[](https://gitter.im/jlevy/the-art-of-command-line?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
Fluency on the command line is a skill often neglected or considered arcane, but it improves your flexibility and productivity as an engineer in both obvious and subtle ways. This is a selection of notes and tips on using the command-line that we've found useful when working on Linux. Some tips are elementary, and some are fairly specific, sophisticated, or obscure. This page is not long, but if you can use and recall all the items here, you know a lot.
- This guide is both for beginners and the experienced. The goals are *breadth* (everything important), *specificity* (give concrete examples of the most common case), and *brevity* (avoid things that aren't essential or digressions you can easily look up elsewhere). Every tip is essential in some situation or significantly saves time over alternatives.
- This is written for Linux, with the exception of the "[OS X only](#os-x-only)" section. Many of the other items apply or can be installed on other Unices or OS X (or even Cygwin).
- It includes both "standard" Unix commands as well as ones that require special package installs -- so long as they are important enough to merit inclusion.
- To keep this to one page, content is implicitly included by reference. You're smart enough to look up more detail elsewhere once you know the idea or command to Google. Use `apt-get`/`yum`/`dnf`/`pacman`/`pip`/`brew` (as appropriate) to install new programs.
- Learn basic Bash. Actually, type `man bash` and at least skim the whole thing; it's pretty easy to follow and not that long. Alternate shells can be nice, but Bash is powerful and always available (learning *only* zsh, fish, etc., while tempting on your own laptop, restricts you in many situations, such as using existing servers).
- Learn at least one text-based editor well. Ideally Vim (`vi`), as there's really no competition for random editing in a terminal (even if you use Emacs, a big IDE, or a modern hipster editor most of the time).
- Know how to read documentation with `man` (for the inquisitive, `man man` lists the section numbers, e.g. 1 is "regular" commands, 5 is files/conventions, and 8 are for administration). Find man pages with `apropos`. Know that some commands are not executables, but Bash builtins, and that you can get help on them with `help` and `help -d`.
- Learn about redirection of output and input using `>` and `<` and pipes using `|`. Know `>` overwrites the output file and `>>` appends. Learn about stdout and stderr.
- Learn about file glob expansion with `*` (and perhaps `?` and `[`...`]`) and quoting and the difference between double `"` and single `'` quotes. (See more on variable expansion below.)
- Basic file management: `ls` and `ls -l` (in particular, learn what every column in `ls -l` means), `less`, `head`, `tail` and `tail -f` (or even better, `less +F`), `ln` and `ln -s` (learn the differences and advantages of hard versus soft links), `chown`, `chmod`, `du` (for a quick summary of disk usage: `du -hs *`). For filesystem management, `df`, `mount`, `fdisk`, `mkfs`, `lsblk`. Learn what an inode is (`ls -i` or `df -i`).
- Learn to use `apt-get`, `yum`, `dnf` or `pacman` (depending on distro) to find and install packages. And make sure you have `pip` to install Python-based command-line tools (a few below are easiest to install via `pip`).
- In Bash, use **Tab** to complete arguments or list all available commands and **ctrl-r** to search through command history (after pressing, type to search, press **ctrl-r** repeatedly to cycle through more matches, press **Enter** to execute the found command, or hit the right arrow to put the result in the current line to allow editing).
- In Bash, use **ctrl-w** to delete the last word, and **ctrl-u** to delete all the way back to the start of the line. Use **alt-b** and **alt-f** to move by word, **ctrl-a** to move cursor to beginning of line, **ctrl-e** to move cursor to end of line, **ctrl-k** to kill to the end of the line, **ctrl-l** to clear the screen. See `man readline` for all the default keybindings in Bash. There are a lot. For example **alt-.** cycles through previous arguments, and **alt-*** expands a glob.
- Alternatively, if you love vi-style key-bindings, use `set -o vi` (and `set -o emacs` to put it back).
- For editing long commands, after setting your editor (for example `export EDITOR=vim`), **ctrl-x****ctrl-e** will open the current command in an editor for multi-line editing. Or in vi style, **escape-v**.
- To see recent commands, `history`. There are also many abbreviations such as `!$` (last argument) and `!!` last command, though these are often easily replaced with **ctrl-r** and **alt-.**.
- If you are halfway through typing a command but change your mind, hit **alt-#** to add a `#` at the beginning and enter it as a comment (or use **ctrl-a**, **#**, **enter**). You can then return to it later via command history.
- Use `xargs` (or `parallel`). It's very powerful. Note you can control how many items execute per line (`-L`) as well as parallelism (`-P`). If you're not sure if it'll do the right thing, use `xargs echo` first. Also, `-I{}` is handy. Examples:
- Understand that care is needed when variables and filenames include whitespace. Surround your Bash variables with quotes, e.g. `"$FOO"`. Prefer the `-0` or `-print0` options to enable null characters to delimit filenames, e.g. `locate -0 pattern | xargs -0 ls -al` or `find / -print0 -type d | xargs -0 ls -al`. To iterate on filenames containing whitespace in a for loop, set your IFS to to be a newline only using `IFS=$'\n'`.
- In Bash scripts, use `set -x` (or the variant `set -v`, which logs raw input, including unexpanded variables and comments) for debugging output. Use strict modes unless you have a good reason not to: Use `set -e` to abort on errors (nonzero exit code). Use `set -u` to detect unset variable usages. Consider `set -o pipefail` too, to on errors within pipes, too (though read up on it more if you do, as this topic is a bit subtle). For more involved scripts, also use `trap` on EXIT or ERR. A useful habit is to start a script like this, which will make it detect and abort on common errors and print a message:
- In Bash scripts, subshells (written with parentheses) are convenient ways to group commands. A common example is to temporarily move to a different working directory, e.g.
- In Bash, note there are lots of kinds of variable expansion. Checking a variable exists: `${name:?error message}`. For example, if a Bash script requires a single argument, just write `input_file=${1:?usage: $0 input_file}`. Arithmetic expansion: `i=$(( (i + 1) % 5 ))`. Sequences: `{1..10}`. Trimming of strings: `${var%suffix}` and `${var#prefix}`. For example if `var=foo.pdf`, then `echo ${var%.pdf}.txt` prints `foo.txt`.
- Brace expansion using `{`...`}` can reduce having to re-type similar text and automate combinations of items. This is helpful in examples like `mv foo.{txt,pdf} some-dir` (which moves both files), `cp somefile{,.bak}` (which expands to `cp somefile somefile.bak`) or `mkdir -p test-{a,b,c}/subtest-{1,2,3}` (which expands all possible combinations and creates a directory tree).
- In Bash, redirect both standard output and standard error via: `some-command >logfile 2>&1` or `some-command &>logfile`. Often, to ensure a command does not leave an open file handle to standard input, tying it to the terminal you are in, it is also good practice to add `</dev/null`.
- Use `man ascii` for a good ASCII table, with hex and decimal values. For general encoding info, `man unicode`, `man utf-8`, and `man latin1` are helpful.
- Use `screen` or [`tmux`](https://tmux.github.io/) to multiplex the screen, especially useful on remote ssh sessions and to detach and re-attach to a session. `byobu` can enhance screen or tmux providing more information and easier management. A more minimal alternative for session persistence only is `dtach`.
- It can be useful to make a few optimizations to your ssh configuration; for example, this `~/.ssh/config` contains settings to avoid dropped connections in certain network environments, uses compression (which is helpful with scp over low-bandwidth connections), and multiplex channels to the same server with a local control file:
- A few other options relevant to ssh are security sensitive and should be enabled with care, e.g. per subnet or host or in trusted networks: `StrictHostKeyChecking=no`, `ForwardAgent=yes`
- Consider [`mosh`](https://mosh.mit.edu/) an alternative to ssh that uses UDP, avoiding dropped connections and adding convenience on the road (requires server-side setup).
- To get the permissions on a file in octal form, which is useful for system configuration but not available in `ls` and easy to bungle, use something like
- For interactive selection of values from the output of another command, use [`percol`](https://github.com/mooz/percol) or [`fzf`](https://github.com/junegunn/fzf).
- For running a command with privileges, use `sudo` (for root) or `sudo -u` (for another user). Use `su` or `sudo bash` to actually run a shell as that user. Use `su -` to simulate a fresh login as root or another user.
- To locate a file by name in the current directory, `find . -iname '*something*'` (or similar). To find a file anywhere by name, use `locate something` (but bear in mind `updatedb` may not have indexed recently created files).
- For Amazon S3, [`s3cmd`](https://github.com/s3tools/s3cmd) is convenient and [`s4cmd`](https://github.com/bloomreach/s4cmd) is faster. Amazon's [`aws`](https://github.com/aws/aws-cli) and the improved [`saws`](https://github.com/donnemartin/saws) are essential for other AWS-related tasks.
- Know that locale affects a lot of command line tools in subtle ways, including sorting order (collation) and performance. Most Linux installations will set `LANG` or other locale variables to a local setting like US English. But be aware sorting will change if you change locale. And know i18n routines can make sort or other commands run *many times* slower. In some situations (such as the set operations or uniqueness operations below) you can safely ignore slow i18n routines entirely and use traditional byte-based sort order, using `export LC_ALL=C`.
- Know basic `awk` and `sed` for simple data munging. For example, summing all numbers in the third column of a text file: `awk '{ x += $3 } END { print x }'`. This is probably 3X faster and 3X shorter than equivalent Python.
- To rename multiple files and/or search and replace within files, try [`repren`](https://github.com/jlevy/repren). (In some cases the `rename` command also allows multiple renames, but be careful as its functionality is not the same on all Linux distributions.)
- As the man page says, `rsync` really is a fast and extraordinarily versatile file copying tool. It's known for synchronizing between machines but is equally useful locally. It also is among the [fastest ways](https://web.archive.org/web/20130929001850/http://linuxnote.net/jianingy/en/linux/a-fast-way-to-remove-huge-number-of-files.html) to delete large numbers of files:
- Know `sort`'s options. For numbers, use `-n`, or `-h` for handling human-readable numbers (e.g. from `du -h`). Know how keys work (`-t` and `-k`). In particular, watch out that you need to write `-k1,1` to sort by only the first field; `-k1` means sort according to the whole line. Stable sort (`sort -s`) can be useful. For example, to sort first by field 2, then secondarily by field 1, you can use `sort -k1,1 | sort -s -k2,2`.
- If you ever need to write a tab literal in a command line in Bash (e.g. for the -t argument to sort), press **ctrl-v****[Tab]** or write `$'\t'` (the latter is better as you can copy/paste it).
- The standard tools for patching source code are `diff` and `patch`. See also `diffstat` for summary statistics of a diff and `sdiff` for a side-by-side diff. Note `diff -r` works for entire directories. Use `diff -r tree1 tree2 | diffstat` for a summary of changes. Use `vimdiff` to compare and edit files.
- To convert text encodings, try `iconv`. Or `uconv` for more advanced use; it supports some advanced Unicode things. For example, this command lowercases and removes all accents (by expanding and dropping them):
- To know current cpu/disk status, the classic tools are `top` (or the better `htop`), `iostat`, and `iotop`. Use `iostat -mxz 15` for basic CPU and detailed per-partition disk stats and performance insight.
- For network connection details, use `netstat` and `ss`.
- For a quick overview of what's happening on a system, `dstat` is especially useful. For broadest overview with details, use [`glances`](https://github.com/nicolargo/glances).
- To know memory status, run and understand the output of `free` and `vmstat`. In particular, be aware the "cached" value is memory held by the Linux kernel as file cache, so effectively counts toward the "free" value.
- Java system debugging is a different kettle of fish, but a simple trick on Oracle's and some other JVMs is that you can run `kill -3 <pid>` and a full stack trace and heap summary (including generational garbage collection details, which can be highly informative) will be dumped to stderr/logs. The JDK's `jps`, `jstat`, `jstack`, `jmap` are useful. [SJK tools](https://github.com/aragozin/jvm-tools) are more advanced.
- Know about `strace` and `ltrace`. These can be helpful if a program is failing, hanging, or crashing, and you don't know why, or if you want to get a general idea of performance. Note the profiling option (`-c`), and the ability to attach to a running process (`-p`).
- Use `/proc`. It's amazingly helpful sometimes when debugging live problems. Examples: `/proc/cpuinfo`, `/proc/meminfo`, `/proc/cmdline`, `/proc/xxx/cwd`, `/proc/xxx/exe`, `/proc/xxx/fd/`, `/proc/xxx/smaps` (where `xxx` is the process id or pid).
- For deeper systems and performance analyses, look at `stap` ([SystemTap](https://sourceware.org/systemtap/wiki)), [`perf`](https://en.wikipedia.org/wiki/Perf_(Linux)), and [`sysdig`](https://github.com/draios/sysdig).
- It is remarkably helpful sometimes that you can do set intersection, union, and difference of text files via `sort`/`uniq`. Suppose `a` and `b` are text files that are already uniqued. This is fast, and works on files of arbitrary size, up to many gigabytes. (Sort is not limited by memory, though you may need to use the `-T` option if `/tmp` is on a small root partition.) See also the note about `LC_ALL` above and `sort`'s `-u` option (left out for clarity below).
- Use `grep . *` to quickly examine the contents of all files in a directory (so each line is paired with the filename), or `head -100 *` (so each file has a heading). This can be useful for directories filled with config settings like those in `/sys`, `/proc`, `/etc`.
- Say you have a text file, like a web server log, and a certain value that appears on some lines, such as an `acct_id` parameter that is present in the URL. If you want a tally of how many requests for each `acct_id`:
- To continuously monitor changes, use `watch`, e.g. check changes to files in a directory with `watch -d -n 2 'ls -rtlh | tail'` or to network settings while troubleshooting your wifi settings with `watch -d -n 2 ifconfig`.
-`sponge`: read all input before writing it, useful for reading from then writing to the same file, e.g., `grep -v something some-file | sponge some-file`
- To enable the Option key in OS X Terminal as an alt key (such as used in the commands above like **alt-b**, **alt-f**, etc.), open Preferences -> Profiles -> Keyboard and select "Use Option as Meta key".
- Be aware OS X is based on BSD Unix, and many commands (for example `ps`, `ls`, `tail`, `awk`, `sed`) have many subtle variations from Linux, which is largely influenced by System V-style Unix and GNU tools. You can often tell the difference by noting a man page has the heading "BSD General Commands Manual." In some cases GNU versions can be installed, too (such as `gawk` and `gsed` for GNU awk and sed). If writing cross-platform Bash scripts, avoid such commands (for example, consider Python or `perl`) or test carefully.
- [shellcheck](https://github.com/koalaman/shellcheck): A shell script static analysis tool. Essentially, lint for bash/sh/zsh.
- [Filenames and Pathnames in Shell](http://www.dwheeler.com/essays/filenames-in-shell.html): The sadly complex minutiae on how to handle filenames correctly in shell scripts.
With the exception of very small tasks, code is written so others can read it. With power comes responsibility. The fact you *can* do something in Bash doesn't necessarily mean you should! ;)
