Improve performance when a file has many matching lines

[krobelus]

We open and close a file for every line. This can be needlessly slow if one file has lots of matching lines. I wonder if there's an elegant way of optimizing that.

[jtrv]

Sorry to leave this without a response, it's been on the back of my mind and I was unsure of how to approach this.

Ideally I would like to diff the grep buffer and only replace the inserted lines in the diff. I'm not quite sure how to best implement this cleanly.

Alternatively we could pipe the buffer out to a faster tool for replacement like awk which should be much faster than scripting kakoune to open files and replace lines the way we are now.

https://unix.stackexchange.com/questions/677578/sed-replace-line-number-with-new-text-and-variables-in-text-file

I still need to look into this more...

[krobelus]

No worries. I actually don't know what kind of solution I want. I rarely need it these days (usually lsp-rename does the job) but it can be handy sometimes