# complete [![Build Status](https://travis-ci.org/posener/complete.svg?branch=master)](https://travis-ci.org/posener/complete) [![codecov](https://codecov.io/gh/posener/complete/branch/master/graph/badge.svg)](https://codecov.io/gh/posener/complete) [![GoDoc](https://godoc.org/github.com/posener/complete?status.svg)](http://godoc.org/github.com/posener/complete) [![Go Report Card](https://goreportcard.com/badge/github.com/posener/complete)](https://goreportcard.com/report/github.com/posener/complete) A tool for bash writing bash completion in go. Writing bash completion scripts is a hard work. This package provides an easy way to create bash completion scripts for any command, and also an easy way to install/uninstall the completion of the command. ## go command bash completion In [gocomplete](./gocomplete) there is an example for bash completion for the `go` command line. This is an example that uses the `complete` package on the `go` command - the `complete` package can also be used to implement any completions, see [Usage](#usage). ### Install 1. Type in your shell: ``` go get -u github.com/posener/complete/gocomplete gocomplete -install ``` 2. Restart your shell Uninstall by `gocomplete -uninstall` ### Features - Complete `go` command, including sub commands and all flags. - Complete packages names or `.go` files when necessary. - Complete test names after `-run` flag. ## complete package Supported shells: - [x] bash - [x] zsh ### Usage Assuming you have program called `run` and you want to have bash completion for it, meaning, if you type `run` then space, then press the `Tab` key, the shell will suggest relevant complete options. In that case, we will create a program called `runcomplete`, a go program, with a `func main()` and so, that will make the completion of the `run` program. Once the `runcomplete` will be in a binary form, we could `runcomplete -install` and that will add to our shell all the bash completion options for `run`. So here it is: ```go import "github.com/posener/complete" func main() { // create a Command object, that represents the command we want // to complete. run := complete.Command{ // Sub defines a list of sub commands of the program, // this is recursive, since every command is of type command also. Sub: complete.Commands{ // add a build sub command "build": complete.Command { // define flags of the build sub command Flags: complete.Flags{ // build sub command has a flag '-cpus', which // expects number of cpus after it. in that case // anything could complete this flag. "-cpus": complete.PredictAnything, }, }, }, // define flags of the 'run' main command Flags: complete.Flags{ // a flag -o, which expects a file ending with .out after // it, the tab completion will auto complete for files matching // the given pattern. "-o": complete.PredictFiles("*.out"), }, // define global flags of the 'run' main command // those will show up also when a sub command was entered in the // command line GlobalFlags: complete.Flags{ // a flag '-h' which does not expects anything after it "-h": complete.PredictNothing, }, } // run the command completion, as part of the main() function. // this triggers the autocompletion when needed. // name must be exactly as the binary that we want to complete. complete.New("run", run).Run() } ``` ### Self completing program In case that the program that we want to complete is written in go we can make it self completing. Here is an [example](./example/self/main.go)