Converting XML to (like) XPATH. [List Repository](https://github.com/td-shi/ShellScriptsOnGist)

toXPATH.sh

#!/bin/bash --posix
# -*- coding:utf-8 -*-

# === Coding shell scripting Memo ==============================================
# ${<name>#<pattern>} :: matching delete with shortest by forword.
# ${<name>##<pattern>} :: matching delete with longest by forword.
# ${<name>%<pattern>} :: matching delete with shortest by backword.
# ${<name>%%<pattern>} :: mathing delete with longest by backword.
# ${<name>/<before>/<after>} :: replace only first matching.
# ${<name>//<before>/<after>} :: replace all matching.
# ${<name>:-<value>} :: if no yet set value, return value.
# ${<name>:=<value>} :: if no yet set value, return value and set.

# ". <shell script>" is to keep current shell and take over environment.

# === Initialize shell environment =============================================
#set -u                        # Just stop undefined values.
#set -e                        # Just stop error.
#set -x                        # Debug running command.

umask 0022
export LC_ALL=C
export LANG=C
PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/bin:${PATH+:}${PATH-}"
type command >/dev/null 2>&1 && type getconf >/dev/null 2>&1 &&
export PATH="$(command -p getconf PATH):${PATH}"
export UNIX_STD=2003          # to make HP-UX conform to POSIX

# === Define the functions for printing usage and error message ================
usage_and_exit(){
    cat <<-"    USAGE" 1>&2
    # About

    toXPATH.sh convert to (like) XPATH from XML.

    # Usage

    toXPATH.sh [options] [<XML file>]

    + [sample.xml](https://www.w3schools.com/xml/schema_example.asp)

    ```
    $> toXPATH.sh sample.xml
    /shiporder[1]@orderid 889923
    /shiporder[1]@xmlns:xsi http://www.w3.org/2001/XMLSchema-instance
    /shiporder[1]@xsi:noNamespaceSchemaLocation shiporder.xsd
    /shiporder[1]/orderperson[1] John Smith
    /shiporder[1]/shipto[1]/name[1] Ola Nordmann
    /shiporder[1]/shipto[1]/address[1] Langgt 23
    /shiporder[1]/shipto[1]/city[1] 4000 Stavanger
    /shiporder[1]/shipto[1]/country[1] Norway
    /shiporder[1]/item[1]/title[1] Empire Burlesque
    /shiporder[1]/item[1]/note[1] Special Edition
    /shiporder[1]/item[1]/price[1] 10.90
    /shiporder[1]/item[2]/title[1] Hide your heart
    /shiporder[1]/item[2]/price[1] 9.90
    ```

    # Options

    + -h |--help |--version
      - This help.

    # Version

    2020-02-24T20:10:10 0.02

    # LICENSE

    [CC0(Public domain)](https://creativecommons.org/publicdomain/zero/1.0/legalcode)

    # Author

    2020 TD

    USAGE

    exit 1
}
error_exit() {
    ${2+:} false && echo "${0##*/}: $2" 1>&2
    exit "$1"
}
# === Initialize parameters ====================================================
# Detect home directory of this app. and define more
#Homedir="$(d=${0%/*}/; [ "_$d" = "_$0/" ] && d='./'; cd "$d.."; pwd)"
#PATH="$Homedir/<Add Dir>:$PATH" # for additional command
LF=$( printf '\\\012_');LF=${LF%_} # for sed command.
NROOT="NROOT"
#. "$Homedir/<shell script config-file.>"        # configration value.
input=''
# === Confirm that the required commands exist =================================
# --- cURL or Wget (exsample)
#if   type curl    >/dev/null 2>&1; then
#  CMD_CURL='curl'
#elif type wget    >/dev/null 2>&1; then
#  CMD_WGET='wget'
#else
#  error_exit 1 'No HTTP-GET/POST command found.'
#fi
# === Print usage and exit if one of the help options is set ===================
case "$# ${1:-}" in
    '1 -h'|'1 --help'|'1 --version') usage_and_exit;;
esac
# === Read options =============================================================
while :; do
    case "${1:-}" in
        --|-)
            break
            ;;
        --*|-*)
            error_exit 1 'Invalid option'
            ;;
        *)
            break
            ;;
    esac
done
# === Require parameters check =================================================
#printf '%s\n' "${zzz}" | grep -Eq '^$|^-?[0-9.]+,-?[0-9.]+$' || {
#  error_exit 1 'Invalid -l,--location option'
#}
# === Last parameter ===========================================================
case $# in
 0) input=$(cat -)
    ;;
 1) case "${1:-}" in
      '--') usage_and_exit;;
       '-') input=$(cat -)    ;;
         *) input=$(cat "$1"); NROOT="$1"   ;;
    esac
    ;;
 *) case "$1" in '--') shift;; esac
    input=$(cat "$@")
    ;;
esac                         # Escape 0x0A to 0x1E
# === Define funcitons =========================================================
# === Main routine =============================================================
printf "%s" "${input}"                                                        |\
    sed "/<?[^?]*?>/d"                                                        |\
    sed "s/^[[:blank:]]*/ /"                                                  |\
    awk '{printf $0;}'                                                        |\
    sed "s:<\([^ ]*\)\(\( [^ ]*\)\{0,\}\)/>:<\1\2></\1>:g"                    |\
    sed 's/\(<[^>]*>\)/'"${LF}"'\1'"${LF}"'/g'                                |\
    sed "s/ >/>/g"                                                            |\
    sed 's/<!--/'"${LF}"'<!--'"${LF}"'/g'                                     |\
    sed 's/-->/-->'"${LF}"'/g'                                                |\
    sed '/^<!--/,/-->$/d'                                                     |\
    sed 's/ *$/'"${LF}"'/g'                                                   |\
    sed "s/^ *//g"                                                            |\
    sed "/^$/d"                                                               |\
    awk '
function chgPathString( pos){
    for (M = 0; M < N; M++){
        pos = index(StackHead[M], " ");
        tmp = sprintf("/%s", substr(StackHead[M], 1, pos));
        tmp = tmp sprintf("[%d]", substr(StackHead[M], pos + 1));
        gsub(/ /, "", tmp);
        printf("%s", tmp);
    }
    return;
}
function chgOptString(option, term){
    sub(/^[ >]+/, "", option);
    match(option, /^[^ =]+=\"[^\"]*\"/);
    if (-1 != RLENGTH) {
        term = substr(option, RSTART, RLENGTH);
        sub(/^[^ =]+=\"[^\"]*\"/, "", option);
        sub(/=/, " ", term);
        gsub(/\"/, "", term);
        chgPathString();
        printf("@%s\n", term);
        chgOptString(option);
    }
    return;
}
BEGIN{ StackHead[0] = "'"${NROOT}"' 1"; StackBody[0] = ""; N = 1; }
# Part of Text.
/^[^< ].+$/{
    chgPathString();
    printf(" %s\n", $0);
}
# Open Tag.
/^<[^\/].+$/{
    sub(/^</,"");
    sub(/>$/,"");
    stag = $1;
    sub(/^[^ ]+/,"");
    record = $0;
    if (index(StackBody[N-1], stag)) {
        search = stag " [0-9]+";
        match(StackBody[N-1], search);
        pp = substr(StackBody[N-1], RSTART, RLENGTH);
        sub(search, "", StackBody[N-1]);
        match(pp, /[0-9]+/);
        pp = substr(pp, RSTART, RLENGTH);
        pp = pp + 1;
        StackBody[N-1] = StackBody[N-1] " " stag " " pp;
        StackHead[N] = stag " " pp;
        StackBody[N] = "";
    }
    else {
        StackBody[N-1] = StackBody[N-1] " " stag " 1";
        StackHead[N] = stag " 1";
        StackBody[N] = "";
    }
    sub(/^ +/, "", StackBody[N-1])
    N = N + 1;
    chgOptString(record);
}
# Close Tag.
/^<\/.+$/ { StackHead[N] = ""; StackBody[N] = ""; N = N - 1; }
'                                                                             |\
    sed 's:^/'"${NROOT}"'\[[0-9]*\]::g'
# === End shell script =========================================================
exit 0