… | |
… | |
4 | |
4 | |
5 | =head1 SYNOPSIS |
5 | =head1 SYNOPSIS |
6 | |
6 | |
7 | use JSON::XS; |
7 | use JSON::XS; |
8 | |
8 | |
9 | # exported functions, croak on error |
9 | # exported functions, they croak on error |
|
|
10 | # and expect/generate UTF-8 |
10 | |
11 | |
11 | $utf8_encoded_json_text = to_json $perl_hash_or_arrayref; |
12 | $utf8_encoded_json_text = to_json $perl_hash_or_arrayref; |
12 | $perl_hash_or_arrayref = from_json $utf8_encoded_json_text; |
13 | $perl_hash_or_arrayref = from_json $utf8_encoded_json_text; |
13 | |
14 | |
|
|
15 | # objToJson and jsonToObj aliases to to_json and from_json |
|
|
16 | # are exported for compatibility to the JSON module, |
|
|
17 | # but should not be used in new code. |
|
|
18 | |
14 | # oo-interface |
19 | # OO-interface |
15 | |
20 | |
16 | $coder = JSON::XS->new->ascii->pretty->allow_nonref; |
21 | $coder = JSON::XS->new->ascii->pretty->allow_nonref; |
17 | $pretty_printed_unencoded = $coder->encode ($perl_scalar); |
22 | $pretty_printed_unencoded = $coder->encode ($perl_scalar); |
18 | $perl_scalar = $coder->decode ($unicode_json_text); |
23 | $perl_scalar = $coder->decode ($unicode_json_text); |
19 | |
24 | |
… | |
… | |
36 | |
41 | |
37 | =head2 FEATURES |
42 | =head2 FEATURES |
38 | |
43 | |
39 | =over 4 |
44 | =over 4 |
40 | |
45 | |
41 | =item * correct handling of unicode issues |
46 | =item * correct unicode handling |
42 | |
47 | |
43 | This module knows how to handle Unicode, and even documents how and when |
48 | This module knows how to handle Unicode, and even documents how and when |
44 | it does so. |
49 | it does so. |
45 | |
50 | |
46 | =item * round-trip integrity |
51 | =item * round-trip integrity |
47 | |
52 | |
48 | When you serialise a perl data structure using only datatypes supported |
53 | When you serialise a perl data structure using only datatypes supported |
49 | by JSON, the deserialised data structure is identical on the Perl level. |
54 | by JSON, the deserialised data structure is identical on the Perl level. |
50 | (e.g. the string "2.0" doesn't suddenly become "2"). |
55 | (e.g. the string "2.0" doesn't suddenly become "2" just because it looks |
|
|
56 | like a number). |
51 | |
57 | |
52 | =item * strict checking of JSON correctness |
58 | =item * strict checking of JSON correctness |
53 | |
59 | |
54 | There is no guessing, no generating of illegal JSON texts by default, |
60 | There is no guessing, no generating of illegal JSON texts by default, |
55 | and only JSON is accepted as input by default (the latter is a security |
61 | and only JSON is accepted as input by default (the latter is a security |
… | |
… | |
66 | interface. |
72 | interface. |
67 | |
73 | |
68 | =item * reasonably versatile output formats |
74 | =item * reasonably versatile output formats |
69 | |
75 | |
70 | You can choose between the most compact guarenteed single-line format |
76 | You can choose between the most compact guarenteed single-line format |
71 | possible (nice for simple line-based protocols), a pure-ascii format (for |
77 | possible (nice for simple line-based protocols), a pure-ascii format |
72 | when your transport is not 8-bit clean), or a pretty-printed format (for |
78 | (for when your transport is not 8-bit clean, still supports the whole |
73 | when you want to read that stuff). Or you can combine those features in |
79 | unicode range), or a pretty-printed format (for when you want to read that |
74 | whatever way you like. |
80 | stuff). Or you can combine those features in whatever way you like. |
75 | |
81 | |
76 | =back |
82 | =back |
77 | |
83 | |
78 | =cut |
84 | =cut |
79 | |
85 | |
80 | package JSON::XS; |
86 | package JSON::XS; |
81 | |
87 | |
|
|
88 | use strict; |
|
|
89 | |
82 | BEGIN { |
90 | BEGIN { |
83 | $VERSION = '0.31'; |
91 | our $VERSION = '1.12'; |
84 | @ISA = qw(Exporter); |
92 | our @ISA = qw(Exporter); |
85 | |
93 | |
86 | @EXPORT = qw(to_json from_json); |
94 | our @EXPORT = qw(to_json from_json objToJson jsonToObj); |
87 | require Exporter; |
95 | require Exporter; |
88 | |
96 | |
89 | require XSLoader; |
97 | require XSLoader; |
90 | XSLoader::load JSON::XS::, $VERSION; |
98 | XSLoader::load JSON::XS::, $VERSION; |
91 | } |
99 | } |
… | |
… | |
120 | $perl_scalar = JSON::XS->new->utf8->decode ($json_text) |
128 | $perl_scalar = JSON::XS->new->utf8->decode ($json_text) |
121 | |
129 | |
122 | except being faster. |
130 | except being faster. |
123 | |
131 | |
124 | =back |
132 | =back |
|
|
133 | |
125 | |
134 | |
126 | =head1 OBJECT-ORIENTED INTERFACE |
135 | =head1 OBJECT-ORIENTED INTERFACE |
127 | |
136 | |
128 | The object oriented interface lets you configure your own encoding or |
137 | The object oriented interface lets you configure your own encoding or |
129 | decoding style, within the limits of supported formats. |
138 | decoding style, within the limits of supported formats. |
… | |
… | |
145 | |
154 | |
146 | If C<$enable> is true (or missing), then the C<encode> method will not |
155 | If C<$enable> is true (or missing), then the C<encode> method will not |
147 | generate characters outside the code range C<0..127> (which is ASCII). Any |
156 | generate characters outside the code range C<0..127> (which is ASCII). Any |
148 | unicode characters outside that range will be escaped using either a |
157 | unicode characters outside that range will be escaped using either a |
149 | single \uXXXX (BMP characters) or a double \uHHHH\uLLLLL escape sequence, |
158 | single \uXXXX (BMP characters) or a double \uHHHH\uLLLLL escape sequence, |
150 | as per RFC4627. |
159 | as per RFC4627. The resulting encoded JSON text can be treated as a native |
|
|
160 | unicode string, an ascii-encoded, latin1-encoded or UTF-8 encoded string, |
|
|
161 | or any other superset of ASCII. |
151 | |
162 | |
152 | If C<$enable> is false, then the C<encode> method will not escape Unicode |
163 | If C<$enable> is false, then the C<encode> method will not escape Unicode |
153 | characters unless required by the JSON syntax. This results in a faster |
164 | characters unless required by the JSON syntax. This results in a faster |
154 | and more compact format. |
165 | and more compact format. |
155 | |
166 | |
… | |
… | |
275 | => "Hello, World!" |
286 | => "Hello, World!" |
276 | |
287 | |
277 | =item $json = $json->shrink ([$enable]) |
288 | =item $json = $json->shrink ([$enable]) |
278 | |
289 | |
279 | Perl usually over-allocates memory a bit when allocating space for |
290 | Perl usually over-allocates memory a bit when allocating space for |
280 | strings. This flag optionally resizes strings generated by either |
291 | strings. This flag optionally resizes strings generated by either |
281 | C<encode> or C<decode> to their minimum size possible. This can save |
292 | C<encode> or C<decode> to their minimum size possible. This can save |
282 | memory when your JSON texts are either very very long or you have many |
293 | memory when your JSON texts are either very very long or you have many |
283 | short strings. It will also try to downgrade any strings to octet-form |
294 | short strings. It will also try to downgrade any strings to octet-form |
284 | if possible: perl stores strings internally either in an encoding called |
295 | if possible: perl stores strings internally either in an encoding called |
285 | UTF-X or in octet-form. The latter cannot store everything but uses less |
296 | UTF-X or in octet-form. The latter cannot store everything but uses less |
286 | space in general. |
297 | space in general (and some buggy Perl or C code might even rely on that |
|
|
298 | internal representation being used). |
287 | |
299 | |
|
|
300 | The actual definition of what shrink does might change in future versions, |
|
|
301 | but it will always try to save space at the expense of time. |
|
|
302 | |
288 | If C<$enable> is true (or missing), the string returned by C<encode> will be shrunk-to-fit, |
303 | If C<$enable> is true (or missing), the string returned by C<encode> will |
289 | while all strings generated by C<decode> will also be shrunk-to-fit. |
304 | be shrunk-to-fit, while all strings generated by C<decode> will also be |
|
|
305 | shrunk-to-fit. |
290 | |
306 | |
291 | If C<$enable> is false, then the normal perl allocation algorithms are used. |
307 | If C<$enable> is false, then the normal perl allocation algorithms are used. |
292 | If you work with your data, then this is likely to be faster. |
308 | If you work with your data, then this is likely to be faster. |
293 | |
309 | |
294 | In the future, this setting might control other things, such as converting |
310 | In the future, this setting might control other things, such as converting |
295 | strings that look like integers or floats into integers or floats |
311 | strings that look like integers or floats into integers or floats |
296 | internally (there is no difference on the Perl level), saving space. |
312 | internally (there is no difference on the Perl level), saving space. |
|
|
313 | |
|
|
314 | =item $json = $json->max_depth ([$maximum_nesting_depth]) |
|
|
315 | |
|
|
316 | Sets the maximum nesting level (default C<512>) accepted while encoding |
|
|
317 | or decoding. If the JSON text or Perl data structure has an equal or |
|
|
318 | higher nesting level then this limit, then the encoder and decoder will |
|
|
319 | stop and croak at that point. |
|
|
320 | |
|
|
321 | Nesting level is defined by number of hash- or arrayrefs that the encoder |
|
|
322 | needs to traverse to reach a given point or the number of C<{> or C<[> |
|
|
323 | characters without their matching closing parenthesis crossed to reach a |
|
|
324 | given character in a string. |
|
|
325 | |
|
|
326 | Setting the maximum depth to one disallows any nesting, so that ensures |
|
|
327 | that the object is only a single hash/object or array. |
|
|
328 | |
|
|
329 | The argument to C<max_depth> will be rounded up to the next nearest power |
|
|
330 | of two. |
|
|
331 | |
|
|
332 | See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
297 | |
333 | |
298 | =item $json_text = $json->encode ($perl_scalar) |
334 | =item $json_text = $json->encode ($perl_scalar) |
299 | |
335 | |
300 | Converts the given Perl data structure (a simple scalar or a reference |
336 | Converts the given Perl data structure (a simple scalar or a reference |
301 | to a hash or array) to its JSON representation. Simple scalars will be |
337 | to a hash or array) to its JSON representation. Simple scalars will be |
… | |
… | |
313 | Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes |
349 | Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes |
314 | C<1>, C<false> becomes C<0> and C<null> becomes C<undef>. |
350 | C<1>, C<false> becomes C<0> and C<null> becomes C<undef>. |
315 | |
351 | |
316 | =back |
352 | =back |
317 | |
353 | |
|
|
354 | |
318 | =head1 MAPPING |
355 | =head1 MAPPING |
319 | |
356 | |
320 | This section describes how JSON::XS maps Perl values to JSON values and |
357 | This section describes how JSON::XS maps Perl values to JSON values and |
321 | vice versa. These mappings are designed to "do the right thing" in most |
358 | vice versa. These mappings are designed to "do the right thing" in most |
322 | circumstances automatically, preserving round-tripping characteristics |
359 | circumstances automatically, preserving round-tripping characteristics |
… | |
… | |
375 | =over 4 |
412 | =over 4 |
376 | |
413 | |
377 | =item hash references |
414 | =item hash references |
378 | |
415 | |
379 | Perl hash references become JSON objects. As there is no inherent ordering |
416 | Perl hash references become JSON objects. As there is no inherent ordering |
380 | in hash keys, they will usually be encoded in a pseudo-random order that |
417 | in hash keys (or JSON objects), they will usually be encoded in a |
381 | can change between runs of the same program but stays generally the same |
418 | pseudo-random order that can change between runs of the same program but |
382 | within a single run of a program. JSON::XS can optionally sort the hash |
419 | stays generally the same within a single run of a program. JSON::XS can |
383 | keys (determined by the I<canonical> flag), so the same datastructure |
420 | optionally sort the hash keys (determined by the I<canonical> flag), so |
384 | will serialise to the same JSON text (given same settings and version of |
421 | the same datastructure will serialise to the same JSON text (given same |
385 | JSON::XS), but this incurs a runtime overhead. |
422 | settings and version of JSON::XS), but this incurs a runtime overhead |
|
|
423 | and is only rarely useful, e.g. when you want to compare some JSON text |
|
|
424 | against another for equality. |
386 | |
425 | |
387 | =item array references |
426 | =item array references |
388 | |
427 | |
389 | Perl array references become JSON arrays. |
428 | Perl array references become JSON arrays. |
|
|
429 | |
|
|
430 | =item other references |
|
|
431 | |
|
|
432 | Other unblessed references are generally not allowed and will cause an |
|
|
433 | exception to be thrown, except for references to the integers C<0> and |
|
|
434 | C<1>, which get turned into C<false> and C<true> atoms in JSON. You can |
|
|
435 | also use C<JSON::XS::false> and C<JSON::XS::true> to improve readability. |
|
|
436 | |
|
|
437 | to_json [\0,JSON::XS::true] # yields [false,true] |
390 | |
438 | |
391 | =item blessed objects |
439 | =item blessed objects |
392 | |
440 | |
393 | Blessed objects are not allowed. JSON::XS currently tries to encode their |
441 | Blessed objects are not allowed. JSON::XS currently tries to encode their |
394 | underlying representation (hash- or arrayref), but this behaviour might |
442 | underlying representation (hash- or arrayref), but this behaviour might |
… | |
… | |
427 | $x *= 1; # same thing, the choise is yours. |
475 | $x *= 1; # same thing, the choise is yours. |
428 | |
476 | |
429 | You can not currently output JSON booleans or force the type in other, |
477 | You can not currently output JSON booleans or force the type in other, |
430 | less obscure, ways. Tell me if you need this capability. |
478 | less obscure, ways. Tell me if you need this capability. |
431 | |
479 | |
432 | =item circular data structures |
|
|
433 | |
|
|
434 | Those will be encoded until memory or stackspace runs out. |
|
|
435 | |
|
|
436 | =back |
480 | =back |
|
|
481 | |
437 | |
482 | |
438 | =head1 COMPARISON |
483 | =head1 COMPARISON |
439 | |
484 | |
440 | As already mentioned, this module was created because none of the existing |
485 | As already mentioned, this module was created because none of the existing |
441 | JSON modules could be made to work correctly. First I will describe the |
486 | JSON modules could be made to work correctly. First I will describe the |
… | |
… | |
527 | tables. They have been generated with the help of the C<eg/bench> program |
572 | tables. They have been generated with the help of the C<eg/bench> program |
528 | in the JSON::XS distribution, to make it easy to compare on your own |
573 | in the JSON::XS distribution, to make it easy to compare on your own |
529 | system. |
574 | system. |
530 | |
575 | |
531 | First comes a comparison between various modules using a very short JSON |
576 | First comes a comparison between various modules using a very short JSON |
532 | string (83 bytes), showing the number of encodes/decodes per second |
577 | string: |
533 | (JSON::XS is the functional interface, while JSON::XS/2 is the OO |
578 | |
|
|
579 | {"method": "handleMessage", "params": ["user1", "we were just talking"], "id": null} |
|
|
580 | |
|
|
581 | It shows the number of encodes/decodes per second (JSON::XS uses the |
|
|
582 | functional interface, while JSON::XS/2 uses the OO interface with |
534 | interface with pretty-printing and hashkey sorting enabled). Higher is |
583 | pretty-printing and hashkey sorting enabled). Higher is better: |
535 | better: |
|
|
536 | |
584 | |
537 | module | encode | decode | |
585 | module | encode | decode | |
538 | -----------|------------|------------| |
586 | -----------|------------|------------| |
539 | JSON | 14006 | 6820 | |
587 | JSON | 11488.516 | 7823.035 | |
540 | JSON::DWIW | 200937 | 120386 | |
588 | JSON::DWIW | 94708.054 | 129094.260 | |
541 | JSON::PC | 85065 | 129366 | |
589 | JSON::PC | 63884.157 | 128528.212 | |
542 | JSON::Syck | 59898 | 44232 | |
590 | JSON::Syck | 34898.677 | 42096.911 | |
543 | JSON::XS | 1171478 | 342435 | |
591 | JSON::XS | 654027.064 | 396423.669 | |
544 | JSON::XS/2 | 730760 | 328714 | |
592 | JSON::XS/2 | 371564.190 | 371725.613 | |
545 | -----------+------------+------------+ |
593 | -----------+------------+------------+ |
546 | |
594 | |
547 | That is, JSON::XS is 6 times faster than than JSON::DWIW and about 80 |
595 | That is, JSON::XS is more than six times faster than JSON::DWIW on |
|
|
596 | encoding, more than three times faster on decoding, and about thirty times |
548 | times faster than JSON, even with pretty-printing and key sorting. |
597 | faster than JSON, even with pretty-printing and key sorting. |
549 | |
598 | |
550 | Using a longer test string (roughly 18KB, generated from Yahoo! Locals |
599 | Using a longer test string (roughly 18KB, generated from Yahoo! Locals |
551 | search API (http://nanoref.com/yahooapis/mgPdGg): |
600 | search API (http://nanoref.com/yahooapis/mgPdGg): |
552 | |
601 | |
553 | module | encode | decode | |
602 | module | encode | decode | |
554 | -----------|------------|------------| |
603 | -----------|------------|------------| |
555 | JSON | 673 | 38 | |
604 | JSON | 273.023 | 44.674 | |
556 | JSON::DWIW | 5271 | 770 | |
605 | JSON::DWIW | 1089.383 | 1145.704 | |
557 | JSON::PC | 9901 | 2491 | |
606 | JSON::PC | 3097.419 | 2393.921 | |
558 | JSON::Syck | 2360 | 786 | |
607 | JSON::Syck | 514.060 | 843.053 | |
559 | JSON::XS | 37398 | 3202 | |
608 | JSON::XS | 6479.668 | 3636.364 | |
560 | JSON::XS/2 | 13765 | 3153 | |
609 | JSON::XS/2 | 3774.221 | 3599.124 | |
561 | -----------+------------+------------+ |
610 | -----------+------------+------------+ |
562 | |
611 | |
563 | Again, JSON::XS leads by far in the encoding case, while still beating |
612 | Again, JSON::XS leads by far. |
564 | every other module in the decoding case. |
|
|
565 | |
613 | |
566 | On large strings containing lots of unicode characters, some modules |
614 | On large strings containing lots of high unicode characters, some modules |
567 | (such as JSON::PC) decode faster than JSON::XS, but the result will be |
615 | (such as JSON::PC) seem to decode faster than JSON::XS, but the result |
568 | broken due to missing unicode handling. Others refuse to decode or encode |
616 | will be broken due to missing (or wrong) unicode handling. Others refuse |
569 | properly, so it was impossible to prepare a fair comparison table for that |
617 | to decode or encode properly, so it was impossible to prepare a fair |
570 | case. |
618 | comparison table for that case. |
571 | |
619 | |
572 | =head1 RESOURCE LIMITS |
|
|
573 | |
620 | |
574 | JSON::XS does not impose any limits on the size of JSON texts or Perl |
621 | =head1 SECURITY CONSIDERATIONS |
575 | values they represent - if your machine can handle it, JSON::XS will |
622 | |
576 | encode or decode it. Future versions might optionally impose structure |
623 | When you are using JSON in a protocol, talking to untrusted potentially |
577 | depth and memory use resource limits. |
624 | hostile creatures requires relatively few measures. |
|
|
625 | |
|
|
626 | First of all, your JSON decoder should be secure, that is, should not have |
|
|
627 | any buffer overflows. Obviously, this module should ensure that and I am |
|
|
628 | trying hard on making that true, but you never know. |
|
|
629 | |
|
|
630 | Second, you need to avoid resource-starving attacks. That means you should |
|
|
631 | limit the size of JSON texts you accept, or make sure then when your |
|
|
632 | resources run out, thats just fine (e.g. by using a separate process that |
|
|
633 | can crash safely). The size of a JSON text in octets or characters is |
|
|
634 | usually a good indication of the size of the resources required to decode |
|
|
635 | it into a Perl structure. |
|
|
636 | |
|
|
637 | Third, JSON::XS recurses using the C stack when decoding objects and |
|
|
638 | arrays. The C stack is a limited resource: for instance, on my amd64 |
|
|
639 | machine with 8MB of stack size I can decode around 180k nested arrays but |
|
|
640 | only 14k nested JSON objects (due to perl itself recursing deeply on croak |
|
|
641 | to free the temporary). If that is exceeded, the program crashes. to be |
|
|
642 | conservative, the default nesting limit is set to 512. If your process |
|
|
643 | has a smaller stack, you should adjust this setting accordingly with the |
|
|
644 | C<max_depth> method. |
|
|
645 | |
|
|
646 | And last but least, something else could bomb you that I forgot to think |
|
|
647 | of. In that case, you get to keep the pieces. I am always open for hints, |
|
|
648 | though... |
|
|
649 | |
578 | |
650 | |
579 | =head1 BUGS |
651 | =head1 BUGS |
580 | |
652 | |
581 | While the goal of this module is to be correct, that unfortunately does |
653 | While the goal of this module is to be correct, that unfortunately does |
582 | not mean its bug-free, only that I think its design is bug-free. It is |
654 | not mean its bug-free, only that I think its design is bug-free. It is |
583 | still very young and not well-tested. If you keep reporting bugs they will |
655 | still relatively early in its development. If you keep reporting bugs they |
584 | be fixed swiftly, though. |
656 | will be fixed swiftly, though. |
585 | |
657 | |
586 | =cut |
658 | =cut |
|
|
659 | |
|
|
660 | sub true() { \1 } |
|
|
661 | sub false() { \0 } |
587 | |
662 | |
588 | 1; |
663 | 1; |
589 | |
664 | |
590 | =head1 AUTHOR |
665 | =head1 AUTHOR |
591 | |
666 | |