Source for file JSON.php
Documentation is available at JSON.php
/* vim: set expandtab tabstop=4 shiftwidth=4 softtabstop=4: */
* Converts to and from JSON format.
* JSON (JavaScript Object Notation) is a lightweight data-interchange
* format. It is easy for humans to read and write. It is easy for machines
* to parse and generate. It is based on a subset of the JavaScript
* Programming Language, Standard ECMA-262 3rd Edition - December 1999.
* This feature can also be found in Python. JSON is a text format that is
* completely language independent but uses conventions that are familiar
* to programmers of the C-family of languages, including C, C++, C#, Java,
* JavaScript, Perl, TCL, and many others. These properties make JSON an
* ideal data-interchange language.
* This package provides a simple encoder and decoder for JSON notation. It
* is intended for use with client-side Javascript applications that make
* use of HTTPRequest to perform server communication functions - data can
* be encoded into JSON notation for use in a client-side javascript, or
* decoded from incoming Javascript requests. JSON format is native to
* Javascript, and can be directly eval()'ed with no further parsing
* All strings should be in ASCII or UTF-8 format!
* LICENSE: Redistribution and use in source and binary forms, with or
* without modification, are permitted provided that the following
* conditions are met: Redistributions of source code must retain the
* above copyright notice, this list of conditions and the following
* disclaimer. Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following disclaimer
* in the documentation and/or other materials provided with the
* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED
* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN
* NO EVENT SHALL CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
* BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
* OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
* ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
* TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
* USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
* @author Michal Migurski <mike-json@teczno.com>
* @author Matt Knapp <mdknapp[at]gmail[dot]com>
* @author Brett Stimmerman <brettstimmerman[at]gmail[dot]com>
* @copyright 2005 Michal Migurski
* @version CVS: $Id: JSON.php,v 1.31 2006/06/28 05:54:17 migurski Exp $
* @license http://www.opensource.org/licenses/bsd-license.php
* @link http://pear.php.net/pepr/pepr-proposal-show.php?id=198
* Marker constant for Services_JSON::decode(), used to flag stack state
define('SERVICES_JSON_SLICE', 1);
* Marker constant for Services_JSON::decode(), used to flag stack state
define('SERVICES_JSON_IN_STR', 2);
* Marker constant for Services_JSON::decode(), used to flag stack state
define('SERVICES_JSON_IN_ARR', 3);
* Marker constant for Services_JSON::decode(), used to flag stack state
define('SERVICES_JSON_IN_OBJ', 4);
* Marker constant for Services_JSON::decode(), used to flag stack state
define('SERVICES_JSON_IN_CMT', 5);
* Behavior switch for Services_JSON::decode()
define('SERVICES_JSON_LOOSE_TYPE', 16);
* Behavior switch for Services_JSON::decode()
define('SERVICES_JSON_SUPPRESS_ERRORS', 32);
* Converts to and from JSON format.
* // create a new instance of Services_JSON
* $json = new Services_JSON();
* // convert a complexe value to JSON notation, and send it to the browser
* $value = array('foo', 'bar', array(1, 2, 'baz'), array(3, array(4)));
* $output = $json->encode($value);
* // prints: ["foo","bar",[1,2,"baz"],[3,[4]]]
* // accept incoming POST data, assumed to be in JSON notation
* $input = file_get_contents('php://input', 1000000);
* $value = $json->decode($input);
* constructs a new JSON instance
* @param int $use object behavior flags; combine with boolean-OR
* - SERVICES_JSON_LOOSE_TYPE: loose typing.
* "{...}" syntax creates associative arrays
* instead of objects in decode().
* - SERVICES_JSON_SUPPRESS_ERRORS: error suppression.
* Values which can't be encoded (e.g. resources)
* appear as NULL instead of throwing errors.
* By default, a deeply-nested resource will
* bubble up with an error, so all return values
* from encode() should be checked with isError()
* convert a string from one UTF-16 char to one UTF-8 char
* Normally should be handled by mb_convert_encoding, but
* provides a slower PHP-only method for installations
* that lack the multibye string extension.
* @param string $utf16 UTF-16 character
* @return string UTF-8 character
function utf162utf8($utf16)
// oh please oh please oh please oh please oh please
$bytes = (ord($utf16{0}) << 8) | ord($utf16{1});
case ((0x7F & $bytes) == $bytes):
// this case should never be reached, because we are in ASCII range
// see: http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
return chr(0x7F & $bytes);
case (0x07FF & $bytes) == $bytes:
// return a 2-byte UTF-8 character
// see: http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
return chr(0xC0 | (($bytes >> 6) & 0x1F))
. chr(0x80 | ($bytes & 0x3F));
case (0xFFFF & $bytes) == $bytes:
// return a 3-byte UTF-8 character
// see: http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
return chr(0xE0 | (($bytes >> 12) & 0x0F))
. chr(0x80 | (($bytes >> 6) & 0x3F))
. chr(0x80 | ($bytes & 0x3F));
// ignoring UTF-32 for now, sorry
* convert a string from one UTF-8 char to one UTF-16 char
* Normally should be handled by mb_convert_encoding, but
* provides a slower PHP-only method for installations
* that lack the multibye string extension.
* @param string $utf8 UTF-8 character
* @return string UTF-16 character
function utf82utf16($utf8)
// oh please oh please oh please oh please oh please
// this case should never be reached, because we are in ASCII range
// see: http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
// return a UTF-16 character from a 2-byte UTF-8 char
// see: http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
return chr(0x07 & (ord($utf8{0}) >> 2))
. chr((0xC0 & (ord($utf8{0}) << 6))
| (0x3F & ord($utf8{1})));
// return a UTF-16 character from a 3-byte UTF-8 char
// see: http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
return chr((0xF0 & (ord($utf8{0}) << 4))
| (0x0F & (ord($utf8{1}) >> 2)))
. chr((0xC0 & (ord($utf8{1}) << 6))
| (0x7F & ord($utf8{2})));
// ignoring UTF-32 for now, sorry
* encodes an arbitrary variable into JSON format
* @param mixed $var any number, boolean, string, array, or object to be encoded.
* see argument 1 to Services_JSON() above for array-parsing behavior.
* if var is a strng, note that encode() always expects it
* to be in ASCII or UTF-8 format!
* @return mixed JSON string representation of input var or an error if a problem occurs
return $var ? 'true' : 'false';
// STRINGS ARE EXPECTED TO BE IN ASCII OR UTF-8 FORMAT
* Iterate over every character in the string,
* escaping with a slash or encoding to UTF-8 where necessary
for ($c = 0; $c < $strlen_var; ++ $c) {
$ord_var_c = ord($var{$c});
// double quote, slash, slosh
case (($ord_var_c >= 0x20) && ($ord_var_c <= 0x7F)):
// characters U-00000000 - U-0000007F (same as ASCII)
case (($ord_var_c & 0xE0) == 0xC0):
// characters U-00000080 - U-000007FF, mask 110XXXXX
// see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$char = pack('C*', $ord_var_c, ord($var{$c + 1}));
$utf16 = $this->utf82utf16($char);
case (($ord_var_c & 0xF0) == 0xE0):
// characters U-00000800 - U-0000FFFF, mask 1110XXXX
// see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$char = pack('C*', $ord_var_c,
$utf16 = $this->utf82utf16($char);
case (($ord_var_c & 0xF8) == 0xF0):
// characters U-00010000 - U-001FFFFF, mask 11110XXX
// see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$char = pack('C*', $ord_var_c,
$utf16 = $this->utf82utf16($char);
case (($ord_var_c & 0xFC) == 0xF8):
// characters U-00200000 - U-03FFFFFF, mask 111110XX
// see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$char = pack('C*', $ord_var_c,
$utf16 = $this->utf82utf16($char);
case (($ord_var_c & 0xFE) == 0xFC):
// characters U-04000000 - U-7FFFFFFF, mask 1111110X
// see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$char = pack('C*', $ord_var_c,
$utf16 = $this->utf82utf16($char);
* As per JSON spec if any array key is not an integer
* we must treat the the whole array as an object. We
* also try to catch a sparsely populated associative
* array with numeric keys here because some JS engines
* will create an array with empty indexes up to
* max_index which can cause memory issues and because
* the keys, which may be relevant, will be remapped
* As per the ECMA and JSON specification an object may
* have any string as a property. Unfortunately due to
* a hole in the ECMA specification if the key is a
* ECMA reserved word or starts with a digit the
* parameter is only accessible using ECMAScript's
// treat as a JSON object
$properties = array_map(array($this, 'name_value'),
foreach($properties as $property) {
return '{' . join(',', $properties) . '}';
// treat it like a regular array
$elements = array_map(array($this, 'encode'), $var);
foreach($elements as $element) {
return '[' . join(',', $elements) . ']';
$properties = array_map(array($this, 'name_value'),
foreach($properties as $property) {
return '{' . join(',', $properties) . '}';
* array-walking function for use in generating JSON-formatted name-value pairs
* @param string $name name of key to use
* @param mixed $value reference to an array element to be encoded
* @return string JSON-formatted name-value pair, like '"name":value'
function name_value($name, $value)
$encoded_value = $this->encode($value);
* reduce a string by removing leading and trailing comments and whitespace
* @param $str string string value to strip of comments and whitespace
* @return string string value stripped of comments and whitespace
function reduce_string($str)
// eliminate single line comments in '// ...' form
// eliminate multi-line comments in '/* ... */' form, at start of string
// eliminate multi-line comments in '/* ... */' form, at end of string
// eliminate extraneous space
* decodes a JSON string into appropriate variable
* @param string $str JSON-formatted string
* @return mixed number, boolean, string, array, or object
* corresponding to given JSON input string.
* See argument 1 to Services_JSON() above for object-output behavior.
* Note that decode() always returns strings
* in ASCII or UTF-8 format!
$str = $this->reduce_string($str);
// Lookie-loo, it's a number
// This would work on its own, but I'm trying to be
// good about returning integers where appropriate:
// Return float or int, as appropriate
return ((float) $str == (integer) $str)
} elseif (preg_match('/^("|\').*(\1)$/s', $str, $m) && $m[1] == $m[2]) {
// STRINGS RETURNED IN UTF-8 FORMAT
for ($c = 0; $c < $strlen_chrs; ++ $c) {
$substr_chrs_c_2 = substr($chrs, $c, 2);
$ord_chrs_c = ord($chrs{$c});
case $substr_chrs_c_2 == '\b':
case $substr_chrs_c_2 == '\t':
case $substr_chrs_c_2 == '\n':
case $substr_chrs_c_2 == '\f':
case $substr_chrs_c_2 == '\r':
case $substr_chrs_c_2 == '\\"':
case $substr_chrs_c_2 == '\\\'':
case $substr_chrs_c_2 == '\\\\':
case $substr_chrs_c_2 == '\\/':
if (($delim == '"' && $substr_chrs_c_2 != '\\\'') ||
($delim == "'" && $substr_chrs_c_2 != '\\"')) {
// single, escaped unicode character
$utf8 .= $this->utf162utf8($utf16);
case ($ord_chrs_c >= 0x20) && ($ord_chrs_c <= 0x7F):
case ($ord_chrs_c & 0xE0) == 0xC0:
// characters U-00000080 - U-000007FF, mask 110XXXXX
//see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$utf8 .= substr($chrs, $c, 2);
case ($ord_chrs_c & 0xF0) == 0xE0:
// characters U-00000800 - U-0000FFFF, mask 1110XXXX
// see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$utf8 .= substr($chrs, $c, 3);
case ($ord_chrs_c & 0xF8) == 0xF0:
// characters U-00010000 - U-001FFFFF, mask 11110XXX
// see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$utf8 .= substr($chrs, $c, 4);
case ($ord_chrs_c & 0xFC) == 0xF8:
// characters U-00200000 - U-03FFFFFF, mask 111110XX
// see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$utf8 .= substr($chrs, $c, 5);
case ($ord_chrs_c & 0xFE) == 0xFC:
// characters U-04000000 - U-7FFFFFFF, mask 1111110X
// see http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
$utf8 .= substr($chrs, $c, 6);
// array, or object notation
$chrs = $this->reduce_string($chrs);
//print("\nparsing {$chrs}\n");
for ($c = 0; $c <= $strlen_chrs; ++ $c) {
$substr_chrs_c_2 = substr($chrs, $c, 2);
// found a comma that is not inside a string, array, etc.,
// OR we've reached the end of the character list
$slice = substr($chrs, $top['where'], ($c - $top['where']));
//print("Found split at {$c}: ".substr($chrs, $top['where'], (1 + $c - $top['where']))."\n");
// we are in an array, so just push an element onto the stack
// we are in an object, so figure
// out the property name and set an
// element in an associative array,
if (preg_match('/^\s*(["\'].*[^\\\]["\'])\s*:\s*(\S.*),?$/Uis', $slice, $parts)) {
$key = $this->decode($parts[1]);
$val = $this->decode($parts[2]);
} elseif (preg_match('/^\s*(\w+)\s*:\s*(\S.*),?$/Uis', $slice, $parts)) {
// name:value pair, where name is unquoted
$val = $this->decode($parts[2]);
// found a quote, and we are not inside a string
//print("Found start of string at {$c}\n");
} elseif (($chrs{$c} == $top['delim']) &&
// found a quote, we're in a string, and it's not escaped
// we know that it's not escaped becase there is _not_ an
// odd number of backslashes at the end of the string so far
//print("Found end of string at {$c}: ".substr($chrs, $top['where'], (1 + 1 + $c - $top['where']))."\n");
} elseif (($chrs{$c} == '[') &&
// found a left-bracket, and we are in an array, object, or slice
//print("Found start of array at {$c}\n");
// found a right-bracket, and we're in an array
//print("Found end of array at {$c}: ".substr($chrs, $top['where'], (1 + $c - $top['where']))."\n");
} elseif (($chrs{$c} == '{') &&
// found a left-brace, and we are in an array, object, or slice
//print("Found start of object at {$c}\n");
// found a right-brace, and we're in an object
//print("Found end of object at {$c}: ".substr($chrs, $top['where'], (1 + $c - $top['where']))."\n");
} elseif (($substr_chrs_c_2 == '/*') &&
// found a comment start, and we are in an array, object, or slice
//print("Found start of comment at {$c}\n");
// found a comment end, and we're in one now
for ($i = $top['where']; $i <= $c; ++ $i)
//print("Found end of comment at {$c}: ".substr($chrs, $top['where'], (1 + $c - $top['where']))."\n");
* @todo Ultimately, this should just call PEAR::isError()
function isError($data, $code = null)
return PEAR::isError($data, $code);
$mode = null, $options = null, $userinfo = null)
parent::PEAR_Error($message, $code, $mode, $options, $userinfo);
* @todo Ultimately, this class shall be descended from PEAR_Error
$mode = null, $options = null, $userinfo = null)
|