Question: is there a way to configure Tonsky's indentation rule? #21

Activity

1. markx

   changed the title [-]Question: is there a way to configure Tonsky's indent rule?[/-] [+]Question: is there a way to configure Tonsky's indentation rule?[/+] on Dec 21, 2021
2. axvr commented on Dec 21, 2021

   

   axvr

   on Dec 21, 2021

   Last edited by axvr

   Member

   More actions

   From a quick scan of that page, it looks like clojurefmt doesn’t exist (yet?). But if it does exist, you can tell Vim to use it by setting formatprg like this:

   autocmd! FileType clojure setlocal formatprg=clojurefmt

   This will set the program used by the gq operator. So to format the entire file using clojurefmt you would type gggqG in normal mode.

   The downside to this is that formatting won’t happen automatically, but you could write a little Vim script to auto-format on file save.
3. axvr

   closed this as completedon Dec 22, 2021
4. markx commented on Dec 23, 2021

   

   markx

   on Dec 23, 2021

   Last edited by markx

   Author

   More actions

   Thanks for the tip about formatprg. With it I can use external tools like cljstyles to format the file.

   However this is not what I was looking for originally. I was looking for a way to adjust the vim indent config so that it auto-indents following Tonsky's indentation rule. This way the indentation would be correct while I type, so I don't need to save the file constantly.

   After some exploring, I found the indentation for lispwords basically works in my setting (I think mostly just set shiftwidth=2), so I just add everything to listpwordsby using:

   let g:clojure_fuzzy_indent_patterns = ['^\S\+']
   

   This works in most cases but sometimes doesn't. I haven't figured out why yet.

   Is there a better to achieve this?
5. axvr commented on Dec 23, 2021

   

   axvr

   on Dec 23, 2021

   Member

   More actions

   Oh, OK I think I understand now.

   Do you have an example of where your g:clojure_fuzzy_indent_patterns value isn't working properly?

   Maybe you could try this one (it should be slightly more accurate):

   let g:clojure_fuzzy_indent_patterns = ['^[\k\d]\+']
6. axvr

   reopened this on Dec 23, 2021
7. markx commented on Dec 24, 2021

   

   markx

   on Dec 24, 2021

   Last edited by markx

   Author

   More actions

   E.g. for this sometimes it doesn't work. (Sometimes it does work. Not sure why. )

   (+
    1
    1)
   

   Desired result would be:

   (+
     1
     1)
   

   Also let g:clojure_fuzzy_indent_patterns = ['^[\k\d]\+'] doesn't seem to work for me :(
   What does \k match?
8. axvr commented on Dec 27, 2021

   

   axvr

   on Dec 27, 2021

   Last edited by axvr

   Member

   More actions

   Thanks for the example. I'll look into this when I get a chance (hopefully soon).

   (The \k matches characters in the iskeyword option. This value lists (most of) the characters allowed in Clojure symbols.)
9. axvr

   self-assigned this

   on Dec 27, 2021
10. Deraen commented on Feb 4, 2022

    

    Deraen

    on Feb 4, 2022

    Last edited by Deraen

    Member

    More actions

    g:clojure_align_subforms option controls whether to indent function calls with one (tonsky's post) or two spaces ("emacs"/clojure-style-guide). The first, two spaces, is the default.

    If you haven't changed the option, check your indentexpr, formatexpr, formatprg, you might another plugin which is indenting function call forms with one space.
11. axvr

    added a commit that references this issue on Apr 27, 2023

    Begin rewrite of indentation code

    ...

    7d6bdd7
12. axvr

    added

    indentationAffects indentationAffects indentation

    on Jun 23, 2023
13. axvr

    linked a pull request that will close this issueRewrite indentation code #31on Jun 23, 2023
14. axvr

    added this to the Indentation overhaul milestone on Jun 23, 2023
15. garrett-hopper commented on May 8, 2025

    

    garrett-hopper

    on May 8, 2025

    Last edited by garrett-hopper

    More actions

    Looks like the (+<CR>...) issue is due to s:current_word calling searchpos('\v>'...) when it's on the symbol in question. This search fails when the symbol is a single character. Potentially making it move back a character before searching would be sufficient?

    RE: clojure_fuzzy_indent_patterns, I've been using this pattern which I adapted from the tree-sitter grammar

    local suffixonly = ":#'0-9\\/" -- not allowed as first character
    local forbidden = {
      [[\[\]\n\r\t \\]], -- brackets, whitespace, slashes
      '(){}"@~^;`,', -- parens, braces, special characters
      suffixonly,
    }
    local prefix = string.format("[^%s]", table.concat(forbidden))
    local suffix = string.format([[\(%s\|[%s]\)*]], prefix, suffixonly)
    local symbol = prefix .. suffix
    vim.g["clojure_fuzzy_indent_patterns"] = { symbol }

    It's slightly more accurate than relying on \k in the pattern. Still, I don't think there's an easy userland fix for the single character symbols.

    Edit: Ha, just saw #31; wish I'd looked closer before spending time looking into this. 😆