1XS(3) User Contributed Perl Documentation XS(3)
2
6 Cpanel::JSON::XS - cPanel fork of JSON::XS, fast and correct
7 serializing
8
10 use Cpanel::JSON::XS;
11 # exported functions, they croak on error
13 # and expect/generate UTF-8
14 $utf8_encoded_json_text = encode_json $perl_hash_or_arrayref;
16 $perl_hash_or_arrayref = decode_json $utf8_encoded_json_text;
17 # OO-interface
19 $coder = Cpanel::JSON::XS->new->ascii->pretty->allow_nonref;
21 $pretty_printed_unencoded = $coder->encode ($perl_scalar);
22 $perl_scalar = $coder->decode ($unicode_json_text);
23 # Note that 5.6 misses most smart utf8 and encoding functionalities
25 # of newer releases.
26 # Note that L<JSON::MaybeXS> will automatically use Cpanel::JSON::XS
28 # if available, at virtually no speed overhead either, so you should
29 # be able to just:
30 use JSON::MaybeXS;
32 # and do the same things, except that you have a pure-perl fallback now.
34 Note that this module will be replaced by a new JSON::Safe module soon,
36 with the same API just guaranteed safe defaults.
37
39 This module converts Perl data structures to JSON and vice versa. Its
40 primary goal is to be correct and its secondary goal is to be fast. To
41 reach the latter goal it was written in C.
42 As this is the n-th-something JSON module on CPAN, what was the reason
44 to write yet another JSON module? While it seems there are many JSON
45 modules, none of them correctly handle all corner cases, and in most
46 cases their maintainers are unresponsive, gone missing, or not
47 listening to bug reports for other reasons.
48 See below for the cPanel fork.
50 See MAPPING, below, on how Cpanel::JSON::XS maps perl values to JSON
52 values and vice versa.
53 FEATURES
55 · correct Unicode handling
56 This module knows how to handle Unicode with Perl version higher
58 than 5.8.5, documents how and when it does so, and even documents
59 what "correct" means.
60 · round-trip integrity
62 When you serialize a perl data structure using only data types
64 supported by JSON and Perl, the deserialized data structure is
65 identical on the Perl level. (e.g. the string "2.0" doesn't
66 suddenly become "2" just because it looks like a number). There are
67 minor exceptions to this, read the MAPPING section below to learn
68 about those.
69 · strict checking of JSON correctness
71 There is no guessing, no generating of illegal JSON texts by
73 default, and only JSON is accepted as input by default. the latter
74 is a security feature.
75 · fast
77 Compared to other JSON modules and other serializers such as
79 Storable, this module usually compares favourably in terms of
80 speed, too.
81 · simple to use
83 This module has both a simple functional interface as well as an
85 object oriented interface.
86 · reasonably versatile output formats
88 You can choose between the most compact guaranteed-single-line
90 format possible (nice for simple line-based protocols), a pure-
91 ASCII format (for when your transport is not 8-bit clean, still
92 supports the whole Unicode range), or a pretty-printed format (for
93 when you want to read that stuff). Or you can combine those
94 features in whatever way you like.
95 cPanel fork
97 Since the original author MLEHMANN has no public bugtracker, this
98 cPanel fork sits now on github.
99 src repo: <https://github.com/rurban/Cpanel-JSON-XS> original:
101 <http://cvs.schmorp.de/JSON-XS/>
102 RT: <https://github.com/rurban/Cpanel-JSON-XS/issues> or
104 <https://rt.cpan.org/Public/Dist/Display.html?Queue=Cpanel-JSON-XS>
105 Changes to JSON::XS
107 - stricter decode_json() as documented. non-refs are disallowed.
109 added a 2nd optional argument. decode() honors now allow_nonref.
110 - fixed encode of numbers for dual-vars. Different string
112 representations are preserved, but numbers with temporary strings
113 which represent the same number are here treated as numbers, not
114 strings. Cpanel::JSON::XS is a bit slower, but preserves numeric
115 types better.
116 - numbers ending with .0 stray numbers, are not converted to
118 integers. [#63] dual-vars which are represented as number not
119 integer (42+"bar" != 5.8.9) are now encoded as number (=> 42.0)
120 because internally it's now a NOK type. However !!1 which is
121 wrongly encoded in 5.8 as "1"/1.0 is still represented as integer.
122 - different handling of inf/nan. Default now to null, optionally with
124 stringify_infnan() to "inf"/"nan". [#28, #32]
125 - added "binary" extension, non-JSON and non JSON parsable, allows
127 "\xNN" and "\NNN" sequences.
128 - 5.6.2 support; sacrificing some utf8 features (assuming bytes
130 all-over), no multi-byte unicode characters with 5.6.
131 - interop for true/false overloading. JSON::XS, JSON::PP and Mojo::JSON
133 representations for booleans are accepted and JSON::XS accepts
134 Cpanel::JSON::XS booleans [#13, #37]
135 Fixed overloading of booleans. Cpanel::JSON::XS::true stringifies
136 again
137 to "1", not "true", analog to all other JSON modules.
138 - native boolean mapping of yes and no to true and false, as in
140 YAML::XS.
141 In perl "!0" is yes, "!1" is no.
142 The JSON value true maps to 1, false maps to 0. [#39]
143 - support arbitrary stringification with encode, with convert_blessed
145 and allow_blessed.
146 - ithread support. Cpanel::JSON::XS is thread-safe, JSON::XS not
148 - is_bool can be called as method, JSON::XS::is_bool not.
150 - performance optimizations for threaded Perls
152 - relaxed mode, allowing many popular extensions
154 - additional fixes for:
156 - [cpan #88061] AIX atof without USE_LONG_DOUBLE
158 - #10 unshare_hek crash
160 - #7, #29 avoid re-blessing where possible. It fails in JSON::XS for
162 READONLY values, i.e. restricted hashes.
163 - #41 overloading of booleans, use the object not the reference.
165 - #62 -Dusequadmath conversion and no SEGV.
167 - #72 parsing of values followed \0, like 1\0 does fail.
169 - #72 parsing of illegal unicode or non-unicode characters.
171 - #96 locale-insensitive numeric conversion
173 - #154 numeric conversion fixed since 5.22, using the same strtold as perl5.
175 - public maintenance and bugtracker
177 - use ppport.h, sanify XS.xs comment styles, harness C coding style
179 - common::sense is optional. When available it is not used in the
181 published production module, just during development and testing.
182 - extended testsuite, passes all http://seriot.ch/parsing_json.html
184 tests. In fact it is the only know JSON decoder which does so,
185 while also being the fastest.
186 - support many more options and methods from JSON::PP:
188 stringify_infnan, allow_unknown, allow_stringify, allow_barekey,
189 encode_stringify, allow_bignum, allow_singlequote, sort_by
190 (partially), escape_slash, convert_blessed, ... optional
191 decode_json(, allow_nonref) arg.
192 relaxed implements allow_dupkeys.
193 - support all 5 unicode BOM's: UTF-8, UTF-16LE, UTF-16BE, UTF-32LE,
195 UTF-32BE, encoding internally to UTF-8.
196
198 The following convenience methods are provided by this module. They are
199 exported by default:
200 $json_text = encode_json $perl_scalar, [json_type]
202 Converts the given Perl data structure to a UTF-8 encoded, binary
203 string (that is, the string contains octets only). Croaks on error.
204 This function call is functionally identical to:
206 $json_text = Cpanel::JSON::XS->new->utf8->encode ($perl_scalar, $json_type)
208 Except being faster.
210 For the type argument see Cpanel::JSON::XS::Type.
212 $perl_scalar = decode_json $json_text [, $allow_nonref [, my $json_type
214 ] ]
215 The opposite of "encode_json": expects an UTF-8 (binary) string of
216 an json reference and tries to parse that as an UTF-8 encoded JSON
217 text, returning the resulting reference. Croaks on error.
218 This function call is functionally identical to:
220 $perl_scalar = Cpanel::JSON::XS->new->utf8->decode ($json_text, $json_type)
222 except being faster.
224 Note that older decode_json versions in Cpanel::JSON::XS older than
226 3.0116 and JSON::XS did not set allow_nonref but allowed them due
227 to a bug in the decoder.
228 If the new optional $allow_nonref argument is set and not false,
230 the allow_nonref option will be set and the function will act is
231 described as in the relaxed RFC 7159 allowing all values such as
232 objects, arrays, strings, numbers, "null", "true", and "false".
233 For the type argument see Cpanel::JSON::XS::Type.
235 $is_boolean = Cpanel::JSON::XS::is_bool $scalar
237 Returns true if the passed scalar represents either
238 "JSON::XS::true" or "JSON::XS::false", two constants that act like
239 1 and 0, respectively and are used to represent JSON "true" and
240 "false" values in Perl.
241 See MAPPING, below, for more information on how JSON values are
243 mapped to Perl.
244
246 from_json
247 from_json has been renamed to decode_json
248 to_json
250 to_json has been renamed to encode_json
251
253 Since this often leads to confusion, here are a few very clear words on
254 how Unicode works in Perl, modulo bugs.
255 1. Perl strings can store characters with ordinal values > 255.
257 This enables you to store Unicode characters as single characters
258 in a Perl string - very natural.
259 2. Perl does not associate an encoding with your strings.
261 ... until you force it to, e.g. when matching it against a regex,
262 or printing the scalar to a file, in which case Perl either
263 interprets your string as locale-encoded text, octets/binary, or as
264 Unicode, depending on various settings. In no case is an encoding
265 stored together with your data, it is use that decides encoding,
266 not any magical meta data.
267 3. The internal utf-8 flag has no meaning with regards to the encoding
269 of your string.
270 4. A "Unicode String" is simply a string where each character can be
271 validly interpreted as a Unicode code point.
272 If you have UTF-8 encoded data, it is no longer a Unicode string,
273 but a Unicode string encoded in UTF-8, giving you a binary string.
274 5. A string containing "high" (> 255) character values is not a UTF-8
276 string.
277 6. Unicode noncharacters only warn, as in core.
278 The 66 Unicode noncharacters U+FDD0..U+FDEF, and U+*FFFE, U+*FFFF
279 just warn, see <http://www.unicode.org/versions/corrigendum9.html>.
280 But illegal surrogate pairs fail to parse.
281 7. Raw non-Unicode characters above U+10FFFF are disallowed.
283 Raw non-Unicode characters outside the valid unicode range fail to
284 parse, because "A string is a sequence of zero or more Unicode
285 characters" RFC 7159 section 1 and "JSON text SHALL be encoded in
286 Unicode RFC 7159 section 8.1. We use now the UTF8_DISALLOW_SUPER
287 flag when parsing unicode.
288 I hope this helps :)
290
292 The object oriented interface lets you configure your own encoding or
293 decoding style, within the limits of supported formats.
294 $json = new Cpanel::JSON::XS
296 Creates a new JSON object that can be used to de/encode JSON
297 strings. All boolean flags described below are by default disabled.
298 The mutators for flags all return the JSON object again and thus
300 calls can be chained:
301 my $json = Cpanel::JSON::XS->new->utf8->space_after->encode ({a => [1,2]})
303 => {"a": [1, 2]}
304 $json = $json->ascii ([$enable])
306 $enabled = $json->get_ascii
307 If $enable is true (or missing), then the "encode" method will not
308 generate characters outside the code range 0..127 (which is ASCII).
309 Any Unicode characters outside that range will be escaped using
310 either a single "\uXXXX" (BMP characters) or a double
311 "\uHHHH\uLLLLL" escape sequence, as per RFC4627. The resulting
312 encoded JSON text can be treated as a native Unicode string, an
313 ascii-encoded, latin1-encoded or UTF-8 encoded string, or any other
314 superset of ASCII.
315 If $enable is false, then the "encode" method will not escape
317 Unicode characters unless required by the JSON syntax or other
318 flags. This results in a faster and more compact format.
319 See also the section ENCODING/CODESET FLAG NOTES later in this
321 document.
322 The main use for this flag is to produce JSON texts that can be
324 transmitted over a 7-bit channel, as the encoded JSON texts will
325 not contain any 8 bit characters.
326 Cpanel::JSON::XS->new->ascii (1)->encode ([chr 0x10401])
328 => ["\ud801\udc01"]
329 $json = $json->latin1 ([$enable])
331 $enabled = $json->get_latin1
332 If $enable is true (or missing), then the "encode" method will
333 encode the resulting JSON text as latin1 (or ISO-8859-1), escaping
334 any characters outside the code range 0..255. The resulting string
335 can be treated as a latin1-encoded JSON text or a native Unicode
336 string. The "decode" method will not be affected in any way by this
337 flag, as "decode" by default expects Unicode, which is a strict
338 superset of latin1.
339 If $enable is false, then the "encode" method will not escape
341 Unicode characters unless required by the JSON syntax or other
342 flags.
343 See also the section ENCODING/CODESET FLAG NOTES later in this
345 document.
346 The main use for this flag is efficiently encoding binary data as
348 JSON text, as most octets will not be escaped, resulting in a
349 smaller encoded size. The disadvantage is that the resulting JSON
350 text is encoded in latin1 (and must correctly be treated as such
351 when storing and transferring), a rare encoding for JSON. It is
352 therefore most useful when you want to store data structures known
353 to contain binary data efficiently in files or databases, not when
354 talking to other JSON encoders/decoders.
355 Cpanel::JSON::XS->new->latin1->encode (["\x{89}\x{abc}"]
357 => ["\x{89}\\u0abc"] # (perl syntax, U+abc escaped, U+89 not)
358 $json = $json->binary ([$enable])
360 $enabled = $json = $json->get_binary
361 If the $enable argument is true (or missing), then the "encode"
362 method will not try to detect an UTF-8 encoding in any JSON string,
363 it will strictly interpret it as byte sequence. The result might
364 contain new "\xNN" sequences, which is unparsable JSON. The
365 "decode" method forbids "\uNNNN" sequences and accepts "\xNN" and
366 octal "\NNN" sequences.
367 There is also a special logic for perl 5.6 and utf8. 5.6 encodes
369 any string to utf-8 automatically when seeing a codepoint >= 0x80
370 and < 0x100. With the binary flag enabled decode the perl utf8
371 encoded string to the original byte encoding and encode this with
372 "\xNN" escapes. This will result to the same encodings as with
373 newer perls. But note that binary multi-byte codepoints with 5.6
374 will result in "illegal unicode character in binary string" errors,
375 unlike with newer perls.
376 If $enable is false, then the "encode" method will smartly try to
378 detect Unicode characters unless required by the JSON syntax or
379 other flags and hex and octal sequences are forbidden.
380 See also the section ENCODING/CODESET FLAG NOTES later in this
382 document.
383 The main use for this flag is to avoid the smart unicode detection
385 and possible double encoding. The disadvantage is that the
386 resulting JSON text is encoded in new "\xNN" and in latin1
387 characters and must correctly be treated as such when storing and
388 transferring, a rare encoding for JSON. It will produce non-
389 readable JSON strings in the browser. It is therefore most useful
390 when you want to store data structures known to contain binary data
391 efficiently in files or databases, not when talking to other JSON
392 encoders/decoders. The binary decoding method can also be used
393 when an encoder produced a non-JSON conformant hex or octal
394 encoding "\xNN" or "\NNN".
395 Cpanel::JSON::XS->new->binary->encode (["\x{89}\x{abc}"])
397 5.6: Error: malformed or illegal unicode character in binary string
398 >=5.8: ['\x89\xe0\xaa\xbc']
399 Cpanel::JSON::XS->new->binary->encode (["\x{89}\x{bc}"])
401 => ["\x89\xbc"]
402 Cpanel::JSON::XS->new->binary->decode (["\x89\ua001"])
404 Error: malformed or illegal unicode character in binary string
405 Cpanel::JSON::XS->new->decode (["\x89"])
407 Error: illegal hex character in non-binary string
408 $json = $json->utf8 ([$enable])
410 $enabled = $json->get_utf8
411 If $enable is true (or missing), then the "encode" method will
412 encode the JSON result into UTF-8, as required by many protocols,
413 while the "decode" method expects to be handled an UTF-8-encoded
414 string. Please note that UTF-8-encoded strings do not contain any
415 characters outside the range 0..255, they are thus useful for
416 bytewise/binary I/O. In future versions, enabling this option might
417 enable autodetection of the UTF-16 and UTF-32 encoding families, as
418 described in RFC4627.
419 If $enable is false, then the "encode" method will return the JSON
421 string as a (non-encoded) Unicode string, while "decode" expects
422 thus a Unicode string. Any decoding or encoding (e.g. to UTF-8 or
423 UTF-16) needs to be done yourself, e.g. using the Encode module.
424 See also the section ENCODING/CODESET FLAG NOTES later in this
426 document.
427 Example, output UTF-16BE-encoded JSON:
429 use Encode;
431 $jsontext = encode "UTF-16BE", Cpanel::JSON::XS->new->encode ($object);
432 Example, decode UTF-32LE-encoded JSON:
434 use Encode;
436 $object = Cpanel::JSON::XS->new->decode (decode "UTF-32LE", $jsontext);
437 $json = $json->pretty ([$enable])
439 This enables (or disables) all of the "indent", "space_before" and
440 "space_after" (and in the future possibly more) flags in one call
441 to generate the most readable (or most compact) form possible.
442 Example, pretty-print some simple structure:
444 my $json = Cpanel::JSON::XS->new->pretty(1)->encode ({a => [1,2]})
446 =>
447 {
448 "a" : [
449 1,
450 2
451 ]
452 }
453 $json = $json->indent ([$enable])
455 $enabled = $json->get_indent
456 If $enable is true (or missing), then the "encode" method will use
457 a multiline format as output, putting every array member or
458 object/hash key-value pair into its own line, indenting them
459 properly.
460 If $enable is false, no newlines or indenting will be produced, and
462 the resulting JSON text is guaranteed not to contain any
463 "newlines".
464 This setting has no effect when decoding JSON texts.
466 $json = $json->indent_length([$number_of_spaces])
468 $length = $json->get_indent_length()
469 Set the indent length (default 3). This option is only useful when
470 you also enable indent or pretty. The acceptable range is from 0
471 (no indentation) to 15
472 $json = $json->space_before ([$enable])
474 $enabled = $json->get_space_before
475 If $enable is true (or missing), then the "encode" method will add
476 an extra optional space before the ":" separating keys from values
477 in JSON objects.
478 If $enable is false, then the "encode" method will not add any
480 extra space at those places.
481 This setting has no effect when decoding JSON texts. You will also
483 most likely combine this setting with "space_after".
484 Example, space_before enabled, space_after and indent disabled:
486 {"key" :"value"}
488 $json = $json->space_after ([$enable])
490 $enabled = $json->get_space_after
491 If $enable is true (or missing), then the "encode" method will add
492 an extra optional space after the ":" separating keys from values
493 in JSON objects and extra whitespace after the "," separating key-
494 value pairs and array members.
495 If $enable is false, then the "encode" method will not add any
497 extra space at those places.
498 This setting has no effect when decoding JSON texts.
500 Example, space_before and indent disabled, space_after enabled:
502 {"key": "value"}
504 $json = $json->relaxed ([$enable])
506 $enabled = $json->get_relaxed
507 If $enable is true (or missing), then "decode" will accept some
508 extensions to normal JSON syntax (see below). "encode" will not be
509 affected in anyway. Be aware that this option makes you accept
510 invalid JSON texts as if they were valid!. I suggest only to use
511 this option to parse application-specific files written by humans
512 (configuration files, resource files etc.)
513 If $enable is false (the default), then "decode" will only accept
515 valid JSON texts.
516 Currently accepted extensions are:
518 · list items can have an end-comma
520 JSON separates array elements and key-value pairs with commas.
522 This can be annoying if you write JSON texts manually and want
523 to be able to quickly append elements, so this extension
524 accepts comma at the end of such items not just between them:
525 [
527 1,
528 2, <- this comma not normally allowed
529 ]
530 {
531 "k1": "v1",
532 "k2": "v2", <- this comma not normally allowed
533 }
534 · shell-style '#'-comments
536 Whenever JSON allows whitespace, shell-style comments are
538 additionally allowed. They are terminated by the first
539 carriage-return or line-feed character, after which more white-
540 space and comments are allowed.
541 [
543 1, # this comment not allowed in JSON
544 # neither this one...
545 ]
546 · literal ASCII TAB characters in strings
548 Literal ASCII TAB characters are now allowed in strings (and
550 treated as "\t") in relaxed mode. Despite JSON mandates, that
551 TAB character is substituted for "\t" sequence.
552 [
554 "Hello\tWorld",
555 "Hello<TAB>World", # literal <TAB> would not normally be allowed
556 ]
557 · allow_singlequote
559 Single quotes are accepted instead of double quotes. See the
561 "allow_singlequote" option.
562 { "foo":'bar' }
564 { 'foo':"bar" }
565 { 'foo':'bar' }
566 · allow_barekey
568 Accept unquoted object keys instead of with mandatory double
570 quotes. See the "allow_barekey" option.
571 { foo:"bar" }
573 · allow_dupkeys
575 Allow decoding of duplicate keys in hashes. By default
577 duplicate keys are forbidden. See
578 <http://seriot.ch/parsing_json.php#24>: RFC 7159 section 4:
579 "The names within an object should be unique." See the
580 "allow_dupkeys" option.
581 $json = $json->canonical ([$enable])
583 $enabled = $json->get_canonical
584 If $enable is true (or missing), then the "encode" method will
585 output JSON objects by sorting their keys. This is adding a
586 comparatively high overhead.
587 If $enable is false, then the "encode" method will output key-value
589 pairs in the order Perl stores them (which will likely change
590 between runs of the same script, and can change even within the
591 same run from 5.18 onwards).
592 This option is useful if you want the same data structure to be
594 encoded as the same JSON text (given the same overall settings). If
595 it is disabled, the same hash might be encoded differently even if
596 contains the same data, as key-value pairs have no inherent
597 ordering in Perl.
598 This setting has no effect when decoding JSON texts.
600 This setting has currently no effect on tied hashes.
602 $json = $json->sort_by (undef, 0, 1 or a block)
604 This currently only (un)sets the "canonical" option, and ignores
605 custom sort blocks.
606 This setting has no effect when decoding JSON texts.
608 This setting has currently no effect on tied hashes.
610 $json = $json->escape_slash ([$enable])
612 $enabled = $json->get_escape_slash
613 According to the JSON Grammar, the forward slash character (U+002F)
614 "/" need to be escaped. But by default strings are encoded without
615 escaping slashes in all perl JSON encoders.
616 If $enable is true (or missing), then "encode" will escape slashes,
618 "\/".
619 This setting has no effect when decoding JSON texts.
621 $json = $json->unblessed_bool ([$enable])
623 $enabled = $json->get_unblessed_bool
624 $json = $json->unblessed_bool([$enable])
625 If $enable is true (or missing), then "decode" will return Perl
627 non-object boolean variables (1 and 0) for JSON booleans ("true"
628 and "false"). If $enable is false, then "decode" will return
629 "Cpanel::JSON::XS::Boolean" objects for JSON booleans.
630 $json = $json->allow_singlequote ([$enable])
632 $enabled = $json->get_allow_singlequote
633 $json = $json->allow_singlequote([$enable])
634 If $enable is true (or missing), then "decode" will accept JSON
636 strings quoted by single quotations that are invalid JSON format.
637 $json->allow_singlequote->decode({"foo":'bar'});
639 $json->allow_singlequote->decode({'foo':"bar"});
640 $json->allow_singlequote->decode({'foo':'bar'});
641 This is also enabled with "relaxed". As same as the "relaxed"
643 option, this option may be used to parse application-specific files
644 written by humans.
645 $json = $json->allow_barekey ([$enable])
647 $enabled = $json->get_allow_barekey
648 $json = $json->allow_barekey([$enable])
649 If $enable is true (or missing), then "decode" will accept bare
651 keys of JSON object that are invalid JSON format.
652 Same as with the "relaxed" option, this option may be used to parse
654 application-specific files written by humans.
655 $json->allow_barekey->decode('{foo:"bar"}');
657 $json = $json->allow_bignum ([$enable])
659 $enabled = $json->get_allow_bignum
660 $json = $json->allow_bignum([$enable])
661 If $enable is true (or missing), then "decode" will convert the big
663 integer Perl cannot handle as integer into a Math::BigInt object
664 and convert a floating number (any) into a Math::BigFloat.
665 On the contrary, "encode" converts "Math::BigInt" objects and
667 "Math::BigFloat" objects into JSON numbers with "allow_blessed"
668 enable.
669 $json->allow_nonref->allow_blessed->allow_bignum;
671 $bigfloat = $json->decode('2.000000000000000000000000001');
672 print $json->encode($bigfloat);
673 # => 2.000000000000000000000000001
674 See "MAPPING" about the normal conversion of JSON number.
676 $json = $json->allow_bigint ([$enable])
678 This option is obsolete and replaced by allow_bignum.
679 $json = $json->allow_nonref ([$enable])
681 $enabled = $json->get_allow_nonref
682 If $enable is true (or missing), then the "encode" method can
683 convert a non-reference into its corresponding string, number or
684 null JSON value, which is an extension to RFC4627. Likewise,
685 "decode" will accept those JSON values instead of croaking.
686 If $enable is false, then the "encode" method will croak if it
688 isn't passed an arrayref or hashref, as JSON texts must either be
689 an object or array. Likewise, "decode" will croak if given
690 something that is not a JSON object or array.
691 Example, encode a Perl scalar as JSON value with enabled
693 "allow_nonref", resulting in an invalid JSON text:
694 Cpanel::JSON::XS->new->allow_nonref->encode ("Hello, World!")
696 => "Hello, World!"
697 $json = $json->allow_unknown ([$enable])
699 $enabled = $json->get_allow_unknown
700 If $enable is true (or missing), then "encode" will not throw an
701 exception when it encounters values it cannot represent in JSON
702 (for example, filehandles) but instead will encode a JSON "null"
703 value. Note that blessed objects are not included here and are
704 handled separately by c<allow_nonref>.
705 If $enable is false (the default), then "encode" will throw an
707 exception when it encounters anything it cannot encode as JSON.
708 This option does not affect "decode" in any way, and it is
710 recommended to leave it off unless you know your communications
711 partner.
712 $json = $json->allow_stringify ([$enable])
714 $enabled = $json->get_allow_stringify
715 If $enable is true (or missing), then "encode" will stringify the
716 non-object perl value or reference. Note that blessed objects are
717 not included here and are handled separately by "allow_blessed" and
718 "convert_blessed". String references are stringified to the string
719 value, other references as in perl.
720 This option does not affect "decode" in any way.
722 This option is special to this module, it is not supported by other
724 encoders. So it is not recommended to use it.
725 $json = $json->require_types ([$enable])
727 $enable = $json->get_require_types
728 $json = $json->require_types([$enable])
729 If $enable is true (or missing), then "encode" will require either
731 enabled "type_all_string" or second argument with supplied JSON
732 types. See Cpanel::JSON::XS::Type. When "type_all_string" is not
733 enabled or second argument is not provided (or is undef), then
734 "encode" croaks. It also croaks when the type for provided
735 structure in "encode" is incomplete.
736 $json = $json->type_all_string ([$enable])
738 $enable = $json->get_type_all_string
739 $json = $json->type_all_string([$enable])
740 If $enable is true (or missing), then "encode" will always produce
742 stable deterministic JSON string types in resulted output.
743 When $enable is false, then result of encoded JSON output may be
745 different for different Perl versions and may depends on loaded
746 modules.
747 This is useful it you need deterministic JSON types, independently
749 of used Perl version and other modules, but do not want to write
750 complicated type definitions for Cpanel::JSON::XS::Type.
751 $json = $json->allow_dupkeys ([$enable])
753 $enabled = $json->get_allow_dupkeys
754 If $enable is true (or missing), then the "decode" method will not
755 die when it encounters duplicate keys in a hash. "allow_dupkeys"
756 is also enabled in the "relaxed" mode.
757 The JSON spec allows duplicate name in objects but recommends to
759 disable it, however with Perl hashes they are impossible, parsing
760 JSON in Perl silently ignores duplicate names, using the last value
761 found.
762 See <http://seriot.ch/parsing_json.php#24>: RFC 7159 section 4:
764 "The names within an object should be unique."
765 $json = $json->allow_blessed ([$enable])
767 $enabled = $json->get_allow_blessed
768 If $enable is true (or missing), then the "encode" method will not
769 barf when it encounters a blessed reference. Instead, the value of
770 the convert_blessed option will decide whether "null"
771 ("convert_blessed" disabled or no "TO_JSON" method found) or a
772 representation of the object ("convert_blessed" enabled and
773 "TO_JSON" method found) is being encoded. Has no effect on
774 "decode".
775 If $enable is false (the default), then "encode" will throw an
777 exception when it encounters a blessed object.
778 This setting has no effect on "decode".
780 $json = $json->convert_blessed ([$enable])
782 $enabled = $json->get_convert_blessed
783 If $enable is true (or missing), then "encode", upon encountering a
784 blessed object, will check for the availability of the "TO_JSON"
785 method on the object's class. If found, it will be called in scalar
786 context and the resulting scalar will be encoded instead of the
787 object. If no "TO_JSON" method is found, a stringification overload
788 method is tried next. If both are not found, the value of
789 "allow_blessed" will decide what to do.
790 The "TO_JSON" method may safely call die if it wants. If "TO_JSON"
792 returns other blessed objects, those will be handled in the same
793 way. "TO_JSON" must take care of not causing an endless recursion
794 cycle (== crash) in this case. The same care must be taken with
795 calling encode in stringify overloads (even if this works by luck
796 in older perls) or other callbacks. The name of "TO_JSON" was
797 chosen because other methods called by the Perl core (== not by the
798 user of the object) are usually in upper case letters and to avoid
799 collisions with any "to_json" function or method.
800 If $enable is false (the default), then "encode" will not consider
802 this type of conversion.
803 This setting has no effect on "decode".
805 $json = $json->allow_tags ([$enable])
807 $enabled = $json->get_allow_tags
808 See "OBJECT SERIALIZATION" for details.
809 If $enable is true (or missing), then "encode", upon encountering a
811 blessed object, will check for the availability of the "FREEZE"
812 method on the object's class. If found, it will be used to
813 serialize the object into a nonstandard tagged JSON value (that
814 JSON decoders cannot decode).
815 It also causes "decode" to parse such tagged JSON values and
817 deserialize them via a call to the "THAW" method.
818 If $enable is false (the default), then "encode" will not consider
820 this type of conversion, and tagged JSON values will cause a parse
821 error in "decode", as if tags were not part of the grammar.
822 $json = $json->filter_json_object ([$coderef->($hashref)])
824 When $coderef is specified, it will be called from "decode" each
825 time it decodes a JSON object. The only argument is a reference to
826 the newly-created hash. If the code references returns a single
827 scalar (which need not be a reference), this value (i.e. a copy of
828 that scalar to avoid aliasing) is inserted into the deserialized
829 data structure. If it returns an empty list (NOTE: not "undef",
830 which is a valid scalar), the original deserialized hash will be
831 inserted. This setting can slow down decoding considerably.
832 When $coderef is omitted or undefined, any existing callback will
834 be removed and "decode" will not change the deserialized hash in
835 any way.
836 Example, convert all JSON objects into the integer 5:
838 my $js = Cpanel::JSON::XS->new->filter_json_object (sub { 5 });
840 # returns [5]
841 $js->decode ('[{}]')
842 # throw an exception because allow_nonref is not enabled
843 # so a lone 5 is not allowed.
844 $js->decode ('{"a":1, "b":2}');
845 $json = $json->filter_json_single_key_object ($key [=>
847 $coderef->($value)])
848 Works remotely similar to "filter_json_object", but is only called
849 for JSON objects having a single key named $key.
850 This $coderef is called before the one specified via
852 "filter_json_object", if any. It gets passed the single value in
853 the JSON object. If it returns a single value, it will be inserted
854 into the data structure. If it returns nothing (not even "undef"
855 but the empty list), the callback from "filter_json_object" will be
856 called next, as if no single-key callback were specified.
857 If $coderef is omitted or undefined, the corresponding callback
859 will be disabled. There can only ever be one callback for a given
860 key.
861 As this callback gets called less often then the
863 "filter_json_object" one, decoding speed will not usually suffer as
864 much. Therefore, single-key objects make excellent targets to
865 serialize Perl objects into, especially as single-key JSON objects
866 are as close to the type-tagged value concept as JSON gets (it's
867 basically an ID/VALUE tuple). Of course, JSON does not support this
868 in any way, so you need to make sure your data never looks like a
869 serialized Perl hash.
870 Typical names for the single object key are "__class_whatever__",
872 or "$__dollars_are_rarely_used__$" or "}ugly_brace_placement", or
873 even things like "__class_md5sum(classname)__", to reduce the risk
874 of clashing with real hashes.
875 Example, decode JSON objects of the form "{ "__widget__" => <id> }"
877 into the corresponding $WIDGET{<id>} object:
878 # return whatever is in $WIDGET{5}:
880 Cpanel::JSON::XS
881 ->new
882 ->filter_json_single_key_object (__widget__ => sub {
883 $WIDGET{ $_[0] }
884 })
885 ->decode ('{"__widget__": 5')
886 # this can be used with a TO_JSON method in some "widget" class
888 # for serialization to json:
889 sub WidgetBase::TO_JSON {
890 my ($self) = @_;
891 unless ($self->{id}) {
893 $self->{id} = ..get..some..id..;
894 $WIDGET{$self->{id}} = $self;
895 }
896 { __widget__ => $self->{id} }
898 }
899 $json = $json->shrink ([$enable])
901 $enabled = $json->get_shrink
902 Perl usually over-allocates memory a bit when allocating space for
903 strings. This flag optionally resizes strings generated by either
904 "encode" or "decode" to their minimum size possible. This can save
905 memory when your JSON texts are either very very long or you have
906 many short strings. It will also try to downgrade any strings to
907 octet-form if possible: perl stores strings internally either in an
908 encoding called UTF-X or in octet-form. The latter cannot store
909 everything but uses less space in general (and some buggy Perl or C
910 code might even rely on that internal representation being used).
911 The actual definition of what shrink does might change in future
913 versions, but it will always try to save space at the expense of
914 time.
915 If $enable is true (or missing), the string returned by "encode"
917 will be shrunk-to-fit, while all strings generated by "decode" will
918 also be shrunk-to-fit.
919 If $enable is false, then the normal perl allocation algorithms are
921 used. If you work with your data, then this is likely to be
922 faster.
923 In the future, this setting might control other things, such as
925 converting strings that look like integers or floats into integers
926 or floats internally (there is no difference on the Perl level),
927 saving space.
928 $json = $json->max_depth ([$maximum_nesting_depth])
930 $max_depth = $json->get_max_depth
931 Sets the maximum nesting level (default 512) accepted while
932 encoding or decoding. If a higher nesting level is detected in JSON
933 text or a Perl data structure, then the encoder and decoder will
934 stop and croak at that point.
935 Nesting level is defined by number of hash- or arrayrefs that the
937 encoder needs to traverse to reach a given point or the number of
938 "{" or "[" characters without their matching closing parenthesis
939 crossed to reach a given character in a string.
940 Setting the maximum depth to one disallows any nesting, so that
942 ensures that the object is only a single hash/object or array.
943 If no argument is given, the highest possible setting will be used,
945 which is rarely useful.
946 Note that nesting is implemented by recursion in C. The default
948 value has been chosen to be as large as typical operating systems
949 allow without crashing.
950 See SECURITY CONSIDERATIONS, below, for more info on why this is
952 useful.
953 $json = $json->max_size ([$maximum_string_size])
955 $max_size = $json->get_max_size
956 Set the maximum length a JSON text may have (in bytes) where
957 decoding is being attempted. The default is 0, meaning no limit.
958 When "decode" is called on a string that is longer then this many
959 bytes, it will not attempt to decode the string but throw an
960 exception. This setting has no effect on "encode" (yet).
961 If no argument is given, the limit check will be deactivated (same
963 as when 0 is specified).
964 See "SECURITY CONSIDERATIONS", below, for more info on why this is
966 useful.
967 $json->stringify_infnan ([$infnan_mode = 1])
969 $infnan_mode = $json->get_stringify_infnan
970 Get or set how Cpanel::JSON::XS encodes "inf", "-inf" or "nan" for
971 numeric values. Also qnan, snan or negative nan on some platforms.
972 "null": infnan_mode = 0. Similar to most JSON modules in other
974 languages. Always null.
975 stringified: infnan_mode = 1. As in Mojo::JSON. Platform specific
977 strings. Stringified via sprintf(%g), with double quotes.
978 inf/nan: infnan_mode = 2. As in JSON::XS, and older releases.
980 Passes through platform dependent values, invalid JSON. Stringified
981 via sprintf(%g), but without double quotes.
982 "inf/-inf/nan": infnan_mode = 3. Platform independent inf/nan/-inf
984 strings. No QNAN/SNAN/negative NAN support, unified to "nan". Much
985 easier to detect, but may conflict with valid strings.
986 $json_text = $json->encode ($perl_scalar, $json_type)
988 Converts the given Perl data structure (a simple scalar or a
989 reference to a hash or array) to its JSON representation. Simple
990 scalars will be converted into JSON string or number sequences,
991 while references to arrays become JSON arrays and references to
992 hashes become JSON objects. Undefined Perl values (e.g. "undef")
993 become JSON "null" values. Neither "true" nor "false" values will
994 be generated.
995 For the type argument see Cpanel::JSON::XS::Type.
997 $perl_scalar = $json->decode ($json_text, my $json_type)
999 The opposite of "encode": expects a JSON text and tries to parse
1000 it, returning the resulting simple scalar or reference. Croaks on
1001 error.
1002 JSON numbers and strings become simple Perl scalars. JSON arrays
1004 become Perl arrayrefs and JSON objects become Perl hashrefs. "true"
1005 becomes 1, "false" becomes 0 and "null" becomes "undef".
1006 For the type argument see Cpanel::JSON::XS::Type.
1008 ($perl_scalar, $characters) = $json->decode_prefix ($json_text)
1010 This works like the "decode" method, but instead of raising an
1011 exception when there is trailing garbage after the first JSON
1012 object, it will silently stop parsing there and return the number
1013 of characters consumed so far.
1014 This is useful if your JSON texts are not delimited by an outer
1016 protocol and you need to know where the JSON text ends.
1017 Cpanel::JSON::XS->new->decode_prefix ("[1] the tail")
1019 => ([1], 3)
1020 $json->to_json ($perl_hash_or_arrayref)
1022 Deprecated method for perl 5.8 and newer. Use encode_json instead.
1023 $json->from_json ($utf8_encoded_json_text)
1025 Deprecated method for perl 5.8 and newer. Use decode_json instead.
1026
1028 In some cases, there is the need for incremental parsing of JSON texts.
1029 While this module always has to keep both JSON text and resulting Perl
1030 data structure in memory at one time, it does allow you to parse a JSON
1031 stream incrementally. It does so by accumulating text until it has a
1032 full JSON object, which it then can decode. This process is similar to
1033 using "decode_prefix" to see if a full JSON object is available, but is
1034 much more efficient (and can be implemented with a minimum of method
1035 calls).
1036 Cpanel::JSON::XS will only attempt to parse the JSON text once it is
1038 sure it has enough text to get a decisive result, using a very simple
1039 but truly incremental parser. This means that it sometimes won't stop
1040 as early as the full parser, for example, it doesn't detect mismatched
1041 parentheses. The only thing it guarantees is that it starts decoding as
1042 soon as a syntactically valid JSON text has been seen. This means you
1043 need to set resource limits (e.g. "max_size") to ensure the parser will
1044 stop parsing in the presence if syntax errors.
1045 The following methods implement this incremental parser.
1047 [void, scalar or list context] = $json->incr_parse ([$string])
1049 This is the central parsing function. It can both append new text
1050 and extract objects from the stream accumulated so far (both of
1051 these functions are optional).
1052 If $string is given, then this string is appended to the already
1054 existing JSON fragment stored in the $json object.
1055 After that, if the function is called in void context, it will
1057 simply return without doing anything further. This can be used to
1058 add more text in as many chunks as you want.
1059 If the method is called in scalar context, then it will try to
1061 extract exactly one JSON object. If that is successful, it will
1062 return this object, otherwise it will return "undef". If there is a
1063 parse error, this method will croak just as "decode" would do (one
1064 can then use "incr_skip" to skip the erroneous part). This is the
1065 most common way of using the method.
1066 And finally, in list context, it will try to extract as many
1068 objects from the stream as it can find and return them, or the
1069 empty list otherwise. For this to work, there must be no separators
1070 between the JSON objects or arrays, instead they must be
1071 concatenated back-to-back. If an error occurs, an exception will be
1072 raised as in the scalar context case. Note that in this case, any
1073 previously-parsed JSON texts will be lost.
1074 Example: Parse some JSON arrays/objects in a given string and
1076 return them.
1077 my @objs = Cpanel::JSON::XS->new->incr_parse ("[5][7][1,2]");
1079 $lvalue_string = $json->incr_text (>5.8 only)
1081 This method returns the currently stored JSON fragment as an
1082 lvalue, that is, you can manipulate it. This only works when a
1083 preceding call to "incr_parse" in scalar context successfully
1084 returned an object, and 2. only with Perl >= 5.8
1085 Under all other circumstances you must not call this function (I
1087 mean it. although in simple tests it might actually work, it will
1088 fail under real world conditions). As a special exception, you can
1089 also call this method before having parsed anything.
1090 This function is useful in two cases: a) finding the trailing text
1092 after a JSON object or b) parsing multiple JSON objects separated
1093 by non-JSON text (such as commas).
1094 $json->incr_skip
1096 This will reset the state of the incremental parser and will remove
1097 the parsed text from the input buffer so far. This is useful after
1098 "incr_parse" died, in which case the input buffer and incremental
1099 parser state is left unchanged, to skip the text parsed so far and
1100 to reset the parse state.
1101 The difference to "incr_reset" is that only text until the parse
1103 error occurred is removed.
1104 $json->incr_reset
1106 This completely resets the incremental parser, that is, after this
1107 call, it will be as if the parser had never parsed anything.
1108 This is useful if you want to repeatedly parse JSON objects and
1110 want to ignore any trailing data, which means you have to reset the
1111 parser after each successful decode.
1112 LIMITATIONS
1114 All options that affect decoding are supported, except "allow_nonref".
1115 The reason for this is that it cannot be made to work sensibly: JSON
1116 objects and arrays are self-delimited, i.e. you can concatenate them
1117 back to back and still decode them perfectly. This does not hold true
1118 for JSON numbers, however.
1119 For example, is the string 1 a single JSON number, or is it simply the
1121 start of 12? Or is 12 a single JSON number, or the concatenation of 1
1122 and 2? In neither case you can tell, and this is why Cpanel::JSON::XS
1123 takes the conservative route and disallows this case.
1124 EXAMPLES
1126 Some examples will make all this clearer. First, a simple example that
1127 works similarly to "decode_prefix": We want to decode the JSON object
1128 at the start of a string and identify the portion after the JSON
1129 object:
1130 my $text = "[1,2,3] hello";
1132 my $json = new Cpanel::JSON::XS;
1134 my $obj = $json->incr_parse ($text)
1136 or die "expected JSON object or array at beginning of string";
1137 my $tail = $json->incr_text;
1139 # $tail now contains " hello"
1140 Easy, isn't it?
1142 Now for a more complicated example: Imagine a hypothetical protocol
1144 where you read some requests from a TCP stream, and each request is a
1145 JSON array, without any separation between them (in fact, it is often
1146 useful to use newlines as "separators", as these get interpreted as
1147 whitespace at the start of the JSON text, which makes it possible to
1148 test said protocol with "telnet"...).
1149 Here is how you'd do it (it is trivial to write this in an event-based
1151 manner):
1152 my $json = new Cpanel::JSON::XS;
1154 # read some data from the socket
1156 while (sysread $socket, my $buf, 4096) {
1157 # split and decode as many requests as possible
1159 for my $request ($json->incr_parse ($buf)) {
1160 # act on the $request
1161 }
1162 }
1163 Another complicated example: Assume you have a string with JSON objects
1165 or arrays, all separated by (optional) comma characters (e.g. "[1],[2],
1166 [3]"). To parse them, we have to skip the commas between the JSON
1167 texts, and here is where the lvalue-ness of "incr_text" comes in
1168 useful:
1169 my $text = "[1],[2], [3]";
1171 my $json = new Cpanel::JSON::XS;
1172 # void context, so no parsing done
1174 $json->incr_parse ($text);
1175 # now extract as many objects as possible. note the
1177 # use of scalar context so incr_text can be called.
1178 while (my $obj = $json->incr_parse) {
1179 # do something with $obj
1180 # now skip the optional comma
1182 $json->incr_text =~ s/^ \s* , //x;
1183 }
1184 Now lets go for a very complex example: Assume that you have a gigantic
1186 JSON array-of-objects, many gigabytes in size, and you want to parse
1187 it, but you cannot load it into memory fully (this has actually
1188 happened in the real world :).
1189 Well, you lost, you have to implement your own JSON parser. But
1191 Cpanel::JSON::XS can still help you: You implement a (very simple)
1192 array parser and let JSON decode the array elements, which are all full
1193 JSON objects on their own (this wouldn't work if the array elements
1194 could be JSON numbers, for example):
1195 my $json = new Cpanel::JSON::XS;
1197 # open the monster
1199 open my $fh, "<bigfile.json"
1200 or die "bigfile: $!";
1201 # first parse the initial "["
1203 for (;;) {
1204 sysread $fh, my $buf, 65536
1205 or die "read error: $!";
1206 $json->incr_parse ($buf); # void context, so no parsing
1207 # Exit the loop once we found and removed(!) the initial "[".
1209 # In essence, we are (ab-)using the $json object as a simple scalar
1210 # we append data to.
1211 last if $json->incr_text =~ s/^ \s* \[ //x;
1212 }
1213 # now we have the skipped the initial "[", so continue
1215 # parsing all the elements.
1216 for (;;) {
1217 # in this loop we read data until we got a single JSON object
1218 for (;;) {
1219 if (my $obj = $json->incr_parse) {
1220 # do something with $obj
1221 last;
1222 }
1223 # add more data
1225 sysread $fh, my $buf, 65536
1226 or die "read error: $!";
1227 $json->incr_parse ($buf); # void context, so no parsing
1228 }
1229 # in this loop we read data until we either found and parsed the
1231 # separating "," between elements, or the final "]"
1232 for (;;) {
1233 # first skip whitespace
1234 $json->incr_text =~ s/^\s*//;
1235 # if we find "]", we are done
1237 if ($json->incr_text =~ s/^\]//) {
1238 print "finished.\n";
1239 exit;
1240 }
1241 # if we find ",", we can continue with the next element
1243 if ($json->incr_text =~ s/^,//) {
1244 last;
1245 }
1246 # if we find anything else, we have a parse error!
1248 if (length $json->incr_text) {
1249 die "parse error near ", $json->incr_text;
1250 }
1251 # else add more data
1253 sysread $fh, my $buf, 65536
1254 or die "read error: $!";
1255 $json->incr_parse ($buf); # void context, so no parsing
1256 }
1257 This is a complex example, but most of the complexity comes from the
1259 fact that we are trying to be correct (bear with me if I am wrong, I
1260 never ran the above example :).
1261
1263 Detect all unicode Byte Order Marks on decode. Which are UTF-8,
1264 UTF-16LE, UTF-16BE, UTF-32LE and UTF-32BE.
1265 The BOM encoding is set only for one specific decode call, it does not
1267 change the state of the JSON object.
1268 Warning: With perls older than 5.20 you need load the Encode module
1270 before loading a multibyte BOM, i.e. >= UTF-16. Otherwise an error is
1271 thrown. This is an implementation limitation and might get fixed later.
1272 See <https://tools.ietf.org/html/rfc7159#section-8.1> "JSON text SHALL
1274 be encoded in UTF-8, UTF-16, or UTF-32."
1275 "Implementations MUST NOT add a byte order mark to the beginning of a
1277 JSON text", "implementations (...) MAY ignore the presence of a byte
1278 order mark rather than treating it as an error".
1279 See also <http://www.unicode.org/faq/utf_bom.html#BOM>.
1281 Beware that Cpanel::JSON::XS is currently the only JSON module which
1283 does accept and decode a BOM.
1284 The latest JSON spec
1286 <https://www.greenbytes.de/tech/webdav/rfc8259.html#character.encoding>
1287 forbid the usage of UTF-16 or UTF-32, the character encoding is UTF-8.
1288 Thus in subsequent updates BOM's of UTF-16 or UTF-32 will throw an
1289 error.
1290
1292 This section describes how Cpanel::JSON::XS maps Perl values to JSON
1293 values and vice versa. These mappings are designed to "do the right
1294 thing" in most circumstances automatically, preserving round-tripping
1295 characteristics (what you put in comes out as something equivalent).
1296 For the more enlightened: note that in the following descriptions,
1298 lowercase perl refers to the Perl interpreter, while uppercase Perl
1299 refers to the abstract Perl language itself.
1300 JSON -> PERL
1302 object
1303 A JSON object becomes a reference to a hash in Perl. No ordering of
1304 object keys is preserved (JSON does not preserve object key
1305 ordering itself).
1306 array
1308 A JSON array becomes a reference to an array in Perl.
1309 string
1311 A JSON string becomes a string scalar in Perl - Unicode codepoints
1312 in JSON are represented by the same codepoints in the Perl string,
1313 so no manual decoding is necessary.
1314 number
1316 A JSON number becomes either an integer, numeric (floating point)
1317 or string scalar in perl, depending on its range and any fractional
1318 parts. On the Perl level, there is no difference between those as
1319 Perl handles all the conversion details, but an integer may take
1320 slightly less memory and might represent more values exactly than
1321 floating point numbers.
1322 If the number consists of digits only, Cpanel::JSON::XS will try to
1324 represent it as an integer value. If that fails, it will try to
1325 represent it as a numeric (floating point) value if that is
1326 possible without loss of precision. Otherwise it will preserve the
1327 number as a string value (in which case you lose roundtripping
1328 ability, as the JSON number will be re-encoded to a JSON string).
1329 Numbers containing a fractional or exponential part will always be
1331 represented as numeric (floating point) values, possibly at a loss
1332 of precision (in which case you might lose perfect roundtripping
1333 ability, but the JSON number will still be re-encoded as a JSON
1334 number).
1335 Note that precision is not accuracy - binary floating point values
1337 cannot represent most decimal fractions exactly, and when
1338 converting from and to floating point, "Cpanel::JSON::XS" only
1339 guarantees precision up to but not including the least significant
1340 bit.
1341 true, false
1343 When "unblessed_bool" is set to true, then JSON "true" becomes 1
1344 and JSON "false" becomes 0.
1345 Otherwise these JSON atoms become "Cpanel::JSON::XS::true" and
1347 "Cpanel::JSON::XS::false", respectively. They are
1348 "JSON::PP::Boolean" objects and are overloaded to act almost
1349 exactly like the numbers 1 and 0. You can check whether a scalar is
1350 a JSON boolean by using the "Cpanel::JSON::XS::is_bool" function.
1351 The other round, from perl to JSON, "!0" which is represented as
1353 "yes" becomes "true", and "!1" which is represented as "no" becomes
1354 "false".
1355 Via Cpanel::JSON::XS::Type you can now even force negation in
1357 "encode", without overloading of "!":
1358 my $false = Cpanel::JSON::XS::false;
1360 print($json->encode([!$false], [JSON_TYPE_BOOL]));
1361 => [true]
1362 null
1364 A JSON null atom becomes "undef" in Perl.
1365 shell-style comments ("# text")
1367 As a nonstandard extension to the JSON syntax that is enabled by
1368 the "relaxed" setting, shell-style comments are allowed. They can
1369 start anywhere outside strings and go till the end of the line.
1370 tagged values ("(tag)value").
1372 Another nonstandard extension to the JSON syntax, enabled with the
1373 "allow_tags" setting, are tagged values. In this implementation,
1374 the tag must be a perl package/class name encoded as a JSON string,
1375 and the value must be a JSON array encoding optional constructor
1376 arguments.
1377 See "OBJECT SERIALIZATION", below, for details.
1379 PERL -> JSON
1381 The mapping from Perl to JSON is slightly more difficult, as Perl is a
1382 truly typeless language, so we can only guess which JSON type is meant
1383 by a Perl value.
1384 hash references
1386 Perl hash references become JSON objects. As there is no inherent
1387 ordering in hash keys (or JSON objects), they will usually be
1388 encoded in a pseudo-random order that can change between runs of
1389 the same program but stays generally the same within a single run
1390 of a program. Cpanel::JSON::XS can optionally sort the hash keys
1391 (determined by the canonical flag), so the same datastructure will
1392 serialize to the same JSON text (given same settings and version of
1393 Cpanel::JSON::XS), but this incurs a runtime overhead and is only
1394 rarely useful, e.g. when you want to compare some JSON text against
1395 another for equality.
1396 array references
1398 Perl array references become JSON arrays.
1399 other references
1401 Other unblessed references are generally not allowed and will cause
1402 an exception to be thrown, except for references to the integers 0
1403 and 1, which get turned into "false" and "true" atoms in JSON.
1404 With the option "allow_stringify", you can ignore the exception and
1406 return the stringification of the perl value.
1407 With the option "allow_unknown", you can ignore the exception and
1409 return "null" instead.
1410 encode_json [\"x"] # => cannot encode reference to scalar 'SCALAR(0x..)'
1412 # unless the scalar is 0 or 1
1413 encode_json [\0, \1] # yields [false,true]
1414 allow_stringify->encode_json [\"x"] # yields "x" unlike JSON::PP
1416 allow_unknown->encode_json [\"x"] # yields null as in JSON::PP
1417 Cpanel::JSON::XS::true, Cpanel::JSON::XS::false
1419 These special values become JSON true and JSON false values,
1420 respectively. You can also use "\1" and "\0" or "!0" and "!1"
1421 directly if you want.
1422 encode_json [Cpanel::JSON::XS::false, Cpanel::JSON::XS::true] # yields [false,true]
1424 encode_json [!1, !0], [JSON_TYPE_BOOL, JSON_TYPE_BOOL] # yields [false,true]
1425 eq/ne comparisons with true, false:
1427 false is eq to the empty string or the string 'false' or the
1429 special empty string "!!0" or "!1", i.e. "SV_NO", or the numbers 0
1430 or 0.0.
1431 true is eq to the string 'true' or to the special string "!0" (i.e.
1433 "SV_YES") or to the numbers 1 or 1.0.
1434 blessed objects
1436 Blessed objects are not directly representable in JSON, but
1437 "Cpanel::JSON::XS" allows various optional ways of handling
1438 objects. See "OBJECT SERIALIZATION", below, for details.
1439 See the "allow_blessed" and "convert_blessed" methods on various
1441 options on how to deal with this: basically, you can choose between
1442 throwing an exception, encoding the reference as if it weren't
1443 blessed, use the objects overloaded stringification method or
1444 provide your own serializer method.
1445 simple scalars
1447 Simple Perl scalars (any scalar that is not a reference) are the
1448 most difficult objects to encode: Cpanel::JSON::XS will encode
1449 undefined scalars or inf/nan as JSON "null" values and other
1450 scalars to either number or string in non-deterministic way which
1451 may be affected or changed by Perl version or any other loaded Perl
1452 module.
1453 If you want to have stable and deterministic types in JSON encoder
1455 then use Cpanel::JSON::XS::Type.
1456 Alternative way for deterministic types is to use "type_all_string"
1458 method when all perl scalars are encoded to JSON strings.
1459 Non-deterministic behavior is following: scalars that have last
1461 been used in a string context before encoding as JSON strings, and
1462 anything else as number value:
1463 # dump as number
1465 encode_json [2] # yields [2]
1466 encode_json [-3.0e17] # yields [-3e+17]
1467 my $value = 5; encode_json [$value] # yields [5]
1468 # used as string, but the two representations are for the same number
1470 print $value;
1471 encode_json [$value] # yields [5]
1472 # used as different string (non-matching dual-var)
1474 my $str = '0 but true';
1475 my $num = 1 + $str;
1476 encode_json [$num, $str] # yields [1,"0 but true"]
1477 # undef becomes null
1479 encode_json [undef] # yields [null]
1480 # inf or nan becomes null, unless you answered
1482 # "Do you want to handle inf/nan as strings" with yes
1483 encode_json [9**9**9] # yields [null]
1484 You can force the type to be a JSON string by stringifying it:
1486 my $x = 3.1; # some variable containing a number
1488 "$x"; # stringified
1489 $x .= ""; # another, more awkward way to stringify
1490 print $x; # perl does it for you, too, quite often
1491 You can force the type to be a JSON number by numifying it:
1493 my $x = "3"; # some variable containing a string
1495 $x += 0; # numify it, ensuring it will be dumped as a number
1496 $x *= 1; # same thing, the choice is yours.
1497 Note that numerical precision has the same meaning as under Perl
1499 (so binary to decimal conversion follows the same rules as in Perl,
1500 which can differ to other languages). Also, your perl interpreter
1501 might expose extensions to the floating point numbers of your
1502 platform, such as infinities or NaN's - these cannot be represented
1503 in JSON, and thus null is returned instead. Optionally you can
1504 configure it to stringify inf and nan values.
1505 OBJECT SERIALIZATION
1507 As JSON cannot directly represent Perl objects, you have to choose
1508 between a pure JSON representation (without the ability to deserialize
1509 the object automatically again), and a nonstandard extension to the
1510 JSON syntax, tagged values.
1511 SERIALIZATION
1513 What happens when "Cpanel::JSON::XS" encounters a Perl object depends
1515 on the "allow_blessed", "convert_blessed" and "allow_tags" settings,
1516 which are used in this order:
1517 1. "allow_tags" is enabled and the object has a "FREEZE" method.
1519 In this case, "Cpanel::JSON::XS" uses the Types::Serialiser object
1520 serialization protocol to create a tagged JSON value, using a
1521 nonstandard extension to the JSON syntax.
1522 This works by invoking the "FREEZE" method on the object, with the
1524 first argument being the object to serialize, and the second
1525 argument being the constant string "JSON" to distinguish it from
1526 other serializers.
1527 The "FREEZE" method can return any number of values (i.e. zero or
1529 more). These values and the paclkage/classname of the object will
1530 then be encoded as a tagged JSON value in the following format:
1531 ("classname")[FREEZE return values...]
1533 e.g.:
1535 ("URI")["http://www.google.com/"]
1537 ("MyDate")[2013,10,29]
1538 ("ImageData::JPEG")["Z3...VlCg=="]
1539 For example, the hypothetical "My::Object" "FREEZE" method might
1541 use the objects "type" and "id" members to encode the object:
1542 sub My::Object::FREEZE {
1544 my ($self, $serializer) = @_;
1545 ($self->{type}, $self->{id})
1547 }
1548 2. "convert_blessed" is enabled and the object has a "TO_JSON" method.
1550 In this case, the "TO_JSON" method of the object is invoked in
1551 scalar context. It must return a single scalar that can be directly
1552 encoded into JSON. This scalar replaces the object in the JSON
1553 text.
1554 For example, the following "TO_JSON" method will convert all URI
1556 objects to JSON strings when serialized. The fact that these values
1557 originally were URI objects is lost.
1558 sub URI::TO_JSON {
1560 my ($uri) = @_;
1561 $uri->as_string
1562 }
1563 3. "convert_blessed" is enabled and the object has a stringification
1565 overload.
1566 In this case, the overloaded "" method of the object is invoked in
1567 scalar context. It must return a single scalar that can be directly
1568 encoded into JSON. This scalar replaces the object in the JSON
1569 text.
1570 For example, the following "" method will convert all URI objects
1572 to JSON strings when serialized. The fact that these values
1573 originally were URI objects is lost.
1574 package URI;
1576 use overload '""' => sub { shift->as_string };
1577 4. "allow_blessed" is enabled.
1579 The object will be serialized as a JSON null value.
1580 5. none of the above
1582 If none of the settings are enabled or the respective methods are
1583 missing, "Cpanel::JSON::XS" throws an exception.
1584 DESERIALIZATION
1586 For deserialization there are only two cases to consider: either
1588 nonstandard tagging was used, in which case "allow_tags" decides, or
1589 objects cannot be automatically be deserialized, in which case you can
1590 use postprocessing or the "filter_json_object" or
1591 "filter_json_single_key_object" callbacks to get some real objects our
1592 of your JSON.
1593 This section only considers the tagged value case: I a tagged JSON
1595 object is encountered during decoding and "allow_tags" is disabled, a
1596 parse error will result (as if tagged values were not part of the
1597 grammar).
1598 If "allow_tags" is enabled, "Cpanel::JSON::XS" will look up the "THAW"
1600 method of the package/classname used during serialization (it will not
1601 attempt to load the package as a Perl module). If there is no such
1602 method, the decoding will fail with an error.
1603 Otherwise, the "THAW" method is invoked with the classname as first
1605 argument, the constant string "JSON" as second argument, and all the
1606 values from the JSON array (the values originally returned by the
1607 "FREEZE" method) as remaining arguments.
1608 The method must then return the object. While technically you can
1610 return any Perl scalar, you might have to enable the "enable_nonref"
1611 setting to make that work in all cases, so better return an actual
1612 blessed reference.
1613 As an example, let's implement a "THAW" function that regenerates the
1615 "My::Object" from the "FREEZE" example earlier:
1616 sub My::Object::THAW {
1618 my ($class, $serializer, $type, $id) = @_;
1619 $class->new (type => $type, id => $id)
1621 }
1622 See the "SECURITY CONSIDERATIONS" section below. Allowing external json
1624 objects being deserialized to perl objects is usually a very bad idea.
1625
1627 The interested reader might have seen a number of flags that signify
1628 encodings or codesets - "utf8", "latin1", "binary" and "ascii". There
1629 seems to be some confusion on what these do, so here is a short
1630 comparison:
1631 "utf8" controls whether the JSON text created by "encode" (and expected
1633 by "decode") is UTF-8 encoded or not, while "latin1" and "ascii" only
1634 control whether "encode" escapes character values outside their
1635 respective codeset range. Neither of these flags conflict with each
1636 other, although some combinations make less sense than others.
1637 Care has been taken to make all flags symmetrical with respect to
1639 "encode" and "decode", that is, texts encoded with any combination of
1640 these flag values will be correctly decoded when the same flags are
1641 used - in general, if you use different flag settings while encoding
1642 vs. when decoding you likely have a bug somewhere.
1643 Below comes a verbose discussion of these flags. Note that a "codeset"
1645 is simply an abstract set of character-codepoint pairs, while an
1646 encoding takes those codepoint numbers and encodes them, in our case
1647 into octets. Unicode is (among other things) a codeset, UTF-8 is an
1648 encoding, and ISO-8859-1 (= latin 1) and ASCII are both codesets and
1649 encodings at the same time, which can be confusing.
1650 "utf8" flag disabled
1652 When "utf8" is disabled (the default), then "encode"/"decode"
1653 generate and expect Unicode strings, that is, characters with high
1654 ordinal Unicode values (> 255) will be encoded as such characters,
1655 and likewise such characters are decoded as-is, no changes to them
1656 will be done, except "(re-)interpreting" them as Unicode codepoints
1657 or Unicode characters, respectively (to Perl, these are the same
1658 thing in strings unless you do funny/weird/dumb stuff).
1659 This is useful when you want to do the encoding yourself (e.g. when
1661 you want to have UTF-16 encoded JSON texts) or when some other
1662 layer does the encoding for you (for example, when printing to a
1663 terminal using a filehandle that transparently encodes to UTF-8 you
1664 certainly do NOT want to UTF-8 encode your data first and have Perl
1665 encode it another time).
1666 "utf8" flag enabled
1668 If the "utf8"-flag is enabled, "encode"/"decode" will encode all
1669 characters using the corresponding UTF-8 multi-byte sequence, and
1670 will expect your input strings to be encoded as UTF-8, that is, no
1671 "character" of the input string must have any value > 255, as UTF-8
1672 does not allow that.
1673 The "utf8" flag therefore switches between two modes: disabled
1675 means you will get a Unicode string in Perl, enabled means you get
1676 an UTF-8 encoded octet/binary string in Perl.
1677 "latin1", "binary" or "ascii" flags enabled
1679 With "latin1" (or "ascii") enabled, "encode" will escape characters
1680 with ordinal values > 255 (> 127 with "ascii") and encode the
1681 remaining characters as specified by the "utf8" flag. With
1682 "binary" enabled, ordinal values > 255 are illegal.
1683 If "utf8" is disabled, then the result is also correctly encoded in
1685 those character sets (as both are proper subsets of Unicode,
1686 meaning that a Unicode string with all character values < 256 is
1687 the same thing as a ISO-8859-1 string, and a Unicode string with
1688 all character values < 128 is the same thing as an ASCII string in
1689 Perl).
1690 If "utf8" is enabled, you still get a correct UTF-8-encoded string,
1692 regardless of these flags, just some more characters will be
1693 escaped using "\uXXXX" then before.
1694 Note that ISO-8859-1-encoded strings are not compatible with UTF-8
1696 encoding, while ASCII-encoded strings are. That is because the
1697 ISO-8859-1 encoding is NOT a subset of UTF-8 (despite the
1698 ISO-8859-1 codeset being a subset of Unicode), while ASCII is.
1699 Surprisingly, "decode" will ignore these flags and so treat all
1701 input values as governed by the "utf8" flag. If it is disabled,
1702 this allows you to decode ISO-8859-1- and ASCII-encoded strings, as
1703 both strict subsets of Unicode. If it is enabled, you can correctly
1704 decode UTF-8 encoded strings.
1705 So neither "latin1", "binary" nor "ascii" are incompatible with the
1707 "utf8" flag - they only govern when the JSON output engine escapes
1708 a character or not.
1709 The main use for "latin1" or "binary" is to relatively efficiently
1711 store binary data as JSON, at the expense of breaking compatibility
1712 with most JSON decoders.
1713 The main use for "ascii" is to force the output to not contain
1715 characters with values > 127, which means you can interpret the
1716 resulting string as UTF-8, ISO-8859-1, ASCII, KOI8-R or most about
1717 any character set and 8-bit-encoding, and still get the same data
1718 structure back. This is useful when your channel for JSON transfer
1719 is not 8-bit clean or the encoding might be mangled in between
1720 (e.g. in mail), and works because ASCII is a proper subset of most
1721 8-bit and multibyte encodings in use in the world.
1722 JSON and ECMAscript
1724 JSON syntax is based on how literals are represented in javascript (the
1725 not-standardized predecessor of ECMAscript) which is presumably why it
1726 is called "JavaScript Object Notation".
1727 However, JSON is not a subset (and also not a superset of course) of
1729 ECMAscript (the standard) or javascript (whatever browsers actually
1730 implement).
1731 If you want to use javascript's "eval" function to "parse" JSON, you
1733 might run into parse errors for valid JSON texts, or the resulting data
1734 structure might not be queryable:
1735 One of the problems is that U+2028 and U+2029 are valid characters
1737 inside JSON strings, but are not allowed in ECMAscript string literals,
1738 so the following Perl fragment will not output something that can be
1739 guaranteed to be parsable by javascript's "eval":
1740 use Cpanel::JSON::XS;
1742 print encode_json [chr 0x2028];
1744 The right fix for this is to use a proper JSON parser in your
1746 javascript programs, and not rely on "eval" (see for example Douglas
1747 Crockford's json2.js parser).
1748 If this is not an option, you can, as a stop-gap measure, simply encode
1750 to ASCII-only JSON:
1751 use Cpanel::JSON::XS;
1753 print Cpanel::JSON::XS->new->ascii->encode ([chr 0x2028]);
1755 Note that this will enlarge the resulting JSON text quite a bit if you
1757 have many non-ASCII characters. You might be tempted to run some
1758 regexes to only escape U+2028 and U+2029, e.g.:
1759 # DO NOT USE THIS!
1761 my $json = Cpanel::JSON::XS->new->utf8->encode ([chr 0x2028]);
1762 $json =~ s/\xe2\x80\xa8/\\u2028/g; # escape U+2028
1763 $json =~ s/\xe2\x80\xa9/\\u2029/g; # escape U+2029
1764 print $json;
1765 Note that this is a bad idea: the above only works for U+2028 and
1767 U+2029 and thus only for fully ECMAscript-compliant parsers. Many
1768 existing javascript implementations, however, have issues with other
1769 characters as well - using "eval" naively simply will cause problems.
1770 Another problem is that some javascript implementations reserve some
1772 property names for their own purposes (which probably makes them non-
1773 ECMAscript-compliant). For example, Iceweasel reserves the "__proto__"
1774 property name for its own purposes.
1775 If that is a problem, you could parse try to filter the resulting JSON
1777 output for these property strings, e.g.:
1778 $json =~ s/"__proto__"\s*:/"__proto__renamed":/g;
1780 This works because "__proto__" is not valid outside of strings, so
1782 every occurrence of ""__proto__"\s*:" must be a string used as property
1783 name.
1784 Unicode non-characters between U+FFFD and U+10FFFF are decoded either
1786 to the recommended U+FFFD REPLACEMENT CHARACTER (see Unicode PR #121:
1787 Recommended Practice for Replacement Characters), or in the binary or
1788 relaxed mode left as is, keeping the illegal non-characters as before.
1789 Raw non-Unicode characters outside the valid unicode range fail now to
1791 parse, because "A string is a sequence of zero or more Unicode
1792 characters" RFC 7159 section 1 and "JSON text SHALL be encoded in
1793 Unicode RFC 7159 section 8.1. We use now the UTF8_DISALLOW_SUPER flag
1794 when parsing unicode.
1795 If you know of other incompatibilities, please let me know.
1797 JSON and YAML
1799 You often hear that JSON is a subset of YAML. in general, there is no
1800 way to configure JSON::XS to output a data structure as valid YAML that
1801 works in all cases. If you really must use Cpanel::JSON::XS to
1802 generate YAML, you should use this algorithm (subject to change in
1803 future versions):
1804 my $to_yaml = Cpanel::JSON::XS->new->utf8->space_after (1);
1806 my $yaml = $to_yaml->encode ($ref) . "\n";
1807 This will usually generate JSON texts that also parse as valid YAML.
1809 SPEED
1811 It seems that JSON::XS is surprisingly fast, as shown in the following
1812 tables. They have been generated with the help of the "eg/bench"
1813 program in the JSON::XS distribution, to make it easy to compare on
1814 your own system.
1815 JSON::XS is with Data::MessagePack and Sereal one of the fastest
1817 serializers, because JSON and JSON::XS do not support backrefs (no
1818 graph structures), only trees. Storable supports backrefs, i.e. graphs.
1819 Data::MessagePack encodes its data binary (as Storable) and supports
1820 only very simple subset of JSON.
1821 First comes a comparison between various modules using a very short
1823 single-line JSON string (also available at
1824 <http://dist.schmorp.de/misc/json/short.json>).
1825 {"method": "handleMessage", "params": ["user1",
1827 "we were just talking"], "id": null, "array":[1,11,234,-5,1e5,1e7,
1828 1, 0]}
1829 It shows the number of encodes/decodes per second (JSON::XS uses the
1831 functional interface, while Cpanel::JSON::XS/2 uses the OO interface
1832 with pretty-printing and hash key sorting enabled, Cpanel::JSON::XS/3
1833 enables shrink. JSON::DWIW/DS uses the deserialize function, while
1834 JSON::DWIW::FJ uses the from_json method). Higher is better:
1835 module | encode | decode |
1837 --------------|------------|------------|
1838 JSON::DWIW/DS | 86302.551 | 102300.098 |
1839 JSON::DWIW/FJ | 86302.551 | 75983.768 |
1840 JSON::PP | 15827.562 | 6638.658 |
1841 JSON::Syck | 63358.066 | 47662.545 |
1842 JSON::XS | 511500.488 | 511500.488 |
1843 JSON::XS/2 | 291271.111 | 388361.481 |
1844 JSON::XS/3 | 361577.931 | 361577.931 |
1845 Storable | 66788.280 | 265462.278 |
1846 --------------+------------+------------+
1847 That is, JSON::XS is almost six times faster than JSON::DWIW on
1849 encoding, about five times faster on decoding, and over thirty to
1850 seventy times faster than JSON's pure perl implementation. It also
1851 compares favourably to Storable for small amounts of data.
1852 Using a longer test string (roughly 18KB, generated from Yahoo! Locals
1854 search API (<http://dist.schmorp.de/misc/json/long.json>).
1855 module | encode | decode |
1857 --------------|------------|------------|
1858 JSON::DWIW/DS | 1647.927 | 2673.916 |
1859 JSON::DWIW/FJ | 1630.249 | 2596.128 |
1860 JSON::PP | 400.640 | 62.311 |
1861 JSON::Syck | 1481.040 | 1524.869 |
1862 JSON::XS | 20661.596 | 9541.183 |
1863 JSON::XS/2 | 10683.403 | 9416.938 |
1864 JSON::XS/3 | 20661.596 | 9400.054 |
1865 Storable | 19765.806 | 10000.725 |
1866 --------------+------------+------------+
1867 Again, JSON::XS leads by far (except for Storable which non-
1869 surprisingly decodes a bit faster).
1870 On large strings containing lots of high Unicode characters, some
1872 modules (such as JSON::PC) seem to decode faster than JSON::XS, but the
1873 result will be broken due to missing (or wrong) Unicode handling.
1874 Others refuse to decode or encode properly, so it was impossible to
1875 prepare a fair comparison table for that case.
1876 For updated graphs see
1878 <https://github.com/Sereal/Sereal/wiki/Sereal-Comparison-Graphs>
1879
1881 As long as you only serialize data that can be directly expressed in
1882 JSON, "Cpanel::JSON::XS" is incapable of generating invalid JSON output
1883 (modulo bugs, but "JSON::XS" has found more bugs in the official JSON
1884 testsuite (1) than the official JSON testsuite has found in "JSON::XS"
1885 (0)). "Cpanel::JSON::XS" is currently the only known JSON decoder
1886 which passes all <http://seriot.ch/parsing_json.html> tests, while
1887 being the fastest also.
1888 When you have trouble decoding JSON generated by this module using
1890 other decoders, then it is very likely that you have an encoding
1891 mismatch or the other decoder is broken.
1892 When decoding, "JSON::XS" is strict by default and will likely catch
1894 all errors. There are currently two settings that change this:
1895 "relaxed" makes "JSON::XS" accept (but not generate) some non-standard
1896 extensions, and "allow_tags" or "allow_blessed" will allow you to
1897 encode and decode Perl objects, at the cost of being totally insecure
1898 and not outputting valid JSON anymore.
1899 JSON-XS-3.01 broke interoperability with JSON-2.90 with booleans. See
1901 JSON.
1902 Cpanel::JSON::XS needs to know the JSON and JSON::XS versions to be
1904 able work with those objects, especially when encoding a booleans like
1905 "{"is_true":true}". So you need to load these modules before.
1906 true/false overloading and boolean representations are supported.
1908 JSON::XS and JSON::PP representations are accepted and older JSON::XS
1910 accepts Cpanel::JSON::XS booleans. All JSON modules JSON, JSON, PP,
1911 JSON::XS, Cpanel::JSON::XS produce JSON::PP::Boolean objects, just Mojo
1912 and JSON::YAJL not. Mojo produces Mojo::JSON::_Bool and
1913 JSON::YAJL::Parser just an unblessed IV.
1914 Cpanel::JSON::XS accepts JSON::PP::Boolean and Mojo::JSON::_Bool
1916 objects as booleans.
1917 I cannot think of any reason to still use JSON::XS anymore.
1919 TAGGED VALUE SYNTAX AND STANDARD JSON EN/DECODERS
1921 When you use "allow_tags" to use the extended (and also nonstandard and
1922 invalid) JSON syntax for serialized objects, and you still want to
1923 decode the generated serialize objects, you can run a regex to replace
1924 the tagged syntax by standard JSON arrays (it only works for "normal"
1925 package names without comma, newlines or single colons). First, the
1926 readable Perl version:
1927 # if your FREEZE methods return no values, you need this replace first:
1929 $json =~ s/\( \s* (" (?: [^\\":,]+|\\.|::)* ") \s* \) \s* \[\s*\]/[$1]/gx;
1930 # this works for non-empty constructor arg lists:
1932 $json =~ s/\( \s* (" (?: [^\\":,]+|\\.|::)* ") \s* \) \s* \[/[$1,/gx;
1933 And here is a less readable version that is easy to adapt to other
1935 languages:
1936 $json =~ s/\(\s*("([^\\":,]+|\\.|::)*")\s*\)\s*\[/[$1,/g;
1938 Here is an ECMAScript version (same regex):
1940 json = json.replace (/\(\s*("([^\\":,]+|\\.|::)*")\s*\)\s*\[/g, "[$1,");
1942 Since this syntax converts to standard JSON arrays, it might be hard to
1944 distinguish serialized objects from normal arrays. You can prepend a
1945 "magic number" as first array element to reduce chances of a collision:
1946 $json =~ s/\(\s*("([^\\":,]+|\\.|::)*")\s*\)\s*\[/["XU1peReLzT4ggEllLanBYq4G9VzliwKF",$1,/g;
1948 And after decoding the JSON text, you could walk the data structure
1950 looking for arrays with a first element of
1951 "XU1peReLzT4ggEllLanBYq4G9VzliwKF".
1952 The same approach can be used to create the tagged format with another
1954 encoder. First, you create an array with the magic string as first
1955 member, the classname as second, and constructor arguments last, encode
1956 it as part of your JSON structure, and then:
1957 $json =~ s/\[\s*"XU1peReLzT4ggEllLanBYq4G9VzliwKF"\s*,\s*("([^\\":,]+|\\.|::)*")\s*,/($1)[/g;
1959 Again, this has some limitations - the magic string must not be encoded
1961 with character escapes, and the constructor arguments must be non-
1962 empty.
1963
1965 Since this module was written, Google has written a new JSON RFC, RFC
1966 7159 (and RFC7158). Unfortunately, this RFC breaks compatibility with
1967 both the original JSON specification on www.json.org and RFC4627.
1968 As far as I can see, you can get partial compatibility when parsing by
1970 using "->allow_nonref". However, consider the security implications of
1971 doing so.
1972 I haven't decided yet when to break compatibility with RFC4627 by
1974 default (and potentially leave applications insecure) and change the
1975 default to follow RFC7159, but application authors are well advised to
1976 call "->allow_nonref(0)" even if this is the current default, if they
1977 cannot handle non-reference values, in preparation for the day when the
1978 default will change.
1979
1981 JSON::XS and Cpanel::JSON::XS are not only fast. JSON is generally the
1982 most secure serializing format, because it is the only one besides
1983 Data::MessagePack, which does not deserialize objects per default. For
1984 all languages, not just perl. The binary variant BSON (MongoDB) does
1985 more but is unsafe.
1986 It is trivial for any attacker to create such serialized objects in
1988 JSON and trick perl into expanding them, thereby triggering certain
1989 methods. Watch <https://www.youtube.com/watch?v=Gzx6KlqiIZE> for an
1990 exploit demo for "CVE-2015-1592 SixApart MovableType Storable Perl Code
1991 Execution" for a deserializer which expands objects. Deserializing
1992 even coderefs (methods, functions) or external data would be considered
1993 the most dangerous.
1994 Security relevant overview of serializers regarding deserializing
1996 objects by default:
1997 Objects Coderefs External Data
1999 Data::Dumper YES YES YES
2001 Storable YES NO (def) NO
2002 Sereal YES NO NO
2003 YAML YES NO NO
2004 B::C YES YES YES
2005 B::Bytecode YES YES YES
2006 BSON YES YES NO
2007 JSON::SL YES NO YES
2008 JSON NO (def) NO NO
2009 Data::MessagePack NO NO NO
2010 XML NO NO YES
2011 Pickle YES YES YES
2013 PHP Deserialize YES NO NO
2014 When you are using JSON in a protocol, talking to untrusted potentially
2016 hostile creatures requires relatively few measures.
2017 First of all, your JSON decoder should be secure, that is, should not
2019 have any buffer overflows. Obviously, this module should ensure that.
2020 Second, you need to avoid resource-starving attacks. That means you
2022 should limit the size of JSON texts you accept, or make sure then when
2023 your resources run out, that's just fine (e.g. by using a separate
2024 process that can crash safely). The size of a JSON text in octets or
2025 characters is usually a good indication of the size of the resources
2026 required to decode it into a Perl structure. While JSON::XS can check
2027 the size of the JSON text, it might be too late when you already have
2028 it in memory, so you might want to check the size before you accept the
2029 string.
2030 Third, Cpanel::JSON::XS recurses using the C stack when decoding
2032 objects and arrays. The C stack is a limited resource: for instance, on
2033 my amd64 machine with 8MB of stack size I can decode around 180k nested
2034 arrays but only 14k nested JSON objects (due to perl itself recursing
2035 deeply on croak to free the temporary). If that is exceeded, the
2036 program crashes. To be conservative, the default nesting limit is set
2037 to 512. If your process has a smaller stack, you should adjust this
2038 setting accordingly with the "max_depth" method.
2039 Also keep in mind that Cpanel::JSON::XS might leak contents of your
2041 Perl data structures in its error messages, so when you serialize
2042 sensitive information you might want to make sure that exceptions
2043 thrown by JSON::XS will not end up in front of untrusted eyes.
2044 If you are using Cpanel::JSON::XS to return packets to consumption by
2046 JavaScript scripts in a browser you should have a look at
2047 <http://blog.archive.jpsykes.com/47/practical-csrf-and-json-security/>
2048 to see whether you are vulnerable to some common attack vectors (which
2049 really are browser design bugs, but it is still you who will have to
2050 deal with it, as major browser developers care only for features, not
2051 about getting security right). You might also want to also look at
2052 Mojo::JSON special escape rules to prevent from XSS attacks.
2053
2055 TL;DR: Due to security concerns, Cpanel::JSON::XS will not allow scalar
2056 data in JSON texts by default - you need to create your own
2057 Cpanel::JSON::XS object and enable "allow_nonref":
2058 my $json = JSON::XS->new->allow_nonref;
2060 $text = $json->encode ($data);
2062 $data = $json->decode ($text);
2063 The long version: JSON being an important and supposedly stable format,
2065 the IETF standardized it as RFC 4627 in 2006. Unfortunately the
2066 inventor of JSON Douglas Crockford unilaterally changed the definition
2067 of JSON in javascript. Rather than create a fork, the IETF decided to
2068 standardize the new syntax (apparently, so I as told, without finding
2069 it very amusing).
2070 The biggest difference between the original JSON and the new JSON is
2072 that the new JSON supports scalars (anything other than arrays and
2073 objects) at the top-level of a JSON text. While this is strictly
2074 backwards compatible to older versions, it breaks a number of protocols
2075 that relied on sending JSON back-to-back, and is a minor security
2076 concern.
2077 For example, imagine you have two banks communicating, and on one side,
2079 the JSON coder gets upgraded. Two messages, such as 10 and 1000 might
2080 then be confused to mean 101000, something that couldn't happen in the
2081 original JSON, because neither of these messages would be valid JSON.
2082 If one side accepts these messages, then an upgrade in the coder on
2084 either side could result in this becoming exploitable.
2085 This module has always allowed these messages as an optional extension,
2087 by default disabled. The security concerns are the reason why the
2088 default is still disabled, but future versions might/will likely
2089 upgrade to the newer RFC as default format, so you are advised to check
2090 your implementation and/or override the default with "->allow_nonref
2091 (0)" to ensure that future versions are safe.
2092
2094 Cpanel::JSON::XS has proper ithreads support, unlike JSON::XS. If you
2095 encounter any bugs with thread support please report them.
2096
2098 While the goal of the Cpanel::JSON::XS module is to be correct, that
2099 unfortunately does not mean it's bug-free, only that the author thinks
2100 its design is bug-free. If you keep reporting bugs and tests they will
2101 be fixed swiftly, though.
2102 Since the JSON::XS author refuses to use a public bugtracker and
2104 prefers private emails, we use the tracker at github, so you might want
2105 to report any issues twice. Once in private to MLEHMANN to be fixed in
2106 JSON::XS and one to our the public tracker. Issues fixed by JSON::XS
2107 with a new release will also be backported to Cpanel::JSON::XS and
2108 5.6.2, as long as cPanel relies on 5.6.2 and Cpanel::JSON::XS as our
2109 serializer of choice.
2110 <https://github.com/rurban/Cpanel-JSON-XS/issues>
2112
2114 This module is available under the same licences as perl, the Artistic
2115 license and the GPL.
2116
2118 The cpanel_json_xs command line utility for quick experiments.
2119 JSON, JSON::XS, JSON::MaybeXS, Mojo::JSON, Mojo::JSON::MaybeXS,
2121 JSON::SL, JSON::DWIW, JSON::YAJL, JSON::Any, Test::JSON,
2122 Locale::Wolowitz, <https://metacpan.org/search?q=JSON>
2123 <https://tools.ietf.org/html/rfc7159>
2125 <https://tools.ietf.org/html/rfc4627>
2127
2129 Reini Urban <rurban@cpan.org>
2130 Marc Lehmann <schmorp@schmorp.de>, http://home.schmorp.de/
2132
2134 Reini Urban <rurban@cpan.org>
2135perl v5.30.1 2020-02-06 XS(3)