.Dd May 18, 2021
.Dt UNSCOOP 1
.Os
.
.Sh NAME
.Nm unscoop
.Nd import IRC logs
.
.Sh SYNOPSIS
.Nm
.Op Fl v
.Op Fl N Ar network
.Op Fl c Ar context
.Op Fl d Ar path
.Op Fl f Ar format
.Ar
.
.Nm
.Fl D
.Op Fl d Ar path
.
.Sh DESCRIPTION
The
.Nm
utility imports plain text IRC logs
into the
.Xr litterbox 1
database.
Network and channel or query names
can be inferred from file paths, so
.Nm
should be run from the log directory root.
.
.Pp
The current path being imported
is printed to standard output.
Progress based on total file size
is printed to standard error.
.
.Pp
The arguments are as follows:
.Bl -tag -width Ds
.It Fl D
Delete duplicate events from overlapping imports
and exit.
This operation requires SQLite version 3.25.0 or newer.
.
.It Fl N Ar network
Override the network name of the imported logs.
.
.It Fl c Ar context
Override the channel or query name
of the imported logs.
Names beginning with
.Ql #
or
.Ql &
are assumed to be channels.
.
.It Fl d Ar path
Set the database path,
which must have been initialized by
.Xr litterbox 1 .
The default path is as in
.Xr litterbox 1 .
.
.It Fl f Ar format
Set the input log format.
The default is
.Cm generic .
.
.It Fl v
Print SQL
.Sy INSERT
statements on standard error.
.El
.
.Pp
The formats are as follows:
.Bl -tag -width Ds
.It Cm catgirl
Import logs from the
.Xr catgirl 1
IRC client.
.Bd -literal -offset indent
find ~/.local/share/catgirl/log \e
	-type f -name '*.log' -print0 |
xargs -0 unscoop -f catgirl
.Ed
.
.It Cm generic
Import logs using generic matchers.
Network and context names are inferred
from paths of the form
.Pa network/context/* .
Events of the following formats are matched:
.Bd -literal -offset indent
[timestamp] <nick> privmsg
[timestamp] -nick- notice
[timestamp] * nick action
.Ed
.
.It Cm irc
Import logs formatted as
IRC protocol messages tagged with
.Sy server-time .
The network and context must be set with
.Fl N
and
.Fl c .
.
.It Cm textual
Import logs from the Textual IRC client.
.Bd -literal -offset indent
find Textual -type f -name '*.txt' \e
	-not -path '*/Console/*' -print0 |
xargs -0 unscoop -f textual
.Ed
.
.It Cm weechat
Import logs from the WeeChat IRC client.
.Bd -literal -offset indent
find ~/.weechat/logs -type f -name 'irc.*.weechatlog' \e
	-not -name 'irc.server.*.weechatlog' -print0 |
xargs -0 unscoop -f weechat
.Ed
.Pp
.Nm
is not able to correctly infer
network and context names
for channel names containing dots.
Import these logs explicitly with
.Fl N
and
.Fl c .
.
.It Cm znc
Import logs from the
.Xr znc 1
.Sy log
module.
.Bd -literal -offset indent
find ~/.znc/moddata/log \e
	~/.znc/users/*/moddata/log \e
	~/.znc/users/*/networks/*/moddata/log \e
	-type f -name '*.log' \e
	-not -path '*/status/*' -print0 |
xargs -0 unscoop -f znc
.Ed
.El
.
.Sh SEE ALSO
.Xr litterbox 1
.
.Sh AUTHORS
.An June McEnroe Aq Mt june@causal.agency
.
.Sh BUGS
Send mail to
.Aq Mt list+litterbox@causal.agency
or join
.Li #ascii.town
on
.Li irc.tilde.chat .