83032b1aa3
Use glib for all allocations
...
- Removes need to check for malloc failure
- Removes need for NULL checks on free
- Adds checks for integer overflows
- Removes some extra memset calls
- Removes chance of mixing libc and glib malloc/free
2014-12-28 06:44:44 -05:00
c4cb1b25ec
Fixup: Detect urls with unicode symbols
...
missed one in 77ecaa6093
2014-12-28 00:40:51 -05:00
77ecaa6093
Detect urls with unicode symbols
...
Fixes #1250
2014-12-28 00:29:58 -05:00
ec7a0d6e13
Fixed some more signed-unsigned-comparison warnings.
2014-12-15 10:25:28 -08:00
c1831cb19d
Minor tidying up
2014-12-08 21:24:59 -08:00
f83d78dd28
Warning cleanup
...
- ignoring const
- declarations after statements
- some C files didnt include own headers (risking them getting out of sync)
Closes #1064
2014-11-02 14:41:20 -05:00
80bdd9ce11
Detect utf8 urls
...
They might not be valid, but like many things they are still used
2014-09-07 19:51:59 -04:00
d1b9118056
Fix DOMAIN and EMAIL patterns.
...
Closes #1107
2014-09-01 17:55:22 -04:00
120a5a1dcd
Match nick last when checking url
...
Closes #796
2013-10-23 06:32:57 -04:00
6d3c176ff8
url_check_line () now recognizes all urls
...
previously it only recognized url with "://" after the scheme.
Urls without a scheme are purposely not recognized by this function.
Fixes #701
Closes #815
2013-10-22 13:28:03 -04:00
a25363f5c3
make_re() now doesn't free the argument.
...
This is good practice and also helped simplifying re_*() functions.
2013-10-22 13:26:30 -04:00
04dbcdffc0
Only underline the channel part if nick prefix
2013-09-23 15:10:00 -04:00
0146a38faa
Fix url detection of channel with nick prefix
...
Closes #769
2013-09-23 14:58:11 -04:00
73d63ca3da
Allow multiple channels in url, fix nick re
...
- Multiple channels have to be separated by a comma
- Fix #nick becoming a url for nick instead of #nick
2013-08-13 07:39:34 +03:00
4f77f046d0
Channel url and cursor fix
...
- Allow colons in channel urls
- Only add the channel part in the context menu
- Fix cursor not changing between separator and url
2013-08-13 04:19:14 +03:00
61c9ad08bf
Fixed memory leak.
2013-08-02 02:45:54 +01:00
6ec7bfad30
Indentation fix.
2013-08-01 16:39:49 +01:00
0e76283c9a
Now nickname and channel matching is server aware: we only match words that
...
have the prefixes the server uses.
The url regex matching was designed to be more versatile. This was needed to
disambiguate matches between nicks and channels, since they may have common
prefixes now (&+!). In case of ambiguity, when the server supports prefixes that
are common to both channels and nicknames, we choose the nickname (the nickname
only matches if there is a user with that nick in the current session).
Fixes #655 .
2013-08-01 16:28:17 +01:00
bca972bdc3
Permit dot-less domain names
...
Fixes #523
2013-07-19 09:01:53 -04:00
d8286a63ec
Add more uri schemes
2013-06-19 18:57:52 -03:00
9854b100fb
Detect Spotify URIs
2013-06-19 05:07:45 -04:00
d88c4ad411
Added a few more useful url schemes.
2013-06-18 12:32:55 +01:00
8b217981ed
Removed unused argument of make_re().
2013-06-18 12:07:46 +01:00
8ce8a359c1
Now url_check_word() returns WORD_HOST6 when a ipv6 address is matched.
2013-06-18 12:01:39 +01:00
55734baed8
Fix IPv6 addr matching in url: must be enclosed in [].
2013-06-17 22:43:18 +01:00
e4b6f4b375
Refined IPv6 address matching in url.c.
2013-06-16 20:16:58 +01:00
32dee82c58
Added support for passwords in userinfo of urls.
2013-06-16 20:00:35 +01:00
2870586cf9
Fixed irc:// and ircs:// url matching: irc://freenode was not working because
...
"freenode" doesn't match a hostname.
2013-06-16 19:11:26 +01:00
cb1975f450
Fixed file:// url matching.
2013-06-16 19:09:50 +01:00
805b33552b
Rework url matching to make it easier to add schemes.
...
The new way allows great control of what a url is composed of.
Added a bunch of new schemes.
2013-06-16 15:36:52 +01:00
6bc05a8bc8
Improved host matching in url:
...
* Refined ipv4 addr
* Added ipv6 addr
2013-06-16 04:22:28 +01:00
e17a9c6503
In do_an_re() put WORD_CHANNEL before WORD_HOST. fixes #440
2013-04-19 12:41:49 -07:00
6b86940b7b
Allow nick starting with digit despite rfc 2812 violation
2013-04-06 11:51:22 -07:00
3fee94473b
Don't clobber data, just adjust url endpoint
2013-03-31 12:14:52 -07:00
503968c1fd
Strip all \r and \n, not just a single \r.
2013-03-29 13:11:17 -07:00
b5f78f484d
Fix relative paths on windows
2013-03-11 13:55:29 -03:00
8e3d039199
Fix bug in url.c at re_host().
...
Unlike the other re_foo() functions it was not checking immediately
and returning if host_ret had already been filled in. This would
causes a memory leak since the previous GRegex would be lost.
2013-02-13 12:52:49 -08:00
ce31c6dce6
only detect full paths on windows.
...
closes #302
2013-01-31 23:58:18 -05:00
8b451edf9c
Merge pull request #377 from RichardHitt/issue376
...
Parse url according to rfc 1738: add separate define for TLD. Closes 376.
2013-01-16 14:06:50 -08:00
7bc0e017f6
Oops. didn't get the DOMAIN slight change last time. Here it is now.
2013-01-15 13:26:11 -08:00
bb617ccacd
Prev called anything a URL! Reverted but slight DOMAIN change.
2013-01-15 13:19:29 -08:00
d5fbea3411
Fix DOMAIN and TLD per rfc 1738, thanks Arnavion
2013-01-15 01:23:37 -08:00
be2d100d2d
Parse url according to rfc 1738: add separate define for TLD.
2013-01-14 12:33:08 -08:00
07b69fd2f8
Oops, Windows doesn't offer strndup(). Use g_strndup().
2013-01-11 16:12:52 -08:00
487ac0a011
Correct nick recognition. Closes 372.
2013-01-11 01:39:21 -08:00
ccc82895ed
Merge branch 'master' of http://github.com/RichardHitt/hexchat
2013-01-08 12:58:05 -08:00
5ec9d8534e
Add "ircs://" to the list of URL schemes
2013-01-08 12:55:20 -08:00
4bf2fbe092
Ywes, nicks can be as short as one character long!
2013-01-03 14:24:04 -08:00
44d920592d
For NICK0 use NICKSPE, not NICKDIG
2013-01-03 14:13:20 -08:00
f4c2c08ccb
Also allow URLs of of the form \.[a-z]+\/
2013-01-02 23:47:58 -08:00
TingPing