Bash

Abstract

A Unix shell provides an interface that lets the user interact with the operating system by running commands. But a shell is also a fairly rich programming language: there are constructs for flow control, alternation, looping, conditionals, basic mathematical operations, named functions, string variables, and two-way communication between the shell and the commands it invokes. Shells can be used interactively, from a terminal or terminal emulator such as xterm, and non-interactively, reading commands from a file. Most modern shells, including bash, provide command-line editing, in which the command line can be manipulated using emacsor vi-like commands while it’s being entered, and various forms of a saved history of commands. Bash processing is much like a shell pipeline: after being read from the terminal or a script, data is passed through a number of stages, transformed at each step, until the shell finally executes a command and collects its return status. This chapter will explore bash’s major components: input processing, parsing, the various word expansions and other command processing, and command execution, from the pipeline perspective. These components act as a pipeline for data read from the keyboard or from a file, turning it into an executed command.