blib/lib/Cpanel/JSON/XS.pm

filter: Statementstmt 74.6% 50 / 67 Branchbran 73.3% 22 / 30 Conditioncond 62.5% 15 / 24 MC/DCmcdc 41.6% 5 / 12 Subroutinesub 61.1% 11 / 18 Podpod 100% 5 / 5 totaltotal 69.2% 108 / 156 CC 23 SCAR 15.4
CC
23
Cov
72%
__ANON__
9.0
to_json
6.0
from_json
6.0
CRAP 4.7
« blib/lib/Cpanel/JSON/XS/Type.pm ↑ Summary blib/lib/Cpanel/JSON/XS/Boolean.pm »
1 package Cpanel::JSON::XS;
2 69 our $VERSION = '4.82';
3 69 our $XS_VERSION = $VERSION;
4 # $VERSION = eval $VERSION;
5  
6 =pod
7  
8 =head1 NAME
9  
10 Cpanel::JSON::XS - cPanel fork of JSON::XS, fast and correct serializing
11  
12 =head1 SYNOPSIS
13  
14 use Cpanel::JSON::XS;
15  
16 # exported functions, they croak on error
17 # and expect/generate UTF-8
18  
19 $utf8_encoded_json_text = encode_json $perl_hash_or_arrayref;
20 $perl_hash_or_arrayref = decode_json $utf8_encoded_json_text;
21  
22 # OO-interface
23  
24 $coder = Cpanel::JSON::XS->new->ascii->pretty->allow_nonref;
25 $pretty_printed_unencoded = $coder->encode ($perl_scalar);
26 $perl_scalar = $coder->decode ($unicode_json_text);
27  
28 # Note that 5.6 misses most smart utf8 and encoding functionalities
29 # of newer releases.
30  
31 # Note that L<JSON::MaybeXS> will automatically use Cpanel::JSON::XS
32 # if available, at virtually no speed overhead either, so you should
33 # be able to just:
34  
35 use JSON::MaybeXS;
36  
37 # and do the same things, except that you have a pure-perl fallback now.
38  
39 Note that this module will be replaced by a new JSON::Safe module soon,
40 with the same API just guaranteed safe defaults.
41  
42 =head1 DESCRIPTION
43  
44 This module converts Perl data structures to JSON and vice versa. Its
45 primary goal is to be I<correct> and its secondary goal is to be
46 I<fast>. To reach the latter goal it was written in C.
47  
48 As this is the n-th-something JSON module on CPAN, what was the reason
49 to write yet another JSON module? While it seems there are many JSON
50 modules, none of them correctly handle all corner cases, and in most cases
51 their maintainers are unresponsive, gone missing, or not listening to bug
52 reports for other reasons.
53  
54 See below for the cPanel fork.
55  
56 See MAPPING, below, on how Cpanel::JSON::XS maps perl values to JSON
57 values and vice versa.
58  
59 =head2 FEATURES
60  
61 =over 4
62  
63 =item * correct Unicode handling
64  
65 This module knows how to handle Unicode with Perl version higher than 5.8.5,
66 documents how and when it does so, and even documents what "correct" means.
67  
68 =item * round-trip integrity
69  
70 When you serialize a perl data structure using only data types supported
71 by JSON and Perl, the deserialized data structure is identical on the Perl
72 level. (e.g. the string "2.0" doesn't suddenly become "2" just because
73 it looks like a number). There I<are> minor exceptions to this, read the
74 MAPPING section below to learn about those.
75  
76 =item * strict checking of JSON correctness
77  
78 There is no guessing, no generating of illegal JSON texts by default,
79 and only JSON is accepted as input by default. the latter is a security
80 feature.
81  
82 =item * fast
83  
84 Compared to other JSON modules and other serializers such as Storable,
85 this module usually compares favourably in terms of speed, too.
86  
87 =item * simple to use
88  
89 This module has both a simple functional interface as well as an object
90 oriented interface.
91  
92 =item * reasonably versatile output formats
93  
94 You can choose between the most compact guaranteed-single-line format
95 possible (nice for simple line-based protocols), a pure-ASCII format
96 (for when your transport is not 8-bit clean, still supports the whole
97 Unicode range), or a pretty-printed format (for when you want to read that
98 stuff). Or you can combine those features in whatever way you like.
99  
100 =back
101  
102 =head2 cPanel fork
103  
104 Since the original author MLEHMANN has no public
105 bugtracker, this cPanel fork sits now on github.
106  
107 src repo: L<https://github.com/rurban/Cpanel-JSON-XS>
108 original: L<http://cvs.schmorp.de/JSON-XS/>
109  
110 RT: L<https://github.com/rurban/Cpanel-JSON-XS/issues>
111 or L<https://rt.cpan.org/Public/Dist/Display.html?Queue=Cpanel-JSON-XS>
112  
113 B<Changes to JSON::XS>
114  
115 - bare hashkeys are now checked for utf8. (GH #209)
116  
117 - stricter decode_json() as documented. non-refs are disallowed.
118 safe by default.
119 added a 2nd optional argument. decode() honors now allow_nonref.
120  
121 - fixed encode of numbers for dual-vars. Different string
122 representations are preserved, but numbers with temporary strings
123 which represent the same number are here treated as numbers, not
124 strings. Cpanel::JSON::XS is a bit slower, but preserves numeric
125 types better.
126  
127 - numbers ending with .0 stray numbers, are not converted to
128 integers. [#63] dual-vars which are represented as number not
129 integer (42+"bar" != 5.8.9) are now encoded as number (=> 42.0)
130 because internally it's now a NOK type. However !!1 which is
131 wrongly encoded in 5.8 as "1"/1.0 is still represented as integer.
132  
133 - different handling of inf/nan. Default now to null, optionally with
134 stringify_infnan() to "inf"/"nan". [#28, #32]
135  
136 - added C<binary> extension, non-JSON and non JSON parsable, allows
137 C<\xNN> and C<\NNN> sequences.
138  
139 - 5.6.2 support; sacrificing some utf8 features (assuming bytes
140 all-over), no multi-byte unicode characters with 5.6.
141  
142 - interop for true/false overloading. JSON::XS, JSON::PP and Mojo::JSON
143 representations for booleans are accepted and JSON::XS accepts
144 Cpanel::JSON::XS booleans [#13, #37]
145 Fixed overloading of booleans. Cpanel::JSON::XS::true stringifies again
146 to "1", not "true", analog to all other JSON modules.
147  
148 - native boolean mapping of yes and no to true and false, as in YAML::XS.
149 In perl C<!0> is yes, C<!1> is no.
150 The JSON value true maps to 1, false maps to 0. [#39]
151  
152 - support arbitrary stringification with encode, with convert_blessed
153 and allow_blessed.
154  
155 - ithread support. Cpanel::JSON::XS is thread-safe, JSON::XS not
156  
157 - is_bool can be called as method, JSON::XS::is_bool not.
158  
159 - performance optimizations for threaded Perls
160  
161 - relaxed mode, allowing many popular extensions
162  
163 - protect our magic object from corruption by wrong or missing external
164 methods, like FREEZE/THAW or serialization with other methods.
165  
166 - additional fixes for:
167  
168 - #208 - no security-relevant out-of-bounds reading of module memory
169 when decoding hash keys without ending ':'
170  
171 - [cpan #88061] AIX atof without USE_LONG_DOUBLE
172  
173 - #10 unshare_hek crash
174  
175 - #7, #29 avoid re-blessing where possible. It fails in JSON::XS for
176 READONLY values, i.e. restricted hashes.
177  
178 - #41 overloading of booleans, use the object not the reference.
179  
180 - #62 -Dusequadmath conversion and no SEGV.
181  
182 - #72 parsing of values followed \0, like 1\0 does fail.
183  
184 - #72 parsing of illegal unicode or non-unicode characters.
185  
186 - #96 locale-insensitive numeric conversion.
187  
188 - #154 numeric conversion fixed since 5.22, using the same strtold as perl5.
189  
190 - #167 sort tied hashes with canonical.
191  
192 - #212 fix utf8 object stringification
193  
194 - public maintenance and bugtracker
195  
196 - use ppport.h, sanify XS.xs comment styles, harness C coding style
197  
198 - common::sense is optional. When available it is not used in the
199 published production module, just during development and testing.
200  
201 - extended testsuite, passes all http://seriot.ch/projects/parsing_json.html
202 tests. In fact it is the only know JSON decoder which does so,
203 while also being the fastest.
204  
205 - support many more options and methods from JSON::PP:
206 stringify_infnan, allow_unknown, allow_stringify, allow_barekey,
207 encode_stringify, allow_bignum, allow_singlequote, dupkeys_as_arrayref,
208 sort_by (partially), escape_slash, convert_blessed, ...
209 optional decode_json(, allow_nonref) arg.
210 relaxed implements allow_dupkeys.
211  
212 - support all 5 unicode L<BOM|/BOM>'s: UTF-8, UTF-16LE, UTF-16BE, UTF-32LE,
213 UTF-32BE, encoding internally to UTF-8.
214  
215 =cut
216  
217 69 our @ISA = qw(Exporter);
218 69 our @EXPORT = qw(encode_json decode_json to_json from_json);
219  
220 0 ▶sub to_json($@) {
Subroutine to_json: not called CC=2 SCAR=17.9
Pod documented
221 0 ▶ if ($] >= 5.008) {
Branch: 1 branch
BranchTrueFalseTotal
if ($] >= 5.008) { } 0 0 0
222 0 require Carp;
223 0 Carp::croak ("Cpanel::JSON::XS::to_json has been renamed to encode_json,".
224 " either downgrade to pre-2.0 versions of Cpanel::JSON::XS or".
225 " rename the call");
226 } else {
227 0 _to_json(@_);
228 }
229 }
230  
231 0 ▶sub from_json($@) {
Subroutine from_json: not called CC=2 SCAR=17.9
Pod documented
232 0 ▶ if ($] >= 5.008) {
Branch: 1 branch
BranchTrueFalseTotal
if ($] >= 5.008) { } 0 0 0
233 0 require Carp;
234 0 Carp::croak ("Cpanel::JSON::XS::from_json has been renamed to decode_json,".
235 " either downgrade to pre-2.0 versions of Cpanel::JSON::XS or".
236 " rename the call");
237 } else {
238 0 _from_json(@_);
239 }
240 }
241  
242 69 ▶use Exporter;
Subroutine BEGIN: called CC=1 SCAR=0.0
243 69 ▶use XSLoader;
Subroutine BEGIN: called CC=1 SCAR=0.0
244  
245 =head1 FUNCTIONAL INTERFACE
246  
247 The following convenience methods are provided by this module. They are
248 exported by default:
249  
250 =over 4
251  
252 =item $json_text = encode_json $perl_scalar, [json_type]
253  
254 Converts the given Perl data structure to a UTF-8 encoded, binary string
255 (that is, the string contains octets only). Croaks on error.
256  
257 This function call is functionally identical to:
258  
259 $json_text = Cpanel::JSON::XS->new->utf8->encode ($perl_scalar, $json_type)
260  
261 Except being faster.
262  
263 For the type argument see L<Cpanel::JSON::XS::Type>.
264  
265 =item $perl_scalar = decode_json $json_text [, $allow_nonref [, my $json_type ] ]
266  
267 The opposite of C<encode_json>: expects an UTF-8 (binary) string of an
268 json reference and tries to parse that as an UTF-8 encoded JSON text,
269 returning the resulting reference. Croaks on error.
270  
271 This function call is functionally identical to:
272  
273 $perl_scalar = Cpanel::JSON::XS->new->utf8->decode ($json_text, $json_type)
274  
275 except being faster.
276  
277 Note that older decode_json versions in Cpanel::JSON::XS older than
278 3.0116 and JSON::XS did not set allow_nonref but allowed them due to a
279 bug in the decoder.
280  
281 Since version 4.42, C<allow_nonref> is enabled by default, matching
282 JSON::XS 4.0+. The 2nd optional C<$allow_nonref> argument can be set
283 to false (C<0>) to disable it. When enabled, the function accepts all
284 JSON values: objects, arrays, strings, numbers, C<null>, C<true>, and
285 C<false>. See L</"OLD" VS. "NEW" JSON (RFC 4627 VS. RFC 7159)> for
286 why you might not want this.
287  
288 For the 3rd optional type argument see L<Cpanel::JSON::XS::Type>.
289  
290 =item $is_boolean = Cpanel::JSON::XS::is_bool $scalar
291  
292 Returns true if the passed scalar represents either C<JSON::PP::true>
293 or C<JSON::PP::false>, two constants that act like C<1> and C<0>,
294 respectively and are used to represent JSON C<true> and C<false>
295 values in Perl. (Also recognizes the booleans produced by L<JSON::XS>.)
296  
297 See MAPPING, below, for more information on how JSON values are mapped
298 to Perl.
299  
300 =back
301  
302 =head1 DEPRECATED FUNCTIONS
303  
304 =over
305  
306 =item from_json
307  
308 from_json has been renamed to decode_json
309  
310 =item to_json
311  
312 to_json has been renamed to encode_json
313  
314 =back
315  
316  
317 =head1 A FEW NOTES ON UNICODE AND PERL
318  
319 Since this often leads to confusion, here are a few very clear words on
320 how Unicode works in Perl, modulo bugs.
321  
322 =over 4
323  
324 =item 1. Perl strings can store characters with ordinal values > 255.
325  
326 This enables you to store Unicode characters as single characters in a
327 Perl string - very natural.
328  
329 =item 2. Perl does I<not> associate an encoding with your strings.
330  
331 ... until you force it to, e.g. when matching it against a regex, or
332 printing the scalar to a file, in which case Perl either interprets
333 your string as locale-encoded text, octets/binary, or as Unicode,
334 depending on various settings. In no case is an encoding stored
335 together with your data, it is I<use> that decides encoding, not any
336 magical meta data.
337  
338 =item 3. The internal utf-8 flag has no meaning with regards to the
339 encoding of your string.
340  
341 =item 4. A "Unicode String" is simply a string where each character
342 can be validly interpreted as a Unicode code point.
343  
344 If you have UTF-8 encoded data, it is no longer a Unicode string, but
345 a Unicode string encoded in UTF-8, giving you a binary string.
346  
347 =item 5. A string containing "high" (> 255) character values is I<not>
348 a UTF-8 string.
349  
350 =item 6. Raw non-Unicode characters below U+10FFFF are allowed.
351  
352 The 66 Unicode noncharacters U+FDD0..U+FDEF, and U+*FFFE, U+*FFFF are
353 allowed without warning, as JSON::PP does, see
354 L<http://www.unicode.org/versions/corrigendum9.html>. But illegal
355 surrogate pairs fail to parse.
356  
357 =item 7. Raw non-Unicode characters above U+10FFFF are disallowed.
358  
359 Raw non-Unicode characters outside the valid unicode range fail to
360 parse, because "A string is a sequence of zero or more Unicode
361 characters" RFC 7159 section 1 and "JSON text SHALL be encoded in
362 Unicode RFC 7159 section 8.1. We use now the UTF8_DISALLOW_SUPER
363 flag when parsing unicode.
364  
365 =item 8. Lone surrogates or illegal surrogate pairs are disallowed.
366  
367 Since RFC 3629, U+D800 through U+DFFF are not legal Unicode values and
368 their UTF-8 encodings must be treated as an invalid byte sequence.
369 RFC 8259 section 8.2 admits the spec allows string values that contain
370 bit sequences that cannot encode Unicode characters and that the
371 behavior of software that receives such values is unpredictable. To
372 avoid introducing non-Unicode strings into Perl we use the
373 UTF8_DISALLOW_SURROGATE flag when parsing Unicode and verify escaped
374 surrogates form valid pairs.
375  
376 =back
377  
378 I hope this helps :)
379  
380  
381 =head1 OBJECT-ORIENTED INTERFACE
382  
383 The object oriented interface lets you configure your own encoding or
384 decoding style, within the limits of supported formats.
385  
386 =over 4
387  
388 =item $json = new Cpanel::JSON::XS
389  
390 Creates a new JSON object that can be used to de/encode JSON
391 strings. All boolean flags described below are by default I<disabled>.
392  
393 The mutators for flags all return the JSON object again and thus calls can
394 be chained:
395  
396 my $json = Cpanel::JSON::XS->new->utf8->space_after->encode ({a => [1,2]})
397 => {"a": [1, 2]}
398  
399 =item $json = $json->ascii ([$enable])
400  
401 =item $enabled = $json->get_ascii
402  
403 If C<$enable> is true (or missing), then the C<encode> method will not
404 generate characters outside the code range C<0..127> (which is ASCII). Any
405 Unicode characters outside that range will be escaped using either a
406 single C<\uXXXX> (BMP characters) or a double C<\uHHHH\uLLLLL> escape sequence,
407 as per RFC4627. The resulting encoded JSON text can be treated as a native
408 Unicode string, an ascii-encoded, latin1-encoded or UTF-8 encoded string,
409 or any other superset of ASCII.
410  
411 If C<$enable> is false, then the C<encode> method will not escape Unicode
412 characters unless required by the JSON syntax or other flags. This results
413 in a faster and more compact format.
414  
415 See also the section I<ENCODING/CODESET FLAG NOTES> later in this
416 document.
417  
418 The main use for this flag is to produce JSON texts that can be
419 transmitted over a 7-bit channel, as the encoded JSON texts will not
420 contain any 8 bit characters.
421  
422 Cpanel::JSON::XS->new->ascii (1)->encode ([chr 0x10401])
423 => ["\ud801\udc01"]
424  
425 =item $json = $json->latin1 ([$enable])
426  
427 =item $enabled = $json->get_latin1
428  
429 If C<$enable> is true (or missing), then the C<encode> method will encode
430 the resulting JSON text as latin1 (or ISO-8859-1), escaping any characters
431 outside the code range C<0..255>. The resulting string can be treated as a
432 latin1-encoded JSON text or a native Unicode string. The C<decode> method
433 will not be affected in any way by this flag, as C<decode> by default
434 expects Unicode, which is a strict superset of latin1.
435  
436 If C<$enable> is false, then the C<encode> method will not escape Unicode
437 characters unless required by the JSON syntax or other flags.
438  
439 See also the section I<ENCODING/CODESET FLAG NOTES> later in this
440 document.
441  
442 The main use for this flag is efficiently encoding binary data as JSON
443 text, as most octets will not be escaped, resulting in a smaller encoded
444 size. The disadvantage is that the resulting JSON text is encoded
445 in latin1 (and must correctly be treated as such when storing and
446 transferring), a rare encoding for JSON. It is therefore most useful when
447 you want to store data structures known to contain binary data efficiently
448 in files or databases, not when talking to other JSON encoders/decoders.
449  
450 Cpanel::JSON::XS->new->latin1->encode (["\x{89}\x{abc}"]
451 => ["\x{89}\\u0abc"] # (perl syntax, U+abc escaped, U+89 not)
452  
453  
454 =item $json = $json->binary ([$enable])
455  
456 =item $enabled = $json = $json->get_binary
457  
458 If the C<$enable> argument is true (or missing), then the C<encode>
459 method will not try to detect an UTF-8 encoding in any JSON string, it
460 will strictly interpret it as byte sequence. The result might contain
461 new C<\xNN> sequences, which is B<unparsable JSON>. The C<decode>
462 method forbids C<\uNNNN> sequences and accepts C<\xNN> and octal
463 C<\NNN> sequences.
464  
465 There is also a special logic for perl 5.6 and utf8. 5.6 encodes any
466 string to utf-8 automatically when seeing a codepoint >= C<0x80> and
467 < C<0x100>. With the binary flag enabled decode the perl utf8 encoded
468 string to the original byte encoding and encode this with C<\xNN>
469 escapes. This will result to the same encodings as with newer
470 perls. But note that binary multi-byte codepoints with 5.6 will
471 result in C<illegal unicode character in binary string> errors,
472 unlike with newer perls.
473  
474 If C<$enable> is false, then the C<encode> method will smartly try to
475 detect Unicode characters unless required by the JSON syntax or other
476 flags and hex and octal sequences are forbidden.
477  
478 See also the section I<ENCODING/CODESET FLAG NOTES> later in this
479 document.
480  
481 The main use for this flag is to avoid the smart unicode detection and
482 possible double encoding. The disadvantage is that the resulting JSON
483 text is encoded in new C<\xNN> and in latin1 characters and must
484 correctly be treated as such when storing and transferring, a rare
485 encoding for JSON. It will produce non-readable JSON strings in the
486 browser. It is therefore most useful when you want to store data
487 structures known to contain binary data efficiently in files or
488 databases, not when talking to other JSON encoders/decoders. The
489 binary decoding method can also be used when an encoder produced a
490 non-JSON conformant hex or octal encoding C<\xNN> or C<\NNN>.
491  
492 Cpanel::JSON::XS->new->binary->encode (["\x{89}\x{abc}"])
493 5.6: Error: malformed or illegal unicode character in binary string
494 >=5.8: ['\x89\xe0\xaa\xbc']
495  
496 Cpanel::JSON::XS->new->binary->encode (["\x{89}\x{bc}"])
497 => ["\x89\xbc"]
498  
499 Cpanel::JSON::XS->new->binary->decode (["\x89\ua001"])
500 Error: malformed or illegal unicode character in binary string
501  
502 Cpanel::JSON::XS->new->decode (["\x89"])
503 Error: illegal hex character in non-binary string
504  
505  
506 =item $json = $json->utf8 ([$enable])
507  
508 =item $enabled = $json->get_utf8
509  
510 If C<$enable> is true (or missing), then the C<encode> method will encode
511 the JSON result into UTF-8, as required by many protocols, while the
512 C<decode> method expects to be handled an UTF-8-encoded string. Please
513 note that UTF-8-encoded strings do not contain any characters outside the
514 range C<0..255>, they are thus useful for bytewise/binary I/O. In future
515 versions, enabling this option might enable autodetection of the UTF-16
516 and UTF-32 encoding families, as described in RFC4627.
517  
518 If C<$enable> is false, then the C<encode> method will return the JSON
519 string as a (non-encoded) Unicode string, while C<decode> expects thus a
520 Unicode string. Any decoding or encoding (e.g. to UTF-8 or UTF-16) needs
521 to be done yourself, e.g. using the Encode module.
522  
523 See also the section I<ENCODING/CODESET FLAG NOTES> later in this
524 document.
525  
526 Example, output UTF-16BE-encoded JSON:
527  
528 use Encode;
529 $jsontext = encode "UTF-16BE", Cpanel::JSON::XS->new->encode ($object);
530  
531 Example, decode UTF-32LE-encoded JSON:
532  
533 use Encode;
534 $object = Cpanel::JSON::XS->new->decode (decode "UTF-32LE", $jsontext);
535  
536 =item $json = $json->pretty ([$enable])
537  
538 This enables (or disables) all of the C<indent>, C<space_before> and
539 C<space_after> (and in the future possibly more) flags in one call to
540 generate the most readable (or most compact) form possible.
541  
542 Example, pretty-print some simple structure:
543  
544 my $json = Cpanel::JSON::XS->new->pretty(1)->encode ({a => [1,2]})
545 =>
546 {
547 "a" : [
548 1,
549 2
550 ]
551 }
552  
553  
554 =item $json = $json->indent ([$enable])
555  
556 =item $enabled = $json->get_indent
557  
558 If C<$enable> is true (or missing), then the C<encode> method will use
559 a multiline format as output, putting every array member or
560 object/hash key-value pair into its own line, indenting them properly.
561  
562 If C<$enable> is false, no newlines or indenting will be produced, and the
563 resulting JSON text is guaranteed not to contain any C<newlines>.
564  
565 This setting has no effect when decoding JSON texts.
566  
567 =item $json = $json->indent_length([$number_of_spaces])
568  
569 =item $length = $json->get_indent_length()
570  
571 Set the indent length (default C<3>).
572 This option is only useful when you also enable indent or pretty.
573 The acceptable range is from 0 (no indentation) to 15
574  
575 =item $json = $json->space_before ([$enable])
576  
577 =item $enabled = $json->get_space_before
578  
579 If C<$enable> is true (or missing), then the C<encode> method will add an extra
580 optional space before the C<:> separating keys from values in JSON objects.
581  
582 If C<$enable> is false, then the C<encode> method will not add any extra
583 space at those places.
584  
585 This setting has no effect when decoding JSON texts. You will also
586 most likely combine this setting with C<space_after>.
587  
588 Example, space_before enabled, space_after and indent disabled:
589  
590 {"key" :"value"}
591  
592 =item $json = $json->space_after ([$enable])
593  
594 =item $enabled = $json->get_space_after
595  
596 If C<$enable> is true (or missing), then the C<encode> method will add
597 an extra optional space after the C<:> separating keys from values in
598 JSON objects and extra whitespace after the C<,> separating key-value
599 pairs and array members.
600  
601 If C<$enable> is false, then the C<encode> method will not add any extra
602 space at those places.
603  
604 This setting has no effect when decoding JSON texts.
605  
606 Example, space_before and indent disabled, space_after enabled:
607  
608 {"key": "value"}
609  
610 =item $json = $json->relaxed ([$enable])
611  
612 =item $enabled = $json->get_relaxed
613  
614 If C<$enable> is true (or missing), then C<decode> will accept some
615 extensions to normal JSON syntax (see below). C<encode> will not be
616 affected in anyway. I<Be aware that this option makes you accept invalid
617 JSON texts as if they were valid!>. I suggest only to use this option to
618 parse application-specific files written by humans (configuration files,
619 resource files etc.)
620  
621 If C<$enable> is false (the default), then C<decode> will only accept
622 valid JSON texts.
623  
624 Currently accepted extensions are:
625  
626 =over 4
627  
628 =item * list items can have an end-comma
629  
630 JSON I<separates> array elements and key-value pairs with commas. This
631 can be annoying if you write JSON texts manually and want to be able to
632 quickly append elements, so this extension accepts comma at the end of
633 such items not just between them:
634  
635 [
636 1,
637 2, <- this comma not normally allowed
638 ]
639 {
640 "k1": "v1",
641 "k2": "v2", <- this comma not normally allowed
642 }
643  
644 =item * shell-style '#'-comments
645  
646 Whenever JSON allows whitespace, shell-style comments are additionally
647 allowed. They are terminated by the first carriage-return or line-feed
648 character, after which more white-space and comments are allowed.
649  
650 [
651 1, # this comment not allowed in JSON
652 # neither this one...
653 ]
654  
655 =item * literal ASCII TAB characters in strings
656  
657 Literal ASCII TAB characters are now allowed in strings (and treated as
658 C<\t>) in relaxed mode. Despite JSON mandates, that TAB character is
659 substituted for "\t" sequence.
660  
661 [
662 "Hello\tWorld",
663 "Hello<TAB>World", # literal <TAB> would not normally be allowed
664 ]
665  
666 =item * allow_singlequote
667  
668 Single quotes are accepted instead of double quotes. See the
669 L</allow_singlequote> option.
670  
671 { "foo":'bar' }
672 { 'foo':"bar" }
673 { 'foo':'bar' }
674  
675 =item * allow_barekey
676  
677 Accept unquoted object keys instead of with mandatory double quotes. See the
678 L</allow_barekey> option.
679  
680 { foo:"bar" }
681  
682 =item * allow_dupkeys
683  
684 Allow decoding of duplicate keys in hashes. By default duplicate keys are forbidden.
685 See L<http://seriot.ch/projects/parsing_json.php#24>:
686 RFC 7159 section 4: "The names within an object should be unique."
687 See the C<allow_dupkeys> option.
688  
689 =back
690  
691 =item $json = $json->canonical ([$enable])
692  
693 =item $enabled = $json->get_canonical
694  
695 If C<$enable> is true (or missing), then the C<encode> method will
696 output JSON objects by sorting their keys. This is adding a
697 comparatively high overhead.
698  
699 If C<$enable> is false, then the C<encode> method will output key-value
700 pairs in the order Perl stores them (which will likely change between runs
701 of the same script, and can change even within the same run from 5.18
702 onwards).
703  
704 This option is useful if you want the same data structure to be encoded as
705 the same JSON text (given the same overall settings). If it is disabled,
706 the same hash might be encoded differently even if contains the same data,
707 as key-value pairs have no inherent ordering in Perl.
708  
709 This setting has no effect when decoding JSON texts.
710  
711 This is now also done with tied hashes, contrary to L<JSON::XS>.
712 But note that with most large tied hashes stored as tree it is advised to
713 sort the iterator already and don't sort the hash output here. Most such
714 iterators are already sorted, as such e.g. L<DB_File> with C<DB_BTREE>.
715  
716 =item $json = $json->sort_by (undef, 0, 1 or a block)
717  
718 This currently only (un)sets the C<canonical> option, and ignores
719 custom sort blocks.
720  
721 This setting has no effect when decoding JSON texts.
722  
723 This setting has currently no effect on tied hashes.
724  
725  
726 =item $json = $json->escape_slash ([$enable])
727  
728 =item $enabled = $json->get_escape_slash
729  
730 According to the JSON Grammar, the I<forward slash> character (U+002F)
731 C<"/"> need to be escaped. But by default strings are encoded without
732 escaping slashes in all perl JSON encoders.
733  
734 If C<$enable> is true (or missing), then C<encode> will escape slashes,
735 C<"\/">.
736  
737 This setting has no effect when decoding JSON texts.
738  
739  
740 =item $json = $json->unblessed_bool ([$enable])
741  
742 =item $enabled = $json->get_unblessed_bool
743  
744 $json = $json->unblessed_bool([$enable])
745  
746 If C<$enable> is true (or missing), then C<decode> will return Perl
747 non-object boolean variables (1 and 0 as numbers or "1" and "" as
748 strings) for JSON booleans (C<true> and C<false>). If C<$enable> is
749 false, then C<decode> will return C<JSON::PP::Boolean> objects for
750 JSON booleans.
751  
752  
753 =item $json = $json->allow_singlequote ([$enable])
754  
755 =item $enabled = $json->get_allow_singlequote
756  
757 $json = $json->allow_singlequote([$enable])
758  
759 If C<$enable> is true (or missing), then C<decode> will accept
760 JSON strings quoted by single quotations that are invalid JSON
761 format.
762  
763 $json->allow_singlequote->decode({"foo":'bar'});
764 $json->allow_singlequote->decode({'foo':"bar"});
765 $json->allow_singlequote->decode({'foo':'bar'});
766  
767 This is also enabled with C<relaxed>.
768 As same as the C<relaxed> option, this option may be used to parse
769 application-specific files written by humans.
770  
771  
772 =item $json = $json->allow_barekey ([$enable])
773  
774 =item $enabled = $json->get_allow_barekey
775  
776 $json = $json->allow_barekey([$enable])
777  
778 If C<$enable> is true (or missing), then C<decode> will accept
779 bare keys of JSON object that are invalid JSON format.
780  
781 Same as with the C<relaxed> option, this option may be used to parse
782 application-specific files written by humans.
783  
784 $json->allow_barekey->decode('{foo:"bar"}');
785  
786 =item $json = $json->allow_bignum ([$enable])
787  
788 =item $enabled = $json->get_allow_bignum
789  
790 $json = $json->allow_bignum([$enable])
791  
792 If C<$enable> is true (or missing), then C<decode> will convert
793 the big integer Perl cannot handle as integer into a L<Math::BigInt>
794 object and convert a floating number (any) into a L<Math::BigFloat>.
795  
796 $int = $json->allow_nonref->allow_bignum->decode(1); # => 1
797 $bigint = $json->allow_bignum->decode('100000000000000000000000000000000000000');
798 $bigfloat = $json->allow_bignum->decode(1.0);
799  
800 On the contrary, C<encode> converts C<Math::BigInt> objects and
801 C<Math::BigFloat> objects into JSON numbers with C<allow_blessed>
802 enable.
803  
804 $json->allow_nonref->allow_blessed->allow_bignum;
805 $bigfloat = $json->decode('2.000000000000000000000000001');
806 print $json->encode($bigfloat);
807 # => 2.000000000000000000000000001
808  
809 See L</MAPPING> about the normal conversion of JSON number.
810  
811  
812 =item $json = $json->allow_bigint ([$enable])
813  
814 This option is obsolete and replaced by allow_bignum.
815  
816  
817 =item $json = $json->allow_nonref ([$enable])
818  
819 =item $enabled = $json->get_allow_nonref
820  
821 If C<$enable> is true (or missing), then the C<encode> method can
822 convert a non-reference into its corresponding string, number or null
823 JSON value, which is an extension to RFC4627. Likewise, C<decode> will
824 accept those JSON values instead of croaking.
825  
826 B<This option is now enabled by default> (since version 4.42), matching
827 the behavior of JSON::XS 4.0+ and the broken RFC 7159. JSON texts
828 consisting of a bare scalar (string, number, C<true>, C<false>, C<null>)
829 are accepted by C<encode> and C<decode> without explicitly requesting it.
830 If you need strict RFC 4627 compliance, you must explicitly disable it:
831  
832 Cpanel::JSON::XS->new->allow_nonref(0);
833  
834 When disabled, C<encode> will croak if it isn't passed an arrayref or
835 hashref, as JSON texts must either be an object or array. Likewise,
836 C<decode> will croak if given something that is not a JSON object or array.
837  
838 =item $json = $json->allow_unknown ([$enable])
839  
840 =item $enabled = $json->get_allow_unknown
841  
842 If C<$enable> is true (or missing), then C<encode> will I<not> throw an
843 exception when it encounters values it cannot represent in JSON (for
844 example, filehandles) but instead will encode a JSON C<null> value. Note
845 that blessed objects are not included here and are handled separately by
846 c<allow_nonref>.
847  
848 If C<$enable> is false (the default), then C<encode> will throw an
849 exception when it encounters anything it cannot encode as JSON.
850  
851 This option does not affect C<decode> in any way, and it is recommended to
852 leave it off unless you know your communications partner.
853  
854 =item $json = $json->allow_stringify ([$enable])
855  
856 =item $enabled = $json->get_allow_stringify
857  
858 If C<$enable> is true (or missing), then C<encode> will stringify the
859 non-object perl value or reference. Note that blessed objects are not
860 included here and are handled separately by C<allow_blessed> and
861 C<convert_blessed>. String references are stringified to the string
862 value, other references as in perl.
863  
864 This option does not affect C<decode> in any way.
865  
866 This option is special to this module, it is not supported by other
867 encoders. So it is not recommended to use it.
868  
869 =item $json = $json->require_types ([$enable])
870  
871 =item $enable = $json->get_require_types
872  
873 $json = $json->require_types([$enable])
874  
875 If C<$enable> is true (or missing), then C<encode> will require
876 either enabled C<type_all_string> or second argument with supplied JSON types.
877 See L<Cpanel::JSON::XS::Type>. When C<type_all_string> is not enabled or
878 second argument is not provided (or is undef), then C<encode>
879 croaks. It also croaks when the type for provided structure in
880 C<encode> is incomplete.
881  
882 =item $json = $json->type_all_string ([$enable])
883  
884 =item $enable = $json->get_type_all_string
885  
886 $json = $json->type_all_string([$enable])
887  
888 If C<$enable> is true (or missing), then C<encode> will always
889 produce stable deterministic JSON string types in resulted output.
890  
891 When C<$enable> is false, then result of encoded JSON output may be
892 different for different Perl versions and may depends on loaded modules.
893  
894 This is useful it you need deterministic JSON types, independently of used
895 Perl version and other modules, but do not want to write complicated type
896 definitions for L<Cpanel::JSON::XS::Type>.
897  
898 When combined with C<allow_blessed> and/or C<convert_blessed>, blessed
899 objects are handled by those options first, B<not> stringified by
900 C<type_all_string>. For example, with C<allow_blessed + type_all_string>,
901 blessed objects are encoded as the JSON value C<null> (not C<"null">).
902  
903 =item $json = $json->allow_dupkeys ([$enable])
904  
905 =item $enabled = $json->get_allow_dupkeys
906  
907 If C<$enable> is true (or missing), then the C<decode> method will not
908 die when it encounters duplicate keys in a hash.
909 C<allow_dupkeys> is also enabled in the C<relaxed> mode.
910  
911 The JSON spec allows duplicate name in objects but recommends to
912 disable it, however with Perl hashes they are impossible, parsing
913 JSON in Perl silently ignores duplicate names, using the last value
914 found.
915  
916 See L<http://seriot.ch/projects/parsing_json.php#24>:
917 RFC 7159 section 4: "The names within an object should be unique."
918  
919 =item $json = $json->dupkeys_as_arrayref ([$enable])
920  
921 =item $enabled = $json->get_dupkeys_as_arrayref
922  
923 If enabled, allow decoding of duplicate keys in hashes and store the
924 values as arrayref in the hash instead. By default duplicate keys are
925 forbidden. Enabling this also enables the L</allow_dupkeys> option,
926 but disabling this does not disable the L</allow_dupkeys> option.
927  
928 Example:
929  
930 $json->dupkeys_as_arrayref;
931 print encode_json ($json->decode ('{"a":"b","a":"c"}'));
932  
933 => {"a":["b","c"]}
934  
935 This changes the result structure, thus cannot be enabled by default.
936 The client must be aware of it. The resulting arrayref is not yet marked somehow
937 (blessed or such).
938  
939 =item $json = $json->allow_blessed ([$enable])
940  
941 =item $enabled = $json->get_allow_blessed
942  
943 If C<$enable> is true (or missing), then the C<encode> method will not
944 barf when it encounters a blessed reference. Instead, the value of the
945 B<convert_blessed> option will decide whether C<null> (C<convert_blessed>
946 disabled or no C<TO_JSON> method found) or a representation of the
947 object (C<convert_blessed> enabled and C<TO_JSON> method found) is being
948 encoded. Has no effect on C<decode>.
949  
950 If C<$enable> is false (the default), then C<encode> will throw an
951 exception when it encounters a blessed object without C<convert_blessed>
952 and a C<TO_JSON> method.
953  
954 This setting has no effect on C<decode>.
955  
956 =item $json = $json->convert_blessed ([$enable])
957  
958 =item $enabled = $json->get_convert_blessed
959  
960 If C<$enable> is true (or missing), then C<encode>, upon encountering a
961 blessed object, will check for the availability of the C<TO_JSON> method
962 on the object's class. If found, it will be called in scalar context
963 and the resulting scalar will be encoded instead of the object. If no
964 C<TO_JSON> method is found, a stringification overload method is tried next.
965 If both are not found, the value of C<allow_blessed> will decide what
966 to do.
967  
968 The C<TO_JSON> method may safely call die if it wants. If C<TO_JSON>
969 returns other blessed objects, those will be handled in the same
970 way. C<TO_JSON> must take care of not causing an endless recursion
971 cycle (== crash) in this case. The same care must be taken with
972 calling encode in stringify overloads (even if this works by luck in
973 older perls) or other callbacks. The name of C<TO_JSON> was chosen
974 because other methods called by the Perl core (== not by the user of
975 the object) are usually in upper case letters and to avoid collisions
976 with any C<to_json> function or method.
977  
978 If C<$enable> is false (the default), then C<encode> will not consider
979 this type of conversion.
980  
981 This setting has no effect on C<decode>.
982  
983 =item $json = $json->allow_tags ([$enable])
984  
985 =item $enabled = $json->get_allow_tags
986  
987 See L<OBJECT SERIALIZATION> for details.
988  
989 If C<$enable> is true (or missing), then C<encode>, upon encountering a
990 blessed object, will check for the availability of the C<FREEZE> method on
991 the object's class. If found, it will be used to serialize the object into
992 a nonstandard tagged JSON value (that JSON decoders cannot decode).
993  
994 It also causes C<decode> to parse such tagged JSON values and deserialize
995 them via a call to the C<THAW> method.
996  
997 If C<$enable> is false (the default), then C<encode> will not consider
998 this type of conversion, and tagged JSON values will cause a parse error
999 in C<decode>, as if tags were not part of the grammar.
1000  
1001 =item $json = $json->filter_json_object ([$coderef->($hashref)])
1002  
1003 When C<$coderef> is specified, it will be called from C<decode> each
1004 time it decodes a JSON object. The only argument is a reference to the
1005 newly-created hash. If the code references returns a single scalar (which
1006 need not be a reference), this value (i.e. a copy of that scalar to avoid
1007 aliasing) is inserted into the deserialized data structure. If it returns
1008 an empty list (NOTE: I<not> C<undef>, which is a valid scalar), the
1009 original deserialized hash will be inserted. This setting can slow down
1010 decoding considerably.
1011  
1012 When C<$coderef> is omitted or undefined, any existing callback will
1013 be removed and C<decode> will not change the deserialized hash in any
1014 way.
1015  
1016 Example, convert all JSON objects into the integer 5:
1017  
1018 my $js = Cpanel::JSON::XS->new->filter_json_object (sub { 5 });
1019 # returns [5]
1020 $js->decode ('[{}]')
1021 # throw an exception because allow_nonref is not enabled
1022 # so a lone 5 is not allowed.
1023 $js->decode ('{"a":1, "b":2}');
1024  
1025 =item $json = $json->filter_json_single_key_object ($key [=> $coderef->($value)])
1026  
1027 Works remotely similar to C<filter_json_object>, but is only called for
1028 JSON objects having a single key named C<$key>.
1029  
1030 This C<$coderef> is called before the one specified via
1031 C<filter_json_object>, if any. It gets passed the single value in the JSON
1032 object. If it returns a single value, it will be inserted into the data
1033 structure. If it returns nothing (not even C<undef> but the empty list),
1034 the callback from C<filter_json_object> will be called next, as if no
1035 single-key callback were specified.
1036  
1037 If C<$coderef> is omitted or undefined, the corresponding callback will be
1038 disabled. There can only ever be one callback for a given key.
1039  
1040 As this callback gets called less often then the C<filter_json_object>
1041 one, decoding speed will not usually suffer as much. Therefore, single-key
1042 objects make excellent targets to serialize Perl objects into, especially
1043 as single-key JSON objects are as close to the type-tagged value concept
1044 as JSON gets (it's basically an ID/VALUE tuple). Of course, JSON does not
1045 support this in any way, so you need to make sure your data never looks
1046 like a serialized Perl hash.
1047  
1048 Typical names for the single object key are C<__class_whatever__>, or
1049 C<$__dollars_are_rarely_used__$> or C<}ugly_brace_placement>, or even
1050 things like C<__class_md5sum(classname)__>, to reduce the risk of clashing
1051 with real hashes.
1052  
1053 Example, decode JSON objects of the form C<< { "__widget__" => <id> } >>
1054 into the corresponding C<< $WIDGET{<id>} >> object:
1055  
1056 # return whatever is in $WIDGET{5}:
1057 Cpanel::JSON::XS
1058 ->new
1059 ->filter_json_single_key_object (__widget__ => sub {
1060 $WIDGET{ $_[0] }
1061 })
1062 ->decode ('{"__widget__": 5')
1063  
1064 # this can be used with a TO_JSON method in some "widget" class
1065 # for serialization to json:
1066 sub WidgetBase::TO_JSON {
1067 my ($self) = @_;
1068  
1069 unless ($self->{id}) {
1070 $self->{id} = ..get..some..id..;
1071 $WIDGET{$self->{id}} = $self;
1072 }
1073  
1074 { __widget__ => $self->{id} }
1075 }
1076  
1077 =item $json = $json->shrink ([$enable])
1078  
1079 =item $enabled = $json->get_shrink
1080  
1081 Perl usually over-allocates memory a bit when allocating space for
1082 strings. This flag optionally resizes strings generated by either
1083 C<encode> or C<decode> to their minimum size possible. This can save
1084 memory when your JSON texts are either very very long or you have many
1085 short strings. It will also try to downgrade any strings to octet-form
1086 if possible: perl stores strings internally either in an encoding called
1087 UTF-X or in octet-form. The latter cannot store everything but uses less
1088 space in general (and some buggy Perl or C code might even rely on that
1089 internal representation being used).
1090  
1091 The actual definition of what shrink does might change in future versions,
1092 but it will always try to save space at the expense of time.
1093  
1094 If C<$enable> is true (or missing), the string returned by C<encode> will
1095 be shrunk-to-fit, while all strings generated by C<decode> will also be
1096 shrunk-to-fit.
1097  
1098 If C<$enable> is false, then the normal perl allocation algorithms are used.
1099 If you work with your data, then this is likely to be faster.
1100  
1101 In the future, this setting might control other things, such as converting
1102 strings that look like integers or floats into integers or floats
1103 internally (there is no difference on the Perl level), saving space.
1104  
1105 =item $json = $json->max_depth ([$maximum_nesting_depth])
1106  
1107 =item $max_depth = $json->get_max_depth
1108  
1109 Sets the maximum nesting level (default C<512>) accepted while encoding
1110 or decoding. If a higher nesting level is detected in JSON text or a Perl
1111 data structure, then the encoder and decoder will stop and croak at that
1112 point.
1113  
1114 Nesting level is defined by number of hash- or arrayrefs that the encoder
1115 needs to traverse to reach a given point or the number of C<{> or C<[>
1116 characters without their matching closing parenthesis crossed to reach a
1117 given character in a string.
1118  
1119 Setting the maximum depth to one disallows any nesting, so that ensures
1120 that the object is only a single hash/object or array.
1121  
1122 If no argument is given, the highest possible setting will be used, which
1123 is rarely useful.
1124  
1125 Note that nesting is implemented by recursion in C. The default value has
1126 been chosen to be as large as typical operating systems allow without
1127 crashing.
1128  
1129 See L</SECURITY CONSIDERATIONS>, below, for more info on why this is useful.
1130  
1131 =item $json = $json->max_size ([$maximum_string_size])
1132  
1133 =item $max_size = $json->get_max_size
1134  
1135 Set the maximum length a JSON text may have (in bytes) where decoding is
1136 being attempted. The default is C<0>, meaning no limit. When C<decode>
1137 is called on a string that is longer then this many bytes, it will not
1138 attempt to decode the string but throw an exception. This setting has no
1139 effect on C<encode> (yet).
1140  
1141 If no argument is given, the limit check will be deactivated (same as when
1142 C<0> is specified).
1143  
1144 See L</SECURITY CONSIDERATIONS>, below, for more info on why this is useful.
1145  
1146 =item $json->stringify_infnan ([$infnan_mode = 1])
1147  
1148 =item $infnan_mode = $json->get_stringify_infnan
1149  
1150 Get or set how Cpanel::JSON::XS encodes C<inf>, C<-inf> or C<nan> for numeric
1151 values. Also qnan, snan or negative nan on some platforms.
1152  
1153 C<null>: infnan_mode = 0. Similar to most JSON modules in other languages.
1154 Always null.
1155  
1156 stringified: infnan_mode = 1. As in Mojo::JSON. Platform specific strings.
1157 Stringified via sprintf(%g), with double quotes.
1158  
1159 inf/nan: infnan_mode = 2. As in JSON::XS, and older releases.
1160 Passes through platform dependent values, invalid JSON. Stringified via
1161 sprintf(%g), but without double quotes.
1162  
1163 "inf/-inf/nan": infnan_mode = 3. Platform independent inf/nan/-inf
1164 strings. No QNAN/SNAN/negative NAN support, unified to "nan". Much
1165 easier to detect, but may conflict with valid strings.
1166  
1167 =item $json_text = $json->encode ($perl_scalar, $json_type)
1168  
1169 Converts the given Perl data structure (a simple scalar or a reference
1170 to a hash or array) to its JSON representation. Simple scalars will be
1171 converted into JSON string or number sequences, while references to
1172 arrays become JSON arrays and references to hashes become JSON
1173 objects. Undefined Perl values (e.g. C<undef>) become JSON C<null>
1174 values. Neither C<true> nor C<false> values will be generated.
1175  
1176 For the type argument see L<Cpanel::JSON::XS::Type>.
1177  
1178 =item $bytes_written = $json->encode_to ($filehandle, $perl_scalar, $json_type)
1179  
1180 Like C<encode>, but writes the JSON representation directly to
1181 C<$filehandle> as it is generated, instead of building the whole result
1182 in memory and returning it as a string. Returns the number of bytes
1183 (octets) written, and croaks (as C<encode> does) if C<$perl_scalar>
1184 cannot be represented, and if the write to C<$filehandle> fails.
1185  
1186 This is useful for very large data structures, where it lowers peak
1187 memory usage and gives the reader on the other end a head start instead
1188 of waiting for the whole document to be assembled first:
1189  
1190 open my $fh, ">", "big.json" or die $!;
1191 $json->encode_to ($fh, $data);
1192 close $fh;
1193  
1194 C<encode_to> always writes raw bytes: character data is emitted using
1195 the same octet encoding C<encode> would produce for the object's current
1196 flags (see C<utf8>, C<ascii>, C<latin1>, C<binary>), regardless of any
1197 PerlIO encoding layer already present on C<$filehandle>. As with
1198 C<encode>, do not additionally push a C<:encoding(UTF-8)> (or similar)
1199 layer onto the handle, or the output will be double-encoded.
1200  
1201 =item $perl_scalar = $json->decode ($json_text, my $json_type)
1202  
1203 The opposite of C<encode>: expects a JSON text and tries to parse it,
1204 returning the resulting simple scalar or reference. Croaks on error.
1205  
1206 JSON numbers and strings become simple Perl scalars. JSON arrays become
1207 Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes
1208 C<1>, C<false> becomes C<0> and C<null> becomes C<undef>.
1209  
1210 For the type argument see L<Cpanel::JSON::XS::Type>.
1211  
1212 =item $perl_scalar = $json->decode_and_validate ($json_text, $schema)
1213  
1214 Like C<decode>, but additionally validates the decoded structure
1215 against C<$schema>, a schema hash built from the type spec building
1216 blocks in L<Cpanel::JSON::XS::Type> (plain C<JSON_TYPE_*> constants
1217 or their short names, C<json_type_arrayof>, C<json_type_hashof>,
1218 C<json_type_anyof>, C<json_type_optional>, ...). Croaks with one line
1219 per mismatch (missing/unexpected hash keys, wrong scalar type, wrong
1220 array length/element type) if C<$data> doesn't match C<$schema>;
1221 otherwise returns the decoded data exactly like C<decode> would.
1222  
1223 use Cpanel::JSON::XS::Type;
1224 my $schema = { id => JSON_TYPE_INT, name => 'Str', tags => json_type_arrayof('Str') };
1225 my $data = $json->decode_and_validate($json_text, $schema);
1226  
1227 See L<Cpanel::JSON::XS::Type/check_type> for the full schema
1228 semantics, including how it differs from the encoder's permissive
1229 hash type spec.
1230  
1231 =item ($perl_scalar, $characters) = $json->decode_prefix ($json_text)
1232  
1233 This works like the C<decode> method, but instead of raising an exception
1234 when there is trailing garbage after the first JSON object, it will
1235 silently stop parsing there and return the number of characters consumed
1236 so far.
1237  
1238 This is useful if your JSON texts are not delimited by an outer protocol
1239 and you need to know where the JSON text ends.
1240  
1241 Cpanel::JSON::XS->new->decode_prefix ("[1] the tail")
1242 => ([1], 3)
1243  
1244 =item $json->to_json ($perl_hash_or_arrayref)
1245  
1246 Deprecated method for perl 5.8 and newer. Use L<encode_json> instead.
1247  
1248 =item $json->from_json ($utf8_encoded_json_text)
1249  
1250 Deprecated method for perl 5.8 and newer. Use L<decode_json> instead.
1251  
1252 =back
1253  
1254  
1255 =head1 INCREMENTAL PARSING
1256  
1257 In some cases, there is the need for incremental parsing of JSON
1258 texts. While this module always has to keep both JSON text and resulting
1259 Perl data structure in memory at one time, it does allow you to parse a
1260 JSON stream incrementally. It does so by accumulating text until it has
1261 a full JSON object, which it then can decode. This process is similar to
1262 using C<decode_prefix> to see if a full JSON object is available, but
1263 is much more efficient (and can be implemented with a minimum of method
1264 calls).
1265  
1266 Cpanel::JSON::XS will only attempt to parse the JSON text once it is
1267 sure it has enough text to get a decisive result, using a very simple
1268 but truly incremental parser. This means that it sometimes won't stop
1269 as early as the full parser, for example, it doesn't detect mismatched
1270 parentheses. The only thing it guarantees is that it starts decoding
1271 as soon as a syntactically valid JSON text has been seen. This means
1272 you need to set resource limits (e.g. C<max_size>) to ensure the
1273 parser will stop parsing in the presence of syntax errors.
1274  
1275 The following methods implement this incremental parser.
1276  
1277 =over 4
1278  
1279 =item [void, scalar or list context] = $json->incr_parse ([$string])
1280  
1281 This is the central parsing function. It can both append new text and
1282 extract objects from the stream accumulated so far (both of these
1283 functions are optional).
1284  
1285 If C<$string> is given, then this string is appended to the already
1286 existing JSON fragment stored in the C<$json> object.
1287  
1288 After that, if the function is called in void context, it will simply
1289 return without doing anything further. This can be used to add more text
1290 in as many chunks as you want.
1291  
1292 If the method is called in scalar context, then it will try to extract
1293 exactly I<one> JSON object. If that is successful, it will return this
1294 object, otherwise it will return C<undef>. If there is a parse error,
1295 this method will croak just as C<decode> would do (one can then use
1296 C<incr_skip> to skip the erroneous part). This is the most common way of
1297 using the method.
1298  
1299 And finally, in list context, it will try to extract as many objects
1300 from the stream as it can find and return them, or the empty list
1301 otherwise. For this to work, there must be no separators between the JSON
1302 objects or arrays, instead they must be concatenated back-to-back. If
1303 an error occurs, an exception will be raised as in the scalar context
1304 case. Note that in this case, any previously-parsed JSON texts will be
1305 lost.
1306  
1307 Example: Parse some JSON arrays/objects in a given string and return
1308 them.
1309  
1310 my @objs = Cpanel::JSON::XS->new->incr_parse ("[5][7][1,2]");
1311  
1312 =item $lvalue_string = $json->incr_text (>5.8 only)
1313  
1314 This method returns the currently stored JSON fragment as an lvalue, that
1315 is, you can manipulate it. This I<only> works when a preceding call to
1316 C<incr_parse> in I<scalar context> successfully returned an object, and
1317 2. only with Perl >= 5.8
1318  
1319 Under all other circumstances you must not call this function (I mean
1320 it. although in simple tests it might actually work, it I<will> fail
1321 under real world conditions). As a special exception, you can also
1322 call this method before having parsed anything.
1323  
1324 This function is useful in two cases: a) finding the trailing text after a
1325 JSON object or b) parsing multiple JSON objects separated by non-JSON text
1326 (such as commas).
1327  
1328 =item $json->incr_skip
1329  
1330 This will reset the state of the incremental parser and will remove
1331 the parsed text from the input buffer so far. This is useful after
1332 C<incr_parse> died, in which case the input buffer and incremental parser
1333 state is left unchanged, to skip the text parsed so far and to reset the
1334 parse state.
1335  
1336 The difference to C<incr_reset> is that only text until the parse error
1337 occurred is removed.
1338  
1339 =item $json->incr_reset
1340  
1341 This completely resets the incremental parser, that is, after this call,
1342 it will be as if the parser had never parsed anything.
1343  
1344 This is useful if you want to repeatedly parse JSON objects and want to
1345 ignore any trailing data, which means you have to reset the parser after
1346 each successful decode.
1347  
1348 =back
1349  
1350 =head2 LIMITATIONS
1351  
1352 All options that affect decoding are supported, except
1353 C<allow_nonref>. The reason for this is that it cannot be made to work
1354 sensibly: JSON objects and arrays are self-delimited, i.e. you can
1355 concatenate them back to back and still decode them perfectly. This
1356 does not hold true for JSON numbers, however.
1357  
1358 For example, is the string C<1> a single JSON number, or is it simply
1359 the start of C<12>? Or is C<12> a single JSON number, or the
1360 concatenation of C<1> and C<2>? In neither case you can tell, and this
1361 is why Cpanel::JSON::XS takes the conservative route and disallows
1362 this case.
1363  
1364 Since C<allow_nonref> became the default (4.42), C<incr_parse> still
1365 rejects a top-level number, C<true>, C<false> or C<null> unless
1366 C<allow_nonref> was enabled explicitly, in which case the ambiguity above
1367 applies.
1368  
1369 =head2 EXAMPLES
1370  
1371 Some examples will make all this clearer. First, a simple example that
1372 works similarly to C<decode_prefix>: We want to decode the JSON object at
1373 the start of a string and identify the portion after the JSON object:
1374  
1375 my $text = "[1,2,3] hello";
1376  
1377 my $json = new Cpanel::JSON::XS;
1378  
1379 my $obj = $json->incr_parse ($text)
1380 or die "expected JSON object or array at beginning of string";
1381  
1382 my $tail = $json->incr_text;
1383 # $tail now contains " hello"
1384  
1385 Easy, isn't it?
1386  
1387 Now for a more complicated example: Imagine a hypothetical protocol where
1388 you read some requests from a TCP stream, and each request is a JSON
1389 array, without any separation between them (in fact, it is often useful to
1390 use newlines as "separators", as these get interpreted as whitespace at
1391 the start of the JSON text, which makes it possible to test said protocol
1392 with C<telnet>...).
1393  
1394 Here is how you'd do it (it is trivial to write this in an event-based
1395 manner):
1396  
1397 my $json = new Cpanel::JSON::XS;
1398  
1399 # read some data from the socket
1400 while (sysread $socket, my $buf, 4096) {
1401  
1402 # split and decode as many requests as possible
1403 for my $request ($json->incr_parse ($buf)) {
1404 # act on the $request
1405 }
1406 }
1407  
1408 Another complicated example: Assume you have a string with JSON objects
1409 or arrays, all separated by (optional) comma characters (e.g. C<[1],[2],
1410 [3]>). To parse them, we have to skip the commas between the JSON texts,
1411 and here is where the lvalue-ness of C<incr_text> comes in useful:
1412  
1413 my $text = "[1],[2], [3]";
1414 my $json = new Cpanel::JSON::XS;
1415  
1416 # void context, so no parsing done
1417 $json->incr_parse ($text);
1418  
1419 # now extract as many objects as possible. note the
1420 # use of scalar context so incr_text can be called.
1421 while (my $obj = $json->incr_parse) {
1422 # do something with $obj
1423  
1424 # now skip the optional comma
1425 $json->incr_text =~ s/^ \s* , //x;
1426 }
1427  
1428 Now lets go for a very complex example: Assume that you have a gigantic
1429 JSON array-of-objects, many gigabytes in size, and you want to parse it,
1430 but you cannot load it into memory fully (this has actually happened in
1431 the real world :).
1432  
1433 Well, you lost, you have to implement your own JSON parser. But
1434 Cpanel::JSON::XS can still help you: You implement a (very simple)
1435 array parser and let JSON decode the array elements, which are all
1436 full JSON objects on their own (this wouldn't work if the array
1437 elements could be JSON numbers, for example):
1438  
1439 my $json = new Cpanel::JSON::XS;
1440  
1441 # open the monster
1442 open my $fh, "<bigfile.json"
1443 or die "bigfile: $!";
1444  
1445 # first parse the initial "["
1446 for (;;) {
1447 sysread $fh, my $buf, 65536
1448 or die "read error: $!";
1449 $json->incr_parse ($buf); # void context, so no parsing
1450  
1451 # Exit the loop once we found and removed(!) the initial "[".
1452 # In essence, we are (ab-)using the $json object as a simple scalar
1453 # we append data to.
1454 last if $json->incr_text =~ s/^ \s* \[ //x;
1455 }
1456  
1457 # now we have the skipped the initial "[", so continue
1458 # parsing all the elements.
1459 for (;;) {
1460 # in this loop we read data until we got a single JSON object
1461 for (;;) {
1462 if (my $obj = $json->incr_parse) {
1463 # do something with $obj
1464 last;
1465 }
1466  
1467 # add more data
1468 sysread $fh, my $buf, 65536
1469 or die "read error: $!";
1470 $json->incr_parse ($buf); # void context, so no parsing
1471 }
1472  
1473 # in this loop we read data until we either found and parsed the
1474 # separating "," between elements, or the final "]"
1475 for (;;) {
1476 # first skip whitespace
1477 $json->incr_text =~ s/^\s*//;
1478  
1479 # if we find "]", we are done
1480 if ($json->incr_text =~ s/^\]//) {
1481 print "finished.\n";
1482 exit;
1483 }
1484  
1485 # if we find ",", we can continue with the next element
1486 if ($json->incr_text =~ s/^,//) {
1487 last;
1488 }
1489  
1490 # if we find anything else, we have a parse error!
1491 if (length $json->incr_text) {
1492 die "parse error near ", $json->incr_text;
1493 }
1494  
1495 # else add more data
1496 sysread $fh, my $buf, 65536
1497 or die "read error: $!";
1498 $json->incr_parse ($buf); # void context, so no parsing
1499 }
1500  
1501 This is a complex example, but most of the complexity comes from the fact
1502 that we are trying to be correct (bear with me if I am wrong, I never ran
1503 the above example :).
1504  
1505 =head1 BOM
1506  
1507 Detect all unicode B<Byte Order Marks> on decode.
1508 Which are UTF-8, UTF-16LE, UTF-16BE, UTF-32LE and UTF-32BE.
1509  
1510 The BOM encoding is set only for one specific decode call, it does not
1511 change the state of the JSON object.
1512  
1513 B<Warning>: With perls older than 5.20 you need load the Encode module
1514 before loading a multibyte BOM, i.e. >= UTF-16. Otherwise an error is
1515 thrown. This is an implementation limitation and might get fixed later.
1516  
1517 See L<https://tools.ietf.org/html/rfc7159#section-8.1>
1518 I<"JSON text SHALL be encoded in UTF-8, UTF-16, or UTF-32.">
1519  
1520 I<"Implementations MUST NOT add a byte order mark to the beginning of a
1521 JSON text", "implementations (...) MAY ignore the presence of a byte
1522 order mark rather than treating it as an error".>
1523  
1524 See also L<http://www.unicode.org/faq/utf_bom.html#BOM>.
1525  
1526 Beware that Cpanel::JSON::XS is currently the only JSON module which
1527 does accept and decode a BOM.
1528  
1529 The latest JSON spec
1530 L<https://www.greenbytes.de/tech/webdav/rfc8259.html#character.encoding>
1531 forbid the usage of UTF-16 or UTF-32, the character encoding is UTF-8.
1532 Thus in subsequent updates BOM's of UTF-16 or UTF-32 will throw an error.
1533  
1534 =head1 MAPPING
1535  
1536 This section describes how Cpanel::JSON::XS maps Perl values to JSON
1537 values and vice versa. These mappings are designed to "do the right
1538 thing" in most circumstances automatically, preserving round-tripping
1539 characteristics (what you put in comes out as something equivalent).
1540  
1541 For the more enlightened: note that in the following descriptions,
1542 lowercase I<perl> refers to the Perl interpreter, while uppercase I<Perl>
1543 refers to the abstract Perl language itself.
1544  
1545  
1546 =head2 JSON -> PERL
1547  
1548 =over 4
1549  
1550 =item object
1551  
1552 A JSON object becomes a reference to a hash in Perl. No ordering of object
1553 keys is preserved (JSON does not preserve object key ordering itself).
1554  
1555 =item array
1556  
1557 A JSON array becomes a reference to an array in Perl.
1558  
1559 =item string
1560  
1561 A JSON string becomes a string scalar in Perl - Unicode codepoints in JSON
1562 are represented by the same codepoints in the Perl string, so no manual
1563 decoding is necessary.
1564  
1565 =item number
1566  
1567 A JSON number becomes either an integer, numeric (floating point) or
1568 string scalar in perl, depending on its range and any fractional parts. On
1569 the Perl level, there is no difference between those as Perl handles all
1570 the conversion details, but an integer may take slightly less memory and
1571 might represent more values exactly than floating point numbers.
1572  
1573 If the number consists of digits only, Cpanel::JSON::XS will try to
1574 represent it as an integer value. If that fails, it will try to
1575 represent it as a numeric (floating point) value if that is possible
1576 without loss of precision. Otherwise it will preserve the number as a
1577 string value (in which case you lose roundtripping ability, as the
1578 JSON number will be re-encoded to a JSON string).
1579  
1580 Numbers containing a fractional or exponential part will always be
1581 represented as numeric (floating point) values, possibly at a loss of
1582 precision (in which case you might lose perfect roundtripping ability, but
1583 the JSON number will still be re-encoded as a JSON number).
1584  
1585 Note that precision is not accuracy - binary floating point values
1586 cannot represent most decimal fractions exactly, and when converting
1587 from and to floating point, C<Cpanel::JSON::XS> only guarantees precision
1588 up to but not including the least significant bit.
1589  
1590 =item true, false
1591  
1592 When C<unblessed_bool> is set to true, then JSON C<true> becomes C<1> and
1593 JSON C<false> becomes C<0>.
1594  
1595 Otherwise these JSON atoms become C<JSON::PP::true> and
1596 C<JSON::PP::false>, respectively. They are C<JSON::PP::Boolean>
1597 objects and are overloaded to act almost exactly like the numbers C<1>
1598 and C<0>. You can check whether a scalar is a JSON boolean by using
1599 the C<Cpanel::JSON::XS::is_bool> function.
1600  
1601 The other round, from perl to JSON, C<!0> which is represented as
1602 C<yes> becomes C<true>, and C<!1> which is represented as
1603 C<no> becomes C<false>.
1604  
1605 Via L<Cpanel::JSON::XS::Type> you can now even force negation in C<encode>,
1606 without overloading of C<!>:
1607  
1608 my $false = Cpanel::JSON::XS::false;
1609 print($json->encode([!$false], [JSON_TYPE_BOOL]));
1610 => [true]
1611  
1612 =item null
1613  
1614 A JSON null atom becomes C<undef> in Perl.
1615  
1616 =item shell-style comments (C<< # I<text> >>)
1617  
1618 As a nonstandard extension to the JSON syntax that is enabled by the
1619 C<relaxed> setting, shell-style comments are allowed. They can start
1620 anywhere outside strings and go till the end of the line.
1621  
1622 =item tagged values (C<< (I<tag>)I<value> >>).
1623  
1624 Another nonstandard extension to the JSON syntax, enabled with the
1625 C<allow_tags> setting, are tagged values. In this implementation, the
1626 I<tag> must be a perl package/class name encoded as a JSON string, and the
1627 I<value> must be a JSON array encoding optional constructor arguments.
1628  
1629 See L<OBJECT SERIALIZATION>, below, for details.
1630  
1631 =back
1632  
1633  
1634 =head2 PERL -> JSON
1635  
1636 The mapping from Perl to JSON is slightly more difficult, as Perl is a
1637 truly typeless language, so we can only guess which JSON type is meant by
1638 a Perl value.
1639  
1640 =over 4
1641  
1642 =item hash references
1643  
1644 Perl hash references become JSON objects. As there is no inherent ordering
1645 in hash keys (or JSON objects), they will usually be encoded in a
1646 pseudo-random order that can change between runs of the same program but
1647 stays generally the same within a single run of a program. Cpanel::JSON::XS can
1648 optionally sort the hash keys (determined by the I<canonical> flag), so
1649 the same datastructure will serialize to the same JSON text (given same
1650 settings and version of Cpanel::JSON::XS), but this incurs a runtime overhead
1651 and is only rarely useful, e.g. when you want to compare some JSON text
1652 against another for equality.
1653  
1654 =item array references
1655  
1656 Perl array references become JSON arrays.
1657  
1658 =item other references
1659  
1660 Other unblessed references are generally not allowed and will cause an
1661 exception to be thrown, except for references to the integers C<0> and
1662 C<1>, which get turned into C<false> and C<true> atoms in JSON.
1663  
1664 With the option C<allow_stringify>, you can ignore the exception and return
1665 the stringification of the perl value.
1666  
1667 With the option C<allow_unknown>, you can ignore the exception and
1668 return C<null> instead.
1669  
1670 encode_json [\"x"] # => cannot encode reference to scalar 'SCALAR(0x..)'
1671 # unless the scalar is 0 or 1
1672 encode_json [\0, \1] # yields [false,true]
1673  
1674 allow_stringify->encode_json [\"x"] # yields "x" unlike JSON::PP
1675 allow_unknown->encode_json [\"x"] # yields null as in JSON::PP
1676  
1677 =item Cpanel::JSON::XS::true, Cpanel::JSON::XS::false
1678  
1679 These special values become JSON true and JSON false values,
1680 respectively. You can also use C<\1> and C<\0> or C<!0> and C<!1>
1681 directly if you want.
1682  
1683 encode_json [Cpanel::JSON::XS::false, Cpanel::JSON::XS::true] # yields [false,true]
1684 encode_json [!1, !0], [JSON_TYPE_BOOL, JSON_TYPE_BOOL] # yields [false,true]
1685  
1686 eq/ne comparisons with true, false:
1687  
1688 false is eq to the empty string or the string 'false' or the special
1689 empty string C<!!0> or C<!1>, i.e. C<SV_NO>, or the numbers 0 or 0.0.
1690  
1691 true is eq to the string 'true' or to the special string C<!0>
1692 (i.e. C<SV_YES>) or to the numbers 1 or 1.0.
1693  
1694 =item blessed objects
1695  
1696 Blessed objects are not directly representable in JSON, but
1697 C<Cpanel::JSON::XS> allows various optional ways of handling
1698 objects. See L<OBJECT SERIALIZATION>, below, for details.
1699  
1700 See the C<allow_blessed> and C<convert_blessed> methods on various
1701 options on how to deal with this: basically, you can choose between
1702 throwing an exception, encoding the reference as if it weren't
1703 blessed, use the objects overloaded stringification method or provide
1704 your own serializer method.
1705  
1706 =item simple scalars
1707  
1708 Simple Perl scalars (any scalar that is not a reference) are the most
1709 difficult objects to encode: Cpanel::JSON::XS will encode undefined
1710 scalars or inf/nan as JSON C<null> values and other scalars to either
1711 number or string in non-deterministic way which may be affected or
1712 changed by Perl version or any other loaded Perl module.
1713  
1714 If you want to have stable and deterministic types in JSON encoder then
1715 use L<Cpanel::JSON::XS::Type>.
1716  
1717 Alternative way for deterministic types is to use C<type_all_string>
1718 method when all perl scalars are encoded to JSON strings.
1719  
1720 Non-deterministic behavior is following: scalars that have last been
1721 used in a string context before encoding as JSON strings, and anything
1722 else as number value:
1723  
1724 # dump as number
1725 encode_json [2] # yields [2]
1726 encode_json [-3.0e17] # yields [-3e+17]
1727 my $value = 5; encode_json [$value] # yields [5]
1728  
1729 # used as string, but the two representations are for the same number
1730 print $value;
1731 encode_json [$value] # yields [5]
1732  
1733 # used as different string (non-matching dual-var)
1734 my $str = '0 but true';
1735 my $num = 1 + $str;
1736 encode_json [$num, $str] # yields [1,"0 but true"]
1737  
1738 # undef becomes null
1739 encode_json [undef] # yields [null]
1740  
1741 # inf or nan becomes null, unless you answered
1742 # "Do you want to handle inf/nan as strings" with yes
1743 encode_json [9**9**9] # yields [null]
1744  
1745 You can force the type to be a JSON string by stringifying it:
1746  
1747 my $x = 3.1; # some variable containing a number
1748 "$x"; # stringified
1749 $x .= ""; # another, more awkward way to stringify
1750 print $x; # perl does it for you, too, quite often
1751  
1752 You can force the type to be a JSON number by numifying it:
1753  
1754 my $x = "3"; # some variable containing a string
1755 $x += 0; # numify it, ensuring it will be dumped as a number
1756 $x *= 1; # same thing, the choice is yours.
1757  
1758 Note that numerical precision has the same meaning as under Perl (so
1759 binary to decimal conversion follows the same rules as in Perl, which
1760 can differ to other languages). Also, your perl interpreter might expose
1761 extensions to the floating point numbers of your platform, such as
1762 infinities or NaN's - these cannot be represented in JSON, and thus
1763 null is returned instead. Optionally you can configure it to stringify
1764 inf and nan values.
1765  
1766 =back
1767  
1768 =head2 OBJECT SERIALIZATION
1769  
1770 As JSON cannot directly represent Perl objects, you have to choose between
1771 a pure JSON representation (without the ability to deserialize the object
1772 automatically again), and a nonstandard extension to the JSON syntax,
1773 tagged values.
1774  
1775 =head3 SERIALIZATION
1776  
1777 What happens when C<Cpanel::JSON::XS> encounters a Perl object depends
1778 on the C<allow_blessed>, C<convert_blessed> and C<allow_tags>
1779 settings, which are used in this order:
1780  
1781 =over 4
1782  
1783 =item 1. C<allow_tags> is enabled and the object has a C<FREEZE> method.
1784  
1785 In this case, C<Cpanel::JSON::XS> uses the L<Types::Serialiser> object
1786 serialization protocol to create a tagged JSON value, using a nonstandard
1787 extension to the JSON syntax.
1788  
1789 This works by invoking the C<FREEZE> method on the object, with the first
1790 argument being the object to serialize, and the second argument being the
1791 constant string C<JSON> to distinguish it from other serializers.
1792  
1793 The C<FREEZE> method can return any number of values (i.e. zero or
1794 more). These values and the paclkage/classname of the object will then be
1795 encoded as a tagged JSON value in the following format:
1796  
1797 ("classname")[FREEZE return values...]
1798  
1799 e.g.:
1800  
1801 ("URI")["http://www.google.com/"]
1802 ("MyDate")[2013,10,29]
1803 ("ImageData::JPEG")["Z3...VlCg=="]
1804  
1805 For example, the hypothetical C<My::Object> C<FREEZE> method might use the
1806 objects C<type> and C<id> members to encode the object:
1807  
1808 sub My::Object::FREEZE {
1809 my ($self, $serializer) = @_;
1810  
1811 ($self->{type}, $self->{id})
1812 }
1813  
1814 =item 2. C<convert_blessed> is enabled and the object has a C<TO_JSON> method.
1815  
1816 In this case, the C<TO_JSON> method of the object is invoked in scalar
1817 context. It must return a single scalar that can be directly encoded into
1818 JSON. This scalar replaces the object in the JSON text.
1819  
1820 For example, the following C<TO_JSON> method will convert all L<URI>
1821 objects to JSON strings when serialized. The fact that these values
1822 originally were L<URI> objects is lost.
1823  
1824 sub URI::TO_JSON {
1825 my ($uri) = @_;
1826 $uri->as_string
1827 }
1828  
1829 =item 3. C<convert_blessed> is enabled and the object has a stringification overload.
1830  
1831 In this case, the overloaded C<""> method of the object is invoked in scalar
1832 context. It must return a single scalar that can be directly encoded into
1833 JSON. This scalar replaces the object in the JSON text.
1834  
1835 For example, the following C<""> method will convert all L<URI>
1836 objects to JSON strings when serialized. The fact that these values
1837 originally were L<URI> objects is lost.
1838  
1839 package URI;
1840 use overload '""' => sub { shift->as_string };
1841  
1842 =item 4. C<allow_blessed> is enabled.
1843  
1844 The object will be serialized as a JSON null value.
1845  
1846 =item 5. none of the above
1847  
1848 If none of the settings are enabled or the respective methods are missing,
1849 C<Cpanel::JSON::XS> throws an exception.
1850  
1851 =back
1852  
1853 =head3 DESERIALIZATION
1854  
1855 For deserialization there are only two cases to consider: either
1856 nonstandard tagging was used, in which case C<allow_tags> decides,
1857 or objects cannot be automatically be deserialized, in which
1858 case you can use postprocessing or the C<filter_json_object> or
1859 C<filter_json_single_key_object> callbacks to get some real objects our of
1860 your JSON.
1861  
1862 This section only considers the tagged value case: I a tagged JSON object
1863 is encountered during decoding and C<allow_tags> is disabled, a parse
1864 error will result (as if tagged values were not part of the grammar).
1865  
1866 If C<allow_tags> is enabled, C<Cpanel::JSON::XS> will look up the C<THAW> method
1867 of the package/classname used during serialization (it will not attempt
1868 to load the package as a Perl module). If there is no such method, the
1869 decoding will fail with an error.
1870  
1871 Otherwise, the C<THAW> method is invoked with the classname as first
1872 argument, the constant string C<JSON> as second argument, and all the
1873 values from the JSON array (the values originally returned by the
1874 C<FREEZE> method) as remaining arguments.
1875  
1876 The method must then return the object. While technically you can return
1877 any Perl scalar, you might have to enable the C<enable_nonref> setting to
1878 make that work in all cases, so better return an actual blessed reference.
1879  
1880 As an example, let's implement a C<THAW> function that regenerates the
1881 C<My::Object> from the C<FREEZE> example earlier:
1882  
1883 sub My::Object::THAW {
1884 my ($class, $serializer, $type, $id) = @_;
1885  
1886 $class->new (type => $type, id => $id)
1887 }
1888  
1889 See the L</SECURITY CONSIDERATIONS> section below. Allowing external
1890 json objects being deserialized to perl objects is usually a very bad
1891 idea.
1892  
1893  
1894 =head1 ENCODING/CODESET FLAG NOTES
1895  
1896 The interested reader might have seen a number of flags that signify
1897 encodings or codesets - C<utf8>, C<latin1>, C<binary> and
1898 C<ascii>. There seems to be some confusion on what these do, so here
1899 is a short comparison:
1900  
1901 C<utf8> controls whether the JSON text created by C<encode> (and expected
1902 by C<decode>) is UTF-8 encoded or not, while C<latin1> and C<ascii> only
1903 control whether C<encode> escapes character values outside their respective
1904 codeset range. Neither of these flags conflict with each other, although
1905 some combinations make less sense than others.
1906  
1907 Care has been taken to make all flags symmetrical with respect to
1908 C<encode> and C<decode>, that is, texts encoded with any combination of
1909 these flag values will be correctly decoded when the same flags are used
1910 - in general, if you use different flag settings while encoding vs. when
1911 decoding you likely have a bug somewhere.
1912  
1913 Below comes a verbose discussion of these flags. Note that a "codeset" is
1914 simply an abstract set of character-codepoint pairs, while an encoding
1915 takes those codepoint numbers and I<encodes> them, in our case into
1916 octets. Unicode is (among other things) a codeset, UTF-8 is an encoding,
1917 and ISO-8859-1 (= latin 1) and ASCII are both codesets I<and> encodings at
1918 the same time, which can be confusing.
1919  
1920 =over 4
1921  
1922 =item C<utf8> flag disabled
1923  
1924 When C<utf8> is disabled (the default), then C<encode>/C<decode> generate
1925 and expect Unicode strings, that is, characters with high ordinal Unicode
1926 values (> 255) will be encoded as such characters, and likewise such
1927 characters are decoded as-is, no changes to them will be done, except
1928 "(re-)interpreting" them as Unicode codepoints or Unicode characters,
1929 respectively (to Perl, these are the same thing in strings unless you do
1930 funny/weird/dumb stuff).
1931  
1932 This is useful when you want to do the encoding yourself (e.g. when you
1933 want to have UTF-16 encoded JSON texts) or when some other layer does
1934 the encoding for you (for example, when printing to a terminal using a
1935 filehandle that transparently encodes to UTF-8 you certainly do NOT want
1936 to UTF-8 encode your data first and have Perl encode it another time).
1937  
1938 =item C<utf8> flag enabled
1939  
1940 If the C<utf8>-flag is enabled, C<encode>/C<decode> will encode all
1941 characters using the corresponding UTF-8 multi-byte sequence, and will
1942 expect your input strings to be encoded as UTF-8, that is, no "character"
1943 of the input string must have any value > 255, as UTF-8 does not allow
1944 that.
1945  
1946 The C<utf8> flag therefore switches between two modes: disabled means you
1947 will get a Unicode string in Perl, enabled means you get an UTF-8 encoded
1948 octet/binary string in Perl.
1949  
1950 =item C<latin1>, C<binary> or C<ascii> flags enabled
1951  
1952 With C<latin1> (or C<ascii>) enabled, C<encode> will escape
1953 characters with ordinal values > 255 (> 127 with C<ascii>) and encode
1954 the remaining characters as specified by the C<utf8> flag.
1955 With C<binary> enabled, ordinal values > 255 are illegal.
1956  
1957 If C<utf8> is disabled, then the result is also correctly encoded in those
1958 character sets (as both are proper subsets of Unicode, meaning that a
1959 Unicode string with all character values < 256 is the same thing as a
1960 ISO-8859-1 string, and a Unicode string with all character values < 128 is
1961 the same thing as an ASCII string in Perl).
1962  
1963 If C<utf8> is enabled, you still get a correct UTF-8-encoded string,
1964 regardless of these flags, just some more characters will be escaped using
1965 C<\uXXXX> then before.
1966  
1967 Note that ISO-8859-1-I<encoded> strings are not compatible with UTF-8
1968 encoding, while ASCII-encoded strings are. That is because the ISO-8859-1
1969 encoding is NOT a subset of UTF-8 (despite the ISO-8859-1 I<codeset> being
1970 a subset of Unicode), while ASCII is.
1971  
1972 Surprisingly, C<decode> will ignore these flags and so treat all input
1973 values as governed by the C<utf8> flag. If it is disabled, this allows you
1974 to decode ISO-8859-1- and ASCII-encoded strings, as both strict subsets of
1975 Unicode. If it is enabled, you can correctly decode UTF-8 encoded strings.
1976  
1977 So neither C<latin1>, C<binary> nor C<ascii> are incompatible with the
1978 C<utf8> flag - they only govern when the JSON output engine escapes a
1979 character or not.
1980  
1981 The main use for C<latin1> or C<binary> is to relatively efficiently
1982 store binary data as JSON, at the expense of breaking compatibility
1983 with most JSON decoders.
1984  
1985 The main use for C<ascii> is to force the output to not contain characters
1986 with values > 127, which means you can interpret the resulting string
1987 as UTF-8, ISO-8859-1, ASCII, KOI8-R or most about any character set and
1988 8-bit-encoding, and still get the same data structure back. This is useful
1989 when your channel for JSON transfer is not 8-bit clean or the encoding
1990 might be mangled in between (e.g. in mail), and works because ASCII is a
1991 proper subset of most 8-bit and multibyte encodings in use in the world.
1992  
1993 =back
1994  
1995  
1996 =head2 JSON and ECMAscript
1997  
1998 JSON syntax is based on how literals are represented in javascript (the
1999 not-standardized predecessor of ECMAscript) which is presumably why it is
2000 called "JavaScript Object Notation".
2001  
2002 However, JSON is not a subset (and also not a superset of course) of
2003 ECMAscript (the standard) or javascript (whatever browsers actually
2004 implement).
2005  
2006 If you want to use javascript's C<eval> function to "parse" JSON, you
2007 might run into parse errors for valid JSON texts, or the resulting data
2008 structure might not be queryable:
2009  
2010 One of the problems is that U+2028 and U+2029 are valid characters inside
2011 JSON strings, but are not allowed in ECMAscript string literals, so the
2012 following Perl fragment will not output something that can be guaranteed
2013 to be parsable by javascript's C<eval>:
2014  
2015 use Cpanel::JSON::XS;
2016  
2017 print encode_json [chr 0x2028];
2018  
2019 The right fix for this is to use a proper JSON parser in your javascript
2020 programs, and not rely on C<eval> (see for example Douglas Crockford's
2021 F<json2.js> parser).
2022  
2023 If this is not an option, you can, as a stop-gap measure, simply encode to
2024 ASCII-only JSON:
2025  
2026 use Cpanel::JSON::XS;
2027  
2028 print Cpanel::JSON::XS->new->ascii->encode ([chr 0x2028]);
2029  
2030 Note that this will enlarge the resulting JSON text quite a bit if you
2031 have many non-ASCII characters. You might be tempted to run some regexes
2032 to only escape U+2028 and U+2029, e.g.:
2033  
2034 # DO NOT USE THIS!
2035 my $json = Cpanel::JSON::XS->new->utf8->encode ([chr 0x2028]);
2036 $json =~ s/\xe2\x80\xa8/\\u2028/g; # escape U+2028
2037 $json =~ s/\xe2\x80\xa9/\\u2029/g; # escape U+2029
2038 print $json;
2039  
2040 Note that I<this is a bad idea>: the above only works for U+2028 and
2041 U+2029 and thus only for fully ECMAscript-compliant parsers. Many existing
2042 javascript implementations, however, have issues with other characters as
2043 well - using C<eval> naively simply I<will> cause problems.
2044  
2045 Another problem is that some javascript implementations reserve
2046 some property names for their own purposes (which probably makes
2047 them non-ECMAscript-compliant). For example, Iceweasel reserves the
2048 C<__proto__> property name for its own purposes.
2049  
2050 If that is a problem, you could parse try to filter the resulting JSON
2051 output for these property strings, e.g.:
2052  
2053 $json =~ s/"__proto__"\s*:/"__proto__renamed":/g;
2054  
2055 This works because C<__proto__> is not valid outside of strings, so every
2056 occurrence of C<"__proto__"\s*:> must be a string used as property name.
2057  
2058 Raw non-Unicode characters outside the valid unicode range fail now to
2059 parse, because "A string is a sequence of zero or more Unicode
2060 characters" RFC 7159 section 1 and "JSON text SHALL be encoded in
2061 Unicode RFC 7159 section 8.1. We use now the UTF8_DISALLOW_SUPER
2062 flag when parsing unicode.
2063  
2064 Since RFC 3629, U+D800 through U+DFFF are not legal Unicode values and
2065 their UTF-8 encodings must be treated as an invalid byte sequence.
2066 RFC 8259 section 8.2 admits the spec allows string values that contain
2067 bit sequences that cannot encode Unicode characters and that the
2068 behavior of software that receives such values is unpredictable. To
2069 avoid introducing non-Unicode strings into Perl we use the
2070 UTF8_DISALLOW_SURROGATE flag when parsing Unicode and verify escaped
2071 surrogates form valid pairs.
2072  
2073 If you know of other incompatibilities, please let me know.
2074  
2075  
2076 =head2 JSON and YAML
2077  
2078 You often hear that JSON is a subset of YAML. I<in general, there is
2079 no way to configure JSON::XS to output a data structure as valid YAML>
2080 that works in all cases. If you really must use Cpanel::JSON::XS to
2081 generate YAML, you should use this algorithm (subject to change in
2082 future versions):
2083  
2084 my $to_yaml = Cpanel::JSON::XS->new->utf8->space_after (1);
2085 my $yaml = $to_yaml->encode ($ref) . "\n";
2086  
2087 This will I<usually> generate JSON texts that also parse as valid
2088 YAML.
2089  
2090  
2091 =head2 SPEED
2092  
2093 It seems that JSON::XS is surprisingly fast, as shown in the following
2094 tables. They have been generated with the help of the C<eg/bench> program
2095 in the JSON::XS distribution, to make it easy to compare on your own
2096 system.
2097  
2098 JSON::XS is with L<Data::MessagePack> and L<Sereal> one of the fastest
2099 serializers, because JSON and JSON::XS do not support backrefs (no
2100 graph structures), only trees. Storable supports backrefs,
2101 i.e. graphs. Data::MessagePack encodes its data binary (as Storable)
2102 and supports only very simple subset of JSON.
2103  
2104 First comes a comparison between various modules using
2105 a very short single-line JSON string (also available at
2106 L<http://dist.schmorp.de/misc/json/short.json>).
2107  
2108 {"method": "handleMessage", "params": ["user1",
2109 "we were just talking"], "id": null, "array":[1,11,234,-5,1e5,1e7,
2110 1, 0]}
2111  
2112 It shows the number of encodes/decodes per second (JSON::XS uses
2113 the functional interface, while Cpanel::JSON::XS/2 uses the OO interface
2114 with pretty-printing and hash key sorting enabled, Cpanel::JSON::XS/3 enables
2115 shrink. JSON::DWIW/DS uses the deserialize function, while JSON::DWIW::FJ
2116 uses the from_json method). Higher is better:
2117  
2118 module | encode | decode |
2119 --------------|------------|------------|
2120 JSON::DWIW/DS | 86302.551 | 102300.098 |
2121 JSON::DWIW/FJ | 86302.551 | 75983.768 |
2122 JSON::PP | 15827.562 | 6638.658 |
2123 JSON::Syck | 63358.066 | 47662.545 |
2124 JSON::XS | 511500.488 | 511500.488 |
2125 JSON::XS/2 | 291271.111 | 388361.481 |
2126 JSON::XS/3 | 361577.931 | 361577.931 |
2127 Storable | 66788.280 | 265462.278 |
2128 --------------+------------+------------+
2129  
2130 That is, JSON::XS is almost six times faster than JSON::DWIW on encoding,
2131 about five times faster on decoding, and over thirty to seventy times
2132 faster than JSON's pure perl implementation. It also compares favourably
2133 to Storable for small amounts of data.
2134  
2135 Using a longer test string (roughly 18KB, generated from Yahoo! Locals
2136 search API (L<http://dist.schmorp.de/misc/json/long.json>).
2137  
2138 module | encode | decode |
2139 --------------|------------|------------|
2140 JSON::DWIW/DS | 1647.927 | 2673.916 |
2141 JSON::DWIW/FJ | 1630.249 | 2596.128 |
2142 JSON::PP | 400.640 | 62.311 |
2143 JSON::Syck | 1481.040 | 1524.869 |
2144 JSON::XS | 20661.596 | 9541.183 |
2145 JSON::XS/2 | 10683.403 | 9416.938 |
2146 JSON::XS/3 | 20661.596 | 9400.054 |
2147 Storable | 19765.806 | 10000.725 |
2148 --------------+------------+------------+
2149  
2150 Again, JSON::XS leads by far (except for Storable which non-surprisingly
2151 decodes a bit faster).
2152  
2153 On large strings containing lots of high Unicode characters, some modules
2154 (such as JSON::PC) seem to decode faster than JSON::XS, but the result
2155 will be broken due to missing (or wrong) Unicode handling. Others refuse
2156 to decode or encode properly, so it was impossible to prepare a fair
2157 comparison table for that case.
2158  
2159 For updated graphs see L<https://github.com/Sereal/Sereal/wiki/Sereal-Comparison-Graphs>
2160  
2161  
2162 =head1 INTEROP with JSON and JSON::XS and other JSON modules
2163  
2164 As long as you only serialize data that can be directly expressed in
2165 JSON, C<Cpanel::JSON::XS> is incapable of generating invalid JSON
2166 output (modulo bugs, but C<JSON::XS> has found more bugs in the
2167 official JSON testsuite (1) than the official JSON testsuite has found
2168 in C<JSON::XS> (0)).
2169 C<Cpanel::JSON::XS> is currently the only known JSON decoder which passes all
2170 L<http://seriot.ch/projects/parsing_json.html> tests, while being the fastest also.
2171  
2172 When you have trouble decoding JSON generated by this module using other
2173 decoders, then it is very likely that you have an encoding mismatch or the
2174 other decoder is broken.
2175  
2176 When decoding, C<JSON::XS> is strict by default and will likely catch
2177 all errors. There are currently two settings that change this:
2178 C<relaxed> makes C<JSON::XS> accept (but not generate) some
2179 non-standard extensions, and C<allow_tags> or C<allow_blessed> will
2180 allow you to encode and decode Perl objects, at the cost of being
2181 totally insecure and not outputting valid JSON anymore.
2182  
2183 JSON-XS-3.01 broke interoperability with JSON-2.90 with booleans. See L<JSON>.
2184  
2185 Cpanel::JSON::XS needs to know the JSON and JSON::XS versions to be able work
2186 with those objects, especially when encoding a booleans like C<{"is_true":true}>.
2187 So you need to load these modules before.
2188  
2189 true/false overloading and boolean representations are supported.
2190  
2191 JSON::XS and JSON::PP representations are accepted and older JSON::XS
2192 accepts Cpanel::JSON::XS booleans. All JSON modules JSON, JSON, PP,
2193 JSON::XS, Cpanel::JSON::XS produce JSON::PP::Boolean objects, just
2194 Mojo and JSON::YAJL not. Mojo produces Mojo::JSON::_Bool and
2195 JSON::YAJL::Parser just an unblessed IV.
2196  
2197 Cpanel::JSON::XS accepts JSON::PP::Boolean and Mojo::JSON::_Bool
2198 objects as booleans.
2199  
2200 I cannot think of any reason to still use JSON::XS anymore.
2201  
2202  
2203 =head2 TAGGED VALUE SYNTAX AND STANDARD JSON EN/DECODERS
2204  
2205 When you use C<allow_tags> to use the extended (and also nonstandard
2206 and invalid) JSON syntax for serialized objects, and you still want to
2207 decode the generated serialize objects, you can run a regex to replace
2208 the tagged syntax by standard JSON arrays (it only works for "normal"
2209 package names without comma, newlines or single colons). First, the
2210 readable Perl version:
2211  
2212 # if your FREEZE methods return no values, you need this replace first:
2213 $json =~ s/\( \s* (" (?: [^\\":,]+|\\.|::)* ") \s* \) \s* \[\s*\]/[$1]/gx;
2214  
2215 # this works for non-empty constructor arg lists:
2216 $json =~ s/\( \s* (" (?: [^\\":,]+|\\.|::)* ") \s* \) \s* \[/[$1,/gx;
2217  
2218 And here is a less readable version that is easy to adapt to other
2219 languages:
2220  
2221 $json =~ s/\(\s*("([^\\":,]+|\\.|::)*")\s*\)\s*\[/[$1,/g;
2222  
2223 Here is an ECMAScript version (same regex):
2224  
2225 json = json.replace (/\(\s*("([^\\":,]+|\\.|::)*")\s*\)\s*\[/g, "[$1,");
2226  
2227 Since this syntax converts to standard JSON arrays, it might be hard to
2228 distinguish serialized objects from normal arrays. You can prepend a
2229 "magic number" as first array element to reduce chances of a collision:
2230  
2231 $json =~ s/\(\s*("([^\\":,]+|\\.|::)*")\s*\)\s*\[/["XU1peReLzT4ggEllLanBYq4G9VzliwKF",$1,/g;
2232  
2233 And after decoding the JSON text, you could walk the data
2234 structure looking for arrays with a first element of
2235 C<XU1peReLzT4ggEllLanBYq4G9VzliwKF>.
2236  
2237 The same approach can be used to create the tagged format with another
2238 encoder. First, you create an array with the magic string as first member,
2239 the classname as second, and constructor arguments last, encode it as part
2240 of your JSON structure, and then:
2241  
2242 $json =~ s/\[\s*"XU1peReLzT4ggEllLanBYq4G9VzliwKF"\s*,\s*("([^\\":,]+|\\.|::)*")\s*,/($1)[/g;
2243  
2244 Again, this has some limitations - the magic string must not be encoded
2245 with character escapes, and the constructor arguments must be non-empty.
2246  
2247  
2248 =head1 RFC7159
2249  
2250 Since this module was written, Google has written a new JSON RFC, RFC 7159
2251 (and RFC7158). Unfortunately, this RFC breaks compatibility with both the
2252 original JSON specification on www.json.org and RFC4627.
2253  
2254 As far as I can see, you can get partial compatibility when parsing by
2255 using C<< ->allow_nonref >>. However, consider the security implications
2256 of doing so.
2257  
2258 I haven't decided yet when to break compatibility with RFC4627 by default
2259 (and potentially leave applications insecure) and change the default to
2260 follow RFC7159, but application authors are well advised to call C<<
2261 ->allow_nonref(0) >> even if this is the current default, if they cannot
2262 handle non-reference values, in preparation for the day when the default
2263 will change.
2264  
2265 =head1 SECURITY CONSIDERATIONS
2266  
2267 JSON::XS and Cpanel::JSON::XS are not only fast. JSON is generally the
2268 most secure serializing format, because it is the only one besides
2269 Data::MessagePack, which does not deserialize objects per default. For
2270 all languages, not just perl. The binary variant BSON (MongoDB) does
2271 more but is unsafe.
2272  
2273 It is trivial for any attacker to create such serialized objects in
2274 JSON and trick perl into expanding them, thereby triggering certain
2275 methods. Watch L<https://www.youtube.com/watch?v=Gzx6KlqiIZE> for an
2276 exploit demo for "CVE-2015-1592 SixApart MovableType Storable Perl
2277 Code Execution" for a deserializer which expands objects.
2278 Deserializing even coderefs (methods, functions) or external
2279 data would be considered the most dangerous.
2280  
2281 Security relevant overview of serializers regarding deserializing
2282 objects by default:
2283  
2284 Objects Coderefs External Data
2285  
2286 Data::Dumper YES YES YES
2287 Storable YES NO (def) NO
2288 Sereal YES NO NO
2289 YAML YES NO NO
2290 B::C YES YES YES
2291 B::Bytecode YES YES YES
2292 BSON YES YES NO
2293 JSON::SL YES NO YES
2294 JSON NO (def) NO NO
2295 Data::MessagePack NO NO NO
2296 XML NO NO YES
2297  
2298 Pickle YES YES YES
2299 PHP Deserialize YES NO NO
2300  
2301 When you are using JSON in a protocol, talking to untrusted potentially
2302 hostile creatures requires relatively few measures.
2303  
2304 First of all, your JSON decoder should be secure, that is, should not have
2305 any buffer overflows. Obviously, this module should ensure that.
2306  
2307 Second, you need to avoid resource-starving attacks. That means you should
2308 limit the size of JSON texts you accept, or make sure then when your
2309 resources run out, that's just fine (e.g. by using a separate process that
2310 can crash safely). The size of a JSON text in octets or characters is
2311 usually a good indication of the size of the resources required to decode
2312 it into a Perl structure. While JSON::XS can check the size of the JSON
2313 text, it might be too late when you already have it in memory, so you
2314 might want to check the size before you accept the string.
2315  
2316 Third, Cpanel::JSON::XS recurses using the C stack when decoding objects and
2317 arrays. The C stack is a limited resource: for instance, on my amd64
2318 machine with 8MB of stack size I can decode around 180k nested arrays but
2319 only 14k nested JSON objects (due to perl itself recursing deeply on croak
2320 to free the temporary). If that is exceeded, the program crashes. To be
2321 conservative, the default nesting limit is set to 512. If your process
2322 has a smaller stack, you should adjust this setting accordingly with the
2323 C<max_depth> method.
2324  
2325 Also keep in mind that Cpanel::JSON::XS might leak contents of your Perl data
2326 structures in its error messages, so when you serialize sensitive
2327 information you might want to make sure that exceptions thrown by JSON::XS
2328 will not end up in front of untrusted eyes.
2329  
2330 If you are using Cpanel::JSON::XS to return packets to consumption
2331 by JavaScript scripts in a browser you should have a look at
2332 L<http://blog.archive.jpsykes.com/47/practical-csrf-and-json-security/> to
2333 see whether you are vulnerable to some common attack vectors (which really
2334 are browser design bugs, but it is still you who will have to deal with
2335 it, as major browser developers care only for features, not about getting
2336 security right). You might also want to also look at L<Mojo::JSON>
2337 special escape rules to prevent from XSS attacks.
2338  
2339 =head1 "OLD" VS. "NEW" JSON (RFC 4627 VS. RFC 7159)
2340  
2341 B<Note: Since version 4.42, Cpanel::JSON::XS defaults to the broken RFC
2342 7159 behavior> (like JSON::XS 4.0+), accepting scalar top-level JSON values
2343 without requiring C<allow_nonref>. This was done for compatibility with the
2344 larger Perl JSON ecosystem, despite the security concerns outlined below.
2345  
2346 If you rely on the old strict behavior, explicitly disable it:
2347  
2348 my $json = Cpanel::JSON::XS->new->allow_nonref(0);
2349  
2350 The rationale for the original (pre-4.42) default, and the reason you might
2351 want to keep it off, follows:
2352  
2353 TL;DR: The original JSON RFC 4627 required JSON texts to be arrays or
2354 objects. The IETF later standardized a broken revision (RFC 7159) that
2355 allows bare scalars at top-level, opening a class of protocol confusion
2356 attacks.
2357  
2358 The long version: JSON being an important and supposedly stable format,
2359 the IETF standardized it as RFC 4627 in 2006. Unfortunately the inventor
2360 of JSON Douglas Crockford unilaterally changed the definition of JSON in
2361 javascript. Rather than create a fork, the IETF decided to standardize the
2362 new syntax (apparently, so I was told, without finding it very amusing).
2363  
2364 The biggest difference between the original JSON and the new JSON is that
2365 the new JSON supports scalars (anything other than arrays and objects) at
2366 the top-level of a JSON text. While this is strictly backwards compatible
2367 the top-level of a JSON text. While this is supposedly backwards compatible
2368 to older versions, it breaks a number of protocols that relied on sending
2369 JSON back-to-back, and is a security concern.
2370  
2371 For example, imagine you have two banks communicating, and on one side,
2372 the JSON coder gets upgraded. Two messages, such as C<10> and C<1000>
2373 might then be confused to mean C<101000>, something that could not happen
2374 in the original JSON, because neither of these messages would be valid
2375 JSON.
2376  
2377 If one side accepts these messages, then an upgrade in the coder on either
2378 side could result in this becoming exploitable.
2379  
2380 The IETF's own I-JSON profile (RFC 7493, also edited by Tim Bray who
2381 authored RFC 7159) acknowledges this mistake: Section 4.1 states that
2382 "there are software implementations, coded to the older specification
2383 [RFC4627], which only accept JSON objects or JSON arrays at the top
2384 level of JSON texts. For maximum interoperability with such
2385 implementations, protocol designers SHOULD NOT use top-level JSON
2386 texts that are neither objects nor arrays."
2387  
2388 =head1 THREADS
2389  
2390 Cpanel::JSON::XS has proper ithreads support, unlike JSON::XS. If you
2391 encounter any bugs with thread support please report them.
2392  
2393 From Version 4.00 - 4.19 you couldn't encode true with threads::shared
2394 magic.
2395  
2396 =head1 BUGS
2397  
2398 While the goal of the Cpanel::JSON::XS module is to be correct, that
2399 unfortunately does not mean it's bug-free, only that the author thinks
2400 its design is bug-free. If you keep reporting bugs and tests they will
2401 be fixed swiftly, though.
2402  
2403 Since the JSON::XS author refuses to use a public bugtracker and
2404 prefers private emails, we use the tracker at B<github>, so you might want
2405 to report any issues twice. Once in private to MLEHMANN to be fixed in
2406 JSON::XS and one to our the public tracker. Issues fixed by JSON::XS
2407 with a new release will also be backported to Cpanel::JSON::XS and
2408 5.6.2, as long as cPanel relies on 5.6.2 and Cpanel::JSON::XS as our
2409 serializer of choice.
2410  
2411 L<https://github.com/rurban/Cpanel-JSON-XS/issues>
2412  
2413 =head1 LICENSE
2414  
2415 This module is available under the same licences as perl, the Artistic
2416 license and the GPL.
2417  
2418 =cut
2419  
2420 0 ▶sub allow_bigint {
Subroutine allow_bigint: not called CC=1 SCAR=6.9
Pod documented
2421 0 Carp::carp("allow_bigint() is obsoleted. use allow_bignum() instead.");
2422 }
2423  
2424 3 ▶sub decode_and_validate {
Subroutine decode_and_validate: called CC=2 SCAR=6.9
Pod documented
2425 3 my ($self, $json_text, $schema) = @_;
2426 3 require Cpanel::JSON::XS::Type;
2427 3 my $data = $self->decode($json_text, my $found_type);
2428 3 my @errors = Cpanel::JSON::XS::Type::check_type($data, $schema, $found_type);
2429 3 ▶ if (@errors) {
Branch: 1 branch
BranchTrueFalseTotal
if (@errors) 2 1 3
2430 2 require Carp;
2431 2 Carp::croak("JSON does not match schema:\n " . join("\n ", @errors));
2432 }
2433 1 return $data;
2434 }
2435  
2436 69 ▶BEGIN {
Subroutine BEGIN: called CC=3 SCAR=11.2
2437 package
2438 JSON::PP::Boolean;
2439  
2440 69 require overload;
2441  
2442 69 local $^W; # silence redefine warnings. no warnings 'redefine' does not help
2443 # These already come with JSON::PP::Boolean. Avoid redefine warning.
2444 69 ▶ if (!defined $JSON::PP::Boolean::VERSION or $JSON::PP::VERSION lt '4.00') {
Branch: 1 branch
BranchTrueFalseTotal
if (not defined $VERSION or $JSON::PP::VERSION lt "4.00") 69 0 69
Condition33%1/3
not defined $VERSION or $JSON::PP::VERSION lt "4.00"
l!l&&r!l&&!r
6900
MC/DC0%0/2
not defined $VERSION or $JSON::PP::VERSION lt "4.00"
not defined $VERSION $JSON::PP::VERSION lt "4.00"
Truth table33%1/3
not(A) or B
A= not defined $VERSIONB= $JSON::PP::VERSION lt "4.00"
ABresult
1X1
011
000
2445 69 &overload::unimport( 'overload', '0+', '++', '--' );
2446 &overload::import( 'overload',
2447 6 ▶ "0+" => sub { ${$_[0]} },
Subroutine __ANON__: called CC=1 SCAR=0.0
2448 0 ▶ "++" => sub { $_[0] = ${$_[0]} + 1 },
Subroutine __ANON__: not called CC=1 SCAR=6.9
2449 0 ▶ "--" => sub { $_[0] = ${$_[0]} - 1 },
Subroutine __ANON__: not called CC=1 SCAR=6.9
2450 69 );
2451 }
2452 # workaround 5.6 reserved keyword warning
2453 69 &overload::unimport( 'overload', '""', 'eq', 'ne' );
2454 &overload::import( 'overload',
2455 35 ▶ '""' => sub { ${$_[0]} == 1 ? '1' : '0' }, # GH 29
Subroutine __ANON__: called CC=2 SCAR=6.9
Branch: 1 branch
BranchTrueFalseTotal
${$_[0];} == 1 ? : 16 19 35
2456 # NOTE: unlike JSON::PP::Boolean which relies on fallback and has no
2457 # explicit eq/ne overload, we provide semantic boolean matching:
2458 # false eq "false", false eq "", false eq !!0, true eq "true".
2459 # JSON::PP would return FALSE for all of these (plain string eq).
2460 36 ▶ 'eq' => sub {
Subroutine __ANON__: called CC=9 SCAR=22.0
2461 36 ▶ my ($obj, $op) = $_[2] ? ($_[1], $_[0]) : ($_[0], $_[1]);
Branch: 1 branch
BranchTrueFalseTotal
$_[2] ? : 0 36 36
2462 36 ▶ if (ref $op) { # if 2nd also blessed might recurse endlessly
Branch: 2 branches
BranchTrueFalseTotal
if (ref $op) { } 6 30 36
elsif (not $op =~ /^[0-9]+$/) { } 18 12 30
2463 6 ▶ return $obj ? 1 == $op : 0 == $op;
Branch: 1 branch
BranchTrueFalseTotal
$obj ? : 3 3 6
2464 }
2465 # if string, only accept numbers or true|false or "" (e.g. !!0 / SV_NO)
2466 elsif ($op !~ /^[0-9]+$/) {
2467 18 ▶ return "$obj" eq '1' ? 'true' eq $op
Branch: 1 branch
BranchTrueFalseTotal
"$obj" eq '1' ? : 7 11 18
Condition100%6/6
"false" eq $op or defined $op and "" eq $op
l!l&&r!l&&!r
254
defined $op and "" eq $op
!ll&&!rl&&r
225
MC/DC100%3/3
"false" eq $op or defined $op and "" eq $op
"false" eq $op defined $op "" eq $op
Truth table100%4/4
A or B and C
A= "false" eq $opB= defined $opC= "" eq $op
ABCresult
1XX1
0111
00X0
0100
2468 : 'false' eq $op || (defined($op) && "" eq $op); # GH #207
2469 }
2470 else {
2471 12 ▶ return $obj ? 1 == $op : 0 == $op;
Branch: 1 branch
BranchTrueFalseTotal
$obj ? : 5 7 12
2472 }
2473 },
2474 6 ▶ 'ne' => sub {
Subroutine __ANON__: called CC=2 SCAR=7.2
2475 6 ▶ my ($obj, $op) = $_[2] ? ($_[1], $_[0]) : ($_[0], $_[1]);
Branch: 1 branch
BranchTrueFalseTotal
$_[2] ? : 0 6 6
2476 #warn "ne obj:$obj op:$op";
2477 6 return !($obj eq $op);
2478 },
2479 69 fallback => 1);
2480 }
2481  
2482 69 our ($true, $false);
2483 69 ▶BEGIN {
Subroutine BEGIN: called CC=4 SCAR=17.3
2484 69 ▶ if ($INC{'JSON/XS.pm'}
Branch: 1 branch
BranchTrueFalseTotal
if ($INC{'JSON/XS.pm'} and $INC{'Types/Serialiser.pm'} and $JSON::XS::VERSION ge '3.00') { } 0 69 69
Condition33%2/6
$INC{"JSON/XS.pm"} and $INC{"Types/Serialiser.pm"} and $JSON::XS::VERSION ge "3.00"
!ll&&!rl&&r
6900
$INC{"JSON/XS.pm"} and $INC{"Types/Serialiser.pm"}
!ll&&!rl&&r
6900
MC/DC0%0/3
$INC{"JSON/XS.pm"} and $INC{"Types/Serialiser.pm"} and $JSON::XS::VERSION ge "3.00"
$INC{"JSON/XS.pm"} $INC{"Types/Serialiser.pm"} $JSON::XS::VERSION ge "3.00"
Truth table25%1/4
A and B and C
A= $INC{"JSON/XS.pm"}B= $INC{"Types/Serialiser.pm"}C= $JSON::XS::VERSION ge "3.00"
ABCresult
0XX0
10X0
1100
1111
2485 and $INC{'Types/Serialiser.pm'}
2486 and $JSON::XS::VERSION ge "3.00") {
2487 0 $true = $Types::Serialiser::true; # readonly if loaded by JSON::XS
2488 0 $false = $Types::Serialiser::false;
2489 } else {
2490 69 $true = do { bless \(my $dummy = 1), "JSON::PP::Boolean" };
2491 69 $false = do { bless \(my $dummy = 0), "JSON::PP::Boolean" };
2492 }
2493 }
2494  
2495 69 ▶BEGIN {
Subroutine BEGIN: called CC=1 SCAR=0.4
2496 69 my $const_true = $true;
2497 69 my $const_false = $false;
2498 0 ▶ *true = sub () { $const_true };
Subroutine __ANON__: not called CC=1 SCAR=1.2
2499 0 ▶ *false = sub () { $const_false };
Subroutine __ANON__: not called CC=1 SCAR=1.2
2500 }
2501  
2502 37 ▶sub is_bool($) {
Subroutine is_bool: called CC=5 SCAR=16.5
Pod documented
2503 37 ▶ shift if @_ == 2; # as method call
Branch: 1 branch
BranchTrueFalseTotal
if @_ == 2 3 34 37
2504 (ref($_[0]) and UNIVERSAL::isa( $_[0], JSON::PP::Boolean::))
2505 37 ▶ or (exists $INC{'Types/Serialiser.pm'} and Types::Serialiser::is_bool($_[0]))
Branch: 1 branch
BranchTrueFalseTotal
ref $_[0] and UNIVERSAL::isa($_[0], "JSON::PP::Boolean") or exists $INC{"Types/Serialiser.pm"} and Types::Serialiser::is_bool($_[0]) 8 29 37
Condition66%6/9
ref $_[0] && UNIVERSAL::isa($_[0], 'JSON::PP::Boolean') || exists $INC{'Types/Serialiser.pm'} && Types::Serialiser::is_bool($_[0])
l!l&&r!l&&!r
8029
ref $_[0] and UNIVERSAL::isa($_[0], "JSON::PP::Boolean")
!ll&&!rl&&r
2818
exists $INC{"Types/Serialiser.pm"} and Types::Serialiser::is_bool($_[0])
!ll&&!rl&&r
2900
MC/DC50%2/4
ref $_[0] && UNIVERSAL::isa($_[0], 'JSON::PP::Boolean') || exists $INC{'Types/Serialiser.pm'} && Types::Serialiser::is_bool($_[0])
ref $_[0] UNIVERSAL::isa($_[0], "JSON::PP::Boolean") exists $INC{"Types/Serialiser.pm"} Types::Serialiser::is_bool($_[0])
Truth table42%3/7
A and B || C and D
A= ref $_[0]B= UNIVERSAL::isa($_[0], "JSON::PP::Boolean")C= exists $INC{"Types/Serialiser.pm"}D= Types::Serialiser::is_bool($_[0])
ABCDresult
11XX1
0X111
10111
0X0X0
0X100
100X0
10100
2506 }
2507  
2508 69 XSLoader::load 'Cpanel::JSON::XS', $XS_VERSION;
2509  
2510 69 1;
2511  
2512 =head1 SEE ALSO
2513  
2514 The F<cpanel_json_xs> command line utility for quick experiments.
2515  
2516 L<JSON::PP>, L<JSON>, L<JSON::XS>, L<JSON::MaybeXS>, L<Mojo::JSON>,
2517 L<Mojo::JSON::MaybeXS>, L<JSON::SL>, L<JSON::DWIW>, L<JSON::YAJL>,
2518 L<JSON::Any>, L<Test::JSON>, L<Locale::Wolowitz>,
2519 L<https://metacpan.org/search?q=JSON>
2520  
2521 L<https://tools.ietf.org/html/rfc7159>
2522  
2523 L<https://tools.ietf.org/html/rfc4627>
2524  
2525  
2526 =head1 AUTHOR
2527  
2528 Reini Urban <reini.urban@gmail.com>
2529  
2530 Marc Lehmann <schmorp@schmorp.de>, http://home.schmorp.de/
2531  
2532 =head1 MAINTAINER
2533  
2534 Reini Urban <reini.urban@gmail.com>
2535  
2536 =cut
2537  
« blib/lib/Cpanel/JSON/XS/Type.pm ↑ Summary blib/lib/Cpanel/JSON/XS/Boolean.pm »