mirror of https://tildegit.org/ben/sedbot
e336fdb008 | ||
---|---|---|
.cfg.example | ||
.gitignore | ||
LICENSE | ||
README.md | ||
account.ini.sample | ||
sedbot.bash | ||
sedbot.service |
README.md
sedbot
forked from clsr/sedbot -
updated for -i
's new argument requirement and daemonized
sedbot is an IRC search-replace bot written using bash and sed.
usage
cp .cfg{.example,}
- adjust as needed
bash sedbot.bash
daemonization
- adjust sedbot.service
mkdir -p ~/.config/systemd/user
cp sedbot.service ~/.config/systemd/user/
systemctl --user daemon-reload
systemctl --user enable --now sedbot
authenticate to services
cp account.ini{.sample,}
- fill in your credentials
- restart the bot (
systemctl --user restart sedbot
)
Only the s command and g and i flags are supported. Multiple regular expressions can be used at once, delimit them with spaces in between the flags and the s next one's s command. The last one may omit the trailing / if it has no options.
Example usage in chat:
<foo> Hello ther!
<foo> s/ther/there
<sedbot> <foo> Hello there!
<foo> I'm programmign right now
<bar> foo: s/gn/ng/
<sedbot> <foo> I'm programming right now
<foo> abcdefghi
<foo> s/\(.\)./\u\1/g s/
<sedbot> <foo> ACEGi
<foo> s/[a-e]//g s/\(.\)\(.\)/\2\1
<sedbot> <foo> gfhi
Note that the bot uses the standard grep (POSIX) regular expressions, i.e. use .\+
and \(foo\)\?
instead of .+
and (foo)?
as you'd do in egrep or other regex engines. Backreferences are written like s/\(.\)/\1
, where \1
matches the first capturing group. Read more about sed regular expressions.