Learning Notes - Linux Commands

This post documents my learning notes of useful Linux commands

Examples used is my VIM key mapping configuration file: ~/.config/nvim/lua/user/keymappings.lua

local opts = { noremap = true, silent = true }
local term_opts = { silent = true }

local keymap = vim.api.nvim_set_keymap

--Remap space as leader key
keymap("", "<Space>", "<Nop>", opts)
vim.g.mapleader = " "
vim.g.maplocalleader = " "

-- Modes
--   normal_mode = "n",
--   insert_mode = "i",
--   visual_mode = "v",
--   visual_block_mode = "x",
--   term_mode = "t",
--   command_mode = "c",

-- Normal --

keymap("n", "<S-e>", "$", opts)
keymap("n", "<S-b>", "^", opts)
keymap("n", "<S-y>", "yg_", opts)

-- remap j/k to gj/gk
keymap("n", "j", "gj", opts)
keymap("n", "k", "gk", opts)
keymap("n", "gj", "j", opts)
keymap("n", "gk", "k", opts)

grep: Pattern Search

grep searches particular regular expression patterns in file(s). This is extremely useful when you need to search for a particular piece of information in one or multiple long text files.

For example, I can search all my keybindings that involves leader key.

grep -i '<leader>' keymappings.lua

This would give me the following:

keymap("n", "<leader>c", ":bd<CR>", opts)
keymap("n", "<leader>e", ":NERDTreeToggle<CR>", opts)
keymap("n", "<leader>tf", ":NERDTreeFind<CR>", opts)
--keymap("n", "<leader>t", ":VimwikiToggleListItem<CR>", opts)
keymap('n', "<leader>ff", ":lua require('telescope.builtin').find_files()<cr>",opts)
keymap('n', "<leader>fg", ":lua require('telescope.builtin').live_grep()<cr>",opts)
keymap('n', "<leader>fb", ":lua require('telescope.builtin').buffers()<cr>",opts)
keymap('n', "<leader>fh", ":lua require('telescope.builtin').help_tags()<cr>",opts)
keymap("n", "<leader>h", ":noh<CR>", opts)
keymap("n", "<leader>s", "1z=<CR>", opts)
keymap("n", "<leader>v", ":vsplit<CR>", opts)

Many times you wound want to search file using the regular expression. For example, I can search all the modes in my configuration file.

grep -P '\-\-.{0,}\-\-' keymappings.lua

This will list out the section lines in the file which are written in a pattern like "-- Mode Name --". Linxu usually support pearl regular expressions.

-- Normal --
-- Insert --
-- Visual --
-- Visual Block --
-- Command Mode --

There are also output options you can use with grep, meaning you can not only just output the line with matched expressions. You can also output n lines before (-B) or n lines after (-A) or both (-C). This is useful when you want to find the context of the line when you search. For example I can search anything associated with vimwiki and return 3 lines before and after the matching line.

grep -i -C 3 'vimwiki' keymappings.lua

Below is the output.

keymap("n", "<leader>c", ":bd<CR>", opts)
keymap("n", "<leader>e", ":NERDTreeToggle<CR>", opts)
keymap("n", "<leader>tf", ":NERDTreeFind<CR>", opts)
--keymap("n", "<leader>t", ":VimwikiToggleListItem<CR>", opts)

-- Telescope
keymap('n', "<leader>ff", ":lua require('telescope.builtin').find_files()<cr>",opts)

sed

sed is used to find & replace patterns in an input stream (text file)

Useful flags:

• -i: inplace, useful when you need to modify the file
• -n: don't output the input stream
• -E: extended regular expressions such as grouping

Useful commands:

• s: replace
• d: delete
• c: change
• p: print

Useful tricks:

• Use line numbers to specify a particular line
• Use comma to specify a block
• Use grouping () in the search part and then reuse it in the replace part

Examples

1. sed to replace grep

sed -n /find_pattern/p input file

2. To find and replace a string (-n to show only affected lines):

# To output the replaced stream
sed -n s/old_string/new_string/p input_file

# To change inplace in the file
sed -i s/old_string/new_string/ input_file

3. To find and replace within a block of text

sed -n /block_start/,/block_end/s/old_string/new_string/p input_file

4. To delete an entire block

sed  /block_start,/block_end/d input_file

5. To change and entire blcok with new contents

sed /block_start/,/block_end/cNew\ Block\ Content input_file

awk

awk is a more powerful tool (more like a programming language) to handle text streams. awk is specifically powerful to deal with csv files.

The basics of awk are:

awk 'the_pattern_to_search_for {the actions to perform}'

Special identifiers:

• $1, $4: column 1 to column 4
• $0: all columns
• $NF: the last column
• FNR == n: the n-th row (both row and column are 1-based)
• NR: line numbers
• ,: similar to sed use , to specify blocks
• BEGIN/END: first/last match

Useful flags:

• -F: to specify column separator (by default is space).
• -i: source a library (such as -i inplace)

Functions can be used inside the pattern part

• search: /regex_pattern/
• length

Functions used inside the action part

• if(...)
• for(i=1;i<=10;i++) ...

Useful tricks

Examples

1. Replacement for grep

# grep lines with 'bin' from /etc/passwd
awk '/bin/' /etc/passwd

# grep the first column
awk -F ":" '{print $1}' /etc/passwd

# grep the first row
awk 'FNR==1' /etc/passwd

2. Replacement for sed

awk -F ":" '/root/ {print $1, "FOUND: ", $2}' /etc/passwd