如何通过命令行(多文件包)在工作目录内运行所有. go 文件

我是新来的。我非常喜欢这种语言,但我很快意识到,由于程序大小的增加,我需要开始分割我的文件。

go run main.go(main.go 是我的 main ()函数的文件)

没有工作,我遇到了一个障碍,一段时间,因为我不知道如何让我的程序工作。

快速的搜索让我找到了

go run main.go other.go ..

通过输入我的 package main包含的所有文件,我可以让程序运行起来。然而,这样做每次都非常麻烦和令人沮丧。

我写下这些自我回答的问题是为了防止像我这样的人再次碰到这个障碍。

87826 次浏览

Unix related systems

go run *.go will be sufficient in most cases.

Continue to the below method if this causes errors.

Windows systems (and in other cases where go run *.go doesn't work)

Token expansion doesn't work in the windows command line and hence the above will not work and display an error. go run *.go may also not work in OSs in some cases due to current compiler limitations.

In these cases, use

go build && foo.exe

where foo.exe is the name of the .exe file produced. If perhaps you have no idea what the name of your executable is, first

go build and check the name of the .exe file produced. Afterwards, use the method that includes the file name.

These 2 methods will build and run all the .go files within your current directory with minimum fuss.

If i understand your question right, you need import other code as libraries.

Little example

./testing/main.go:

package main


import "fmt"
import L "testing/lib"


func main() {
fmt.Println("Hello from main()")
L.Somefunc()
}

./testing/lib/somelib.go:

package lib


import "fmt"


func Somefunc() {
fmt.Println("Hello from Somefunc()")
return
}

To launch - go run main.go

As Nate Finch notes:

Go run is ... really only meant to be used on very small programs, which generally only need a single file.

Even on unix, go run *.go is often not correct. In any project with unit tests (and every project should have unit tests), this will give the error:

go run: cannot run *_test.go files (something_test.go)

It will also ignore build restrictions, so _windows.go files will be compiled (or attempted to be compiled) on Unix, which is not what you want.

There has been a bit of discussion of making go run work like the rest of the go commands, and there's an open CL for it (5164). It's currently under consideration for Go 1.4. In the meantime, the recommended solution on all platforms is:

go build && ./<executable>

The best way to do it is to run it like this:

go run !(*_test).go

It skips all your test files which is exactly what you need to avoid the error.

The other suggestion:

go build && ./<executable>

is a bit annoying. You have to delete the executable all the time to avoid being marked by git. You can put it in gitignore, of course, but I am lazy and this is an extra step.

You can run all .go files, excluding tests, using this bash construction:

go run $(ls -1 *.go | grep -v _test.go)

On Windows I usually just add a little test.bat to all my project directories:

go build
.\{project_name}.exe
go clean

Works well enough. Replace {project_name} with the name of your executable, of course. And once the executable finishes, the batch script moves on to the clean up.

just use this

go run *.go

it will work assuming u don't have any test files

Here is my solution:

go run $(find . -name "*.go" -and -not -name "*_test.go" -maxdepth 1)

I use it with an alias to make it easy to run command line apps

alias gorun='go run $(find . -name "*.go" -and -not -name "*_test.go" -maxdepth 1)'


$ gorun param1 param2

For peoples attempting to use go run combined with go generate a solution can be :

//go:generate sh -c "go run path/*.go"

For window the following works: Open cmd and go to the path where your folder exists. Then type the following command and press Enter.

go build

after this one executable will be created. Then in the command prompt call the executable. If your executable name is Project.exe then type the following and press Enter:

Project.exe

If you are run in terminal of Vscode on Windows AND without knowledge about workspace, just do like this:

go run $(ls *.go)