Json5Decoder::string() - Code Metrics - Inspection of "Drop PHP 5.3 support" - colinodell/json5 - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( 341b80...1751fd )

by Colin

created 2017-11-11 13:55 UTC

Json5Decoder::string() C

↳ Parent: Json5Decoder

Complexity

Conditions	12
Paths	12

Size

Total Lines	44
Code Lines	29

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	24
CRAP Score	14.25

Importance

Changes

Metric	Value
cc	12
eloc	29
nc	12
nop	0
dl	0
loc	44
ccs	24
cts	32
cp	0.75
crap	14.25
rs	5.1612
c	0
b	0
f	0

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 self::fromCharCode($m[1]);
        }, $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 = $this->ch;
        while ($this->next() !== null) {
            if ($this->ch === $delim) {
                $this->next();

                return $string;
            } elseif ($this->ch === '\\') {
                $this->next();
                if ($this->ch === 'u') {
                    $this->next();
                    $hex = $this->match('/^[A-Fa-f0-9]{4}/');
                    if ($hex === null) {
                        break;
                    }
                    $string .= self::fromCharCode($hex);
                } elseif ($this->ch === "\r") {
                    if ($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->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()
    {
        if ($this->ch !== '/') {
            $this->throwSyntaxError('Not an inline comment');
        }

        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()
    {
        if ($this->ch !== '*') {
            $this->throwSyntaxError('Not a block comment');
        }

        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->at, $this->lineNumber, $this->columnNumber);
    }

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

    /**
     * @param string $hex Hex code
     *
     * @return string Unicode character
     */
    private static function fromCharCode($hex)
    {
        return mb_convert_encoding('&#' . hexdec($hex) . ';', 'UTF-8', 'HTML-ENTITIES');
    }

    /**
     * @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 '\b';
            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 $castBigIntToString
50		*/
51	360	private function __construct($json, $associative = false, $depth = 512, $castBigIntToString = false)
52		{
53	360	$this->json = $json;
54	360	$this->associative = $associative;
55	360	$this->maxDepth = $depth;
56	360	$this->castBigIntToString = $castBigIntToString;
57
58	360	$this->length = mb_strlen($json, 'utf-8');
59
60	360	$this->ch = $this->charAt(0);
61	360	}
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 $options Bitmask of JSON decode options.
73		*
74		* @return mixed
75		*/
76	360	public static function decode($source, $associative = false, $depth = 512, $options = 0)
77		{
78	360	$associative = $associative \|\| ($options & JSON_OBJECT_AS_ARRAY);
79	360	$castBigIntToString = $options & JSON_BIGINT_AS_STRING;
80
81	360	$decoder = new self((string)$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	360	$result = $decoder->value();
84	285	$decoder->white();
85	282	if ($decoder->ch) {
86	18	$decoder->throwSyntaxError('Syntax error');
87		}
88
89	264	return $result;
90		}
91
92		/**
93		* @param int $at
94		*
95		* @return string\|null
96		*/
97	360	private function charAt($at)
98		{
99	360	if ($at < 0 \|\| $at >= $this->length) {
100	276	return null;
101		}
102
103	357	return mb_substr($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 $c
113		*
114		* @return null\|string
115		*/
116	330	private function next($c = null)
117		{
118		// If a c parameter is provided, verify that it matches the current character.
119	330	if ($c !== null && $c !== $this->ch) {
120	9	$this->throwSyntaxError(sprintf(
121	9	'Expected %s instead of %s',
122	9	self::renderChar($c),
123	9	self::renderChar($this->ch)
124	6	));
125		}
126
127		// Get the next character. When there are no more characters,
128		// return the empty string.
129	330	if ($this->ch === "\n" \|\| ($this->ch === "\r" && $this->peek() !== "\n")) {
130	264	$this->at++;
131	264	$this->lineNumber++;
132	264	$this->columnNumber = 1;
133	176	} else {
134	291	$this->at++;
135	291	$this->columnNumber++;
136		}
137
138	330	$this->ch = $this->charAt($this->at);
139
140	330	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	210	private function getLineRemainder()
158		{
159		// Line are separated by "\n" or "\r" without an "\n" next
160	210	if ($this->lineCache === null) {
161	210	$this->lineCache = preg_split('/\n\|\r\n?/u', $this->json);
162	140	}
163
164	210	$line = $this->lineCache[$this->lineNumber - 1];
165
166	210	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 $regex
175		*
176		* @return string\|null
177		*/
178	210	private function match($regex)
179		{
180	210	$subject = $this->getLineRemainder();
181
182	210	$matches = [];
183	210	if (!preg_match($regex, $subject, $matches, PREG_OFFSET_CAPTURE)) {
184	111	return null;
185		}
186
187		// PREG_OFFSET_CAPTURE always returns the byte offset, not the char offset, which is annoying
188	198	$offset = mb_strlen(mb_strcut($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	198	$advanceBy = $offset + mb_strlen($matches[0][0], 'utf-8');
193
194	198	$this->at += $advanceBy;
195	198	$this->columnNumber += $advanceBy;
196	198	$this->ch = $this->charAt($this->at);
197
198	198	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	39	private function identifier()
212		{
213		// @codingStandardsIgnoreStart
214		// Be careful when editing this regex, there are a couple Unicode characters in between here -------------vv
215	39	$match = $this->match('/^(?:[\$_\p{L}\p{Nl}]\|\\\\u[0-9A-Fa-f]{4})(?:[\$_\p{L}\p{Nl}\p{Mn}\p{Mc}\p{Nd}\p{Pc}‌‍]\|\\\\u[0-9A-Fa-f]{4})*/u');
216		// @codingStandardsIgnoreEnd
217
218	39	if ($match === null) {
219	9	$this->throwSyntaxError('Bad identifier as unquoted key');
220		}
221
222		// Un-escape escaped Unicode chars
223	30	$unescaped = preg_replace_callback('/\\\\u([0-9A-Fa-f]{4})/', function ($m) {
224	3	return self::fromCharCode($m[1]);
225	30	}, $match);
226
227	30	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 $number + 0;
308		}
309
310	66	private function string()
311		{
312	66	if (!($this->ch === '"' \|\| $this->ch === "'")) {
313		$this->throwSyntaxError('Bad string');
314		}
315
316	66	$string = '';
317
318	66	$delim = $this->ch;
319	66	while ($this->next() !== null) {
320	66	if ($this->ch === $delim) {
321	63	$this->next();
322
323	63	return $string;
324	66	} elseif ($this->ch === '\\') {
325	18	$this->next();
326	18	if ($this->ch === 'u') {
327		$this->next();
328		$hex = $this->match('/^[A-Fa-f0-9]{4}/');
329		if ($hex === null) {
330		break;
331		}
332		$string .= self::fromCharCode($hex);
333	18	} elseif ($this->ch === "\r") {
334	6	if ($this->peek() === "\n") {
335	4	$this->next();
336	2	}
337	16	} elseif (($escapee = self::getEscapee($this->ch)) !== null) {
338	12	$string .= $escapee;
339	8	} else {
340	6	break;
341		}
342	66	} elseif ($this->ch === "\n") {
343		// unescaped newlines are invalid; see:
344		// https://github.com/json5/json5/issues/24
345		// @todo this feels special-cased; are there other invalid unescaped chars?
346	3	break;
347		} else {
348	66	$string .= $this->ch;
349		}
350	44	}
351
352	3	$this->throwSyntaxError('Bad string');
353		}
354
355		/**
356		* Skip an inline comment, assuming this is one.
357		*
358		* The current character should be the second / character in the // pair that begins this inline comment.
359		* To finish the inline comment, we look for a newline or the end of the text.
360		*/
361	36	private function inlineComment()
362		{
363	36	if ($this->ch !== '/') {
364		$this->throwSyntaxError('Not an inline comment');
365		}
366
367		do {
368	36	$this->next();
369	36	if ($this->ch === "\n" \|\| $this->ch === "\r") {
370	33	$this->next();
371
372	33	return;
373		}
374	36	} while ($this->ch !== null);
375	3	}
376
377		/**
378		* Skip a block comment, assuming this is one.
379		*
380		* The current character should be the * character in the /* pair that begins this block comment.
381		* To finish the block comment, we look for an ending */ pair of characters,
382		* but we also watch for the end of text before the comment is terminated.
383		*/
384	21	private function blockComment()
385		{
386	21	if ($this->ch !== '*') {
387		$this->throwSyntaxError('Not a block comment');
388		}
389
390		do {
391	21	$this->next();
392	21	while ($this->ch === '*') {
393	18	$this->next('*');
394	18	if ($this->ch === '/') {
395	18	$this->next('/');
396
397	18	return;
398		}
399	2	}
400	21	} while ($this->ch !== null);
401
402	3	$this->throwSyntaxError('Unterminated block comment');
403		}
404
405		/**
406		* Skip a comment, whether inline or block-level, assuming this is one.
407		*/
408	54	private function comment()
409		{
410		// Comments always begin with a / character.
411	54	if ($this->ch !== '/') {
412		$this->throwSyntaxError('Not a comment');
413		}
414
415	54	$this->next('/');
416
417	54	if ($this->ch === '/') {
418	36	$this->inlineComment();
419	43	} elseif ($this->ch === '*') {
420	21	$this->blockComment();
421	12	} else {
422		$this->throwSyntaxError('Unrecognized comment');
423		}
424	51	}
425
426		/**
427		* Skip whitespace and comments.
428		*
429		* Note that we're detecting comments by only a single / character.
430		* This works since regular expressions are not valid JSON(5), but this will
431		* break if there are other valid values that begin with a / character!
432		*/
433	360	private function white()
434		{
435	360	while ($this->ch !== null) {
436	357	if ($this->ch === '/') {
437	54	$this->comment();
438	355	} elseif (preg_match(self::REGEX_WHITESPACE, $this->ch) === 1) {
439	267	$this->next();
440	178	} else {
441	351	return;
442		}
443	182	}
444	273	}
445
446		/**
447		* Matches true, false, null, etc
448		*/
449	78	private function word()
450		{
451	78	switch ($this->ch) {
452	78	case 't':
453	36	$this->next('t');
454	36	$this->next('r');
455	36	$this->next('u');
456	36	$this->next('e');
457	36	return true;
458	57	case 'f':
459	18	$this->next('f');
460	18	$this->next('a');
461	18	$this->next('l');
462	18	$this->next('s');
463	18	$this->next('e');
464	18	return false;
465	42	case 'n':
466	18	$this->next('n');
467	18	$this->next('u');
468	18	$this->next('l');
469	18	$this->next('l');
470	18	return null;
471	24	case 'I':
472	12	$this->next('I');
473	12	$this->next('n');
474	12	$this->next('f');
475	12	$this->next('i');
476	12	$this->next('n');
477	12	$this->next('i');
478	12	$this->next('t');
479	12	$this->next('y');
480	12	return INF;
481	12	case 'N':
482	3	$this->next('N');
483	3	$this->next('a');
484	3	$this->next('N');
485	3	return NAN;
486	6	}
487
488	9	$this->throwSyntaxError('Unexpected ' . self::renderChar($this->ch));
489		}
490
491	42	private function arr()
492		{
493	42	$arr = [];
494
495	42	if ($this->ch === '[') {
496	42	if (++$this->depth > $this->maxDepth) {
497	3	$this->throwSyntaxError('Maximum stack depth exceeded');
498		}
499
500	42	$this->next('[');
501	42	$this->white();
502	42	while ($this->ch !== null) {
503	42	if ($this->ch === ']') {
504	12	$this->next(']');
505	12	$this->depth--;
506	12	return $arr; // Potentially empty array
507		}
508		// ES5 allows omitting elements in arrays, e.g. [,] and
509		// [,null]. We don't allow this in JSON5.
510	39	if ($this->ch === ',') {
511	6	$this->throwSyntaxError('Missing array element');
512		} else {
513	33	$arr[] = $this->value();
514		}
515	30	$this->white();
516		// If there's no comma after this value, this needs to
517		// be the end of the array.
518	30	if ($this->ch !== ',') {
519	21	$this->next(']');
520	18	$this->depth--;
521	18	return $arr;
522		}
523	15	$this->next(',');
524	15	$this->white();
525	10	}
526		}
527
528		$this->throwSyntaxError('Bad array');
529		}
530
531		/**
532		* Parse an object value
533		*/
534	75	private function obj()
535		{
536	75	$object = $this->associative ? [] : new \stdClass;
537
538	75	if ($this->ch === '{') {
539	75	if (++$this->depth > $this->maxDepth) {
540		$this->throwSyntaxError('Maximum stack depth exceeded');
541		}
542
543	75	$this->next('{');
544	75	$this->white();
545	75	while ($this->ch) {
546	75	if ($this->ch === '}') {
547	21	$this->next('}');
548	21	$this->depth--;
549	21	return $object; // Potentially empty object
550		}
551
552		// Keys can be unquoted. If they are, they need to be
553		// valid JS identifiers.
554	63	if ($this->ch === '"' \|\| $this->ch === "'") {
555	27	$key = $this->string();
556	18	} else {
557	39	$key = $this->identifier();
558		}
559
560	54	$this->white();
561	54	$this->next(':');
562	51	if ($this->associative) {
563	45	$object[$key] = $this->value();
564	30	} else {
565	48	$object->{$key} = $this->value();
566		}
567	51	$this->white();
568		// If there's no comma after this pair, this needs to be
569		// the end of the object.
570	51	if ($this->ch !== ',') {
571	42	$this->next('}');
572	39	$this->depth--;
573	39	return $object;
574		}
575	18	$this->next(',');
576	18	$this->white();
577	12	}
578		}
579
580		$this->throwSyntaxError('Bad object');
581		}
582
583		/**
584		* Parse a JSON value.
585		*
586		* It could be an object, an array, a string, a number,
587		* or a word.
588		*/
589	360	private function value()
590		{
591	360	$this->white();
592	360	switch ($this->ch) {
593	360	case '{':
594	75	return $this->obj();
595	336	case '[':
596	42	return $this->arr();
597	324	case '"':
598	315	case "'":
599	54	return $this->string();
600	279	case '-':
601	264	case '+':
602	249	case '.':
603	102	return $this->number();
604	120	default:
605	180	return is_numeric($this->ch) ? $this->number() : $this->word();
606	120	}
607		}
608
609	96	private function throwSyntaxError($message)
610		{
611	96	throw new SyntaxError($message, $this->at, $this->lineNumber, $this->columnNumber);
612		}
613
614	18	private static function renderChar($chr)
615		{
616	18	return $chr === null ? 'EOF' : "'" . $chr . "'";
617		}
618
619		/**
620		* @param string $hex Hex code
621		*
622		* @return string Unicode character
623		*/
624	3	private static function fromCharCode($hex)
625		{
626	3	return mb_convert_encoding('&#' . hexdec($hex) . ';', 'UTF-8', 'HTML-ENTITIES');
627		}
628
629		/**
630		* @param string $ch
631		*
632		* @return string\|null
633		*/
634	12	private static function getEscapee($ch)
635		{
636		switch ($ch) {
637		// @codingStandardsIgnoreStart
638	12	case "'": return "'";
639	9	case '"': return '"';
640	9	case '\\': return '\\';
641	9	case '/': return '/';
642	9	case "\n": return '';
643		case 'b': return '\b';
644		case 'f': return '\f';
645		case 'n': return '\n';
646		case 'r': return '\r';
647		case 't': return '\t';
648		default: return null;
649		// @codingStandardsIgnoreEnd
650		}
651		}
652		}
653

colinodell / json5

Push — master ( 341b80...1751fd )

Json5Decoder::string() C

Complexity

Size

Duplication

Code Coverage

Importance

How to fix Complexity

Long Method

Duplication Side-by-Side

Filter issues like