… | |
… | |
85 | |
85 | |
86 | package JSON::XS; |
86 | package JSON::XS; |
87 | |
87 | |
88 | use strict; |
88 | use strict; |
89 | |
89 | |
90 | BEGIN { |
|
|
91 | our $VERSION = '1.12'; |
90 | our $VERSION = '1.4'; |
92 | our @ISA = qw(Exporter); |
91 | our @ISA = qw(Exporter); |
93 | |
92 | |
94 | our @EXPORT = qw(to_json from_json objToJson jsonToObj); |
93 | our @EXPORT = qw(to_json from_json objToJson jsonToObj); |
95 | require Exporter; |
|
|
96 | |
94 | |
97 | require XSLoader; |
95 | use Exporter; |
98 | XSLoader::load JSON::XS::, $VERSION; |
96 | use XSLoader; |
99 | } |
|
|
100 | |
97 | |
101 | =head1 FUNCTIONAL INTERFACE |
98 | =head1 FUNCTIONAL INTERFACE |
102 | |
99 | |
103 | The following convinience methods are provided by this module. They are |
100 | The following convinience methods are provided by this module. They are |
104 | exported by default: |
101 | exported by default: |
… | |
… | |
126 | This function call is functionally identical to: |
123 | This function call is functionally identical to: |
127 | |
124 | |
128 | $perl_scalar = JSON::XS->new->utf8->decode ($json_text) |
125 | $perl_scalar = JSON::XS->new->utf8->decode ($json_text) |
129 | |
126 | |
130 | except being faster. |
127 | except being faster. |
|
|
128 | |
|
|
129 | =item $is_boolean = JSON::XS::is_bool $scalar |
|
|
130 | |
|
|
131 | Returns true if the passed scalar represents either JSON::XS::true or |
|
|
132 | JSON::XS::false, two constants that act like C<1> and C<0>, respectively |
|
|
133 | and are used to represent JSON C<true> and C<false> values in Perl. |
|
|
134 | |
|
|
135 | See MAPPING, below, for more information on how JSON values are mapped to |
|
|
136 | Perl. |
131 | |
137 | |
132 | =back |
138 | =back |
133 | |
139 | |
134 | |
140 | |
135 | =head1 OBJECT-ORIENTED INTERFACE |
141 | =head1 OBJECT-ORIENTED INTERFACE |
… | |
… | |
309 | Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>, |
315 | Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>, |
310 | resulting in an invalid JSON text: |
316 | resulting in an invalid JSON text: |
311 | |
317 | |
312 | JSON::XS->new->allow_nonref->encode ("Hello, World!") |
318 | JSON::XS->new->allow_nonref->encode ("Hello, World!") |
313 | => "Hello, World!" |
319 | => "Hello, World!" |
|
|
320 | |
|
|
321 | =item $json = $json->allow_blessed ([$enable]) |
|
|
322 | |
|
|
323 | If C<$enable> is true (or missing), then the C<encode> method will not |
|
|
324 | barf when it encounters a blessed reference. Instead, the value of the |
|
|
325 | B<convert_blessed> option will decide wether C<null> (C<convert_blessed> |
|
|
326 | disabled or no C<to_json> method found) or a representation of the |
|
|
327 | object (C<convert_blessed> enabled and C<to_json> method found) is being |
|
|
328 | encoded. Has no effect on C<decode>. |
|
|
329 | |
|
|
330 | If C<$enable> is false (the default), then C<encode> will throw an |
|
|
331 | exception when it encounters a blessed object. |
|
|
332 | |
|
|
333 | =item $json = $json->convert_blessed ([$enable]) |
|
|
334 | |
|
|
335 | If C<$enable> is true (or missing), then C<encode>, upon encountering a |
|
|
336 | blessed object, will check for the availability of the C<TO_JSON> method |
|
|
337 | on the object's class. If found, it will be called in scalar context |
|
|
338 | and the resulting scalar will be encoded instead of the object. If no |
|
|
339 | C<TO_JSON> method is found, the value of C<allow_blessed> will decide what |
|
|
340 | to do. |
|
|
341 | |
|
|
342 | The C<TO_JSON> method may safely call die if it wants. If C<TO_JSON> |
|
|
343 | returns other blessed objects, those will be handled in the same |
|
|
344 | way. C<TO_JSON> must take care of not causing an endless recursion cycle |
|
|
345 | (== crash) in this case. The name of C<TO_JSON> was chosen because other |
|
|
346 | methods called by the Perl core (== not by the user of the object) are |
|
|
347 | usually in upper case letters and to avoid collisions with the C<to_json> |
|
|
348 | function. |
|
|
349 | |
|
|
350 | This setting does not yet influence C<decode> in any way, but in the |
|
|
351 | future, global hooks might get installed that influence C<decode> and are |
|
|
352 | enabled by this setting. |
|
|
353 | |
|
|
354 | If C<$enable> is false, then the C<allow_blessed> setting will decide what |
|
|
355 | to do when a blessed object is found. |
314 | |
356 | |
315 | =item $json = $json->shrink ([$enable]) |
357 | =item $json = $json->shrink ([$enable]) |
316 | |
358 | |
317 | Perl usually over-allocates memory a bit when allocating space for |
359 | Perl usually over-allocates memory a bit when allocating space for |
318 | strings. This flag optionally resizes strings generated by either |
360 | strings. This flag optionally resizes strings generated by either |
… | |
… | |
351 | given character in a string. |
393 | given character in a string. |
352 | |
394 | |
353 | Setting the maximum depth to one disallows any nesting, so that ensures |
395 | Setting the maximum depth to one disallows any nesting, so that ensures |
354 | that the object is only a single hash/object or array. |
396 | that the object is only a single hash/object or array. |
355 | |
397 | |
356 | The argument to C<max_depth> will be rounded up to the next nearest power |
398 | The argument to C<max_depth> will be rounded up to the next highest power |
357 | of two. |
399 | of two. If no argument is given, the highest possible setting will be |
|
|
400 | used, which is rarely useful. |
|
|
401 | |
|
|
402 | See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
|
|
403 | |
|
|
404 | =item $json = $json->max_size ([$maximum_string_size]) |
|
|
405 | |
|
|
406 | Set the maximum length a JSON text may have (in bytes) where decoding is |
|
|
407 | being attempted. The default is C<0>, meaning no limit. When C<decode> |
|
|
408 | is called on a string longer then this number of characters it will not |
|
|
409 | attempt to decode the string but throw an exception. This setting has no |
|
|
410 | effect on C<encode> (yet). |
|
|
411 | |
|
|
412 | The argument to C<max_size> will be rounded up to the next B<highest> |
|
|
413 | power of two (so may be more than requested). If no argument is given, the |
|
|
414 | limit check will be deactivated (same as when C<0> is specified). |
358 | |
415 | |
359 | See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
416 | See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
360 | |
417 | |
361 | =item $json_text = $json->encode ($perl_scalar) |
418 | =item $json_text = $json->encode ($perl_scalar) |
362 | |
419 | |
… | |
… | |
374 | |
431 | |
375 | JSON numbers and strings become simple Perl scalars. JSON arrays become |
432 | JSON numbers and strings become simple Perl scalars. JSON arrays become |
376 | Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes |
433 | Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes |
377 | C<1>, C<false> becomes C<0> and C<null> becomes C<undef>. |
434 | C<1>, C<false> becomes C<0> and C<null> becomes C<undef>. |
378 | |
435 | |
|
|
436 | =item ($perl_scalar, $characters) = $json->decode_prefix ($json_text) |
|
|
437 | |
|
|
438 | This works like the C<decode> method, but instead of raising an exception |
|
|
439 | when there is trailing garbage after the first JSON object, it will |
|
|
440 | silently stop parsing there and return the number of characters consumed |
|
|
441 | so far. |
|
|
442 | |
|
|
443 | This is useful if your JSON texts are not delimited by an outer protocol |
|
|
444 | (which is not the brightest thing to do in the first place) and you need |
|
|
445 | to know where the JSON text ends. |
|
|
446 | |
|
|
447 | JSON::XS->new->decode_prefix ("[1] the tail") |
|
|
448 | => ([], 3) |
|
|
449 | |
379 | =back |
450 | =back |
380 | |
451 | |
381 | |
452 | |
382 | =head1 MAPPING |
453 | =head1 MAPPING |
383 | |
454 | |
… | |
… | |
387 | (what you put in comes out as something equivalent). |
458 | (what you put in comes out as something equivalent). |
388 | |
459 | |
389 | For the more enlightened: note that in the following descriptions, |
460 | For the more enlightened: note that in the following descriptions, |
390 | lowercase I<perl> refers to the Perl interpreter, while uppcercase I<Perl> |
461 | lowercase I<perl> refers to the Perl interpreter, while uppcercase I<Perl> |
391 | refers to the abstract Perl language itself. |
462 | refers to the abstract Perl language itself. |
|
|
463 | |
392 | |
464 | |
393 | =head2 JSON -> PERL |
465 | =head2 JSON -> PERL |
394 | |
466 | |
395 | =over 4 |
467 | =over 4 |
396 | |
468 | |
… | |
… | |
417 | conversion details, but an integer may take slightly less memory and might |
489 | conversion details, but an integer may take slightly less memory and might |
418 | represent more values exactly than (floating point) numbers. |
490 | represent more values exactly than (floating point) numbers. |
419 | |
491 | |
420 | =item true, false |
492 | =item true, false |
421 | |
493 | |
422 | These JSON atoms become C<0>, C<1>, respectively. Information is lost in |
494 | These JSON atoms become C<JSON::XS::true> and C<JSON::XS::false>, |
423 | this process. Future versions might represent those values differently, |
495 | respectively. They are overloaded to act almost exactly like the numbers |
424 | but they will be guarenteed to act like these integers would normally in |
496 | C<1> and C<0>. You can check wether a scalar is a JSON boolean by using |
425 | Perl. |
497 | the C<JSON::XS::is_bool> function. |
426 | |
498 | |
427 | =item null |
499 | =item null |
428 | |
500 | |
429 | A JSON null atom becomes C<undef> in Perl. |
501 | A JSON null atom becomes C<undef> in Perl. |
430 | |
502 | |
431 | =back |
503 | =back |
|
|
504 | |
432 | |
505 | |
433 | =head2 PERL -> JSON |
506 | =head2 PERL -> JSON |
434 | |
507 | |
435 | The mapping from Perl to JSON is slightly more difficult, as Perl is a |
508 | The mapping from Perl to JSON is slightly more difficult, as Perl is a |
436 | truly typeless language, so we can only guess which JSON type is meant by |
509 | truly typeless language, so we can only guess which JSON type is meant by |
… | |
… | |
461 | C<1>, which get turned into C<false> and C<true> atoms in JSON. You can |
534 | C<1>, which get turned into C<false> and C<true> atoms in JSON. You can |
462 | also use C<JSON::XS::false> and C<JSON::XS::true> to improve readability. |
535 | also use C<JSON::XS::false> and C<JSON::XS::true> to improve readability. |
463 | |
536 | |
464 | to_json [\0,JSON::XS::true] # yields [false,true] |
537 | to_json [\0,JSON::XS::true] # yields [false,true] |
465 | |
538 | |
|
|
539 | =item JSON::XS::true, JSON::XS::false |
|
|
540 | |
|
|
541 | These special values become JSON true and JSON false values, |
|
|
542 | respectively. You cna alos use C<\1> and C<\0> directly if you want. |
|
|
543 | |
466 | =item blessed objects |
544 | =item blessed objects |
467 | |
545 | |
468 | Blessed objects are not allowed. JSON::XS currently tries to encode their |
546 | Blessed objects are not allowed. JSON::XS currently tries to encode their |
469 | underlying representation (hash- or arrayref), but this behaviour might |
547 | underlying representation (hash- or arrayref), but this behaviour might |
470 | change in future versions. |
548 | change in future versions. |
… | |
… | |
591 | |
669 | |
592 | Does not check input for validity. |
670 | Does not check input for validity. |
593 | |
671 | |
594 | =back |
672 | =back |
595 | |
673 | |
|
|
674 | |
|
|
675 | =head2 JSON and YAML |
|
|
676 | |
|
|
677 | You often hear that JSON is a subset (or a close subset) of YAML. This is, |
|
|
678 | however, a mass hysteria and very far from the truth. In general, there is |
|
|
679 | no way to configure JSON::XS to output a data structure as valid YAML. |
|
|
680 | |
|
|
681 | If you really must use JSON::XS to generate YAML, you should use this |
|
|
682 | algorithm (subject to change in future versions): |
|
|
683 | |
|
|
684 | my $to_yaml = JSON::XS->new->utf8->space_after (1); |
|
|
685 | my $yaml = $to_yaml->encode ($ref) . "\n"; |
|
|
686 | |
|
|
687 | This will usually generate JSON texts that also parse as valid |
|
|
688 | YAML. Please note that YAML has hardcoded limits on (simple) object key |
|
|
689 | lengths that JSON doesn't have, so you should make sure that your hash |
|
|
690 | keys are noticably shorter than the 1024 characters YAML allows. |
|
|
691 | |
|
|
692 | There might be other incompatibilities that I am not aware of. In general |
|
|
693 | you should not try to generate YAML with a JSON generator or vice versa, |
|
|
694 | or try to parse JSON with a YAML parser or vice versa: chances are high |
|
|
695 | that you will run into severe interoperability problems. |
|
|
696 | |
|
|
697 | |
596 | =head2 SPEED |
698 | =head2 SPEED |
597 | |
699 | |
598 | It seems that JSON::XS is surprisingly fast, as shown in the following |
700 | It seems that JSON::XS is surprisingly fast, as shown in the following |
599 | tables. They have been generated with the help of the C<eg/bench> program |
701 | tables. They have been generated with the help of the C<eg/bench> program |
600 | in the JSON::XS distribution, to make it easy to compare on your own |
702 | in the JSON::XS distribution, to make it easy to compare on your own |
601 | system. |
703 | system. |
602 | |
704 | |
603 | First comes a comparison between various modules using a very short JSON |
705 | First comes a comparison between various modules using a very short |
604 | string: |
706 | single-line JSON string: |
605 | |
707 | |
606 | {"method": "handleMessage", "params": ["user1", "we were just talking"], "id": null} |
708 | {"method": "handleMessage", "params": ["user1", "we were just talking"], \ |
|
|
709 | "id": null, "array":[1,11,234,-5,1e5,1e7, true, false]} |
607 | |
710 | |
608 | It shows the number of encodes/decodes per second (JSON::XS uses the |
711 | It shows the number of encodes/decodes per second (JSON::XS uses |
609 | functional interface, while JSON::XS/2 uses the OO interface with |
712 | the functional interface, while JSON::XS/2 uses the OO interface |
610 | pretty-printing and hashkey sorting enabled). Higher is better: |
713 | with pretty-printing and hashkey sorting enabled, JSON::XS/3 enables |
|
|
714 | shrink). Higher is better: |
611 | |
715 | |
612 | module | encode | decode | |
716 | module | encode | decode | |
613 | -----------|------------|------------| |
717 | -----------|------------|------------| |
614 | JSON | 11488.516 | 7823.035 | |
718 | JSON | 7645.468 | 4208.613 | |
615 | JSON::DWIW | 94708.054 | 129094.260 | |
719 | JSON::DWIW | 40721.398 | 77101.176 | |
616 | JSON::PC | 63884.157 | 128528.212 | |
720 | JSON::PC | 65948.176 | 78251.940 | |
617 | JSON::Syck | 34898.677 | 42096.911 | |
721 | JSON::Syck | 22844.793 | 26479.192 | |
618 | JSON::XS | 654027.064 | 396423.669 | |
722 | JSON::XS | 388361.481 | 199728.762 | |
619 | JSON::XS/2 | 371564.190 | 371725.613 | |
723 | JSON::XS/2 | 218453.333 | 192399.266 | |
|
|
724 | JSON::XS/3 | 338250.323 | 192399.266 | |
|
|
725 | Storable | 15779.925 | 14169.946 | |
620 | -----------+------------+------------+ |
726 | -----------+------------+------------+ |
621 | |
727 | |
622 | That is, JSON::XS is more than six times faster than JSON::DWIW on |
728 | That is, JSON::XS is about five times faster than JSON::DWIW on encoding, |
623 | encoding, more than three times faster on decoding, and about thirty times |
729 | about three times faster on decoding, and over fourty times faster |
624 | faster than JSON, even with pretty-printing and key sorting. |
730 | than JSON, even with pretty-printing and key sorting. It also compares |
|
|
731 | favourably to Storable for small amounts of data. |
625 | |
732 | |
626 | Using a longer test string (roughly 18KB, generated from Yahoo! Locals |
733 | Using a longer test string (roughly 18KB, generated from Yahoo! Locals |
627 | search API (http://nanoref.com/yahooapis/mgPdGg): |
734 | search API (http://nanoref.com/yahooapis/mgPdGg): |
628 | |
735 | |
629 | module | encode | decode | |
736 | module | encode | decode | |
630 | -----------|------------|------------| |
737 | -----------|------------|------------| |
631 | JSON | 273.023 | 44.674 | |
738 | JSON | 254.685 | 37.665 | |
632 | JSON::DWIW | 1089.383 | 1145.704 | |
739 | JSON::DWIW | 843.343 | 1049.731 | |
633 | JSON::PC | 3097.419 | 2393.921 | |
740 | JSON::PC | 3602.116 | 2307.352 | |
634 | JSON::Syck | 514.060 | 843.053 | |
741 | JSON::Syck | 505.107 | 787.899 | |
635 | JSON::XS | 6479.668 | 3636.364 | |
742 | JSON::XS | 5747.196 | 3690.220 | |
636 | JSON::XS/2 | 3774.221 | 3599.124 | |
743 | JSON::XS/2 | 3968.121 | 3676.634 | |
|
|
744 | JSON::XS/3 | 6105.246 | 3662.508 | |
|
|
745 | Storable | 4417.337 | 5285.161 | |
637 | -----------+------------+------------+ |
746 | -----------+------------+------------+ |
638 | |
747 | |
639 | Again, JSON::XS leads by far. |
748 | Again, JSON::XS leads by far (except for Storable which non-surprisingly |
|
|
749 | decodes faster). |
640 | |
750 | |
641 | On large strings containing lots of high unicode characters, some modules |
751 | On large strings containing lots of high unicode characters, some modules |
642 | (such as JSON::PC) seem to decode faster than JSON::XS, but the result |
752 | (such as JSON::PC) seem to decode faster than JSON::XS, but the result |
643 | will be broken due to missing (or wrong) unicode handling. Others refuse |
753 | will be broken due to missing (or wrong) unicode handling. Others refuse |
644 | to decode or encode properly, so it was impossible to prepare a fair |
754 | to decode or encode properly, so it was impossible to prepare a fair |
… | |
… | |
657 | Second, you need to avoid resource-starving attacks. That means you should |
767 | Second, you need to avoid resource-starving attacks. That means you should |
658 | limit the size of JSON texts you accept, or make sure then when your |
768 | limit the size of JSON texts you accept, or make sure then when your |
659 | resources run out, thats just fine (e.g. by using a separate process that |
769 | resources run out, thats just fine (e.g. by using a separate process that |
660 | can crash safely). The size of a JSON text in octets or characters is |
770 | can crash safely). The size of a JSON text in octets or characters is |
661 | usually a good indication of the size of the resources required to decode |
771 | usually a good indication of the size of the resources required to decode |
662 | it into a Perl structure. |
772 | it into a Perl structure. While JSON::XS can check the size of the JSON |
|
|
773 | text, it might be too late when you already have it in memory, so you |
|
|
774 | might want to check the size before you accept the string. |
663 | |
775 | |
664 | Third, JSON::XS recurses using the C stack when decoding objects and |
776 | Third, JSON::XS recurses using the C stack when decoding objects and |
665 | arrays. The C stack is a limited resource: for instance, on my amd64 |
777 | arrays. The C stack is a limited resource: for instance, on my amd64 |
666 | machine with 8MB of stack size I can decode around 180k nested arrays but |
778 | machine with 8MB of stack size I can decode around 180k nested arrays but |
667 | only 14k nested JSON objects (due to perl itself recursing deeply on croak |
779 | only 14k nested JSON objects (due to perl itself recursing deeply on croak |
… | |
… | |
672 | |
784 | |
673 | And last but least, something else could bomb you that I forgot to think |
785 | And last but least, something else could bomb you that I forgot to think |
674 | of. In that case, you get to keep the pieces. I am always open for hints, |
786 | of. In that case, you get to keep the pieces. I am always open for hints, |
675 | though... |
787 | though... |
676 | |
788 | |
|
|
789 | If you are using JSON::XS to return packets to consumption |
|
|
790 | by javascript scripts in a browser you should have a look at |
|
|
791 | L<http://jpsykes.com/47/practical-csrf-and-json-security> to see wether |
|
|
792 | you are vulnerable to some common attack vectors (which really are browser |
|
|
793 | design bugs, but it is still you who will have to deal with it, as major |
|
|
794 | browser developers care only for features, not about doing security |
|
|
795 | right). |
|
|
796 | |
677 | |
797 | |
678 | =head1 BUGS |
798 | =head1 BUGS |
679 | |
799 | |
680 | While the goal of this module is to be correct, that unfortunately does |
800 | While the goal of this module is to be correct, that unfortunately does |
681 | not mean its bug-free, only that I think its design is bug-free. It is |
801 | not mean its bug-free, only that I think its design is bug-free. It is |
682 | still relatively early in its development. If you keep reporting bugs they |
802 | still relatively early in its development. If you keep reporting bugs they |
683 | will be fixed swiftly, though. |
803 | will be fixed swiftly, though. |
684 | |
804 | |
685 | =cut |
805 | =cut |
686 | |
806 | |
|
|
807 | our $true = do { bless \(my $dummy = 1), "JSON::XS::Boolean" }; |
|
|
808 | our $false = do { bless \(my $dummy = 0), "JSON::XS::Boolean" }; |
|
|
809 | |
687 | sub true() { \1 } |
810 | sub true() { $true } |
688 | sub false() { \0 } |
811 | sub false() { $false } |
|
|
812 | |
|
|
813 | sub is_bool($) { |
|
|
814 | UNIVERSAL::isa $_[0], "JSON::XS::Boolean" |
|
|
815 | # or UNIVERSAL::isa $_[0], "JSON::Literal" |
|
|
816 | } |
|
|
817 | |
|
|
818 | XSLoader::load "JSON::XS", $VERSION; |
|
|
819 | |
|
|
820 | package JSON::XS::Boolean; |
|
|
821 | |
|
|
822 | use overload |
|
|
823 | "0+" => sub { ${$_[0]} }, |
|
|
824 | "++" => sub { $_[0] = ${$_[0]} + 1 }, |
|
|
825 | "--" => sub { $_[0] = ${$_[0]} - 1 }, |
|
|
826 | fallback => 1; |
689 | |
827 | |
690 | 1; |
828 | 1; |
691 | |
829 | |
692 | =head1 AUTHOR |
830 | =head1 AUTHOR |
693 | |
831 | |