Awesome

telescope-egrepify.nvim

telescope-egrepify.nvim is a telescope.nvim-extension that enhances live grepping with opinionated defaults and functionality.

Features

• Extensible prefix-based CLI parsing for ripgrep, e.g. by default #md,lua sorter would look for sorter in files with md or lua extensions
• Better (opinionated) defaults like AND for tokens in prompt: this here expands to "this*.here" in regex matching this, any characters in between, and here
• Custom entry maker to parse ripgrep json to separate filenames as "section titles" (inspired by consult.el), configure line and column numbers, and perform accurate line highlighting
• Opt-in results highlighting with treesitter (needs results_ts_hl = true)

Motivation

Fuzzy-search is often not great at filtering results. 99% of my time, I mostly want to restrict searches by AND (intersection of words in search prompt) and files. ripgrep syntax is not as trivial as fzf but with, e.g., helpers like telescope-egrepify's prefixes enables more precise finding and filtering.

The screenshot shows searching for require only in files with md extension (i.e. markdown files). For more prefixes and configuration, see Prefixes.

Prefixes

The core functionality of telescope-egrepify.nvim are prefixes. What you need to know at a glance:

• Prefixes seamlessly expand user-specific ripgrep flags on-the-fly
• The below prefixes are the builtin-defaults with examples on how they are used
• During search, you can toggle the use of prefixes by hitting <C-z> (z) in insert (normal) mode
• The configuration section shows how to opt out of a default and add another prefix.

-- DEFAULTS
-- filter for file suffixes
-- example prompt: #lua,md $MY_PROMPT
-- searches with ripgrep prompt $MY_PROMPT in files with extensions lua and md
-- i.e. rg --glob="*.{lua,md}" -- $MY_PROMPT
{ ["#"] = {
      -- #$REMAINDER
      -- # is caught prefix
      -- `input` becomes $REMAINDER
      -- in the above example #lua,md -> input: lua,md
      flag = "glob",
      cb = function(input)
          return string.format([[*.{%s}]], input)
      end,
  },
  -- filter for (partial) folder names
  -- example prompt: >conf $MY_PROMPT
  -- searches with ripgrep prompt $MY_PROMPT in paths that have "conf" in folder
  -- i.e. rg --glob="**/conf*/**" -- $MY_PROMPT
  [">"] = {
      flag = "glob",
      cb = function(input)
          return string.format([[**/{%s}*/**]], input)
      end,
  },
  -- filter for (partial) file names
  -- example prompt: &egrep $MY_PROMPT
  -- searches with ripgrep prompt $MY_PROMPT in paths that have "egrep" in file name
  -- i.e. rg --glob="*egrep*" -- $MY_PROMPT
  ["&"] = {
      flag = "glob",
      cb = function(input)
          return string.format([[*{%s}*]], input)
      end,
  }
}

See also Configuration.

Installation

Here is one way to install this extension with lazy.nvim.

{
    "fdschmidt93/telescope-egrepify.nvim",
    dependencies = { "nvim-telescope/telescope.nvim", "nvim-lua/plenary.nvim" }
}

Make sure to

require "telescope".setup({ "$YOUR_TELESCOPE_OPTS" })
require "telescope".load_extension "egrepify"

to appropriately setup the plugin. See Configuration for options to configure the extension.

Usage

You can call this extension with vimscript

:Telescope egrepify

or lua

require "telescope".extensions.egrepify.egrepify {}

Configuration

The below configuration reflects defaults and examples on how to customize telescope-egrepify.nvim to your liking.

local egrep_actions = require "telescope._extensions.egrepify.actions"

require("telescope").setup {
  extensions = {
    egrepify = {
      -- intersect tokens in prompt ala "str1.*str2" that ONLY matches
      -- if str1 and str2 are consecutively in line with anything in between (wildcard)
      AND = true,                     -- default
      permutations = false,           -- opt-in to imply AND & match all permutations of prompt tokens
      lnum = true,                    -- default, not required
      lnum_hl = "EgrepifyLnum",       -- default, not required, links to `Constant`
      col = false,                    -- default, not required
      col_hl = "EgrepifyCol",         -- default, not required, links to `Constant`
      title = true,                   -- default, not required, show filename as title rather than inline
      filename_hl = "EgrepifyFile",   -- default, not required, links to `Title`
      results_ts_hl = true,           -- set to false if you experience latency issues!
      -- suffix = long line, see screenshot
      -- EXAMPLE ON HOW TO ADD PREFIX!
      prefixes = {
        -- ADDED ! to invert matches
        -- example prompt: ! sorter
        -- matches all lines that do not comprise sorter
        -- rg --invert-match -- sorter
        ["!"] = {
          flag = "invert-match",
        },
        -- HOW TO OPT OUT OF PREFIX
        -- ^ is not a default prefix and safe example
        ["^"] = false
      },
      -- default mappings
      mappings = {
        i = {
          -- toggle prefixes, prefixes is default
          ["<C-z>"] = egrep_actions.toggle_prefixes,
          -- toggle AND, AND is default, AND matches tokens and any chars in between
          ["<C-a>"] = egrep_actions.toggle_and,
          -- toggle permutations, permutations of tokens is opt-in
          ["<C-r>"] = egrep_actions.toggle_permutations,
        },
      },
    },
  },
}
require("telescope").load_extension "egrepify"

Mappings

DISCLAIMER

Please consider forking for your own customization or well-formed PRs instead to fix issues or add new features. This extension foremost serves my own needs and turned into a plugin as maybe other users may want to personalize rg via telescope.nvim in similar fashion. Many options are configurable, but I would like to avoid added maintenace burden beyond review (and hopefully merging) of PRs.

Naming

Kudos to ChatGPT:

egrepify combines the concept of "grep" (a common Unix command for searching through files) with the word "Epsilon" (the fifth letter of the Greek alphabet, which can represent "my" in mathematical notation). The resulting word suggests a personalized or customized version of grep.