Shell的字符串处理

[huruji]

在Shell编程快速入门指南一文中已经简单介绍了字符串的变量命名、截取、获取长度等操作，但通常我们对字符串的操作的需求远远不止这些，Shell本身一起已经内置了一些对字符串的操作。

判断和读取字符串

• ${var-default}和${var=default}：如果var没有被声明，则使用$default为其值

  
  a="linux"
  b="win"

c=${d-$b} e=${a-$b}

echo $c echo $e

输出

win

linux

+ **${var:-default}** 和 **${var:=default}**：如果var没有被声明或者为空，则使用$default为其值

例子使用上例
```shell
c=${d-$b}
e=${a-$b}

echo $c
echo $e

# 输出
# win
# linux

*${var+other} 和 ${var:+other}：如果var声明了，那么其值就是$other，否则就是null字符串（echo打印为一个空行）

a="linux"
b="win"

c=${d+$b}
e=${a+$b}

echo $c
echo $e

# 输出
# 
# win

${var?ERR_MSG} 和 ${var:?ERR_MSG}：如果var没有被声明，就会打印ERR_MSG（也就是说会打印出错误信息，用在调试中会非常有用）

a="linux"
b="win"

c=${d?$b}
e=${a:?$b}

echo $c
echo $e

# 输出
# test.sh:line 4: d: win

*${!varprefix} 和 ${!varprefix@}**：匹配之前所有以varprefix开头进行声明的变量，这是一个数组

javaLang="java"
javascriptLang="javascript"
pythonLang="python"
b="java123"

for ele in ${!java*}
do
    echo $ele
done

# 输出
# javaLang
# javascriptLang

字符串长度获取、替换、截取

${#str}：获取长度

os="linux"

echo ${#os}

# 输出
# 5

${str:position:length}：从position位置开始截取长度为length的子串，其中length可以省略，省略则是截取到最后

lang="javascript"

echo ${lang:4}
echo ${lang:4:3}
echo ${lang}

# 输出
# script
# sci
# javascript

${str#substr}：从str开头删除最短匹配$substr的子串，匹配都是从开头匹配的，开头匹配不上则不删除

lang="javascript is good"

echo ${lang#java*}
echo ${lang#java}

echo ${lang#[^b-c]}
echo ${lang#[^j-z]}

# 输出
# script is good
# script is good
# avascript is good
# javascript is good

${str#substr}：从str开头删除最长匹配$substr的子串，匹配都是从开头匹配的，开头匹配不上则不删除

lang="javascript is good, good study"

echo ${lang##*,}
echo ${lang##java}

# 输出
# good study
# script is good, good study

${str%substr}：从str结尾删除最短匹配$substr的子串

lang="javascript is good, good,study"
good="javascript is good, good study"

echo ${lang%,*}
echo ${good%,*}

# 输出
# javascript is good, good
# javascript is good

${str%%substr}：从str结尾删除最长匹配$substr的子串

lang="javascript is good, good,study"
good="javascript is good, good study"

echo ${lang%%,*}
echo ${good%%,*}

# 输出
# javascript is good
# javascript is good

${str/substr/replacement}：使用$replacement来替代第一个匹配的substr子串

lang="java is good, javascript is good"
lang2="python"
bast="the best"

echo ${lang/java/$lang2}
echo ${lang/good/$bast}

# 输出
# python is good, javascript is good
# java is the best, javascript is good

${str//substr/replacement}：使用$replacement来替代所有匹配的substr子串

lang="java is good, javascript is good"
lang2="python"
bast="the best"

echo ${lang//java/$lang2}
echo ${lang//good/$bast}

# 输出
# python is good, pythonscript is good
# java is the best, javascript is the best

${str/#substr/replacement}：如果$str的前缀匹配$substr则使用$replacement替代

lang="java is good, javascript is good"
lang2="python"

echo ${lang/#java/$lang2}

# 输出
# python is good, javascript is good

${str/%substr/replacement}：如果$str的后缀缀匹配$substr则使用$replacement替代

lang="java is good, javascript is good"
bast="the bast"

echo ${lang/%good/$bast}

# 输出
# java is good, javascript is the best

使用实例

列出当前文件夹下所有文件含有的后缀名

function hasItem() {
    arr=$1;
    item=$2;
    for ele in ${arr[@]}
    do
        if [[ $item = $ele ]]; then
            return 0
        fi
    done
    return 1
}
arr=()

for file in `ls`
do
    ext=${file#*.}
    echo ${arr[@]}
    hasItem $arr $ext
    if [ $? -eq 1 ]; then
        length=${#arr[*]}
        if [ $length -eq 0 ]; then
            arr=($ext)
        else
            arr=("${arr[@]}" "${ext}")
        fi
    else
        arr=$arr
    fi
done

for name in ${arr[@]}
do
    echo $name
done