在 Shell 脚本中调用另一个 Shell 脚本的三种方式以及返回值问题

  • Post author:
  • Post category:其他

https://blog.csdn.net/simple_the_best/article/details/76285429

脚本调用:

先来说一下主要以下有几种方式:

  • fork: 如果脚本有执行权限的话,path/to/foo.sh。如果没有,sh path/to/foo.sh。新开启子shell,需要在父shell定义环境变量的变量子shell才可以使用可以继承环境变量。在脚本中定义环境 变量export a=”111″​ 或 delcare -x a="aaaa"​​​​​​
  • execexec path/to/foo.sh 不开新shell使用当前shell,但是父shell的exec之后就不执行了。需要在父shell定义环境变量的变量子shell才可以使用。在脚本中定义环境 变量export a=”111″​ 或 delcare -x a="aaaa"​​​​​​
  • sourcesource path/to/foo.sh 不新开shell使用当前shell所以父shell的环境变量子shell都可以直接使用。

fork

fork 是最普通的, 就是直接在脚本里面用 path/to/foo.sh 来调用 
foo.sh 这个脚本,比如如果是 foo.sh 在当前目录下,就是 ./foo.sh。运行的时候 terminal 会新开一个子 Shell 执行脚本 foo.sh,子 Shell 执行的时候, 父 Shell 还在。子 Shell 执行完毕后返回父 Shell。 子 Shell 从父 Shell 继承环境变量,但是子 Shell 中的环境变量不会带回父 Shell。

exec

exec 与 fork 不同,不需要新开一个子 Shell 来执行被调用的脚本. 被调用的脚本与父脚本在同一个 Shell 内执行。但是使用 exec 调用一个新脚本以后, 父脚本中 exec 行之后的内容就不会再执行了。这是 exec 和 source 的区别.

source

与 fork 的区别是不新开一个子 Shell 来执行被调用的脚本,而是在同一个 Shell 中执行. 所以被调用的脚本中声明的变量和环境变量, 都可以在主脚本中进行获取和使用。

其实从命名上可以感知到其中的细微区别,下面通过两个脚本来体会三种调用方式的不同:

第一个脚本,我们命名为 1.sh:

#!/usr/bin/env bash

A=1

echo "before exec/source/fork: PID for 1.sh = $$"

export A
echo "In 1.sh: variable A=$A"

case $1 in
        --exec)
                echo -e "==> using exec…\n"
                exec ./2.sh ;;
        --source)
                echo -e "==> using source…\n"
                . ./2.sh ;;
        *)
                echo -e "==> using fork by default…\n"
                ./2.sh ;;
esac

echo "after exec/source/fork: PID for 1.sh = $$"
echo -e "In 1.sh: variable A=$A\n"

第二个脚本,我们命名为 2.sh

#!/usr/bin/env bash

echo "PID for 2.sh = $$"
echo "In 2.sh get variable A=$A from 1.sh"

A=2
export A

echo -e "In 2.sh: variable A=$A\n"

注:这两个脚本中的参数 $$ 用于返回脚本的 PID , 也就是进程 ID。这个例子是想通过显示 PID 判断两个脚本是分开执行还是同一进程里执行,也就是是否有新开子 Shell。当执行完脚本 2.sh 后,脚本 1.sh 后面的内容是否还执行。

chmod +x 1.sh 2.sh 给两个脚本加上可执行权限后执行情况:

fork

fork

fork 方式可以看出,两个脚本都执行了,运行顺序为1-2-1,从两者的PID值(1.sh PID=82266, 2.sh PID=82267),可以看出,两个脚本是分成两个进程运行的。

exec

exec

exec 方式运行的结果是,2.sh 执行完成后,不再回到 1.sh。运行顺序为 1-2。从pid值看,两者是在同一进程 PID=82287 中运行的。

source

source

source方式的结果是两者在同一进程里运行。该方式相当于把两个脚本先合并再运行。

Command Explanation
fork 新开一个子 Shell 执行,子 Shell 可以从父 Shell 继承环境变量,但是子 Shell 中的环境变量不会带回给父 Shell。
exec 在同一个 Shell 内执行,但是父脚本中 exec 行之后的内容就不会再执行了
source 在同一个 Shell 中执行,在被调用的脚本中声明的变量和环境变量, 都可以在主脚本中进行获取和使用,相当于合并两个脚本在执行。

参考: 
– 在shell脚本中调用另一个脚本的三种不同方法(fork, exec, source)

 

 

返回值:

https://blog.csdn.net/hongweigg/article/details/78978295

https://blog.csdn.net/hongweigg/article/details/78978295

 

Shell函数返回值,一般有3种方式:return,argv,echo

1) return 语句
shell函数的返回值,可以和其他语言的返回值一样,通过return语句返回。
示例:

?

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

#!/bin/bash -

function mytest()

{

  echo "arg1 = $1"

  if [ $1 = "1" ] ;then

    return 1

  else

    return 0

  fi

}

 

echo

echo "mytest 1"

mytest 1

echo $?     # print return result

 

echo

echo "mytest 0"

mytest 0

echo $?     # print return result

 

echo

echo "mytest 2"

mytest 2

echo $?     # print return result

 

 

echo

echo "mytest 1 = "`mytest 1`

if mytest 1 ; then

  echo "mytest 1"

fi

 

echo

echo "mytest 0 = "`mytest 0`

if mytest 0 ; then

  echo "mytest 0"

fi

 

echo

echo "if fasle" # if 0 is error

if false; then

  echo "mytest 0"

fi

 

 

echo

mytest 1

res=`echo $?`  # get return result

if [ $res = "1" ]; then

  echo "mytest 1"

fi

 

echo

mytest 0

res=`echo $?`  # get return result

if [ $res = "0" ]; then

  echo "mytest 0"

fi

 

 

 

echo

echo "end"

结果:
mytest 1
arg1 = 1
1

mytest 0
arg1 = 0
0

mytest 2
arg1 = 2
0

mytest 1 = arg1 = 1
arg1 = 1

mytest 0 = arg1 = 0
arg1 = 0
mytest 0

if fasle

arg1 = 1
mytest 1

arg1 = 0
mytest 0

end

先定义了一个函数mytest,根据它输入的参数是否为1来return 1或者return 0.
获取函数的返回值通过调用函数,或者最后执行的值获得。
另外,可以直接用函数的返回值用作if的判断。
注意:return只能用来返回整数值,且和c的区别是返回为正确,其他的值为错误。

2) argv全局变量
这种就类似于C语言中的全局变量(或环境变量)。
示例:

?

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

#!/bin/bash -

 

g_var=

function mytest2()

{

  echo "mytest2"

  echo "args $1"

  g_var=$1

 

  return 0

}

 

mytest2 1

echo "return $?"

 

echo

echo "g_var=$g_var"

结果:
mytest2
args 1
return 0

g_var=1

函数mytest2通过修改全局变量的值,来返回结果。

注: 以上两个方法失效的时候
以上介绍的这两种方法在一般情况下都是好使的,但也有例外。
示例:

?

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

#!/bin/bash -

 

 

function mytest3()

{

  grep "123" test.txt | awk -F: '{print $2}' | while read line ;do

    echo "$line"

    if [ $line = "yxb" ]; then

      return # return to pipe only

    fi

  done

 

  echo "mytest3 here "

  return 1      # return to main process

}

 

g_var=

function mytest4()

{

  grep "123" test.txt | awk -F: '{print $2}' | while read line ;do

    echo "$line"

    if [ $line = "yxb" ]; then

      g_var=0

      echo "g_var=0"

      return # return to pipe only

    fi

  done

 

  echo "mytest4 here "

  return 1

}

 

mytest3

echo $?

 

echo

mytest4

echo $?

 

echo

echo "g_var=$g_var"

其中,test.txt 文件中的内容如下:
456:kkk
123:yxb
123:test
结果:
yxb
mytest3 here 
1

yxb
g_var=0
mytest4 here 
1

g_var=
可以看到mytest3在return了以后其实没有直接返回,而是执行了循环体后的语句,同时看到mytest4中也是一样,同时,在mytest4中,对全局变量的修改也无济于事,全局变量的值根本就没有改变。这个是什么原因那?
笔者认为,之所以return语句没有直接返回,是因为return语句是在管道中执行的,管道其实是另一个子进程,而return只是从子进程中返回而已,只是while语句结束了。而函数体之后的语句会继续执行。
同理,全局变量在子进程中进行了修改,但是子进程的修改没有办法反应到父进程中,全局变量只是作为一个环境变量传入子进程,子进程修改自己的环境变量,不会影响到父进程。
因此在写shell函数的时候,用到管道(cmd &后台进程也一样)的时候一定要清楚此刻是从什么地方返回。

3) echo 返回值
其实在shell中,函数的返回值有一个非常安全的返回方式,即通过输出到标准输出返回。因为子进程会继承父进程的标准输出,因此,子进程的输出也就直接反应到父进程。因此不存在上面提到的由于管道导致返回值失效的情况。
在外边只需要获取函数的返回值即可。
示例:

?

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

#!/bin/bash

 

##############################################

# Author : IT-Homer

# Date  : 2012-09-06

# Blog  : http://blog.csdn.net/sunboy_2050

##############################################

 

function mytest5()

{

  grep "123" test.txt | awk -F: '{print $2}' | while read line; do

    if [ $line = "yxb" ]; then

      echo "0"  # value returned first by this function

      return 0

    fi

  done

 

  return 1

}

 

echo '$? = '"$?"

result=$(mytest5)

 

echo "result = $result"

 

echo

if [ -z $result ]    # string is null

then

  echo "no yxb. result is empyt"

else

  echo "have yxb, result is $result"

fi

结果:
$? = 0
result = 0

have yxb, result is 0
这个方式虽然好使,但是有一点一定要注意,不能向标准输出一些不是结果的东西,比如调试信息,这些信息可以重定向到一个文件中解决,特别要注意的是,用到比如grep这样的命令的时候,一定要记得1>/dev/null 2>&1来避免这些命令的输出。

您可能感兴趣的文章: