索性记一下.
第一个例子是常见用法:
//Flag例1
//author:Xiong Chuan Liang
//date:2015-4-10
package main
import (
"flag"
"fmt"
"os"
)
var (
levelFlag = flag.Int("level","级别")
bnFlag int
)
func init() {
flag.IntVar(&bnFlag,"bn",3,"份数")
}
func main() {
flag.Parse()
count := len(os.Args)
fmt.Println("参数总个数:",count)
fmt.Println("参数详情:")
for i := 0 ; i < count ;i++{
fmt.Println(i,":",os.Args[i])
}
fmt.Println("\n参数值:")
fmt.Println("级别:",*levelFlag)
fmt.Println("份数:",bnFlag)
}
/*
运行结果:
C:\TEMP\testflag>go run tf2.go -level 3 -bn=2
参数总个数: 4
参数详情:
0 : C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp\go-build158983983\command-line-arguments\_obj\exe\tf2.exe
1 : -level
2 : 3
3 : -bn=2
参数值:
级别: 3
份数: 2
*/
第二个例子,可以扩展功能:
//Flag例2 //author:Xiong Chuan Liang //date:2015-4-10 package main import ( "flag" "fmt" "os" "time" ) var ( /* 参数解析出错时错误处理方式 switch f.errorHandling { case ContinueOnError: return err case ExitOnError: os.Exit(2) case PanicOnError: panic(err) } */ //flagSet = flag.NewFlagSet(os.Args[0],flag.PanicOnError) flagSet = flag.NewFlagSet(os.Args[0],flag.ExitOnError) //flagSet = flag.NewFlagSet("xcl",flag.ExitOnError) verFlag = flagSet.String("ver","","version") xtimeFlag = flagSet.Duration("time",10*time.Minute,"time Duration") addrFlag = StringArray{} ) func init() { flagSet.Var(&addrFlag,"a","b") } func main() { fmt.Println("os.Args[0]:",os.Args[0]) flagSet.Parse(os.Args[1:]) //flagSet.Parse(os.Args[0:]) fmt.Println("当前命令行参数类型个数:",flagSet.NFlag()) for i := 0; i != flagSet.NArg(); i++ { fmt.Printf("arg[%d]=%s\n",i,flag.Arg(i)) } fmt.Println("\n参数值:") fmt.Println("ver:",*verFlag) fmt.Println("xtimeFlag:",*xtimeFlag) fmt.Println("addrFlag:",addrFlag.String()) for i,param := range flag.Args(){ fmt.Printf("---#%d :%s\n",param) } } type StringArray []string func (s *StringArray) String() string { return fmt.Sprint([]string(*s)) } func (s *StringArray) Set(value string) error { *s = append(*s,value) return nil } /* 运行结果: C:\TEMP\testflag>go run tfs.go -ver 9.0 -a ba -a ca -a d2 -ver 10.0 -time 2m0s os.Args[0]: C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp\go-build341936307\command-line-arguments\_obj\exe\tfs.exe 当前命令行参数类型个数: 3 参数值: ver: 10.0 xtimeFlag: 2m0s addrFlag: [ba ca d2] C:\TEMP\testflag>go run tfs.go -ver 9.0 -a ba -a ca -a d2 -ver 10.0 os.Args[0]: C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp\go-build712958211\command-line-arguments\_obj\exe\tfs.exe 当前命令行参数类型个数: 2 参数值: ver: 10.0 xtimeFlag: 10m0s addrFlag: [ba ca d2] -- flagSet = flag.NewFlagSet(os.Args[0],flag.PanicOnError) 结果: C:\TEMP\testflag>go run tfs.go -ver 9.0 -a ba -a ca -a d2 -ver 10.0 -time 2m0s33 os.Args[0]: C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp\go-build841833143\command-line-arguments\_obj\exe\tfs.exe invalid value "2m0s33" for flag -time: time: missing unit in duration 2m0s33 Usage of C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp\go-build841833143\command-line-arguments\_obj\exe\tfs.exe: -a=[]: b -time=10m0s: time Duration -ver="": version panic: invalid value "2m0s33" for flag -time: time: missing unit in duration 2m0s33 goroutine 1 [running]: flag.(*FlagSet).Parse(0x10b18180,0x10b42008,0xc,0x0,0x0) c:/go/src/flag/flag.go:814 +0xee main.main() C:/TEMP/testflag/tfs.go:41 +0x163 exit status 2 -- flagSet = flag.NewFlagSet(os.Args[0],flag.ExitOnError) 结果: C:\TEMP\testflag>go run tfs.go -ver 9.0 -a ba -a ca -a d2 -ver 10.0 -time 2m0s33 os.Args[0]: C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp\go-build501686683\command-line-arguments\_obj\exe\tfs.exe invalid value "2m0s33" for flag -time: time: missing unit in duration 2m0s33 Usage of C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp\go-build501686683\command-line-arguments\_obj\exe\tfs.exe: -a=[]: b -time=10m0s: time Duration -ver="": version exit status 2 */flagSet类可以让参数处理更加灵活.
其中NewFlagSet:
flagSet = flag.NewFlagSet("xcl",flag.ExitOnError)
NewFlagSet的第一个参数是可以任意定的.但第二个参数,则决定了参数解析出错时错误处理方式.
如果要扩展参数定义,只要实现下面的接口:
type Value interface {
String() string
Set(string) error
}依例子中的 StringArray 一样.就可以实现如:
-a ba -a ca -a d2
addrFlag: [ba ca d2]
这种效果.
要知道,同样的:
-ver 9.0 -ver 10.0最后的结果是 ver: 10.0 即参数只识别最末一次的.
没想到还能通过interface进行扩展,并且其使用比起Linux的getopt()系列函数.flag包这种方式还是更清晰直观.
