Brian S. Stephan
694f051994
allow karma to happen anywhere in a sentence (rather than just the start), switch to re.search() throughout
2010-10-29 13:08:48 -05:00
Brian S. Stephan
b7f2b9bd0e
use re.search() rather than re.match() in a couple places
...
(search() looks for the regex anywhere in the string whereas
match() only looks for the regex starting from the beginning)
2010-10-29 00:30:02 -05:00
Brian S. Stephan
b5e71d677f
first cut at a database-driven fact storage module.
...
may still be a bit brittle
2010-10-29 00:00:55 -05:00
Brian S. Stephan
487e997a55
define a regexp function in sqlite
2010-10-28 23:52:06 -05:00
Brian S. Stephan
8fbd58e2c6
chmod -x modules/Karma.py
2010-10-28 20:56:59 -05:00
Brian S. Stephan
aabc8904ad
properly handle recursion, so that the alias actually works
2010-10-28 20:46:09 -05:00
Brian S. Stephan
0dfe8bcdd6
have karma module use the normal, non-alias-like command ('rank' rather than '!rank')
2010-10-28 20:15:48 -05:00
Brian S. Stephan
352b9e1830
more punctuation in the regex
2010-10-28 20:08:30 -05:00
Brian S. Stephan
5dda9ac1e5
change the format of the pi output a bit
2010-10-28 20:05:08 -05:00
Brian S. Stephan
eeb8ff83bf
fix a lack of encoding bug in a place where it's probably been lingering for a while.
2010-10-27 23:16:07 -05:00
Brian S. Stephan
898a4aa6c0
Revert "remove even more unicode()/encode() calls."
...
turns out there was a reason why i did this one.
This reverts commit 1afa34554f
.
2010-10-27 23:12:46 -05:00
Brian S. Stephan
1afa34554f
remove even more unicode()/encode() calls.
...
it's now to the point that i have no idea what convinced me to
make them in the first damn place.
2010-10-27 22:54:26 -05:00
Brian S. Stephan
aae4dfa062
remove a (hopefully) useless unicode() call
2010-10-27 22:48:52 -05:00
Brian S. Stephan
8384a54961
updating the TODO. there's probably more stuff here that i forgot
2010-10-27 22:40:24 -05:00
Brian S. Stephan
cb7044e64a
this encode('utf-8')/decode('utf-8') seems to unnecessary and accomplishes nothing?
...
still haven't figured out how to get around the whole character replace nonsense though
2010-10-27 22:09:59 -05:00
Brian S. Stephan
7c4125e099
properly use encode() to do the rot13 conversion, ignore conversion failures
2010-10-27 21:56:18 -05:00
Brian S. Stephan
1bc0dd0b2a
TextTransform module. right now it only does rot13
2010-10-27 19:57:18 -05:00
Brian S. Stephan
eefc41a6d7
last_insert_rowid() only makes sense for the current connection, so ORDER BY id DESC instead
2010-10-27 19:32:09 -05:00
Brian S. Stephan
2d8a86950c
remove unneeded ConfigParser
2010-10-27 19:31:43 -05:00
Brian S. Stephan
97893a3b7a
convert Pi.py to use sqlite backend
2010-10-27 18:02:26 -05:00
Brian S. Stephan
1d73deda1c
sqlite related stuff as part of making sqlite the canonical backend
...
* Module.py __init__ sets up sqlite db connection by default
* Module.py __init__ calls init_db() which is empty, expects subclasses to implement as necessary
* Module.py doesn't close sqlite connection by default
Changes call for a couple updates in Karma.py, namely implementing db_init
and excepting sqlite3.Error rather than closing the connection
2010-10-27 17:59:01 -05:00
Brian S. Stephan
70b49ecbcc
whitespace nitpicking
2010-10-25 21:14:35 -05:00
Brian S. Stephan
3d65a6a158
Merge branch 'bloy-sqlite'
...
Resolved conflicts:
modules/Karma.py - re-added support for (long items)
2010-10-25 21:11:40 -05:00
Brian S. Stephan
8cfeef2efd
Revert "Merge branch 'bloy-sqlite'" because i want to merge this the right way
...
This reverts commit 5a76a9866a
.
2010-10-25 21:06:01 -05:00
Brian S. Stephan
c1c29648d7
Revert "whitespace nitpicking" because I want to merge this the right way
...
This reverts commit b9c2be8c54
.
2010-10-25 21:05:33 -05:00
Brian S. Stephan
b9c2be8c54
whitespace nitpicking
2010-10-25 20:50:59 -05:00
Brian S. Stephan
5a76a9866a
Merge branch 'bloy-sqlite'
2010-10-25 20:49:12 -05:00
Brian S. Stephan
5f6ced3066
add ' to karma matcher regex
2010-10-25 19:21:39 -05:00
Brian S. Stephan
9a1086e855
added support for karma for things with spaces, by wrapping it in ()
...
e.g.: (this is one really long thing that i like)++
2010-10-25 19:11:17 -05:00
Brian S. Stephan
2973d904f4
pythonic header
2010-10-25 18:44:28 -05:00
Brian S. Stephan
a4839fd35f
whitespace nitpickery
2010-10-25 18:37:30 -05:00
Brian S. Stephan
69c8922024
Merge branch 'bloy-master'
2010-10-25 18:31:28 -05:00
Brian S. Stephan
ebd89d35cc
(hopefully) properly catch/log URLError, IndexError
2010-10-25 18:23:25 -05:00
Mike Bloy
84ba3107f0
include rank in karma output
2010-10-24 15:13:01 -05:00
Mike Bloy
98f07105bc
karma data store moved to sqlite
2010-10-24 15:13:00 -05:00
Mike Bloy
6a67795b18
modules know how to ask for a version number from the database
2010-10-24 15:11:39 -05:00
Mike Bloy
56d2847285
ignore database files
2010-10-24 15:11:38 -05:00
Mike Bloy
48427ecd21
create the database if it doesn't exist, on startup
2010-10-24 15:11:37 -05:00
Mike Bloy
240612fecf
database connection grabbing for Modules
2010-10-24 15:11:35 -05:00
Mike Bloy
40ab270324
fixed rank trigger for karma
2010-10-24 14:57:23 -05:00
Mike Bloy
e557137f82
fix karma formatting
2010-10-24 09:37:43 -05:00
Mike Bloy
e453778791
clean up karma module
2010-10-24 09:36:15 -05:00
Mike Bloy
b1fe99d5a2
ignore nbproject dir
2010-10-24 08:43:22 -05:00
Mike Bloy
a64c83ce59
working karma module
2010-10-23 22:52:15 -05:00
Brian S. Stephan
a45af7a0a7
Pi -> pi, fixing what I'm going to claim is a typo.
2010-10-20 11:57:31 -05:00
Brian S. Stephan
c3270ef3df
module to calculate pi over time, via the monte carlo method.
...
idea pinched from #linode.
2010-10-18 22:48:51 -05:00
Brian S. Stephan
2a1ef11d95
don't attempt recursion when the subcmd is, in fact, an empty string (i.e. text was '[]')
2010-10-12 11:54:43 -05:00
Brian S. Stephan
95a27cc44b
read brainfile filename from config file, allow default
2010-10-09 20:37:15 -05:00
Brian S. Stephan
05f1ce325a
put MegaHAL module load in __init__, which is now being implemented here
...
obviously, we need to call Module.__init__ as well, and as a nice
side effect, doing all of this lets us not need to do the on_connect
nonsense, which was the wrong thing to do anyway
2010-10-09 19:36:00 -05:00
Brian S. Stephan
52b740a52c
split the register_handlers into a method other than the object constructor
...
the reasoning behind this is that we may want to load one object
but have it connect to multiple servers. this allows that.
2010-10-09 18:52:51 -05:00