Json5Decoder::getEscapee() - Code Metrics - Inspection of "Simply Unicode unescaping by using json_decode() i..." - colinodell/json5 - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( bd17ed...fa6e51 )

by Colin

created 2017-11-11 15:44 UTC

Json5Decoder::getEscapee() B

↳ Parent: Json5Decoder

Complexity

Conditions	11
Paths	11

Size

Total Lines	18
Code Lines	13

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	13
CRAP Score	11

Importance

Changes	2
Bugs	0	Features	0

Metric	Value
cc	11
eloc	13
c	2
b	0
f	0
nc	11
nop	1
dl	0
loc	18
ccs	13
cts	13
cp	1
crap	11
rs	7.1162

How to fix Complexity

<?php

/*
 * This file is part of the colinodell/json5 package.
 *
 * (c) Colin O'Dell <[email protected]>
 *
 * Based on the official JSON5 implementation for JavaScript (https://github.com/json5/json5)
 *  - (c) 2012-2016 Aseem Kishore and others (https://github.com/json5/json5/contributors)
 *
 * For the full copyright and license information, please view the LICENSE
 * file that was distributed with this source code.
 */

namespace ColinODell\Json5;

final class Json5Decoder
{
    const REGEX_WHITESPACE = '/[ \t\r\n\v\f\xA0\x{FEFF}]/u';

    private $json;

    private $at = 0;

    private $lineNumber = 1;

    private $columnNumber = 1;

    private $ch;

    private $associative = false;

    private $maxDepth = 512;

    private $castBigIntToString = false;

    private $depth = 1;

    private $length;

    private $lineCache;

    /**
     * Private constructor.
     *
     * @param string $json
     * @param bool   $associative
     * @param int    $depth
     * @param bool   $castBigIntToString
     */
    private function __construct($json, $associative = false, $depth = 512, $castBigIntToString = false)
    {
        $this->json = $json;
        $this->associative = $associative;
        $this->maxDepth = $depth;
        $this->castBigIntToString = $castBigIntToString;

        $this->length = mb_strlen($json, 'utf-8');

        $this->ch = $this->charAt(0);
    }

    /**
     * Takes a JSON encoded string and converts it into a PHP variable.
     *
     * The parameters exactly match PHP's json_decode() function - see
     * http://php.net/manual/en/function.json-decode.php for more information.
     *
     * @param string $source      The JSON string being decoded.
     * @param bool   $associative When TRUE, returned objects will be converted into associative arrays.
     * @param int    $depth       User specified recursion depth.
     * @param int    $options     Bitmask of JSON decode options.
     *
     * @return mixed
     */
    public static function decode($source, $associative = false, $depth = 512, $options = 0)
    {
        $associative = $associative || ($options & JSON_OBJECT_AS_ARRAY);
        $castBigIntToString = $options & JSON_BIGINT_AS_STRING;

        $decoder = new self((string)$source, $associative, $depth, $castBigIntToString);
function acceptsInteger($int) { }

$x = '123'; // string "123"

// Instead of
acceptsInteger($x);

// we recommend to use
acceptsInteger((integer) $x);

        $result = $decoder->value();
        $decoder->white();
        if ($decoder->ch) {
            $decoder->throwSyntaxError('Syntax error');
        }

        return $result;
    }

    /**
     * @param int $at
     *
     * @return string|null
     */
    private function charAt($at)
    {
        if ($at < 0 || $at >= $this->length) {
            return null;
        }

        return mb_substr($this->json, $at, 1, 'utf-8');
    }

    /**
     * Parse the next character.
     *
     * If $c is given, the next char will only be parsed if the current
     * one matches $c.
     *
     * @param string|null $c
     *
     * @return null|string
     */
    private function next($c = null)
    {
        // If a c parameter is provided, verify that it matches the current character.
        if ($c !== null && $c !== $this->ch) {
            $this->throwSyntaxError(sprintf(
                'Expected %s instead of %s',
                self::renderChar($c),
                self::renderChar($this->ch)
            ));
        }

        // Get the next character. When there are no more characters,
        // return the empty string.
        if ($this->ch === "\n" || ($this->ch === "\r" && $this->peek() !== "\n")) {
            $this->at++;
            $this->lineNumber++;
            $this->columnNumber = 1;
        } else {
            $this->at++;
            $this->columnNumber++;
        }

        $this->ch = $this->charAt($this->at);

        return $this->ch;
    }

    /**
     * Get the next character without consuming it or
     * assigning it to the ch variable.
     *
     * @return mixed
     */
    private function peek()
    {
        return $this->charAt($this->at + 1);
    }

    /**
     * @return string
     */
    private function getLineRemainder()
    {
        // Line are separated by "\n" or "\r" without an "\n" next
        if ($this->lineCache === null) {
            $this->lineCache = preg_split('/\n|\r\n?/u', $this->json);
        }

        $line = $this->lineCache[$this->lineNumber - 1];

        return mb_substr($line, $this->columnNumber - 1);
    }

    /**
     * Attempt to match a regular expression at the current position on the current line.
     *
     * This function will not match across multiple lines.
     *
     * @param string $regex
     *
     * @return string|null
     */
    private function match($regex)
    {
        $subject = $this->getLineRemainder();

        $matches = [];
        if (!preg_match($regex, $subject, $matches, PREG_OFFSET_CAPTURE)) {
            return null;
        }

        // PREG_OFFSET_CAPTURE always returns the byte offset, not the char offset, which is annoying
        $offset = mb_strlen(mb_strcut($subject, 0, $matches[0][1], 'utf-8'), 'utf-8');

        // [0][0] contains the matched text

        // [0][1] contains the index of that match
        $advanceBy = $offset + mb_strlen($matches[0][0], 'utf-8');

        $this->at += $advanceBy;
        $this->columnNumber += $advanceBy;
        $this->ch = $this->charAt($this->at);

        return $matches[0][0];
    }

    /**
     * Parse an identifier.
     *
     * Normally, reserved words are disallowed here, but we
     * only use this for unquoted object keys, where reserved words are allowed,
     * so we don't check for those here. References:
     * - http://es5.github.com/#x7.6
     * - https://developer.mozilla.org/en/Core_JavaScript_1.5_Guide/Core_Language_Features#Variables
     * - http://docstore.mik.ua/orelly/webprog/jscript/ch02_07.htm
     */
    private function identifier()
    {
        // @codingStandardsIgnoreStart
        // Be careful when editing this regex, there are a couple Unicode characters in between here -------------vv
        $match = $this->match('/^(?:[\$_\p{L}\p{Nl}]|\\\\u[0-9A-Fa-f]{4})(?:[\$_\p{L}\p{Nl}\p{Mn}\p{Mc}\p{Nd}\p{Pc}‌‍]|\\\\u[0-9A-Fa-f]{4})*/u');
        // @codingStandardsIgnoreEnd

        if ($match === null) {
            $this->throwSyntaxError('Bad identifier as unquoted key');
        }

        // Un-escape escaped Unicode chars
        $unescaped = preg_replace_callback('/(?:\\\\u[0-9A-Fa-f]{4})+/', function ($m) {
            return json_decode('"'.$m[0].'"');
        }, $match);

        return $unescaped;
    }

    private function number()
    {
        $number = null;
        $sign = '';
        $string = '';
        $base = 10;

        if ($this->ch === '-' || $this->ch === '+') {
            $sign = $this->ch;
            $this->next($this->ch);
        }

        // support for Infinity
        if ($this->ch === 'I') {
            $number = $this->word();
            if ($number === null) {
                $this->throwSyntaxError('Unexpected word for number');
            }

            return ($sign === '-') ? -INF : INF;
        }

        // support for NaN
        if ($this->ch === 'N') {
            $number = $this->word();
            if ($number !== NAN) {
                $this->throwSyntaxError('expected word to be NaN');
            }

            // ignore sign as -NaN also is NaN
            return $number;
        }

        if ($this->ch === '0') {
            $string .= $this->ch;
            $this->next();
            if ($this->ch === 'x' || $this->ch === 'X') {
                $string .= $this->ch;
                $this->next();
                $base = 16;
            } elseif (is_numeric($this->ch)) {
                $this->throwSyntaxError('Octal literal');
            }
        }

        switch ($base) {
            case 10:
                if (($match = $this->match('/^\d*\.?\d*/')) !== null) {
                    $string .= $match;
                }
                if (($match = $this->match('/^[Ee][-+]?\d*/')) !== null) {
                    $string .= $match;
                }
                $number = $string;
                break;
            case 16:
                if (($match = $this->match('/^[A-Fa-f0-9]+/')) !== null) {
                    $string .= $match;
                    $number = hexdec($string);
                    break;
                }
                $this->throwSyntaxError('Bad hex number');
        }

        if ($sign === '-') {
            $number = -$number;
        }

        if (!is_numeric($number) || !is_finite($number)) {
            $this->throwSyntaxError('Bad number');
        }

        if ($this->castBigIntToString) {
            return $number;
        }

        // Adding 0 will automatically cast this to an int or float
        return $number + 0;
    }

    private function string()
    {
        if (!($this->ch === '"' || $this->ch === "'")) {
            $this->throwSyntaxError('Bad string');
        }

        $string = '';

        $delim = $this->ch;
        $this->next();
        while ($this->ch !== null) {
            if ($this->ch === $delim) {
                $this->next();

                return $string;
            } elseif ($this->ch === '\\') {
                if ($unicodeEscaped = $this->match('/^(?:\\\\u[A-Fa-f0-9]{4})+/')) {
                    $string .= json_decode('"'.$unicodeEscaped.'"');
                    continue;
                }

                $this->next();
                if ($this->ch === "\r") {
                    if ($this->peek() === "\n") {
                        $this->next();
                    }
                } elseif (($escapee = self::getEscapee($this->ch)) !== null) {
                    $string .= $escapee;
                } else {
                    break;
                }
            } elseif ($this->ch === "\n") {
                // unescaped newlines are invalid; see:
                // https://github.com/json5/json5/issues/24
                // @todo this feels special-cased; are there other invalid unescaped chars?
                break;
            } else {
                $string .= $this->ch;
            }

            $this->next();
        }

        $this->throwSyntaxError('Bad string');
    }

    /**
     * Skip an inline comment, assuming this is one.
     *
     * The current character should be the second / character in the // pair that begins this inline comment.
     * To finish the inline comment, we look for a newline or the end of the text.
     */
    private function inlineComment()
    {
        do {
            $this->next();
            if ($this->ch === "\n" || $this->ch === "\r") {
                $this->next();

                return;
            }
        } while ($this->ch !== null);
    }

    /**
     * Skip a block comment, assuming this is one.
     *
     * The current character should be the * character in the /* pair that begins this block comment.
     * To finish the block comment, we look for an ending *​/ pair of characters,
     * but we also watch for the end of text before the comment is terminated.
     */
    private function blockComment()
    {
        do {
            $this->next();
            while ($this->ch === '*') {
                $this->next('*');
                if ($this->ch === '/') {
                    $this->next('/');

                    return;
                }
            }
        } while ($this->ch !== null);

        $this->throwSyntaxError('Unterminated block comment');
    }

    /**
     * Skip a comment, whether inline or block-level, assuming this is one.
     */
    private function comment()
    {
        // Comments always begin with a / character.
        if ($this->ch !== '/') {
            $this->throwSyntaxError('Not a comment');
        }

        $this->next('/');

        if ($this->ch === '/') {
            $this->inlineComment();
        } elseif ($this->ch === '*') {
            $this->blockComment();
        } else {
            $this->throwSyntaxError('Unrecognized comment');
        }
    }

    /**
     * Skip whitespace and comments.
     *
     * Note that we're detecting comments by only a single / character.
     * This works since regular expressions are not valid JSON(5), but this will
     * break if there are other valid values that begin with a / character!
     */
    private function white()
    {
        while ($this->ch !== null) {
            if ($this->ch === '/') {
                $this->comment();
            } elseif (preg_match(self::REGEX_WHITESPACE, $this->ch) === 1) {
                $this->next();
            } else {
                return;
            }
        }
    }

    /**
     * Matches true, false, null, etc
     */
    private function word()
    {
        switch ($this->ch) {
            case 't':
                $this->next('t');
                $this->next('r');
                $this->next('u');
                $this->next('e');
                return true;
            case 'f':
                $this->next('f');
                $this->next('a');
                $this->next('l');
                $this->next('s');
                $this->next('e');
                return false;
            case 'n':
                $this->next('n');
                $this->next('u');
                $this->next('l');
                $this->next('l');
                return null;
            case 'I':
                $this->next('I');
                $this->next('n');
                $this->next('f');
                $this->next('i');
                $this->next('n');
                $this->next('i');
                $this->next('t');
                $this->next('y');
                return INF;
            case 'N':
                $this->next('N');
                $this->next('a');
                $this->next('N');
                return NAN;
        }

        $this->throwSyntaxError('Unexpected ' . self::renderChar($this->ch));
    }

    private function arr()
    {
        $arr = [];

        if ($this->ch === '[') {
            if (++$this->depth > $this->maxDepth) {
                $this->throwSyntaxError('Maximum stack depth exceeded');
            }

            $this->next('[');
            $this->white();
            while ($this->ch !== null) {
                if ($this->ch === ']') {
                    $this->next(']');
                    $this->depth--;
                    return $arr; // Potentially empty array
                }
                // ES5 allows omitting elements in arrays, e.g. [,] and
                // [,null]. We don't allow this in JSON5.
                if ($this->ch === ',') {
                    $this->throwSyntaxError('Missing array element');
                } else {
                    $arr[] = $this->value();
                }
                $this->white();
                // If there's no comma after this value, this needs to
                // be the end of the array.
                if ($this->ch !== ',') {
                    $this->next(']');
                    $this->depth--;
                    return $arr;
                }
                $this->next(',');
                $this->white();
            }
        }

        $this->throwSyntaxError('Bad array');
    }

    /**
     * Parse an object value
     */
    private function obj()
    {
        $object = $this->associative ? [] : new \stdClass;

        if ($this->ch === '{') {
            if (++$this->depth > $this->maxDepth) {
                $this->throwSyntaxError('Maximum stack depth exceeded');
            }

            $this->next('{');
            $this->white();
            while ($this->ch) {
                if ($this->ch === '}') {
                    $this->next('}');
                    $this->depth--;
                    return $object; // Potentially empty object
                }

                // Keys can be unquoted. If they are, they need to be
                // valid JS identifiers.
                if ($this->ch === '"' || $this->ch === "'") {
                    $key = $this->string();
                } else {
                    $key = $this->identifier();
                }

                $this->white();
                $this->next(':');
                if ($this->associative) {
                    $object[$key] = $this->value();
                } else {
                    $object->{$key} = $this->value();
                }
                $this->white();
                // If there's no comma after this pair, this needs to be
                // the end of the object.
                if ($this->ch !== ',') {
                    $this->next('}');
                    $this->depth--;
                    return $object;
                }
                $this->next(',');
                $this->white();
            }
        }

        $this->throwSyntaxError('Bad object');
    }

    /**
     * Parse a JSON value.
     *
     * It could be an object, an array, a string, a number,
     * or a word.
     */
    private function value()
    {
        $this->white();
        switch ($this->ch) {
            case '{':
                return $this->obj();
            case '[':
                return $this->arr();
            case '"':
            case "'":
                return $this->string();
            case '-':
            case '+':
            case '.':
                return $this->number();
            default:
                return is_numeric($this->ch) ? $this->number() : $this->word();
        }
    }

    private function throwSyntaxError($message)
    {
        throw new SyntaxError($message, $this->lineNumber, $this->columnNumber);
    }

    private static function renderChar($chr)
    {
        return $chr === null ? 'EOF' : "'" . $chr . "'";
    }

    /**
     * @param string $ch
     *
     * @return string|null
     */
    private static function getEscapee($ch)
    {
        switch ($ch) {
            // @codingStandardsIgnoreStart
            case "'":  return "'";
            case '"':  return '"';
            case '\\': return '\\';
            case '/':  return '/';
            case "\n": return '';
            case 'b':  return chr(8);
            case 'f':  return "\f";
            case 'n':  return "\n";
            case 'r':  return "\r";
            case 't':  return "\t";
            default:   return null;
            // @codingStandardsIgnoreEnd
        }
    }
}


1		<?php
2
3		/*
4		* This file is part of the colinodell/json5 package.
5		*
6		* (c) Colin O'Dell <[email protected]>
7		*
8		* Based on the official JSON5 implementation for JavaScript (https://github.com/json5/json5)
9		* - (c) 2012-2016 Aseem Kishore and others (https://github.com/json5/json5/contributors)
10		*
11		* For the full copyright and license information, please view the LICENSE
12		* file that was distributed with this source code.
13		*/
14
15		namespace ColinODell\Json5;
16
17		final class Json5Decoder
18		{
19		const REGEX_WHITESPACE = '/[ \t\r\n\v\f\xA0\x{FEFF}]/u';
20
21		private $json;
22
23		private $at = 0;
24
25		private $lineNumber = 1;
26
27		private $columnNumber = 1;
28
29		private $ch;
30
31		private $associative = false;
32
33		private $maxDepth = 512;
34
35		private $castBigIntToString = false;
36
37		private $depth = 1;
38
39		private $length;
40
41		private $lineCache;
42
43		/**
44		* Private constructor.
45		*
46		* @param string $json
47		* @param bool $associative
48		* @param int $depth
49		* @param bool $castBigIntToString
50		*/
51	387	private function __construct($json, $associative = false, $depth = 512, $castBigIntToString = false)
52		{
53	387	$this->json = $json;
54	387	$this->associative = $associative;
55	387	$this->maxDepth = $depth;
56	387	$this->castBigIntToString = $castBigIntToString;
57
58	387	$this->length = mb_strlen($json, 'utf-8');
59
60	387	$this->ch = $this->charAt(0);
61	387	}
62
63		/**
64		* Takes a JSON encoded string and converts it into a PHP variable.
65		*
66		* The parameters exactly match PHP's json_decode() function - see
67		* http://php.net/manual/en/function.json-decode.php for more information.
68		*
69		* @param string $source The JSON string being decoded.
70		* @param bool $associative When TRUE, returned objects will be converted into associative arrays.
71		* @param int $depth User specified recursion depth.
72		* @param int $options Bitmask of JSON decode options.
73		*
74		* @return mixed
75		*/
76	387	public static function decode($source, $associative = false, $depth = 512, $options = 0)
77		{
78	387	$associative = $associative \|\| ($options & JSON_OBJECT_AS_ARRAY);
79	387	$castBigIntToString = $options & JSON_BIGINT_AS_STRING;
80
81	387	$decoder = new self((string)$source, $associative, $depth, $castBigIntToString);
		0 ignored issues – show Documentation introduced 2017-11-11 13:06 UTC by Report Bug Copy Issue Report `$castBigIntToString` is of type `integer`, but the function expects a `boolean`. It seems like the type of the argument is not accepted by the function/method which you are calling. In some cases, in particular if PHP’s automatic type-juggling kicks in this might be fine. In other cases, however this might be a bug. We suggest to add an explicit type cast like in the following example: function acceptsInteger($int) { } $x = '123'; // string "123" // Instead of acceptsInteger($x); // we recommend to use acceptsInteger((integer) $x); Loading history...
82
83	387	$result = $decoder->value();
84	300	$decoder->white();
85	297	if ($decoder->ch) {
86	18	$decoder->throwSyntaxError('Syntax error');
87		}
88
89	279	return $result;
90		}
91
92		/**
93		* @param int $at
94		*
95		* @return string\|null
96		*/
97	387	private function charAt($at)
98		{
99	387	if ($at < 0 \|\| $at >= $this->length) {
100	291	return null;
101		}
102
103	384	return mb_substr($this->json, $at, 1, 'utf-8');
104		}
105
106		/**
107		* Parse the next character.
108		*
109		* If $c is given, the next char will only be parsed if the current
110		* one matches $c.
111		*
112		* @param string\|null $c
113		*
114		* @return null\|string
115		*/
116	357	private function next($c = null)
117		{
118		// If a c parameter is provided, verify that it matches the current character.
119	357	if ($c !== null && $c !== $this->ch) {
120	15	$this->throwSyntaxError(sprintf(
121	15	'Expected %s instead of %s',
122	15	self::renderChar($c),
123	15	self::renderChar($this->ch)
124	10	));
125		}
126
127		// Get the next character. When there are no more characters,
128		// return the empty string.
129	357	if ($this->ch === "\n" \|\| ($this->ch === "\r" && $this->peek() !== "\n")) {
130	279	$this->at++;
131	279	$this->lineNumber++;
132	279	$this->columnNumber = 1;
133	186	} else {
134	318	$this->at++;
135	318	$this->columnNumber++;
136		}
137
138	357	$this->ch = $this->charAt($this->at);
139
140	357	return $this->ch;
141		}
142
143		/**
144		* Get the next character without consuming it or
145		* assigning it to the ch variable.
146		*
147		* @return mixed
148		*/
149	12	private function peek()
150		{
151	12	return $this->charAt($this->at + 1);
152		}
153
154		/**
155		* @return string
156		*/
157	231	private function getLineRemainder()
158		{
159		// Line are separated by "\n" or "\r" without an "\n" next
160	231	if ($this->lineCache === null) {
161	231	$this->lineCache = preg_split('/\n\|\r\n?/u', $this->json);
162	154	}
163
164	231	$line = $this->lineCache[$this->lineNumber - 1];
165
166	231	return mb_substr($line, $this->columnNumber - 1);
167		}
168
169		/**
170		* Attempt to match a regular expression at the current position on the current line.
171		*
172		* This function will not match across multiple lines.
173		*
174		* @param string $regex
175		*
176		* @return string\|null
177		*/
178	231	private function match($regex)
179		{
180	231	$subject = $this->getLineRemainder();
181
182	231	$matches = [];
183	231	if (!preg_match($regex, $subject, $matches, PREG_OFFSET_CAPTURE)) {
184	132	return null;
185		}
186
187		// PREG_OFFSET_CAPTURE always returns the byte offset, not the char offset, which is annoying
188	207	$offset = mb_strlen(mb_strcut($subject, 0, $matches[0][1], 'utf-8'), 'utf-8');
189
190		// [0][0] contains the matched text
		0 ignored issues – show Unused Code Comprehensibility introduced 2017-11-10 22:03 UTC by Report Bug Copy Issue Report `40%` of this comment could be valid code. Did you maybe forget this after debugging? Sometimes obsolete code just ends up commented out instead of removed. In this case it is better to remove the code once you have checked you do not need it. The code might also have been commented out for debugging purposes. In this case it is vital that someone uncomments it again or your project may behave in very unexpected ways in production. This check looks for comments that seem to be mostly valid code and reports them. Loading history...
191		// [0][1] contains the index of that match
192	207	$advanceBy = $offset + mb_strlen($matches[0][0], 'utf-8');
193
194	207	$this->at += $advanceBy;
195	207	$this->columnNumber += $advanceBy;
196	207	$this->ch = $this->charAt($this->at);
197
198	207	return $matches[0][0];
199		}
200
201		/**
202		* Parse an identifier.
203		*
204		* Normally, reserved words are disallowed here, but we
205		* only use this for unquoted object keys, where reserved words are allowed,
206		* so we don't check for those here. References:
207		* - http://es5.github.com/#x7.6
208		* - https://developer.mozilla.org/en/Core_JavaScript_1.5_Guide/Core_Language_Features#Variables
209		* - http://docstore.mik.ua/orelly/webprog/jscript/ch02_07.htm
210		*/
211	42	private function identifier()
212		{
213		// @codingStandardsIgnoreStart
214		// Be careful when editing this regex, there are a couple Unicode characters in between here -------------vv
215	42	$match = $this->match('/^(?:[\$_\p{L}\p{Nl}]\|\\\\u[0-9A-Fa-f]{4})(?:[\$_\p{L}\p{Nl}\p{Mn}\p{Mc}\p{Nd}\p{Pc}‌‍]\|\\\\u[0-9A-Fa-f]{4})*/u');
216		// @codingStandardsIgnoreEnd
217
218	42	if ($match === null) {
219	9	$this->throwSyntaxError('Bad identifier as unquoted key');
220		}
221
222		// Un-escape escaped Unicode chars
223	33	$unescaped = preg_replace_callback('/(?:\\\\u[0-9A-Fa-f]{4})+/', function ($m) {
224	6	return json_decode('"'.$m[0].'"');
225	33	}, $match);
226
227	33	return $unescaped;
228		}
229
230	210	private function number()
231		{
232	210	$number = null;
233	210	$sign = '';
234	210	$string = '';
235	210	$base = 10;
236
237	210	if ($this->ch === '-' \|\| $this->ch === '+') {
238	93	$sign = $this->ch;
239	93	$this->next($this->ch);
240	62	}
241
242		// support for Infinity
243	210	if ($this->ch === 'I') {
244	6	$number = $this->word();
245	6	if ($number === null) {
246		$this->throwSyntaxError('Unexpected word for number');
247		}
248
249	6	return ($sign === '-') ? -INF : INF;
250		}
251
252		// support for NaN
253	204	if ($this->ch === 'N') {
254		$number = $this->word();
255		if ($number !== NAN) {
256		$this->throwSyntaxError('expected word to be NaN');
257		}
258
259		// ignore sign as -NaN also is NaN
260		return $number;
261		}
262
263	204	if ($this->ch === '0') {
264	105	$string .= $this->ch;
265	105	$this->next();
266	105	if ($this->ch === 'x' \|\| $this->ch === 'X') {
267	33	$string .= $this->ch;
268	33	$this->next();
269	33	$base = 16;
270	94	} elseif (is_numeric($this->ch)) {
271	30	$this->throwSyntaxError('Octal literal');
272		}
273	50	}
274
275		switch ($base) {
276	174	case 10:
277	144	if (($match = $this->match('/^\d\.?\d/')) !== null) {
278	144	$string .= $match;
279	96	}
280	144	if (($match = $this->match('/^[Ee][-+]?\d*/')) !== null) {
281	45	$string .= $match;
282	30	}
283	144	$number = $string;
284	144	break;
285	33	case 16:
286	33	if (($match = $this->match('/^[A-Fa-f0-9]+/')) !== null) {
287	30	$string .= $match;
288	30	$number = hexdec($string);
289	30	break;
290		}
291	3	$this->throwSyntaxError('Bad hex number');
292		}
293
294	171	if ($sign === '-') {
295	33	$number = -$number;
296	22	}
297
298	171	if (!is_numeric($number) \|\| !is_finite($number)) {
299	3	$this->throwSyntaxError('Bad number');
300		}
301
302	168	if ($this->castBigIntToString) {
303	3	return $number;
304		}
305
306		// Adding 0 will automatically cast this to an int or float
307	165	return $number + 0;
308		}
309
310	84	private function string()
311		{
312	84	if (!($this->ch === '"' \|\| $this->ch === "'")) {
313		$this->throwSyntaxError('Bad string');
314		}
315
316	84	$string = '';
317
318	84	$delim = $this->ch;
319	84	$this->next();
320	84	while ($this->ch !== null) {
321	84	if ($this->ch === $delim) {
322	78	$this->next();
323
324	78	return $string;
325	84	} elseif ($this->ch === '\\') {
326	30	if ($unicodeEscaped = $this->match('/^(?:\\\\u[A-Fa-f0-9]{4})+/')) {
327	6	$string .= json_decode('"'.$unicodeEscaped.'"');
328	6	continue;
329		}
330
331	24	$this->next();
332	24	if ($this->ch === "\r") {
333	6	if ($this->peek() === "\n") {
334	4	$this->next();
335	2	}
336	22	} elseif (($escapee = self::getEscapee($this->ch)) !== null) {
337	15	$string .= $escapee;
338	10	} else {
339	10	break;
340		}
341	84	} elseif ($this->ch === "\n") {
342		// unescaped newlines are invalid; see:
343		// https://github.com/json5/json5/issues/24
344		// @todo this feels special-cased; are there other invalid unescaped chars?
345	3	break;
346		} else {
347	84	$string .= $this->ch;
348		}
349
350	84	$this->next();
351	56	}
352
353	6	$this->throwSyntaxError('Bad string');
354		}
355
356		/**
357		* Skip an inline comment, assuming this is one.
358		*
359		* The current character should be the second / character in the // pair that begins this inline comment.
360		* To finish the inline comment, we look for a newline or the end of the text.
361		*/
362	36	private function inlineComment()
363		{
364		do {
365	36	$this->next();
366	36	if ($this->ch === "\n" \|\| $this->ch === "\r") {
367	33	$this->next();
368
369	33	return;
370		}
371	36	} while ($this->ch !== null);
372	3	}
373
374		/**
375		* Skip a block comment, assuming this is one.
376		*
377		* The current character should be the * character in the /* pair that begins this block comment.
378		* To finish the block comment, we look for an ending */ pair of characters,
379		* but we also watch for the end of text before the comment is terminated.
380		*/
381	21	private function blockComment()
382		{
383		do {
384	21	$this->next();
385	21	while ($this->ch === '*') {
386	18	$this->next('*');
387	18	if ($this->ch === '/') {
388	18	$this->next('/');
389
390	18	return;
391		}
392	2	}
393	21	} while ($this->ch !== null);
394
395	3	$this->throwSyntaxError('Unterminated block comment');
396		}
397
398		/**
399		* Skip a comment, whether inline or block-level, assuming this is one.
400		*/
401	57	private function comment()
402		{
403		// Comments always begin with a / character.
404	57	if ($this->ch !== '/') {
405		$this->throwSyntaxError('Not a comment');
406		}
407
408	57	$this->next('/');
409
410	57	if ($this->ch === '/') {
411	36	$this->inlineComment();
412	46	} elseif ($this->ch === '*') {
413	21	$this->blockComment();
414	12	} else {
415	3	$this->throwSyntaxError('Unrecognized comment');
416		}
417	51	}
418
419		/**
420		* Skip whitespace and comments.
421		*
422		* Note that we're detecting comments by only a single / character.
423		* This works since regular expressions are not valid JSON(5), but this will
424		* break if there are other valid values that begin with a / character!
425		*/
426	387	private function white()
427		{
428	387	while ($this->ch !== null) {
429	384	if ($this->ch === '/') {
430	57	$this->comment();
431	379	} elseif (preg_match(self::REGEX_WHITESPACE, $this->ch) === 1) {
432	282	$this->next();
433	188	} else {
434	375	return;
435		}
436	192	}
437	288	}
438
439		/**
440		* Matches true, false, null, etc
441		*/
442	84	private function word()
443		{
444	84	switch ($this->ch) {
445	84	case 't':
446	36	$this->next('t');
447	36	$this->next('r');
448	36	$this->next('u');
449	36	$this->next('e');
450	36	return true;
451	63	case 'f':
452	18	$this->next('f');
453	18	$this->next('a');
454	18	$this->next('l');
455	18	$this->next('s');
456	18	$this->next('e');
457	18	return false;
458	48	case 'n':
459	18	$this->next('n');
460	18	$this->next('u');
461	18	$this->next('l');
462	18	$this->next('l');
463	18	return null;
464	30	case 'I':
465	15	$this->next('I');
466	15	$this->next('n');
467	12	$this->next('f');
468	12	$this->next('i');
469	12	$this->next('n');
470	12	$this->next('i');
471	12	$this->next('t');
472	12	$this->next('y');
473	12	return INF;
474	15	case 'N':
475	6	$this->next('N');
476	6	$this->next('a');
477	3	$this->next('N');
478	3	return NAN;
479	6	}
480
481	9	$this->throwSyntaxError('Unexpected ' . self::renderChar($this->ch));
482		}
483
484	42	private function arr()
485		{
486	42	$arr = [];
487
488	42	if ($this->ch === '[') {
489	42	if (++$this->depth > $this->maxDepth) {
490	3	$this->throwSyntaxError('Maximum stack depth exceeded');
491		}
492
493	42	$this->next('[');
494	42	$this->white();
495	42	while ($this->ch !== null) {
496	42	if ($this->ch === ']') {
497	12	$this->next(']');
498	12	$this->depth--;
499	12	return $arr; // Potentially empty array
500		}
501		// ES5 allows omitting elements in arrays, e.g. [,] and
502		// [,null]. We don't allow this in JSON5.
503	39	if ($this->ch === ',') {
504	6	$this->throwSyntaxError('Missing array element');
505		} else {
506	33	$arr[] = $this->value();
507		}
508	30	$this->white();
509		// If there's no comma after this value, this needs to
510		// be the end of the array.
511	30	if ($this->ch !== ',') {
512	21	$this->next(']');
513	18	$this->depth--;
514	18	return $arr;
515		}
516	15	$this->next(',');
517	15	$this->white();
518	10	}
519		}
520
521		$this->throwSyntaxError('Bad array');
522		}
523
524		/**
525		* Parse an object value
526		*/
527	81	private function obj()
528		{
529	81	$object = $this->associative ? [] : new \stdClass;
530
531	81	if ($this->ch === '{') {
532	81	if (++$this->depth > $this->maxDepth) {
533		$this->throwSyntaxError('Maximum stack depth exceeded');
534		}
535
536	81	$this->next('{');
537	81	$this->white();
538	81	while ($this->ch) {
539	81	if ($this->ch === '}') {
540	21	$this->next('}');
541	21	$this->depth--;
542	21	return $object; // Potentially empty object
543		}
544
545		// Keys can be unquoted. If they are, they need to be
546		// valid JS identifiers.
547	69	if ($this->ch === '"' \|\| $this->ch === "'") {
548	30	$key = $this->string();
549	20	} else {
550	42	$key = $this->identifier();
551		}
552
553	60	$this->white();
554	60	$this->next(':');
555	57	if ($this->associative) {
556	48	$object[$key] = $this->value();
557	32	} else {
558	54	$object->{$key} = $this->value();
559		}
560	57	$this->white();
561		// If there's no comma after this pair, this needs to be
562		// the end of the object.
563	57	if ($this->ch !== ',') {
564	48	$this->next('}');
565	45	$this->depth--;
566	45	return $object;
567		}
568	21	$this->next(',');
569	21	$this->white();
570	14	}
571		}
572
573		$this->throwSyntaxError('Bad object');
574		}
575
576		/**
577		* Parse a JSON value.
578		*
579		* It could be an object, an array, a string, a number,
580		* or a word.
581		*/
582	387	private function value()
583		{
584	387	$this->white();
585	384	switch ($this->ch) {
586	384	case '{':
587	81	return $this->obj();
588	360	case '[':
589	42	return $this->arr();
590	348	case '"':
591	334	case "'":
592	72	return $this->string();
593	285	case '-':
594	270	case '+':
595	255	case '.':
596	102	return $this->number();
597	124	default:
598	186	return is_numeric($this->ch) ? $this->number() : $this->word();
599	124	}
600		}
601
602	108	private function throwSyntaxError($message)
603		{
604	108	throw new SyntaxError($message, $this->lineNumber, $this->columnNumber);
605		}
606
607	24	private static function renderChar($chr)
608		{
609	24	return $chr === null ? 'EOF' : "'" . $chr . "'";
610		}
611
612		/**
613		* @param string $ch
614		*
615		* @return string\|null
616		*/
617	18	private static function getEscapee($ch)
618		{
619		switch ($ch) {
620		// @codingStandardsIgnoreStart
621	18	case "'": return "'";
622	15	case '"': return '"';
623	15	case '\\': return '\\';
624	15	case '/': return '/';
625	15	case "\n": return '';
626	6	case 'b': return chr(8);
627	6	case 'f': return "\f";
628	6	case 'n': return "\n";
629	6	case 'r': return "\r";
630	6	case 't': return "\t";
631	3	default: return null;
632		// @codingStandardsIgnoreEnd
633	2	}
634		}
635		}
636

colinodell / json5

Push — master ( bd17ed...fa6e51 )

Json5Decoder::getEscapee() B

Complexity

Size

Duplication

Code Coverage

Importance

How to fix Complexity

Long Method

Duplication Side-by-Side

Filter issues like