Improve the way copycat handles stored regex searches

[bruno-]

Improve the way copycat handles stored regex searches

[bruno-]

After looking into this, I have no idea how could this be improved..

Currently users can define "stored searches" by having something like this in .tmux.conf:

set -g "@copycat" "C-x|my_regex_search"

The above will create a binding prefix + C-x that will perform a search. The same mechanism is used for builtin searches eg prefix + C-f, prefix + C-u etc...

The idea was to improve it by leveraging tmux variable appending, but that is broken in tmux:

tmux set -g "@var" "foo"
tmux set -ag "@var" "bar"       # appending "bar" to the variable
tmux show -g "@var"             # => "bar", but should be "foobar"

[soli]

Ok, now that I understand how it works, it is not that bad.

Indeed set -a does not work for user-options, so there might not be any necessity to improve anything. Worst case scenario, it remains possible to tmux set -ag @copycat "$(tmux show -gvq @copycat) foo" to append to an existing (or not) var.

What is needed is the documentation for the various tmux options that copycat uses. I'll try to add this and create a PR tonight.

[bruno-]

This is a good idea:

tmux set -g @copycat "$(tmux show -gvq @copycat) foo"

Another option for handling user defined searches would be to introduce a new variable eg @copycat_searches. @copycat variable can stay the same for "plugin defined" searches.

The thing that gives me pause about both approaches is that user defined regex could quickly become messy. Example:

# users `tmux.conf
set -g @copycat "C-x|my_stored_search_regex_that_is_long_and_messy C-z|another_stored_search_that_makes_the_line_too_long"

Anyway, if you get any idea for this, I'd be glad to hear it. On the other hand, there's no need to rush with implementation...

[soli]

Well it is always possible to write:

set -g @copycat " \
C-x|my_stored_search_regex_that_is_long_and_messy \
C-z|another_stored_search_that_makes_the_line_too_long \
etc."