this post was submitted on 11 Feb 2024
61 points (96.9% liked)

Linux

46819 readers
1011 users here now

From Wikipedia, the free encyclopedia

Linux is a family of open source Unix-like operating systems based on the Linux kernel, an operating system kernel first released on September 17, 1991 by Linus Torvalds. Linux is typically packaged in a Linux distribution (or distro for short).

Distributions include the Linux kernel and supporting system software and libraries, many of which are provided by the GNU Project. Many Linux distributions use the word "Linux" in their name, but the Free Software Foundation uses the name GNU/Linux to emphasize the importance of GNU software, causing some controversy.

Rules

Related Communities

Community icon by Alpár-Etele Méder, licensed under CC BY 3.0

founded 5 years ago
MODERATORS
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
61
Something like kate advanced find/replace (regex, multiple files) which allows saving searches to use again later? (discuss.tchncs.de)
submitted 6 months ago by [email protected] to c/[email protected]
9 comments fedilink hide all child comments
 

I really like advance find and replace in kate editor. You can optionally use regex and operate on multiple files.

Very importantly it has a robust preview changes ability. it is comfortable to use even with lots of hits, lots of files. So you do not need to apply a bunch of changes and hope you considered every permutation as with a cli tool like sed.

One thing that would really improve my life would be a tool like this which allows you to save search queries and options.

Don't work for me:

  • Kate has a popup for history in the fields which is somewhat helpful but limited. When trying out different queries you don't have a way to remember which one actually worked so going by the history just ends in repeating the same errors over and over. Also it doesn't match the "find" and "replace" fields nor does it associate them with the other options like directory, etc.

  • Keeping notes in a text file is of course possible but cumbersome. I would like the computer to do work like that for me.

For single file searches regex101.com (non floss) and regexr.com (GPLv3) are great in-browser tools for learning and you can save the search. But to operate locally on many files, it doesn't work.

Does anyone know any tools that do anything like this? Can find various utilities which operate on file names but I am looking for file content. Certainly this exists ya?

(Post image is screenshot from Kate website of Kate on windows.)

top 9 comments
sorted by: hot top controversial new old
[–] [email protected] 9 points 6 months ago (1 children)

I think this would be a perfect project for writing a kate plugin! https://develop.kde.org/docs/apps/kate/plugin/

Until then, even though you have eliminated sed from your list, I suggest trying sed + grep to check your changes before modifying anything.

[–] [email protected] 2 points 6 months ago (1 children)

Well the odds of me learning C++ anytime soon are pretty low. But if someone else thought it was useful enough to spend time on, I guess starting with the existing plugin which is pictured above would probably be expedient...

That or a stand alone application.

I'm sure this must exist....

How do you use sed and grep to check?

I've been trying to use sed more but having everything in a single command makes it hard to think about when there is complexity. I know it's dumb but having a "find" box, a "replace" box, and toggles for options that are visually distinct from one another really helps me. Especially when it comes to re-using a line later on, I have a hard time following my own code if there are a lot of escapes or fancy regex components. The only reason I've been able to learn anything at all is because of the websites like regex101 and regexr that make everything colorful. Really what I want is a desktop version of those but it's probably too much to ask for.

[–] [email protected] 2 points 6 months ago* (last edited 6 months ago)

I've only done this a handful of times, so I'm probably missing a better way to do it, but:

1 grep with -r to find the thing I want to change in all the files

2 Google how to use sed again 😹

3 write a sed command to make the necessary change in just one file

4 Google bash loop syntax again 🙃

5 use a bash loop to pass all the files into my sed command, but have sed print the result to stdout, not change the files, and pipe that output into my grep command that shows what I want to change

6 sometimes no output means it works, sometimes the grep regex is general enough that I can see the change. Either way if it works run the bash loop with sed again but make sed modify the files in place.

It truly is a pain until I get the bash and sed syntax loaded in my brain again

Edit: I forgot the most important part: putting the check and modify steps into a script so that you have it again when you need it 😅

[–] [email protected] 6 points 6 months ago (1 children)

Oh the whole previewing of changes thing, I guess I'd probably just use git, to easily roll back.

[–] [email protected] 4 points 6 months ago (1 children)

I've tried it a couple times and can work in some contexts but becomes overly-complex in others. Example, if your fileset is not easily constrained by a single git repo. e.g. if there are multiple repos already existing. Or there are submodules. Or there are repos that have ignore files, but you want to include the files in the changes. Or there are a lot of files that choke git.

Plus it doesn't really facilitate showing the changes that easily. I guess then you immediately go look at it with a diff tool and try to ascertain if it is screwed up or not? The kate component is nice because it shows you a list of changed files by filename/location, which you can expand to lines, and which you can easily open the whole file. Highlights the matches. Very quickly flick though everything for manual error checking. I haven't been able to find any diff tool that is as easy to use. (Would love to learn of one.)

And it still doesn't address the whole concept of saving the query.. I guess if you would write an individual bash script for every query? Then have a directory of those to somehow riffle through when needed?

All this sounds like something a computer would be great at managing but I am shit at managing.

[–] [email protected] 3 points 6 months ago

Why not run sed and pipe to diff to preview changes?

You'd still have to manually copy out the command line to a notes file, but I don't think that that's too terrible. You could use a terminal-integrated snippets palette to make it a little smoother.

I'm not aware of any program that does exactly everything you want it to, so you might write your own or extend an existing one, as mentioned.

[–] [email protected] 1 points 6 months ago (1 children)

Something similar to VSCode's ability to open searches in an editor? Looks like you can save and reopen them too.

[–] [email protected] 1 points 6 months ago

oooooh a .code-search files sounds lovely

[–] [email protected] 0 points 6 months ago

grep and .bash_history