#!/bin/bash
# -*- mode:shell-script; coding:utf-8; -*-
#
# Created: <Thu Sep 10 15:11:11 2015>
# Last Updated: <2015-September-10 15:11:14>
#
## =======================================================
## json parsing
json_throw () {
  echo "$*" >&2
  exit 1
}

BRIEF=0
LEAFONLY=1
PRUNE=0

json_awk_egrep () {
  local pattern_string=$1

  gawk '{
    while ($0) {
      start=match($0, pattern);
      token=substr($0, start, RLENGTH);
      print token;
      $0=substr($0, start+RLENGTH);
    }
  }' pattern=$pattern_string
}

json_tokenize () {
  local GREP
  local ESCAPE
  local CHAR

  if echo "test string" | egrep -ao --color=never "test" &>/dev/null
  then
    GREP='egrep -ao --color=never'
  else
    GREP='egrep -ao'
  fi

  if echo "test string" | egrep -o "test" &>/dev/null
  then
    ESCAPE='(\\[^u[:cntrl:]]|\\u[0-9a-fA-F]{4})'
    CHAR='[^[:cntrl:]"\\]'
  else
    GREP=json_awk_egrep
    ESCAPE='(\\\\[^u[:cntrl:]]|\\u[0-9a-fA-F]{4})'
    CHAR='[^[:cntrl:]"\\\\]'
  fi

  local STRING="\"$CHAR*($ESCAPE$CHAR*)*\""
  local NUMBER='-?(0|[1-9][0-9]*)([.][0-9]*)?([eE][+-]?[0-9]*)?'
  local KEYWORD='null|false|true'
  local SPACE='[[:space:]]+'

  $GREP "$STRING|$NUMBER|$KEYWORD|$SPACE|." | egrep -v "^$SPACE$"
}

json_parse_array () {
  local index=0
  local ary=''
  read -r token
  case "$token" in
    ']') ;;
    *)
      while :
      do
        json_parse_value "$1" "$index"
        index=$((index+1))
        ary="$ary""$value"
        read -r token
        case "$token" in
          ']') break ;;
          ',') ary="$ary," ;;
          *) json_throw "EXPECTED , or ] GOT ${token:-EOF}" ;;
        esac
        read -r token
      done
      ;;
  esac
  [ "$BRIEF" -eq 0 ] && value=`printf '[%s]' "$ary"` || value=
  :
}

json_parse_object () {
  local key
  local obj=''
  read -r token
  case "$token" in
    '}') ;;
    *)
      while :
      do
        case "$token" in
          '"'*'"') key=$token ;;
          *) json_throw "EXPECTED string GOT ${token:-EOF}" ;;
        esac
        read -r token
        case "$token" in
          ':') ;;
          *) json_throw "EXPECTED : GOT ${token:-EOF}" ;;
        esac
        read -r token
        json_parse_value "$1" "$key"
        obj="$obj$key:$value"
        read -r token
        case "$token" in
          '}') break ;;
          ',') obj="$obj," ;;
          *) json_throw "EXPECTED , or } GOT ${token:-EOF}" ;;
        esac
        read -r token
      done
    ;;
  esac
  [ "$BRIEF" -eq 0 ] && value=`printf '{%s}' "$obj"` || value=
  :
}

json_parse_value () {
  local jpath="${1:+$1,}$2" isleaf=0 isempty=0 print=0
  case "$token" in
    '{') json_parse_object "$jpath" ;;
    '[') json_parse_array  "$jpath" ;;
    # At this point, the only valid single-character tokens are digits.
    ''|[!0-9]) json_throw "EXPECTED value GOT ${token:-EOF}" ;;
    *) value=$token
       isleaf=1
       [ "$value" = '""' ] && isempty=1
       ;;
  esac
  [ "$value" = '' ] && return
  [ "$LEAFONLY" -eq 0 ] && [ "$PRUNE" -eq 0 ] && print=1
  [ "$LEAFONLY" -eq 1 ] && [ "$isleaf" -eq 1 ] && [ $PRUNE -eq 0 ] && print=1
  [ "$LEAFONLY" -eq 0 ] && [ "$PRUNE" -eq 1 ] && [ "$isempty" -eq 0 ] && print=1
  [ "$LEAFONLY" -eq 1 ] && [ "$isleaf" -eq 1 ] && \
    [ $PRUNE -eq 1 ] && [ $isempty -eq 0 ] && print=1
  [ "$print" -eq 1 ] && printf "[%s]\t%s\n" "$jpath" "$value"
  :
}

json_parse () {
  read -r token
  json_parse_value
  read -r token
  case "$token" in
    '') ;;
    *) json_throw "EXPECTED EOF GOT $token" ;;
  esac
}

#usage:
# cat  example.json | json_tokenize | json_parse

## =======================================================