Json5Decoder::nextOrFail() - Code Metrics - Inspection of "Merge pull request #2 from colinodell/optimization..." - colinodell/json5 - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Push — master ( 2bb3fb...2d1db6 )

by Colin

created 2018-01-14 14:20 UTC

Json5Decoder::nextOrFail() A

↳ Parent: Json5Decoder

Complexity

Conditions	2
Paths	2

Size

Total Lines	12
Code Lines	7

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	8
CRAP Score	2.0054

Importance

Changes

Metric	Value
cc	2
eloc	7
c	0
b	0
f	0
nc	2
nop	1
dl	0
loc	12
ccs	8
cts	9
cp	0.8889
crap	2.0054
rs	9.4285

<?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
{
    private $at = 0;

    private $lineNumber = 1;

    private $columnNumber = 1;

    private $ch;

    private $chArr;

    private $associative = false;

    private $maxDepth = 512;

    private $castBigIntToString = false;

    private $depth = 1;

    private $length;

    private $remainderCache;

    private $remainderCacheAt;

    /**
     * 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->associative = $associative;
        $this->maxDepth = $depth;
        $this->castBigIntToString = $castBigIntToString;

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

        $this->chArr = preg_split('//u', $json, null, PREG_SPLIT_NO_EMPTY);
        $this->ch = $this->charAt(0);

        $this->remainderCache = $json;
        $this->remainderCacheAt = 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)
    {
        // Try parsing with json_decode first, since that's much faster
        // We only attempt this on PHP 7+ because 5.x doesn't parse some edge cases correctly
        if (PHP_VERSION_ID >= 700000) {
            $result = json_decode($source, $associative, $depth, $options);
            if (json_last_error() === JSON_ERROR_NONE) {
                return $result;
            }
        }

        // Fall back to JSON5 if that fails
        $associative = $associative === true || ($associative === null && $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 >= $this->length) {
            return null;
        }

        return $this->chArr[$at];
    }

    /**
     * Parse the next character.
     *
     * @return null|string
     */
    private function next()
    {
        // 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;
    }

    /**
     * Parse the next character if it matches $c or fail.
     *
     * @param string $c
     *
     * @return string|null
     */
    private function nextOrFail($c)
    {
        if ($c !== $this->ch) {
            $this->throwSyntaxError(sprintf(
                'Expected %s instead of %s',
                self::renderChar($c),
                self::renderChar($this->ch)
            ));
        }

        return $this->next();
    }

    /**
     * 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);
    }

    /**
     * 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->getRemainder();

        $matches = [];
        if (!preg_match($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();
        }

        // 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 ($base) {
            case 10:
                if ((is_numeric($this->ch) || $this->ch === '.') && ($match = $this->match('/^\d*\.?\d*/')) !== null) {

                    $string .= $match;
                }
                if (($this->ch === 'E' || $this->ch === 'e') && ($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()
    {
        $string = '';

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

                return $string;
            }

            if ($this->ch === '\\') {
                if ($this->peek() === 'u' && $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 === '*') {
                $this->nextOrFail('*');
                if ($this->ch === '/') {
                    $this->nextOrFail('/');

                    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.
        $this->nextOrFail('/');

        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('/[ \t\r\n\v\f\xA0\x{FEFF}]/u', $this->ch) === 1) {
                $this->next();
            } else {
                return;
            }
        }
    }

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

                $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->nextOrFail(']');
                    $this->depth--;
                    return $arr;
                }
                $this->nextOrFail(',');
                $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->nextOrFail('{');
            $this->white();
            while ($this->ch) {
                if ($this->ch === '}') {
                    $this->nextOrFail('}');
                    $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();
                $this->nextOrFail(':');
                if ($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->nextOrFail('}');
                    $this->depth--;
                    return $object;
                }
                $this->nextOrFail(',');
                $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
        }
    }

    /**
     * Returns everything from $this->at onwards.
     *
     * Utilizes a cache so we don't have to continuously parse through UTF-8
     * data that was earlier in the string which we don't even care about.
     *
     * @return string
     */
    private function getRemainder()
    {
        if ($this->remainderCacheAt === $this->at) {
            return $this->remainderCache;
        }

        $subject = mb_substr($this->remainderCache, $this->at - $this->remainderCacheAt);
        $this->remainderCache = $subject;
        $this->remainderCacheAt = $this->at;

        return $subject;
    }
}


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			private $at = 0;
20
21			private $lineNumber = 1;
22
23			private $columnNumber = 1;
24
25			private $ch;
26
27			private $chArr;
28
29			private $associative = false;
30
31			private $maxDepth = 512;
32
33			private $castBigIntToString = false;
34
35			private $depth = 1;
36
37			private $length;
38
39			private $remainderCache;
40
41			private $remainderCacheAt;
42
43			/**
44			* Private constructor.
45			*
46			* @param string $json
47			* @param bool $associative
48			* @param int $depth
49			* @param bool $castBigIntToString
50			*/
51	396		private function __construct($json, $associative = false, $depth = 512, $castBigIntToString = false)
52			{
53	396		$this->associative = $associative;
54	396		$this->maxDepth = $depth;
55	396		$this->castBigIntToString = $castBigIntToString;
56
57	396		$this->length = mb_strlen($json, 'utf-8');
58
59	396		$this->chArr = preg_split('//u', $json, null, PREG_SPLIT_NO_EMPTY);
60	396		$this->ch = $this->charAt(0);
61
62	396		$this->remainderCache = $json;
63	396		$this->remainderCacheAt = 0;
64	396		}
65
66			/**
67			* Takes a JSON encoded string and converts it into a PHP variable.
68			*
69			* The parameters exactly match PHP's json_decode() function - see
70			* http://php.net/manual/en/function.json-decode.php for more information.
71			*
72			* @param string $source The JSON string being decoded.
73			* @param bool $associative When TRUE, returned objects will be converted into associative arrays.
74			* @param int $depth User specified recursion depth.
75			* @param int $options Bitmask of JSON decode options.
76			*
77			* @return mixed
78			*/
79	396		public static function decode($source, $associative = false, $depth = 512, $options = 0)
80			{
81			// Try parsing with json_decode first, since that's much faster
82			// We only attempt this on PHP 7+ because 5.x doesn't parse some edge cases correctly
83	396		if (PHP_VERSION_ID >= 700000) {
84			$result = json_decode($source, $associative, $depth, $options);
85			if (json_last_error() === JSON_ERROR_NONE) {
86			return $result;
87			}
88			}
89
90			// Fall back to JSON5 if that fails
91	396		$associative = $associative === true \|\| ($associative === null && $options & JSON_OBJECT_AS_ARRAY);
92	396		$castBigIntToString = $options & JSON_BIGINT_AS_STRING;
93
94	396		$decoder = new self((string)$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...
95
96	396		$result = $decoder->value();
97	303		$decoder->white();
98	300		if ($decoder->ch) {
99	18		$decoder->throwSyntaxError('Syntax error');
100			}
101
102	282		return $result;
103			}
104
105			/**
106			* @param int $at
107			*
108			* @return string\|null
109			*/
110	396		private function charAt($at)
111			{
112	396		if ($at >= $this->length) {
113	297		return null;
114			}
115
116	393		return $this->chArr[$at];
117			}
118
119			/**
120			* Parse the next character.
121			*
122			* @return null\|string
123			*/
124	366		private function next()
125			{
126			// Get the next character. When there are no more characters,
127			// return the empty string.
128	366		if ($this->ch === "\n" \|\| ($this->ch === "\r" && $this->peek() !== "\n")) {
129	282		$this->at++;
130	282		$this->lineNumber++;
131	282		$this->columnNumber = 1;
132	188		} else {
133	327		$this->at++;
134	327		$this->columnNumber++;
135			}
136
137	366		$this->ch = $this->charAt($this->at);
138
139	366		return $this->ch;
140			}
141
142			/**
143			* Parse the next character if it matches $c or fail.
144			*
145			* @param string $c
146			*
147			* @return string\|null
148			*/
149	168		private function nextOrFail($c)
150			{
151	168		if ($c !== $this->ch) {
152	18		$this->throwSyntaxError(sprintf(
153	18		'Expected %s instead of %s',
154	18		self::renderChar($c),
155	18		self::renderChar($this->ch)
156	12		));
157			}
158
159	168		return $this->next();
160			}
161
162			/**
163			* Get the next character without consuming it or
164			* assigning it to the ch variable.
165			*
166			* @return mixed
167			*/
168	36		private function peek()
169			{
170	36		return $this->charAt($this->at + 1);
171			}
172
173			/**
174			* Attempt to match a regular expression at the current position on the current line.
175			*
176			* This function will not match across multiple lines.
177			*
178			* @param string $regex
179			*
180			* @return string\|null
181			*/
182	210		private function match($regex)
183			{
184	210		$subject = $this->getRemainder();
185
186	210		$matches = [];
187	210		if (!preg_match($regex, $subject, $matches, PREG_OFFSET_CAPTURE)) {
188	15		return null;
189			}
190
191			// PREG_OFFSET_CAPTURE always returns the byte offset, not the char offset, which is annoying
192	195		$offset = mb_strlen(mb_strcut($subject, 0, $matches[0][1], 'utf-8'), 'utf-8');
193
194			// [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...
195			// [0][1] contains the index of that match
196	195		$advanceBy = $offset + mb_strlen($matches[0][0], 'utf-8');
197
198	195		$this->at += $advanceBy;
199	195		$this->columnNumber += $advanceBy;
200	195		$this->ch = $this->charAt($this->at);
201
202	195		return $matches[0][0];
203			}
204
205			/**
206			* Parse an identifier.
207			*
208			* Normally, reserved words are disallowed here, but we
209			* only use this for unquoted object keys, where reserved words are allowed,
210			* so we don't check for those here. References:
211			* - http://es5.github.com/#x7.6
212			* - https://developer.mozilla.org/en/Core_JavaScript_1.5_Guide/Core_Language_Features#Variables
213			* - http://docstore.mik.ua/orelly/webprog/jscript/ch02_07.htm
214			*/
215	42		private function identifier()
216			{
217			// @codingStandardsIgnoreStart
218			// Be careful when editing this regex, there are a couple Unicode characters in between here -------------vv
219	42		$match = $this->match('/^(?:[\$_\p{L}\p{Nl}]\|\\\\u[0-9A-Fa-f]{4})(?:[\$_\p{L}\p{Nl}\p{Mn}\p{Mc}\p{Nd}\p{Pc}‌‍]\|\\\\u[0-9A-Fa-f]{4})*/u');
220			// @codingStandardsIgnoreEnd
221
222	42		if ($match === null) {
223	9		$this->throwSyntaxError('Bad identifier as unquoted key');
224			}
225
226			// Un-escape escaped Unicode chars
227	33		$unescaped = preg_replace_callback('/(?:\\\\u[0-9A-Fa-f]{4})+/', function ($m) {
228	6		return json_decode('"'.$m[0].'"');
229	33		}, $match);
230
231	33		return $unescaped;
232			}
233
234	210		private function number()
235			{
236	210		$number = null;
237	210		$sign = '';
238	210		$string = '';
239	210		$base = 10;
240
241	210		if ($this->ch === '-' \|\| $this->ch === '+') {
242	93		$sign = $this->ch;
243	93		$this->next();
244	62		}
245
246			// support for Infinity
247	210		if ($this->ch === 'I') {
248	6		$number = $this->word();
249	6		if ($number === null) {
250			$this->throwSyntaxError('Unexpected word for number');
251			}
252
253	6		return ($sign === '-') ? -INF : INF;
254			}
255
256			// support for NaN
257	204		if ($this->ch === 'N') {
258			$number = $this->word();
259			if ($number !== NAN) {
260			$this->throwSyntaxError('expected word to be NaN');
261			}
262
263			// ignore sign as -NaN also is NaN
264			return $number;
265			}
266
267	204		if ($this->ch === '0') {
268	105		$string .= $this->ch;
269	105		$this->next();
270	105		if ($this->ch === 'x' \|\| $this->ch === 'X') {
271	33		$string .= $this->ch;
272	33		$this->next();
273	33		$base = 16;
274	94		} elseif (is_numeric($this->ch)) {
275	30		$this->throwSyntaxError('Octal literal');
276			}
277	50		}
278
279			switch ($base) {
280	174		case 10:
281	144	View Code Duplication	if ((is_numeric($this->ch) \|\| $this->ch === '.') && ($match = $this->match('/^\d\.?\d/')) !== null) {
			0 ignored issues – show Duplication introduced 2018-01-13 23:13 UTC by Report Bug Copy Issue Report This code seems to be duplicated across your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
282	129		$string .= $match;
283	86		}
284	144	View Code Duplication	if (($this->ch === 'E' \|\| $this->ch === 'e') && ($match = $this->match('/^[Ee][-+]?\d*/')) !== null) {
			0 ignored issues – show Duplication introduced 2018-01-13 23:13 UTC by Report Bug Copy Issue Report This code seems to be duplicated across your project. Duplicated code is one of the most pungent code smells. If you need to duplicate the same code in three or more different places, we strongly encourage you to look into extracting the code into a single class or operation. You can also find more detailed suggestions in the “Code” section of your repository. Loading history...
285	45		$string .= $match;
286	30		}
287	144		$number = $string;
288	144		break;
289	33		case 16:
290	33		if (($match = $this->match('/^[A-Fa-f0-9]+/')) !== null) {
291	30		$string .= $match;
292	30		$number = hexdec($string);
293	30		break;
294			}
295	3		$this->throwSyntaxError('Bad hex number');
296			}
297
298	171		if ($sign === '-') {
299	33		$number = -$number;
300	22		}
301
302	171		if (!is_numeric($number) \|\| !is_finite($number)) {
303	3		$this->throwSyntaxError('Bad number');
304			}
305
306	168		if ($this->castBigIntToString) {
307	3		return $number;
308			}
309
310			// Adding 0 will automatically cast this to an int or float
311	165		return $number + 0;
312			}
313
314	93		private function string()
315			{
316	93		$string = '';
317
318	93		$delim = $this->ch;
319	93		$this->next();
320	93		while ($this->ch !== null) {
321	93		if ($this->ch === $delim) {
322	87		$this->next();
323
324	87		return $string;
325			}
326
327	93		if ($this->ch === '\\') {
328	30		if ($this->peek() === 'u' && $unicodeEscaped = $this->match('/^(?:\\\\u[A-Fa-f0-9]{4})+/')) {
329	6		$string .= json_decode('"'.$unicodeEscaped.'"');
330	6		continue;
331			}
332
333	24		$this->next();
334	24		if ($this->ch === "\r") {
335	6		if ($this->peek() === "\n") {
336	4		$this->next();
337	2		}
338	22		} elseif (($escapee = self::getEscapee($this->ch)) !== null) {
339	15		$string .= $escapee;
340	10		} else {
341	10		break;
342			}
343	93		} elseif ($this->ch === "\n") {
344			// unescaped newlines are invalid; see:
345			// https://github.com/json5/json5/issues/24
346			// @todo this feels special-cased; are there other invalid unescaped chars?
347	3		break;
348			} else {
349	93		$string .= $this->ch;
350			}
351
352	93		$this->next();
353	62		}
354
355	6		$this->throwSyntaxError('Bad string');
356			}
357
358			/**
359			* Skip an inline comment, assuming this is one.
360			*
361			* The current character should be the second / character in the // pair that begins this inline comment.
362			* To finish the inline comment, we look for a newline or the end of the text.
363			*/
364	36		private function inlineComment()
365			{
366			do {
367	36		$this->next();
368	36		if ($this->ch === "\n" \|\| $this->ch === "\r") {
369	33		$this->next();
370
371	33		return;
372			}
373	36		} while ($this->ch !== null);
374	3		}
375
376			/**
377			* Skip a block comment, assuming this is one.
378			*
379			* The current character should be the * character in the /* pair that begins this block comment.
380			* To finish the block comment, we look for an ending */ pair of characters,
381			* but we also watch for the end of text before the comment is terminated.
382			*/
383	21		private function blockComment()
384			{
385			do {
386	21		$this->next();
387	21		while ($this->ch === '*') {
388	18		$this->nextOrFail('*');
389	18		if ($this->ch === '/') {
390	18		$this->nextOrFail('/');
391
392	18		return;
393			}
394	2		}
395	21		} while ($this->ch !== null);
396
397	3		$this->throwSyntaxError('Unterminated block comment');
398			}
399
400			/**
401			* Skip a comment, whether inline or block-level, assuming this is one.
402			*/
403	57		private function comment()
404			{
405			// Comments always begin with a / character.
406	57		$this->nextOrFail('/');
407
408	57		if ($this->ch === '/') {
409	36		$this->inlineComment();
410	46		} elseif ($this->ch === '*') {
411	21		$this->blockComment();
412	12		} else {
413	3		$this->throwSyntaxError('Unrecognized comment');
414			}
415	51		}
416
417			/**
418			* Skip whitespace and comments.
419			*
420			* Note that we're detecting comments by only a single / character.
421			* This works since regular expressions are not valid JSON(5), but this will
422			* break if there are other valid values that begin with a / character!
423			*/
424	396		private function white()
425			{
426	396		while ($this->ch !== null) {
427	393		if ($this->ch === '/') {
428	57		$this->comment();
429	388		} elseif (preg_match('/[ \t\r\n\v\f\xA0\x{FEFF}]/u', $this->ch) === 1) {
430	291		$this->next();
431	194		} else {
432	384		return;
433			}
434	198		}
435	294		}
436
437			/**
438			* Matches true, false, null, etc
439			*/
440	84		private function word()
441			{
442	84		switch ($this->ch) {
443	84		case 't':
444	36		$this->nextOrFail('t');
445	36		$this->nextOrFail('r');
446	36		$this->nextOrFail('u');
447	36		$this->nextOrFail('e');
448	36		return true;
449	63		case 'f':
450	18		$this->nextOrFail('f');
451	18		$this->nextOrFail('a');
452	18		$this->nextOrFail('l');
453	18		$this->nextOrFail('s');
454	18		$this->nextOrFail('e');
455	18		return false;
456	48		case 'n':
457	18		$this->nextOrFail('n');
458	18		$this->nextOrFail('u');
459	18		$this->nextOrFail('l');
460	18		$this->nextOrFail('l');
461	18		return null;
462	30		case 'I':
463	15		$this->nextOrFail('I');
464	15		$this->nextOrFail('n');
465	12		$this->nextOrFail('f');
466	12		$this->nextOrFail('i');
467	12		$this->nextOrFail('n');
468	12		$this->nextOrFail('i');
469	12		$this->nextOrFail('t');
470	12		$this->nextOrFail('y');
471	12		return INF;
472	15		case 'N':
473	6		$this->nextOrFail('N');
474	6		$this->nextOrFail('a');
475	3		$this->nextOrFail('N');
476	3		return NAN;
477	6		}
478
479	9		$this->throwSyntaxError('Unexpected ' . self::renderChar($this->ch));
480			}
481
482	42		private function arr()
483			{
484	42		$arr = [];
485
486	42		if ($this->ch === '[') {
487	42		if (++$this->depth > $this->maxDepth) {
488	3		$this->throwSyntaxError('Maximum stack depth exceeded');
489			}
490
491	42		$this->nextOrFail('[');
492	42		$this->white();
493	42		while ($this->ch !== null) {
494	42		if ($this->ch === ']') {
495	12		$this->nextOrFail(']');
496	12		$this->depth--;
497	12		return $arr; // Potentially empty array
498			}
499			// ES5 allows omitting elements in arrays, e.g. [,] and
500			// [,null]. We don't allow this in JSON5.
501	39		if ($this->ch === ',') {
502	6		$this->throwSyntaxError('Missing array element');
503			}
504
505	33		$arr[] = $this->value();
506
507	30		$this->white();
508			// If there's no comma after this value, this needs to
509			// be the end of the array.
510	30		if ($this->ch !== ',') {
511	21		$this->nextOrFail(']');
512	18		$this->depth--;
513	18		return $arr;
514			}
515	15		$this->nextOrFail(',');
516	15		$this->white();
517	10		}
518			}
519
520			$this->throwSyntaxError('Bad array');
521			}
522
523			/**
524			* Parse an object value
525			*/
526	90		private function obj()
527			{
528	90		$object = $this->associative ? [] : new \stdClass;
529
530	90		if ($this->ch === '{') {
531	90		if (++$this->depth > $this->maxDepth) {
532	3		$this->throwSyntaxError('Maximum stack depth exceeded');
533			}
534
535	90		$this->nextOrFail('{');
536	90		$this->white();
537	90		while ($this->ch) {
538	90		if ($this->ch === '}') {
539	21		$this->nextOrFail('}');
540	21		$this->depth--;
541	21		return $object; // Potentially empty object
542			}
543
544			// Keys can be unquoted. If they are, they need to be
545			// valid JS identifiers.
546	78		if ($this->ch === '"' \|\| $this->ch === "'") {
547	39		$key = $this->string();
548	26		} else {
549	42		$key = $this->identifier();
550			}
551
552	69		$this->white();
553	69		$this->nextOrFail(':');
554	66		if ($this->associative) {
555	45		$object[$key] = $this->value();
556	30		} else {
557	66		$object->{$key} = $this->value();
558			}
559	63		$this->white();
560			// If there's no comma after this pair, this needs to be
561			// the end of the object.
562	63		if ($this->ch !== ',') {
563	54		$this->nextOrFail('}');
564	48		$this->depth--;
565	48		return $object;
566			}
567	21		$this->nextOrFail(',');
568	21		$this->white();
569	14		}
570			}
571
572			$this->throwSyntaxError('Bad object');
573			}
574
575			/**
576			* Parse a JSON value.
577			*
578			* It could be an object, an array, a string, a number,
579			* or a word.
580			*/
581	396		private function value()
582			{
583	396		$this->white();
584	393		switch ($this->ch) {
585	393		case '{':
586	90		return $this->obj();
587	366		case '[':
588	42		return $this->arr();
589	354		case '"':
590	338		case "'":
591	78		return $this->string();
592	285		case '-':
593	270		case '+':
594	255		case '.':
595	102		return $this->number();
596	124		default:
597	186		return is_numeric($this->ch) ? $this->number() : $this->word();
598	124		}
599			}
600
601	114		private function throwSyntaxError($message)
602			{
603	114		throw new SyntaxError($message, $this->lineNumber, $this->columnNumber);
604			}
605
606	27		private static function renderChar($chr)
607			{
608	27		return $chr === null ? 'EOF' : "'" . $chr . "'";
609			}
610
611			/**
612			* @param string $ch
613			*
614			* @return string\|null
615			*/
616	18		private static function getEscapee($ch)
617			{
618			switch ($ch) {
619			// @codingStandardsIgnoreStart
620	18		case "'": return "'";
621	15		case '"': return '"';
622	15		case '\\': return '\\';
623	15		case '/': return '/';
624	15		case "\n": return '';
625	6		case 'b': return chr(8);
626	6		case 'f': return "\f";
627	6		case 'n': return "\n";
628	6		case 'r': return "\r";
629	6		case 't': return "\t";
630	3		default: return null;
631			// @codingStandardsIgnoreEnd
632	2		}
633			}
634
635			/**
636			* Returns everything from $this->at onwards.
637			*
638			* Utilizes a cache so we don't have to continuously parse through UTF-8
639			* data that was earlier in the string which we don't even care about.
640			*
641			* @return string
642			*/
643	210		private function getRemainder()
644			{
645	210		if ($this->remainderCacheAt === $this->at) {
646	66		return $this->remainderCache;
647			}
648
649	186		$subject = mb_substr($this->remainderCache, $this->at - $this->remainderCacheAt);
650	186		$this->remainderCache = $subject;
651	186		$this->remainderCacheAt = $this->at;
652
653	186		return $subject;
654			}
655			}
656

colinodell / json5

Push — master ( 2bb3fb...2d1db6 )

Json5Decoder::nextOrFail() A

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like