and there's no need to have it more than once.
* A few more minor cleanups before merging this into mainline.Wilmer van der Gaast2011-12-26-2/+3
|
* An empty password is still a password, don't refuse accounts for that.Wilmer van der Gaast2011-12-24-1/+1
|
* Merging mainline.Wilmer van der Gaast2011-12-24-28/+227
|\
| * Fix compatibility with old GnuTLS versions, but with a warning. SeeWilmer van der Gaast2011-12-24-0/+4
| | | | | | | | | | http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2009-1417 for details.
| * tls_verify correction: Don't fail cert verification in non-GnuTLS modulesWilmer van der Gaast2011-12-23-10/+6
| | | | | | | | | | unless "cafile" setting is enabled.
| * Merging SSL certificate verification for GnuTLS, with help from AopicieR.Wilmer van der Gaast2011-12-23-28/+224
| |\
| | * Add verify argument to ssl_connect() so HTTPS-based stuff is also secure.Wilmer van der Gaast2011-12-19-8/+21
| | | | | | | | | | | | | | | (Think of Twitter, but also MSN/Yahoo! authentication.)
| | * Move conversion of status codes to status messages into SSL libs.Wilmer van der Gaast2011-12-19-0/+50
| | |
| | * Initial merge of tls_verify patch from AopicieR.Wilmer van der Gaast2011-12-19-20/+154
| | |
| * | Catch condition=G_IO_NVAL from glib's event handler, which should preventWilmer van der Gaast2011-12-19-0/+3
| |/ | | | | | | | | some crashes on unclean shutdowns of connections.
* | Fixed a bug that probably (can't test this now since it's down) broke OAuthWilmer van der Gaast2011-12-21-0/+2
| | | | | | | | | | setup for identi.ca. Turning on oauth for identi.ca accounts by default now.
* | Use sha1_hmac() instead of reimplementing the algorithm in oauth_sign().Wilmer van der Gaast2011-12-21-42/+9
| |
* | NULL-checking in rfc822_get_header() and OAuth response handling.Wilmer van der Gaast2011-12-20-2/+5
| |
* | When updating the XMPP password field with OAuth data, try harder to preserveWilmer van der Gaast2011-12-19-0/+1
| | | | | | | | | | existing data. (Like refresh tokens which we'll need again on next login.)
* | Give a list of SASL mechanisms supported by a server when reporting we don'tWilmer van der Gaast2011-12-19-1/+1
| | | | | | | | | | support any of them.
* | Facebook OAuth2 should now be fully usable.Wilmer van der Gaast2011-12-19-1/+23
| |
* | Moving msn_findheader() to lib/misc.c as get_rfc822_header() so I can use itWilmer van der Gaast2011-12-19-3/+53
| | | | | | | | | | in OAuth as well. (Need it to find the Content-Type: header.)
* | More generic OAuth support now. Should work well for all GTalk accounts nowWilmer van der Gaast2011-12-18-26/+9
| | | | | | | | | | | | and somewhat for MS Messenger. The fb part needs different parsing of the authorize request, and possibly some other work.
* | Mainline merge.Wilmer van der Gaast2011-12-17-46/+116
|\|
| * Support HTTP/1.1 redirect status codes and use HTTPS for OAuth setup. ThisWilmer van der Gaast2011-12-11-6/+22
| | | | | | | | | | | | is required for identi.ca and really should be done for Twitter as well. Twitter OAuth is still broken though, it seems to disagree about signatures.
| * Debug output tweaks: Try to send everything to stderr, and add ifdef toWilmer van der Gaast2011-12-04-11/+13
| | | | | | | | | | enable printing of all SSL traffic.
| * Somewhat improve debug logging of HTTP/SOAP stuff.Wilmer van der Gaast2011-12-02-2/+10
| |
| * GnuTLS now also needs ssl_pending() implemented. Bug #860.Wilmer van der Gaast2011-11-14-17/+28
| |
| * Cleanup of http_client fix. Use g_strdup_printf and completely avoid strcat,Wilmer van der Gaast2011-11-13-14/+21
| | | | | | | | | | truncate request if we switched from POST to GET.
| * Undoing old workaround for MSN troubles and added more proper fix. TheWilmer van der Gaast2011-11-12-4/+7
| | | | | | | | | | | | http_client module needs some refactoring though. Will do that later.. This should hopefully fix bug #850.
| * Strip illegal characters in generated XML streams so Jabber servers won'tWilmer van der Gaast2011-10-20-2/+25
| | | | | | | | | | | | | | disconnect people who don't understand how to disable stupid IRC client features. Based very loosely on a patch and discussion submitted by Artem Savkov on bug #552.
* | Had to change the OAuth secret for GTalk.Wilmer van der Gaast2011-08-04-1/+1
| |
* | OAuth code cleanup.Wilmer van der Gaast2011-07-31-0/+3
| |
* | Facebook authentication. This isn't really OAuth in the end: FB doesn'tWilmer van der Gaast2011-07-31-0/+1
| | | | | | | | | | | | really support desktop app OAuth in a way that would work with BitlBee. Plus, it's only OAuth-compliant by, err, name?
* | oauth2 changes to address http://twitter.com/Wilmer/status/96715400124968960Wilmer van der Gaast2011-07-31-8/+22
| |
* | Export oauth_params_parse().Wilmer van der Gaast2011-07-31-1/+2
| |
* | Having written the same stupid code (ASCII MD5 hashes) 205762 times, time toWilmer van der Gaast2011-07-31-0/+11
| | | | | | | | | | have a function for it..
* | Working OAuth2 support. Needs some more debugging (error handling is notWilmer van der Gaast2011-07-26-34/+158
| | | | | | | | | | great and imc_logout() gets (rightfully) confused when jabber_data is empty).
* | Nothing useful yet, this just generates an auth URL. Things to do: AbilityWilmer van der Gaast2011-07-22-1/+112
|/ | | | | | | | | | to process the answer. This is hard because the GTalk server will time out very quickly which means we lose our state/scope/etc. (And the ability to even receive the answer, at least if I'd do this the same way the Twitter module does it.) And then, get the access token and use it, of course. :-)
* Fixed dumb file descriptor leak.Wilmer van der Gaast2011-07-12-0/+1
|
* Fixed use of g_iconv() in do_iconv(). This was done wrong for ages andWilmer van der Gaast2011-07-02-4/+4
| | | | | | apparently screwed up charset handling in OSCAR pretty badly.. Found and fixed by dalias (#813). May or may not also fix other bugs.
* Send Connection: close headers so webservers won't try keepalive connectionsWilmer van der Gaast2011-04-18-0/+2
| | | | | which http_client really can't deal with.