| 1 | #!/bin/sh
|
| 2 |
|
| 3 | throw() {
|
| 4 | echo "$*" >&2
|
| 5 | exit 1
|
| 6 | }
|
| 7 |
|
| 8 | BRIEF=0
|
| 9 | LEAFONLY=0
|
| 10 | PRUNE=0
|
| 11 | NO_HEAD=0
|
| 12 | NORMALIZE_SOLIDUS=0
|
| 13 |
|
| 14 | usage() {
|
| 15 | echo
|
| 16 | echo "Usage: JSON.sh [-b] [-l] [-p] [-s] [-h]"
|
| 17 | echo
|
| 18 | echo "-p - Prune empty. Exclude fields with empty values."
|
| 19 | echo "-l - Leaf only. Only show leaf nodes, which stops data duplication."
|
| 20 | echo "-b - Brief. Combines 'Leaf only' and 'Prune empty' options."
|
| 21 | echo "-n - No-head. Do not show nodes that have no path (lines that start with [])."
|
| 22 | echo "-s - Remove escaping of the solidus symbol (straight slash)."
|
| 23 | echo "-h - This help text."
|
| 24 | echo
|
| 25 | }
|
| 26 |
|
| 27 | parse_options() {
|
| 28 | set -- "$@"
|
| 29 | local ARGN=$#
|
| 30 | while [ "$ARGN" -ne 0 ]
|
| 31 | do
|
| 32 | case $1 in
|
| 33 | -h) usage
|
| 34 | exit 0
|
| 35 | ;;
|
| 36 | -b) BRIEF=1
|
| 37 | LEAFONLY=1
|
| 38 | PRUNE=1
|
| 39 | ;;
|
| 40 | -l) LEAFONLY=1
|
| 41 | ;;
|
| 42 | -p) PRUNE=1
|
| 43 | ;;
|
| 44 | -n) NO_HEAD=1
|
| 45 | ;;
|
| 46 | -s) NORMALIZE_SOLIDUS=1
|
| 47 | ;;
|
| 48 | ?*) echo "ERROR: Unknown option."
|
| 49 | usage
|
| 50 | exit 0
|
| 51 | ;;
|
| 52 | esac
|
| 53 | shift 1
|
| 54 | ARGN=$((ARGN-1))
|
| 55 | done
|
| 56 | }
|
| 57 |
|
| 58 | awk_egrep () {
|
| 59 | local pattern_string=$1
|
| 60 |
|
| 61 | gawk '{
|
| 62 | while ($0) {
|
| 63 | start=match($0, pattern);
|
| 64 | token=substr($0, start, RLENGTH);
|
| 65 | print token;
|
| 66 | $0=substr($0, start+RLENGTH);
|
| 67 | }
|
| 68 | }' pattern="$pattern_string"
|
| 69 | }
|
| 70 |
|
| 71 | tokenize () {
|
| 72 | local GREP
|
| 73 | local ESCAPE
|
| 74 | local CHAR
|
| 75 |
|
| 76 | if echo "test string" | egrep -ao --color=never "test" >/dev/null 2>&1
|
| 77 | then
|
| 78 | GREP='egrep -ao --color=never'
|
| 79 | else
|
| 80 | GREP='egrep -ao'
|
| 81 | fi
|
| 82 |
|
| 83 | if echo "test string" | egrep -o "test" >/dev/null 2>&1
|
| 84 | then
|
| 85 | ESCAPE='(\\[^u[:cntrl:]]|\\u[0-9a-fA-F]{4})'
|
| 86 | CHAR='[^[:cntrl:]"\\]'
|
| 87 | else
|
| 88 | GREP=awk_egrep
|
| 89 | ESCAPE='(\\\\[^u[:cntrl:]]|\\u[0-9a-fA-F]{4})'
|
| 90 | CHAR='[^[:cntrl:]"\\\\]'
|
| 91 | fi
|
| 92 |
|
| 93 | local STRING="\"$CHAR*($ESCAPE$CHAR*)*\""
|
| 94 | local NUMBER='-?(0|[1-9][0-9]*)([.][0-9]*)?([eE][+-]?[0-9]*)?'
|
| 95 | local KEYWORD='null|false|true'
|
| 96 | local SPACE='[[:space:]]+'
|
| 97 |
|
| 98 | # Force zsh to expand $A into multiple words
|
| 99 | local is_wordsplit_disabled=$(unsetopt 2>/dev/null | grep -c '^shwordsplit$')
|
| 100 | if [ $is_wordsplit_disabled != 0 ]; then setopt shwordsplit; fi
|
| 101 | $GREP "$STRING|$NUMBER|$KEYWORD|$SPACE|." | egrep -v "^$SPACE$"
|
| 102 | if [ $is_wordsplit_disabled != 0 ]; then unsetopt shwordsplit; fi
|
| 103 | }
|
| 104 |
|
| 105 | parse_array () {
|
| 106 | local index=0
|
| 107 | local ary=''
|
| 108 | read -r token
|
| 109 | case "$token" in
|
| 110 | ']') ;;
|
| 111 | *)
|
| 112 | while :
|
| 113 | do
|
| 114 | parse_value "$1" "$index"
|
| 115 | index=$((index+1))
|
| 116 | ary="$ary""$value"
|
| 117 | read -r token
|
| 118 | case "$token" in
|
| 119 | ']') break ;;
|
| 120 | ',') ary="$ary," ;;
|
| 121 | *) throw "EXPECTED , or ] GOT ${token:-EOF}" ;;
|
| 122 | esac
|
| 123 | read -r token
|
| 124 | done
|
| 125 | ;;
|
| 126 | esac
|
| 127 | [ "$BRIEF" -eq 0 ] && value=$(printf '[%s]' "$ary") || value=
|
| 128 | :
|
| 129 | }
|
| 130 |
|
| 131 | parse_object () {
|
| 132 | local key
|
| 133 | local obj=''
|
| 134 | read -r token
|
| 135 | case "$token" in
|
| 136 | '}') ;;
|
| 137 | *)
|
| 138 | while :
|
| 139 | do
|
| 140 | case "$token" in
|
| 141 | '"'*'"') key=$token ;;
|
| 142 | *) throw "EXPECTED string GOT ${token:-EOF}" ;;
|
| 143 | esac
|
| 144 | read -r token
|
| 145 | case "$token" in
|
| 146 | ':') ;;
|
| 147 | *) throw "EXPECTED : GOT ${token:-EOF}" ;;
|
| 148 | esac
|
| 149 | read -r token
|
| 150 | parse_value "$1" "$key"
|
| 151 | obj="$obj$key:$value"
|
| 152 | read -r token
|
| 153 | case "$token" in
|
| 154 | '}') break ;;
|
| 155 | ',') obj="$obj," ;;
|
| 156 | *) throw "EXPECTED , or } GOT ${token:-EOF}" ;;
|
| 157 | esac
|
| 158 | read -r token
|
| 159 | done
|
| 160 | ;;
|
| 161 | esac
|
| 162 | [ "$BRIEF" -eq 0 ] && value=$(printf '{%s}' "$obj") || value=
|
| 163 | :
|
| 164 | }
|
| 165 |
|
| 166 | parse_value () {
|
| 167 | local jpath="${1:+$1,}$2" isleaf=0 isempty=0 print=0
|
| 168 | case "$token" in
|
| 169 | '{') parse_object "$jpath" ;;
|
| 170 | '[') parse_array "$jpath" ;;
|
| 171 | # At this point, the only valid single-character tokens are digits.
|
| 172 | ''|[!0-9]) throw "EXPECTED value GOT ${token:-EOF}" ;;
|
| 173 | *) value=$token
|
| 174 | # if asked, replace solidus ("\/") in json strings with normalized value: "/"
|
| 175 | [ "$NORMALIZE_SOLIDUS" -eq 1 ] && value=$(echo "$value" | sed 's#\\/#/#g')
|
| 176 | isleaf=1
|
| 177 | [ "$value" = '""' ] && isempty=1
|
| 178 | ;;
|
| 179 | esac
|
| 180 | [ "$value" = '' ] && return
|
| 181 | [ "$NO_HEAD" -eq 1 ] && [ -z "$jpath" ] && return
|
| 182 |
|
| 183 | [ "$LEAFONLY" -eq 0 ] && [ "$PRUNE" -eq 0 ] && print=1
|
| 184 | [ "$LEAFONLY" -eq 1 ] && [ "$isleaf" -eq 1 ] && [ $PRUNE -eq 0 ] && print=1
|
| 185 | [ "$LEAFONLY" -eq 0 ] && [ "$PRUNE" -eq 1 ] && [ "$isempty" -eq 0 ] && print=1
|
| 186 | [ "$LEAFONLY" -eq 1 ] && [ "$isleaf" -eq 1 ] && \
|
| 187 | [ $PRUNE -eq 1 ] && [ $isempty -eq 0 ] && print=1
|
| 188 | [ "$print" -eq 1 ] && printf "[%s]\t%s\n" "$jpath" "$value"
|
| 189 | :
|
| 190 | }
|
| 191 |
|
| 192 | parse () {
|
| 193 | read -r token
|
| 194 | parse_value
|
| 195 | read -r token
|
| 196 | case "$token" in
|
| 197 | '') ;;
|
| 198 | *) throw "EXPECTED EOF GOT $token" ;;
|
| 199 | esac
|
| 200 | }
|
| 201 |
|
| 202 | if ([ "$0" = "$BASH_SOURCE" ] || ! [ -n "$BASH_SOURCE" ]);
|
| 203 | then
|
| 204 | parse_options "$@"
|
| 205 | tokenize | parse
|
| 206 | fi
|
| 207 |
|
| 208 | # vi: expandtab sw=2 ts=2
|