DP-S-Script/Dps_A/BaseClass/JsonClass/JsonClass.nut

702 lines
20 KiB
Plaintext
Raw Permalink Normal View History

2024-09-16 17:05:26 +08:00
/**
* JSON Parser
*
* @author Mikhail Yurasov <mikhail@electricimp.com>
* @package JSONParser
* @version 1.0.1
*/
/**
* JSON Parser
* @package JSONParser
*/
class JSONParser {
// should be the same for all components within JSONParser package
static version = "1.0.1";
/**
* Parse JSON string into data structure
*
* @param {string} str
* @param {function({string} value[, "number"|"string"])|null} converter
* @return {*}
*/
function parse(str, converter = null) {
local state;
local stack = []
local container;
local key;
local value;
// actions for string tokens
local string = {
go = function() {
state = "ok";
},
firstokey = function() {
key = value;
state = "colon";
},
okey = function() {
key = value;
state = "colon";
},
ovalue = function() {
value = this._convert(value, "string", converter);
state = "ocomma";
}.bindenv(this),
firstavalue = function() {
value = this._convert(value, "string", converter);
state = "acomma";
}.bindenv(this),
avalue = function() {
value = this._convert(value, "string", converter);
state = "acomma";
}.bindenv(this)
};
// the actions for number tokens
local number = {
go = function() {
state = "ok";
},
ovalue = function() {
value = this._convert(value, "number", converter);
state = "ocomma";
}.bindenv(this),
firstavalue = function() {
value = this._convert(value, "number", converter);
state = "acomma";
}.bindenv(this),
avalue = function() {
value = this._convert(value, "number", converter);
state = "acomma";
}.bindenv(this)
};
// action table
// describes where the state machine will go from each given state
local action = {
"{": {
go = function() {
stack.push({
state = "ok"
});
container = {};
state = "firstokey";
},
ovalue = function() {
stack.push({
container = container,
state = "ocomma",
key = key
});
container = {};
state = "firstokey";
},
firstavalue = function() {
stack.push({
container = container,
state = "acomma"
});
container = {};
state = "firstokey";
},
avalue = function() {
stack.push({
container = container,
state = "acomma"
});
container = {};
state = "firstokey";
}
},
"}": {
firstokey = function() {
local pop = stack.pop();
value = container;
container = ("container" in pop) ? pop.container : null;
key = ("key" in pop) ? pop.key : null;
state = pop.state;
},
ocomma = function() {
local pop = stack.pop();
container[key] <- value;
value = container;
container = ("container" in pop) ? pop.container : null;
key = ("key" in pop) ? pop.key : null;
state = pop.state;
}
},
"[": {
go = function() {
stack.push({
state = "ok"
});
container = [];
state = "firstavalue";
},
ovalue = function() {
stack.push({
container = container,
state = "ocomma",
key = key
});
container = [];
state = "firstavalue";
},
firstavalue = function() {
stack.push({
container = container,
state = "acomma"
});
container = [];
state = "firstavalue";
},
avalue = function() {
stack.push({
container = container,
state = "acomma"
});
container = [];
state = "firstavalue";
}
},
"]": {
firstavalue = function() {
local pop = stack.pop();
value = container;
container = ("container" in pop) ? pop.container : null;
key = ("key" in pop) ? pop.key : null;
state = pop.state;
},
acomma = function() {
local pop = stack.pop();
container.push(value);
value = container;
container = ("container" in pop) ? pop.container : null;
key = ("key" in pop) ? pop.key : null;
state = pop.state;
}
},
":": {
colon = function() {
// Check if the key already exists
// NOTE previous code used 'if (key in container)...'
// but this finds table ('container') member methods too
local err = false;
foreach(akey, avalue in container) {
if (akey == key) err = true;
break
}
if (err) throw "Duplicate key \"" + key + "\"";
state = "ovalue";
}
},
",": {
ocomma = function() {
container[key] <- value;
state = "okey";
},
acomma = function() {
container.push(value);
state = "avalue";
}
},
"true": {
go = function() {
value = true;
state = "ok";
},
ovalue = function() {
value = true;
state = "ocomma";
},
firstavalue = function() {
value = true;
state = "acomma";
},
avalue = function() {
value = true;
state = "acomma";
}
},
"false": {
go = function() {
value = false;
state = "ok";
},
ovalue = function() {
value = false;
state = "ocomma";
},
firstavalue = function() {
value = false;
state = "acomma";
},
avalue = function() {
value = false;
state = "acomma";
}
},
"null": {
go = function() {
value = null;
state = "ok";
},
ovalue = function() {
value = null;
state = "ocomma";
},
firstavalue = function() {
value = null;
state = "acomma";
},
avalue = function() {
value = null;
state = "acomma";
}
}
};
//
state = "go";
stack = [];
// current tokenizeing position
local start = 0;
try {
local
result,
token,
tokenizer = _JSONTokenizer();
while (token = tokenizer.nextToken(str, start)) {
if ("ptfn" == token.type) {
// punctuation/true/false/null
action[token.value][state]();
} else if ("number" == token.type) {
// number
value = token.value;
number[state]();
} else if ("string" == token.type) {
// string
value = tokenizer.unescape(token.value);
string[state]();
}
start += token.length;
}
} catch (e) {
state = e;
}
// check is the final state is not ok
// or if there is somethign left in the str
if (state != "ok" || regexp("[^\\s]").capture(str, start)) {
local min = @(a, b) a< b ? a : b;
local near = str.slice(start, min(str.len(), start + 10));
throw "JSON Syntax Error near `" + near + "`";
}
return value;
}
/**
* Convert strings/numbers
* Uses custom converter function
*
* @param {string} value
* @param {string} type
* @param {function|null} converter
*/
function _convert(value, type, converter) {
if ("function" == typeof converter) {
// # of params for converter function
local parametercCount = 2;
// .getinfos() is missing on ei platform
if ("getinfos" in converter) {
parametercCount = converter.getinfos().parameters.len() -
1 /* "this" is also included */ ;
}
if (parametercCount == 1) {
return converter(value);
} else if (parametercCount == 2) {
return converter(value, type);
} else {
throw "Error: converter function must take 1 or 2 parameters"
}
} else if ("number" == type) {
return (value.find(".") == null && value.find("e") == null && value.find("E") == null) ? value.tointeger() : value.tofloat();
} else {
return value;
}
}
}
/**
* JSON Tokenizer
* @package JSONParser
*/
class _JSONTokenizer {
_ptfnRegex = null;
_numberRegex = null;
_stringRegex = null;
_ltrimRegex = null;
_unescapeRegex = null;
constructor() {
// punctuation/true/false/null
this._ptfnRegex = regexp("^(?:\\,|\\:|\\[|\\]|\\{|\\}|true|false|null)");
// numbers
this._numberRegex = regexp("^(?:\\-?\\d+(?:\\.\\d*)?(?:[eE][+\\-]?\\d+)?)");
// strings
this._stringRegex = regexp("^(?:\\\"((?:[^\\r\\n\\t\\\\\\\"]|\\\\(?:[\"\\\\\\/trnfb]|u[0-9a-fA-F]{4}))*)\\\")");
// ltrim pattern
this._ltrimRegex = regexp("^[\\s\\t\\n\\r]*");
// string unescaper tokenizer pattern
this._unescapeRegex = regexp("\\\\(?:(?:u\\d{4})|[\\\"\\\\/bfnrt])");
}
/**
* Get next available token
* @param {string} str
* @param {integer} start
* @return {{type,value,length}|null}
*/
function nextToken(str, start = 0) {
local
m,
type,
token,
value,
length,
whitespaces;
// count # of left-side whitespace chars
whitespaces = this._leadingWhitespaces(str, start);
start += whitespaces;
if (m = this._ptfnRegex.capture(str, start)) {
// punctuation/true/false/null
value = str.slice(m[0].begin, m[0].end);
type = "ptfn";
} else if (m = this._numberRegex.capture(str, start)) {
// number
value = str.slice(m[0].begin, m[0].end);
type = "number";
} else if (m = this._stringRegex.capture(str, start)) {
// string
value = str.slice(m[1].begin, m[1].end);
type = "string";
} else {
return null;
}
token = {
type = type,
value = value,
length = m[0].end - m[0].begin + whitespaces
};
return token;
}
/**
* Count # of left-side whitespace chars
* @param {string} str
* @param {integer} start
* @return {integer} number of leading spaces
*/
function _leadingWhitespaces(str, start) {
local r = this._ltrimRegex.capture(str, start);
if (r) {
return r[0].end - r[0].begin;
} else {
return 0;
}
}
// unesacape() replacements table
_unescapeReplacements = {
"b": "\b",
"f": "\f",
"n": "\n",
"r": "\r",
"t": "\t"
};
/**
* Unesacape string escaped per JSON standard
* @param {string} str
* @return {string}
*/
function unescape(str) {
local start = 0;
local res = "";
while (start< str.len()) {
local m = this._unescapeRegex.capture(str, start);
if (m) {
local token = str.slice(m[0].begin, m[0].end);
// append chars before match
local pre = str.slice(start, m[0].begin);
res += pre;
if (token.len() == 6) {
// unicode char in format \uhhhh, where hhhh is hex char code
// todo: convert \uhhhh chars
res += token;
} else {
// escaped char
// @see http://www.json.org/
local char = token.slice(1);
if (char in this._unescapeReplacements) {
res += this._unescapeReplacements[char];
} else {
res += char;
}
}
} else {
// append the rest of the source string
res += str.slice(start);
break;
}
start = m[0].end;
}
return res;
}
}
// Copyright (c) 2017 Electric Imp
// This file is licensed under the MIT License
// http://opensource.org/licenses/MIT
class JSONEncoder {
static VERSION = "2.0.0";
// max structure depth
// anything above probably has a cyclic ref
static _maxDepth = 32;
/**
* Encode value to JSON
* @param {table|array|*} value
* @returns {string}
*/
function encode(value) {
return this._encode(value);
}
/**
* @param {table|array} val
* @param {integer=0} depth current depth level
* @private
*/
function _encode(val, depth = 0) {
// detect cyclic reference
if (depth > this._maxDepth) {
throw "Possible cyclic reference";
}
local
r = "",
s = "",
i = 0;
switch (typeof val) {
case "table":
case "class":
s = "";
// serialize properties, but not functions
foreach(k, v in val) {
if (typeof v != "function") {
s += ",\"" + k + "\":" + this._encode(v, depth + 1);
}
}
s = s.len() > 0 ? s.slice(1) : s;
r += "{" + s + "}";
break;
case "array":
s = "";
for (i = 0; i< val.len(); i++) {
s += "," + this._encode(val[i], depth + 1);
}
s = (i > 0) ? s.slice(1) : s;
r += "[" + s + "]";
break;
case "integer":
case "float":
case "bool":
r += val;
break;
case "null":
r += "null";
break;
case "instance":
if ("_serializeRaw" in val && typeof val._serializeRaw == "function") {
// include value produced by _serializeRaw()
r += val._serializeRaw().tostring();
} else if ("_serialize" in val && typeof val._serialize == "function") {
// serialize instances by calling _serialize method
r += this._encode(val._serialize(), depth + 1);
} else {
s = "";
try {
// iterate through instances which implement _nexti meta-method
foreach(k, v in val) {
s += ",\"" + k + "\":" + this._encode(v, depth + 1);
}
} catch (e) {
// iterate through instances w/o _nexti
// serialize properties, but not functions
foreach(k, v in val.getclass()) {
if (typeof v != "function") {
s += ",\"" + k + "\":" + this._encode(val[k], depth + 1);
}
}
}
s = s.len() > 0 ? s.slice(1) : s;
r += "{" + s + "}";
}
break;
case "blob":
// This is a workaround for a known bug:
// on device side Blob.tostring() returns null
// (instaead of an empty string)
r += "\"" + (val.len() ? this._escape(val.tostring()) : "") + "\"";
break;
// strings and all other
default:
r += "\"" + this._escape(val.tostring()) + "\"";
break;
}
return r;
}
/**
* Escape strings according to http://www.json.org/ spec
* @param {string} str
*/
function _escape(str) {
local res = "";
for (local i = 0; i< str.len(); i++) {
local ch1 = (str[i] & 0xFF);
if ((ch1 & 0x80) == 0x00) {
// 7-bit Ascii
ch1 = format("%c", ch1);
if (ch1 == "\"") {
res += "\\\"";
} else if (ch1 == "\\") {
res += "\\\\";
} else if (ch1 == "/") {
res += "\\/";
} else if (ch1 == "\b") {
res += "\\b";
} else if (ch1 == "\f") {
res += "\\f";
} else if (ch1 == "\n") {
res += "\\n";
} else if (ch1 == "\r") {
res += "\\r";
} else if (ch1 == "\t") {
res += "\\t";
} else if (ch1 == "\0") {
res += "\\u0000";
} else {
res += ch1;
}
} else {
if ((ch1 & 0xE0) == 0xC0) {
// 110xxxxx = 2-byte unicode
local ch2 = (str[++i] & 0xFF);
res += format("%c%c", ch1, ch2);
} else if ((ch1 & 0xF0) == 0xE0) {
// 1110xxxx = 3-byte unicode
local ch2 = (str[++i] & 0xFF);
local ch3 = (str[++i] & 0xFF);
res += format("%c%c%c", ch1, ch2, ch3);
} else if ((ch1 & 0xF8) == 0xF0) {
// 11110xxx = 4 byte unicode
local ch2 = (str[++i] & 0xFF);
local ch3 = (str[++i] & 0xFF);
local ch4 = (str[++i] & 0xFF);
res += format("%c%c%c%c", ch1, ch2, ch3, ch4);
}
}
}
return res;
}
}