在所有情况下获取脚本或源脚本的“真实路径”:

fullname=$(readlink $0)  # Take care of  symbolic links
dirname=${fullname%/*}       # Get (most of the time) the dirname
realpath=$(dirname $BASH_SOURCE) # TO handle sourced scripts
[ "$realpath" = '.' ] && realpath=${dirname:-.}

下面是要生成的bash脚本(在新创建的“workdir”子目录中，在当前目录的“mytest”中)，该bash脚本将派生另一个脚本，该脚本将调用bash定义的函数....测试了许多方法来启动它们:

#!/bin/bash
##############################################################

ret=0

fullname=$(readlink $0)  # Take care of  symbolic links
dirname=${fullname%/*}       # Get (most of the time) the dirname
realpath=$(dirname $BASH_SOURCE) # TO handle sourced scripts
[ "$realpath" = '.' ] && realpath=${dirname:-.}

fullname_withoutextension=${fullname%.*}

mkdir -p workdir
cat <<'EOD' > workdir/_script_.sh
#!/bin/bash
##############################################################

ret=0

fullname=$(readlink $0)  # Take care of  symbolic links
dirname=${fullname%/*}       # Get (most of the time) the dirname
realpath=$(dirname $BASH_SOURCE) # TO handle sourced scripts
[ "$realpath" = '.' ] && realpath=${dirname:-.}

fullname_withoutextension=${fullname%.*}

echo
echo "# ------------- RESULTS ------------- #"
echo "# path to me (\$0)----------->  ${0}     "
echo "# arguments called with ---->  ${@}     "
echo "# \$1 ----------------------->  $1       "
echo "# \$2 ----------------------->  $2       "
echo "# path to me (\$fullname)---->  ${fullname} "
echo "# parent path(\${0%/*})------>  ${0%/*}  "
echo "# parent path(\$dirname)----->  ${dirname} "
echo "# my name ----\${0##*/}------>  ${0##*/} "
echo "# my source -\${BASH_SOURCE}->  ${BASH_SOURCE} "
echo "# parent path(from BASH_SOURCE) -> $(dirname $BASH_SOURCE)"
echo "# my function name -\${FUNCNAME[0]}------>  ${FUNCNAME[0]}"
echo "# my source or script real path (realpath)------------------>  $realpath"
echo
[ "$realpath" = "workdir" ] || ret=1
[ $ret = 0 ] || echo "*******************************************************"
[ $ret = 0 ] || echo "***********   ERROR  **********************************"
[ $ret = 0 ] || echo "*******************************************************"

show_params () {
        echo
        echo "# --- RESULTS FROM show_params() ---- #"
        echo "# path to me (\$0)----------->  ${0}     "
        echo "# arguments called with ---->  ${@}     "
        echo "# \$1 ----------------------->  $1       "
        echo "# \$2 ----------------------->  $2       "
        echo "# path to me (\$fullname)---->  ${fullname} "
        echo "# parent path(\${0%/*})------>  ${0%/*}  "
        echo "# parent path(\$dirname)----->  ${dirname} "
        echo "# my name ----\${0##*/}------>  ${0##*/} "
        echo "# my source -\${BASH_SOURCE}->  ${BASH_SOURCE} "
        echo "# parent path(from BASH_SOURCE) -> $(dirname $BASH_SOURCE)"
        echo "# my function name -\${FUNCNAME[0]}------>  ${FUNCNAME[0]}"
        echo "# my source or script real path (realpath)------------------>  $realpath"
        echo
        [ "$realpath" = "workdir" ] || ret=1
        [ $ret = 0 ] || echo "*******************************************************"
        [ $ret = 0 ] || echo "***********   ERROR  **********************************"
        [ $ret = 0 ] || echo "*******************************************************"

}
show_params "$@"

EOD

cat workdir/_script_.sh > workdir/_side_by_side_script_sourced.inc

cat <<'EOD' >> workdir/_script_.sh

echo "# . $realpath/_side_by_side_script_sourced.inc 'hello there' 'william'"
. $realpath/_side_by_side_script_sourced.inc 'hello there' 'william'

[ $ret = 0 ] || echo "*******************************************************"
[ $ret = 0 ] || echo "***********   ERROR  **********************************"
[ $ret = 0 ] || echo "*******************************************************"
EOD

chmod +x  workdir/_script_.sh
[ -L _mytest_ ] && rm _mytest_
ln -s workdir/_script_.sh _mytest_

# ------------- CALLED ------------- #

called_by () {
        echo '=========================================================================='
        echo " Called by : " "$@"
        echo '=========================================================================='
        eval "$@"
}

called_by bash _mytest_
called_by ./_mytest_

called_by bash workdir/_script_.sh
called_by workdir/_script_.sh
called_by . workdir/_script_.sh


# ------------- RESULTS ------------- #

echo
echo
[ $ret = 0 ] || echo "*******************************************************"
[ $ret = 0 ] || echo "***********   ERROR  **********************************"
[ $ret = 0 ] || echo "*******************************************************"
echo
[ $ret = 0 ] && echo ".... location of scripts (\$realpath) should always be equal to $realpath, for all test cases at date".
echo

# ------------- END ------------- #

这些答案对于它们所陈述的情况是正确的，但如果您使用'source'关键字从另一个脚本运行脚本(以便它在同一个shell中运行)，仍然存在一个问题。在本例中，您将获得调用脚本的$0。在这种情况下，我认为不可能获得脚本本身的名称。

这是一个边缘情况，不应该太当真。如果你直接从另一个脚本运行脚本(没有'source')，使用$0可以工作。

me=`basename "$0"`

对于读取symlink1，这通常不是你想要的(你通常不想以这种方式迷惑用户)，尝试:

me="$(basename "$(test -L "$0" && readlink "$0" || echo "$0")")"

在我看来，这会产生令人困惑的输出。“我跑了foo.sh，但它说我跑了bar.sh!?”一定是虫子!”此外，使用不同名称的符号链接的目的之一是根据其名称提供不同的功能(例如某些平台上的gzip和gunzip)。

1也就是说，为了解析符号链接，当用户执行foo.sh时，它实际上是一个到bar.sh的符号链接，你希望使用解析的名称bar.sh而不是foo.sh。

this="$(dirname "$(realpath "$BASH_SOURCE")")"

这将解析符号链接(realpath会这样做)，处理空格(双引号会这样做)，并将找到当前脚本名称，即使是在来源(. ./myscript)或被其他脚本调用时($BASH_SOURCE会这样做)。在所有这些之后，最好将其保存在环境变量中以便重用或方便复制到其他地方(this=)…

$BASH_SOURCE在获取脚本时给出正确答案。

然而，这包括路径，所以只获得脚本文件名，使用:

$(basename $BASH_SOURCE) 

简短，清晰，简单，在my_script.sh中

#!/bin/bash

running_file_name=$(basename "$0")

echo "You are running '$running_file_name' file."

把:

./my_script.sh
You are running 'my_script.sh' file.