* Fix inaccuracy in users._add docstring
Also don't track disconnected users by nick anymore. But that seems less
major than fixing the docstring.
* Fixes
* Fix docstring
Also allow prefixing commands by their role name to remove ambiguity
should a person be multiple roles. For example, "seer see foo" and
"augur see foo" will now work if a person is both seer and augur
(whereas normal see foo would be ambiguous). A player will be directed
to use the unambiguous prefixed version if we detect that a role command
will fire multiple times for them (note: coming soon).
For sanity reasons, these role prefixes are implemented as exclusive
commands, meaning no other commands or command aliases may use the same
name. Clone needs to be special-cased in this regard, as clone is both a
role name and a command name.
This doesn't show the variables for a few specific functions, specifically those which have a '_ignore_locals_' local variable set to a truthy value. For now, those functions are handle_error.__call__ (which is present in every traceback ever, due to its nature), cmd.caller and command.caller (one or the other is present in a large number of tracebacks, again due to their nature).
This makes debugging easier when the error is not in the innermost frame, but rather in one or two frames before, and the locals are lost.
Causing way too many bugs, need to use a more sane method of breaking
the cmd API, and only do so when it's actually ready to go. My
personal preference is to make the new API use @command instead of @cmd.
This reverts commits c90d35e6c0a9ba96692e9d73bbe27efca405d542 and 5f5966a8b49e5214c82d806ac43a2553754fdee6.
Since we have the UUID now, we can delete them at any time if needed,
they don't have to expire automatically, which can be annoying when
trying to look up an old error (although it's also in errors.log).
Now, local variables from the innermost frame (i.e. where the error was raised) are printed with the traceback, and the latter includes all of the call stack (instead of just up to the innermost error handler).
Also add framework so that custom roles can modify what roles are part
of a pack of tetrahedron wolves more easily, but no other part has been
done to help that.
For now, seer and wild child are done. There are some aspects of those
roles that are still incomplete, namely:
- sorcerer and alpha wolf handling of seer/oracle/augur is still in
wolfgame.py instead of via events
- wild child does not modify !stats
All hostmask and account comparisons are now case-insensitive (nicks
still aren't, related to #217 -- changing nick sensitivity would break
everything in numerous places).
Also, refactor some things into other files where it makes sense to do
so, because putting unrelated things into the same commit is fun.
- fwarn/warn commands to view and manipulate warnings
- fstasis can now only decrease stasis, not add to it
- refreshdb command can sync bot game state with what is in the db
(including expiring any unexpired stasis or warnings)
- stasis now expires
- tempban is still not implemented and will not be implemented as
part of the PR (it will come later, if ever)
- sanctions can be automatically applied after warnings cross a
certain threshold; some defaults are configured
- Always apply host-based allows/denies even in var.ACCOUNTS_ONLY is set
- Fix display of mixed-mode allows/denies for a particular user
- Make -acc and -host switches operate more intelligently on both the
bare command as well as when given an argument
This allows for things like making ADMINS/OWNERS take a full hostmask,
or fallowing/fdenying a full hostmask with wildcards. It also allows for
a distinction between two people that share a hostmask but have a
different ident.
stasis now checks all matching stasis and returns the highest amount
fallow/fdeny have two new options -acc and -host to explicitly specify
what type the argument is
-cmds in fallow/fdeny show commands from all matching masks, to bring it
in line with how actual allow/deny parsing works, before it would only
show the first match
Right now the only thing that still tracks accounts if DISABLE_ACCOUNTS
is True is stats stuff (in sqlite), as I'm unsure of how we want to
handle that for when accounts aren't enabled; likely track by nick but
unsure on that right now.
This further implements issue #112 but doesn't fix it entirely yet.
_collections is the C implementation (and where defaultdict lies), and
is as such already loaded when the interpreter is launched. collections
imports a bunch of other useless stuff we're not interested in.
