|
|
1 | =encoding utf-8 |
|
|
2 | |
1 | =head1 NAME |
3 | =head1 NAME |
2 | |
4 | |
3 | JSON::XS - JSON serialising/deserialising, done correctly and fast |
5 | JSON::XS - JSON serialising/deserialising, done correctly and fast |
|
|
6 | |
|
|
7 | JSON::XS - 正しくて高速な JSON シリアライザ/デシリアライザ |
|
|
8 | (http://fleur.hio.jp/perldoc/mix/lib/JSON/XS.html) |
4 | |
9 | |
5 | =head1 SYNOPSIS |
10 | =head1 SYNOPSIS |
6 | |
11 | |
7 | use JSON::XS; |
12 | use JSON::XS; |
8 | |
13 | |
9 | # exported functions, they croak on error |
14 | # exported functions, they croak on error |
10 | # and expect/generate UTF-8 |
15 | # and expect/generate UTF-8 |
11 | |
16 | |
12 | $utf8_encoded_json_text = to_json $perl_hash_or_arrayref; |
17 | $utf8_encoded_json_text = to_json $perl_hash_or_arrayref; |
13 | $perl_hash_or_arrayref = from_json $utf8_encoded_json_text; |
18 | $perl_hash_or_arrayref = from_json $utf8_encoded_json_text; |
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 | |
19 | |
19 | # OO-interface |
20 | # OO-interface |
20 | |
21 | |
21 | $coder = JSON::XS->new->ascii->pretty->allow_nonref; |
22 | $coder = JSON::XS->new->ascii->pretty->allow_nonref; |
22 | $pretty_printed_unencoded = $coder->encode ($perl_scalar); |
23 | $pretty_printed_unencoded = $coder->encode ($perl_scalar); |
… | |
… | |
85 | |
86 | |
86 | package JSON::XS; |
87 | package JSON::XS; |
87 | |
88 | |
88 | use strict; |
89 | use strict; |
89 | |
90 | |
90 | BEGIN { |
|
|
91 | our $VERSION = '1.12'; |
91 | our $VERSION = '1.5'; |
92 | our @ISA = qw(Exporter); |
92 | our @ISA = qw(Exporter); |
93 | |
93 | |
94 | our @EXPORT = qw(to_json from_json objToJson jsonToObj); |
94 | our @EXPORT = qw(to_json from_json); |
95 | require Exporter; |
|
|
96 | |
95 | |
97 | require XSLoader; |
96 | use Exporter; |
98 | XSLoader::load JSON::XS::, $VERSION; |
97 | use XSLoader; |
99 | } |
|
|
100 | |
98 | |
101 | =head1 FUNCTIONAL INTERFACE |
99 | =head1 FUNCTIONAL INTERFACE |
102 | |
100 | |
103 | The following convinience methods are provided by this module. They are |
101 | The following convinience methods are provided by this module. They are |
104 | exported by default: |
102 | exported by default: |
105 | |
103 | |
106 | =over 4 |
104 | =over 4 |
107 | |
105 | |
108 | =item $json_text = to_json $perl_scalar |
106 | =item $json_text = to_json $perl_scalar |
109 | |
107 | |
110 | Converts the given Perl data structure (a simple scalar or a reference to |
108 | Converts the given Perl data structure to a UTF-8 encoded, binary string |
111 | a hash or array) to a UTF-8 encoded, binary string (that is, the string contains |
109 | (that is, the string contains octets only). Croaks on error. |
112 | octets only). Croaks on error. |
|
|
113 | |
110 | |
114 | This function call is functionally identical to: |
111 | This function call is functionally identical to: |
115 | |
112 | |
116 | $json_text = JSON::XS->new->utf8->encode ($perl_scalar) |
113 | $json_text = JSON::XS->new->utf8->encode ($perl_scalar) |
117 | |
114 | |
118 | except being faster. |
115 | except being faster. |
119 | |
116 | |
120 | =item $perl_scalar = from_json $json_text |
117 | =item $perl_scalar = from_json $json_text |
121 | |
118 | |
122 | The opposite of C<to_json>: expects an UTF-8 (binary) string and tries to |
119 | The opposite of C<to_json>: expects an UTF-8 (binary) string and tries |
123 | parse that as an UTF-8 encoded JSON text, returning the resulting simple |
120 | to parse that as an UTF-8 encoded JSON text, returning the resulting |
124 | scalar or reference. Croaks on error. |
121 | reference. Croaks on error. |
125 | |
122 | |
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. |
131 | |
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. |
|
|
137 | |
132 | =back |
138 | =back |
|
|
139 | |
|
|
140 | |
|
|
141 | =head1 A FEW NOTES ON UNICODE AND PERL |
|
|
142 | |
|
|
143 | Since this often leads to confusion, here are a few very clear words on |
|
|
144 | how Unicode works in Perl, modulo bugs. |
|
|
145 | |
|
|
146 | =over 4 |
|
|
147 | |
|
|
148 | =item 1. Perl strings can store characters with ordinal values > 255. |
|
|
149 | |
|
|
150 | This enables you to store unicode characters as single characters in a |
|
|
151 | Perl string - very natural. |
|
|
152 | |
|
|
153 | =item 2. Perl does I<not> associate an encoding with your strings. |
|
|
154 | |
|
|
155 | Unless you force it to, e.g. when matching it against a regex, or printing |
|
|
156 | the scalar to a file, in which case Perl either interprets your string as |
|
|
157 | locale-encoded text, octets/binary, or as Unicode, depending on various |
|
|
158 | settings. In no case is an encoding stored together with your data, it is |
|
|
159 | I<use> that decides encoding, not any magical metadata. |
|
|
160 | |
|
|
161 | =item 3. The internal utf-8 flag has no meaning with regards to the |
|
|
162 | encoding of your string. |
|
|
163 | |
|
|
164 | Just ignore that flag unless you debug a Perl bug, a module written in |
|
|
165 | XS or want to dive into the internals of perl. Otherwise it will only |
|
|
166 | confuse you, as, despite the name, it says nothing about how your string |
|
|
167 | is encoded. You can have unicode strings with that flag set, with that |
|
|
168 | flag clear, and you can have binary data with that flag set and that flag |
|
|
169 | clear. Other possibilities exist, too. |
|
|
170 | |
|
|
171 | If you didn't know about that flag, just the better, pretend it doesn't |
|
|
172 | exist. |
|
|
173 | |
|
|
174 | =item 4. A "Unicode String" is simply a string where each character can be |
|
|
175 | validly interpreted as a Unicode codepoint. |
|
|
176 | |
|
|
177 | If you have UTF-8 encoded data, it is no longer a Unicode string, but a |
|
|
178 | Unicode string encoded in UTF-8, giving you a binary string. |
|
|
179 | |
|
|
180 | =item 5. A string containing "high" (> 255) character values is I<not> a UTF-8 string. |
|
|
181 | |
|
|
182 | Its a fact. Learn to live with it. |
|
|
183 | |
|
|
184 | =back |
|
|
185 | |
|
|
186 | I hope this helps :) |
133 | |
187 | |
134 | |
188 | |
135 | =head1 OBJECT-ORIENTED INTERFACE |
189 | =head1 OBJECT-ORIENTED INTERFACE |
136 | |
190 | |
137 | The object oriented interface lets you configure your own encoding or |
191 | The object oriented interface lets you configure your own encoding or |
… | |
… | |
159 | as per RFC4627. The resulting encoded JSON text can be treated as a native |
213 | 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, |
214 | unicode string, an ascii-encoded, latin1-encoded or UTF-8 encoded string, |
161 | or any other superset of ASCII. |
215 | or any other superset of ASCII. |
162 | |
216 | |
163 | If C<$enable> is false, then the C<encode> method will not escape Unicode |
217 | If C<$enable> is false, then the C<encode> method will not escape Unicode |
164 | characters unless required by the JSON syntax. This results in a faster |
218 | characters unless required by the JSON syntax or other flags. This results |
165 | and more compact format. |
219 | in a faster and more compact format. |
|
|
220 | |
|
|
221 | The main use for this flag is to produce JSON texts that can be |
|
|
222 | transmitted over a 7-bit channel, as the encoded JSON texts will not |
|
|
223 | contain any 8 bit characters. |
166 | |
224 | |
167 | JSON::XS->new->ascii (1)->encode ([chr 0x10401]) |
225 | JSON::XS->new->ascii (1)->encode ([chr 0x10401]) |
168 | => ["\ud801\udc01"] |
226 | => ["\ud801\udc01"] |
|
|
227 | |
|
|
228 | =item $json = $json->latin1 ([$enable]) |
|
|
229 | |
|
|
230 | If C<$enable> is true (or missing), then the C<encode> method will encode |
|
|
231 | the resulting JSON text as latin1 (or iso-8859-1), escaping any characters |
|
|
232 | outside the code range C<0..255>. The resulting string can be treated as a |
|
|
233 | latin1-encoded JSON text or a native unicode string. The C<decode> method |
|
|
234 | will not be affected in any way by this flag, as C<decode> by default |
|
|
235 | expects unicode, which is a strict superset of latin1. |
|
|
236 | |
|
|
237 | If C<$enable> is false, then the C<encode> method will not escape Unicode |
|
|
238 | characters unless required by the JSON syntax or other flags. |
|
|
239 | |
|
|
240 | The main use for this flag is efficiently encoding binary data as JSON |
|
|
241 | text, as most octets will not be escaped, resulting in a smaller encoded |
|
|
242 | size. The disadvantage is that the resulting JSON text is encoded |
|
|
243 | in latin1 (and must correctly be treated as such when storing and |
|
|
244 | transfering), a rare encoding for JSON. It is therefore most useful when |
|
|
245 | you want to store data structures known to contain binary data efficiently |
|
|
246 | in files or databases, not when talking to other JSON encoders/decoders. |
|
|
247 | |
|
|
248 | JSON::XS->new->latin1->encode (["\x{89}\x{abc}"] |
|
|
249 | => ["\x{89}\\u0abc"] # (perl syntax, U+abc escaped, U+89 not) |
169 | |
250 | |
170 | =item $json = $json->utf8 ([$enable]) |
251 | =item $json = $json->utf8 ([$enable]) |
171 | |
252 | |
172 | If C<$enable> is true (or missing), then the C<encode> method will encode |
253 | If C<$enable> is true (or missing), then the C<encode> method will encode |
173 | the JSON result into UTF-8, as required by many protocols, while the |
254 | the JSON result into UTF-8, as required by many protocols, while the |
… | |
… | |
249 | |
330 | |
250 | Example, space_before and indent disabled, space_after enabled: |
331 | Example, space_before and indent disabled, space_after enabled: |
251 | |
332 | |
252 | {"key": "value"} |
333 | {"key": "value"} |
253 | |
334 | |
|
|
335 | =item $json = $json->relaxed ([$enable]) |
|
|
336 | |
|
|
337 | If C<$enable> is true (or missing), then C<decode> will accept some |
|
|
338 | extensions to normal JSON syntax (see below). C<encode> will not be |
|
|
339 | affected in anyway. I<Be aware that this option makes you accept invalid |
|
|
340 | JSON texts as if they were valid!>. I suggest only to use this option to |
|
|
341 | parse application-specific files written by humans (configuration files, |
|
|
342 | resource files etc.) |
|
|
343 | |
|
|
344 | If C<$enable> is false (the default), then C<decode> will only accept |
|
|
345 | valid JSON texts. |
|
|
346 | |
|
|
347 | Currently accepted extensions are: |
|
|
348 | |
|
|
349 | =over 4 |
|
|
350 | |
|
|
351 | =item * list items can have an end-comma |
|
|
352 | |
|
|
353 | JSON I<separates> array elements and key-value pairs with commas. This |
|
|
354 | can be annoying if you write JSON texts manually and want to be able to |
|
|
355 | quickly append elements, so this extension accepts comma at the end of |
|
|
356 | such items not just between them: |
|
|
357 | |
|
|
358 | [ |
|
|
359 | 1, |
|
|
360 | 2, <- this comma not normally allowed |
|
|
361 | ] |
|
|
362 | { |
|
|
363 | "k1": "v1", |
|
|
364 | "k2": "v2", <- this comma not normally allowed |
|
|
365 | } |
|
|
366 | |
|
|
367 | =item * shell-style '#'-comments |
|
|
368 | |
|
|
369 | Whenever JSON allows whitespace, shell-style comments are additionally |
|
|
370 | allowed. They are terminated by the first carriage-return or line-feed |
|
|
371 | character, after which more white-space and comments are allowed. |
|
|
372 | |
|
|
373 | [ |
|
|
374 | 1, # this comment not allowed in JSON |
|
|
375 | # neither this one... |
|
|
376 | ] |
|
|
377 | |
|
|
378 | =back |
|
|
379 | |
254 | =item $json = $json->canonical ([$enable]) |
380 | =item $json = $json->canonical ([$enable]) |
255 | |
381 | |
256 | If C<$enable> is true (or missing), then the C<encode> method will output JSON objects |
382 | If C<$enable> is true (or missing), then the C<encode> method will output JSON objects |
257 | by sorting their keys. This is adding a comparatively high overhead. |
383 | by sorting their keys. This is adding a comparatively high overhead. |
258 | |
384 | |
… | |
… | |
282 | Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>, |
408 | Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>, |
283 | resulting in an invalid JSON text: |
409 | resulting in an invalid JSON text: |
284 | |
410 | |
285 | JSON::XS->new->allow_nonref->encode ("Hello, World!") |
411 | JSON::XS->new->allow_nonref->encode ("Hello, World!") |
286 | => "Hello, World!" |
412 | => "Hello, World!" |
|
|
413 | |
|
|
414 | =item $json = $json->allow_blessed ([$enable]) |
|
|
415 | |
|
|
416 | If C<$enable> is true (or missing), then the C<encode> method will not |
|
|
417 | barf when it encounters a blessed reference. Instead, the value of the |
|
|
418 | B<convert_blessed> option will decide wether C<null> (C<convert_blessed> |
|
|
419 | disabled or no C<to_json> method found) or a representation of the |
|
|
420 | object (C<convert_blessed> enabled and C<to_json> method found) is being |
|
|
421 | encoded. Has no effect on C<decode>. |
|
|
422 | |
|
|
423 | If C<$enable> is false (the default), then C<encode> will throw an |
|
|
424 | exception when it encounters a blessed object. |
|
|
425 | |
|
|
426 | =item $json = $json->convert_blessed ([$enable]) |
|
|
427 | |
|
|
428 | If C<$enable> is true (or missing), then C<encode>, upon encountering a |
|
|
429 | blessed object, will check for the availability of the C<TO_JSON> method |
|
|
430 | on the object's class. If found, it will be called in scalar context |
|
|
431 | and the resulting scalar will be encoded instead of the object. If no |
|
|
432 | C<TO_JSON> method is found, the value of C<allow_blessed> will decide what |
|
|
433 | to do. |
|
|
434 | |
|
|
435 | The C<TO_JSON> method may safely call die if it wants. If C<TO_JSON> |
|
|
436 | returns other blessed objects, those will be handled in the same |
|
|
437 | way. C<TO_JSON> must take care of not causing an endless recursion cycle |
|
|
438 | (== crash) in this case. The name of C<TO_JSON> was chosen because other |
|
|
439 | methods called by the Perl core (== not by the user of the object) are |
|
|
440 | usually in upper case letters and to avoid collisions with the C<to_json> |
|
|
441 | function. |
|
|
442 | |
|
|
443 | This setting does not yet influence C<decode> in any way, but in the |
|
|
444 | future, global hooks might get installed that influence C<decode> and are |
|
|
445 | enabled by this setting. |
|
|
446 | |
|
|
447 | If C<$enable> is false, then the C<allow_blessed> setting will decide what |
|
|
448 | to do when a blessed object is found. |
|
|
449 | |
|
|
450 | =item $json = $json->filter_json_object ([$coderef->($hashref)]) |
|
|
451 | |
|
|
452 | When C<$coderef> is specified, it will be called from C<decode> each |
|
|
453 | time it decodes a JSON object. The only argument is a reference to the |
|
|
454 | newly-created hash. If the code references returns a single scalar (which |
|
|
455 | need not be a reference), this value (i.e. a copy of that scalar to avoid |
|
|
456 | aliasing) is inserted into the deserialised data structure. If it returns |
|
|
457 | an empty list (NOTE: I<not> C<undef>, which is a valid scalar), the |
|
|
458 | original deserialised hash will be inserted. This setting can slow down |
|
|
459 | decoding considerably. |
|
|
460 | |
|
|
461 | When C<$coderef> is omitted or undefined, any existing callback will |
|
|
462 | be removed and C<decode> will not change the deserialised hash in any |
|
|
463 | way. |
|
|
464 | |
|
|
465 | Example, convert all JSON objects into the integer 5: |
|
|
466 | |
|
|
467 | my $js = JSON::XS->new->filter_json_object (sub { 5 }); |
|
|
468 | # returns [5] |
|
|
469 | $js->decode ('[{}]') |
|
|
470 | # throw an exception because allow_nonref is not enabled |
|
|
471 | # so a lone 5 is not allowed. |
|
|
472 | $js->decode ('{"a":1, "b":2}'); |
|
|
473 | |
|
|
474 | =item $json = $json->filter_json_single_key_object ($key [=> $coderef->($value)]) |
|
|
475 | |
|
|
476 | Works remotely similar to C<filter_json_object>, but is only called for |
|
|
477 | JSON objects having a single key named C<$key>. |
|
|
478 | |
|
|
479 | This C<$coderef> is called before the one specified via |
|
|
480 | C<filter_json_object>, if any. It gets passed the single value in the JSON |
|
|
481 | object. If it returns a single value, it will be inserted into the data |
|
|
482 | structure. If it returns nothing (not even C<undef> but the empty list), |
|
|
483 | the callback from C<filter_json_object> will be called next, as if no |
|
|
484 | single-key callback were specified. |
|
|
485 | |
|
|
486 | If C<$coderef> is omitted or undefined, the corresponding callback will be |
|
|
487 | disabled. There can only ever be one callback for a given key. |
|
|
488 | |
|
|
489 | As this callback gets called less often then the C<filter_json_object> |
|
|
490 | one, decoding speed will not usually suffer as much. Therefore, single-key |
|
|
491 | objects make excellent targets to serialise Perl objects into, especially |
|
|
492 | as single-key JSON objects are as close to the type-tagged value concept |
|
|
493 | as JSON gets (its basically an ID/VALUE tuple). Of course, JSON does not |
|
|
494 | support this in any way, so you need to make sure your data never looks |
|
|
495 | like a serialised Perl hash. |
|
|
496 | |
|
|
497 | Typical names for the single object key are C<__class_whatever__>, or |
|
|
498 | C<$__dollars_are_rarely_used__$> or C<}ugly_brace_placement>, or even |
|
|
499 | things like C<__class_md5sum(classname)__>, to reduce the risk of clashing |
|
|
500 | with real hashes. |
|
|
501 | |
|
|
502 | Example, decode JSON objects of the form C<< { "__widget__" => <id> } >> |
|
|
503 | into the corresponding C<< $WIDGET{<id>} >> object: |
|
|
504 | |
|
|
505 | # return whatever is in $WIDGET{5}: |
|
|
506 | JSON::XS |
|
|
507 | ->new |
|
|
508 | ->filter_json_single_key_object (__widget__ => sub { |
|
|
509 | $WIDGET{ $_[0] } |
|
|
510 | }) |
|
|
511 | ->decode ('{"__widget__": 5') |
|
|
512 | |
|
|
513 | # this can be used with a TO_JSON method in some "widget" class |
|
|
514 | # for serialisation to json: |
|
|
515 | sub WidgetBase::TO_JSON { |
|
|
516 | my ($self) = @_; |
|
|
517 | |
|
|
518 | unless ($self->{id}) { |
|
|
519 | $self->{id} = ..get..some..id..; |
|
|
520 | $WIDGET{$self->{id}} = $self; |
|
|
521 | } |
|
|
522 | |
|
|
523 | { __widget__ => $self->{id} } |
|
|
524 | } |
287 | |
525 | |
288 | =item $json = $json->shrink ([$enable]) |
526 | =item $json = $json->shrink ([$enable]) |
289 | |
527 | |
290 | Perl usually over-allocates memory a bit when allocating space for |
528 | Perl usually over-allocates memory a bit when allocating space for |
291 | strings. This flag optionally resizes strings generated by either |
529 | strings. This flag optionally resizes strings generated by either |
… | |
… | |
324 | given character in a string. |
562 | given character in a string. |
325 | |
563 | |
326 | Setting the maximum depth to one disallows any nesting, so that ensures |
564 | Setting the maximum depth to one disallows any nesting, so that ensures |
327 | that the object is only a single hash/object or array. |
565 | that the object is only a single hash/object or array. |
328 | |
566 | |
329 | The argument to C<max_depth> will be rounded up to the next nearest power |
567 | The argument to C<max_depth> will be rounded up to the next highest power |
330 | of two. |
568 | of two. If no argument is given, the highest possible setting will be |
|
|
569 | used, which is rarely useful. |
|
|
570 | |
|
|
571 | See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
|
|
572 | |
|
|
573 | =item $json = $json->max_size ([$maximum_string_size]) |
|
|
574 | |
|
|
575 | Set the maximum length a JSON text may have (in bytes) where decoding is |
|
|
576 | being attempted. The default is C<0>, meaning no limit. When C<decode> |
|
|
577 | is called on a string longer then this number of characters it will not |
|
|
578 | attempt to decode the string but throw an exception. This setting has no |
|
|
579 | effect on C<encode> (yet). |
|
|
580 | |
|
|
581 | The argument to C<max_size> will be rounded up to the next B<highest> |
|
|
582 | power of two (so may be more than requested). If no argument is given, the |
|
|
583 | limit check will be deactivated (same as when C<0> is specified). |
331 | |
584 | |
332 | See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
585 | See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
333 | |
586 | |
334 | =item $json_text = $json->encode ($perl_scalar) |
587 | =item $json_text = $json->encode ($perl_scalar) |
335 | |
588 | |
… | |
… | |
347 | |
600 | |
348 | JSON numbers and strings become simple Perl scalars. JSON arrays become |
601 | JSON numbers and strings become simple Perl scalars. JSON arrays become |
349 | Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes |
602 | Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes |
350 | C<1>, C<false> becomes C<0> and C<null> becomes C<undef>. |
603 | C<1>, C<false> becomes C<0> and C<null> becomes C<undef>. |
351 | |
604 | |
|
|
605 | =item ($perl_scalar, $characters) = $json->decode_prefix ($json_text) |
|
|
606 | |
|
|
607 | This works like the C<decode> method, but instead of raising an exception |
|
|
608 | when there is trailing garbage after the first JSON object, it will |
|
|
609 | silently stop parsing there and return the number of characters consumed |
|
|
610 | so far. |
|
|
611 | |
|
|
612 | This is useful if your JSON texts are not delimited by an outer protocol |
|
|
613 | (which is not the brightest thing to do in the first place) and you need |
|
|
614 | to know where the JSON text ends. |
|
|
615 | |
|
|
616 | JSON::XS->new->decode_prefix ("[1] the tail") |
|
|
617 | => ([], 3) |
|
|
618 | |
352 | =back |
619 | =back |
353 | |
620 | |
354 | |
621 | |
355 | =head1 MAPPING |
622 | =head1 MAPPING |
356 | |
623 | |
… | |
… | |
361 | |
628 | |
362 | For the more enlightened: note that in the following descriptions, |
629 | For the more enlightened: note that in the following descriptions, |
363 | lowercase I<perl> refers to the Perl interpreter, while uppcercase I<Perl> |
630 | lowercase I<perl> refers to the Perl interpreter, while uppcercase I<Perl> |
364 | refers to the abstract Perl language itself. |
631 | refers to the abstract Perl language itself. |
365 | |
632 | |
|
|
633 | |
366 | =head2 JSON -> PERL |
634 | =head2 JSON -> PERL |
367 | |
635 | |
368 | =over 4 |
636 | =over 4 |
369 | |
637 | |
370 | =item object |
638 | =item object |
… | |
… | |
382 | are represented by the same codepoints in the Perl string, so no manual |
650 | are represented by the same codepoints in the Perl string, so no manual |
383 | decoding is necessary. |
651 | decoding is necessary. |
384 | |
652 | |
385 | =item number |
653 | =item number |
386 | |
654 | |
387 | A JSON number becomes either an integer or numeric (floating point) |
655 | A JSON number becomes either an integer, numeric (floating point) or |
388 | scalar in perl, depending on its range and any fractional parts. On the |
656 | string scalar in perl, depending on its range and any fractional parts. On |
389 | Perl level, there is no difference between those as Perl handles all the |
657 | the Perl level, there is no difference between those as Perl handles all |
390 | conversion details, but an integer may take slightly less memory and might |
658 | the conversion details, but an integer may take slightly less memory and |
391 | represent more values exactly than (floating point) numbers. |
659 | might represent more values exactly than (floating point) numbers. |
|
|
660 | |
|
|
661 | If the number consists of digits only, JSON::XS will try to represent |
|
|
662 | it as an integer value. If that fails, it will try to represent it as |
|
|
663 | a numeric (floating point) value if that is possible without loss of |
|
|
664 | precision. Otherwise it will preserve the number as a string value. |
|
|
665 | |
|
|
666 | Numbers containing a fractional or exponential part will always be |
|
|
667 | represented as numeric (floating point) values, possibly at a loss of |
|
|
668 | precision. |
|
|
669 | |
|
|
670 | This might create round-tripping problems as numbers might become strings, |
|
|
671 | but as Perl is typeless there is no other way to do it. |
392 | |
672 | |
393 | =item true, false |
673 | =item true, false |
394 | |
674 | |
395 | These JSON atoms become C<0>, C<1>, respectively. Information is lost in |
675 | These JSON atoms become C<JSON::XS::true> and C<JSON::XS::false>, |
396 | this process. Future versions might represent those values differently, |
676 | respectively. They are overloaded to act almost exactly like the numbers |
397 | but they will be guarenteed to act like these integers would normally in |
677 | C<1> and C<0>. You can check wether a scalar is a JSON boolean by using |
398 | Perl. |
678 | the C<JSON::XS::is_bool> function. |
399 | |
679 | |
400 | =item null |
680 | =item null |
401 | |
681 | |
402 | A JSON null atom becomes C<undef> in Perl. |
682 | A JSON null atom becomes C<undef> in Perl. |
403 | |
683 | |
404 | =back |
684 | =back |
|
|
685 | |
405 | |
686 | |
406 | =head2 PERL -> JSON |
687 | =head2 PERL -> JSON |
407 | |
688 | |
408 | The mapping from Perl to JSON is slightly more difficult, as Perl is a |
689 | The mapping from Perl to JSON is slightly more difficult, as Perl is a |
409 | truly typeless language, so we can only guess which JSON type is meant by |
690 | truly typeless language, so we can only guess which JSON type is meant by |
… | |
… | |
434 | C<1>, which get turned into C<false> and C<true> atoms in JSON. You can |
715 | 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. |
716 | also use C<JSON::XS::false> and C<JSON::XS::true> to improve readability. |
436 | |
717 | |
437 | to_json [\0,JSON::XS::true] # yields [false,true] |
718 | to_json [\0,JSON::XS::true] # yields [false,true] |
438 | |
719 | |
|
|
720 | =item JSON::XS::true, JSON::XS::false |
|
|
721 | |
|
|
722 | These special values become JSON true and JSON false values, |
|
|
723 | respectively. You can also use C<\1> and C<\0> directly if you want. |
|
|
724 | |
439 | =item blessed objects |
725 | =item blessed objects |
440 | |
726 | |
441 | Blessed objects are not allowed. JSON::XS currently tries to encode their |
727 | Blessed objects are not allowed. JSON::XS currently tries to encode their |
442 | underlying representation (hash- or arrayref), but this behaviour might |
728 | underlying representation (hash- or arrayref), but this behaviour might |
443 | change in future versions. |
729 | change in future versions. |
… | |
… | |
564 | |
850 | |
565 | Does not check input for validity. |
851 | Does not check input for validity. |
566 | |
852 | |
567 | =back |
853 | =back |
568 | |
854 | |
|
|
855 | |
|
|
856 | =head2 JSON and YAML |
|
|
857 | |
|
|
858 | You often hear that JSON is a subset (or a close subset) of YAML. This is, |
|
|
859 | however, a mass hysteria and very far from the truth. In general, there is |
|
|
860 | no way to configure JSON::XS to output a data structure as valid YAML. |
|
|
861 | |
|
|
862 | If you really must use JSON::XS to generate YAML, you should use this |
|
|
863 | algorithm (subject to change in future versions): |
|
|
864 | |
|
|
865 | my $to_yaml = JSON::XS->new->utf8->space_after (1); |
|
|
866 | my $yaml = $to_yaml->encode ($ref) . "\n"; |
|
|
867 | |
|
|
868 | This will usually generate JSON texts that also parse as valid |
|
|
869 | YAML. Please note that YAML has hardcoded limits on (simple) object key |
|
|
870 | lengths that JSON doesn't have, so you should make sure that your hash |
|
|
871 | keys are noticably shorter than the 1024 characters YAML allows. |
|
|
872 | |
|
|
873 | There might be other incompatibilities that I am not aware of. In general |
|
|
874 | you should not try to generate YAML with a JSON generator or vice versa, |
|
|
875 | or try to parse JSON with a YAML parser or vice versa: chances are high |
|
|
876 | that you will run into severe interoperability problems. |
|
|
877 | |
|
|
878 | |
569 | =head2 SPEED |
879 | =head2 SPEED |
570 | |
880 | |
571 | It seems that JSON::XS is surprisingly fast, as shown in the following |
881 | It seems that JSON::XS is surprisingly fast, as shown in the following |
572 | tables. They have been generated with the help of the C<eg/bench> program |
882 | tables. They have been generated with the help of the C<eg/bench> program |
573 | in the JSON::XS distribution, to make it easy to compare on your own |
883 | in the JSON::XS distribution, to make it easy to compare on your own |
574 | system. |
884 | system. |
575 | |
885 | |
576 | First comes a comparison between various modules using a very short JSON |
886 | First comes a comparison between various modules using a very short |
577 | string: |
887 | single-line JSON string: |
578 | |
888 | |
579 | {"method": "handleMessage", "params": ["user1", "we were just talking"], "id": null} |
889 | {"method": "handleMessage", "params": ["user1", "we were just talking"], \ |
|
|
890 | "id": null, "array":[1,11,234,-5,1e5,1e7, true, false]} |
580 | |
891 | |
581 | It shows the number of encodes/decodes per second (JSON::XS uses the |
892 | It shows the number of encodes/decodes per second (JSON::XS uses |
582 | functional interface, while JSON::XS/2 uses the OO interface with |
893 | the functional interface, while JSON::XS/2 uses the OO interface |
583 | pretty-printing and hashkey sorting enabled). Higher is better: |
894 | with pretty-printing and hashkey sorting enabled, JSON::XS/3 enables |
|
|
895 | shrink). Higher is better: |
584 | |
896 | |
|
|
897 | Storable | 15779.925 | 14169.946 | |
|
|
898 | -----------+------------+------------+ |
585 | module | encode | decode | |
899 | module | encode | decode | |
586 | -----------|------------|------------| |
900 | -----------|------------|------------| |
587 | JSON | 11488.516 | 7823.035 | |
901 | JSON | 4990.842 | 4088.813 | |
588 | JSON::DWIW | 94708.054 | 129094.260 | |
902 | JSON::DWIW | 51653.990 | 71575.154 | |
589 | JSON::PC | 63884.157 | 128528.212 | |
903 | JSON::PC | 65948.176 | 74631.744 | |
590 | JSON::Syck | 34898.677 | 42096.911 | |
904 | JSON::PP | 8931.652 | 3817.168 | |
591 | JSON::XS | 654027.064 | 396423.669 | |
905 | JSON::Syck | 24877.248 | 27776.848 | |
592 | JSON::XS/2 | 371564.190 | 371725.613 | |
906 | JSON::XS | 388361.481 | 227951.304 | |
|
|
907 | JSON::XS/2 | 227951.304 | 218453.333 | |
|
|
908 | JSON::XS/3 | 338250.323 | 218453.333 | |
|
|
909 | Storable | 16500.016 | 135300.129 | |
593 | -----------+------------+------------+ |
910 | -----------+------------+------------+ |
594 | |
911 | |
595 | That is, JSON::XS is more than six times faster than JSON::DWIW on |
912 | That is, JSON::XS is about five times faster than JSON::DWIW on encoding, |
596 | encoding, more than three times faster on decoding, and about thirty times |
913 | about three times faster on decoding, and over fourty times faster |
597 | faster than JSON, even with pretty-printing and key sorting. |
914 | than JSON, even with pretty-printing and key sorting. It also compares |
|
|
915 | favourably to Storable for small amounts of data. |
598 | |
916 | |
599 | Using a longer test string (roughly 18KB, generated from Yahoo! Locals |
917 | Using a longer test string (roughly 18KB, generated from Yahoo! Locals |
600 | search API (http://nanoref.com/yahooapis/mgPdGg): |
918 | search API (http://nanoref.com/yahooapis/mgPdGg): |
601 | |
919 | |
602 | module | encode | decode | |
920 | module | encode | decode | |
603 | -----------|------------|------------| |
921 | -----------|------------|------------| |
604 | JSON | 273.023 | 44.674 | |
922 | JSON | 55.260 | 34.971 | |
605 | JSON::DWIW | 1089.383 | 1145.704 | |
923 | JSON::DWIW | 825.228 | 1082.513 | |
606 | JSON::PC | 3097.419 | 2393.921 | |
924 | JSON::PC | 3571.444 | 2394.829 | |
607 | JSON::Syck | 514.060 | 843.053 | |
925 | JSON::PP | 210.987 | 32.574 | |
608 | JSON::XS | 6479.668 | 3636.364 | |
926 | JSON::Syck | 552.551 | 787.544 | |
609 | JSON::XS/2 | 3774.221 | 3599.124 | |
927 | JSON::XS | 5780.463 | 4854.519 | |
|
|
928 | JSON::XS/2 | 3869.998 | 4798.975 | |
|
|
929 | JSON::XS/3 | 5862.880 | 4798.975 | |
|
|
930 | Storable | 4445.002 | 5235.027 | |
610 | -----------+------------+------------+ |
931 | -----------+------------+------------+ |
611 | |
932 | |
612 | Again, JSON::XS leads by far. |
933 | Again, JSON::XS leads by far (except for Storable which non-surprisingly |
|
|
934 | decodes faster). |
613 | |
935 | |
614 | On large strings containing lots of high unicode characters, some modules |
936 | On large strings containing lots of high unicode characters, some modules |
615 | (such as JSON::PC) seem to decode faster than JSON::XS, but the result |
937 | (such as JSON::PC) seem to decode faster than JSON::XS, but the result |
616 | will be broken due to missing (or wrong) unicode handling. Others refuse |
938 | will be broken due to missing (or wrong) unicode handling. Others refuse |
617 | to decode or encode properly, so it was impossible to prepare a fair |
939 | to decode or encode properly, so it was impossible to prepare a fair |
… | |
… | |
630 | Second, you need to avoid resource-starving attacks. That means you should |
952 | 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 |
953 | 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 |
954 | 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 |
955 | 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 |
956 | usually a good indication of the size of the resources required to decode |
635 | it into a Perl structure. |
957 | it into a Perl structure. While JSON::XS can check the size of the JSON |
|
|
958 | text, it might be too late when you already have it in memory, so you |
|
|
959 | might want to check the size before you accept the string. |
636 | |
960 | |
637 | Third, JSON::XS recurses using the C stack when decoding objects and |
961 | 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 |
962 | 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 |
963 | 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 |
964 | only 14k nested JSON objects (due to perl itself recursing deeply on croak |
… | |
… | |
645 | |
969 | |
646 | And last but least, something else could bomb you that I forgot to think |
970 | 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, |
971 | of. In that case, you get to keep the pieces. I am always open for hints, |
648 | though... |
972 | though... |
649 | |
973 | |
|
|
974 | If you are using JSON::XS to return packets to consumption |
|
|
975 | by javascript scripts in a browser you should have a look at |
|
|
976 | L<http://jpsykes.com/47/practical-csrf-and-json-security> to see wether |
|
|
977 | you are vulnerable to some common attack vectors (which really are browser |
|
|
978 | design bugs, but it is still you who will have to deal with it, as major |
|
|
979 | browser developers care only for features, not about doing security |
|
|
980 | right). |
|
|
981 | |
650 | |
982 | |
651 | =head1 BUGS |
983 | =head1 BUGS |
652 | |
984 | |
653 | While the goal of this module is to be correct, that unfortunately does |
985 | While the goal of this module is to be correct, that unfortunately does |
654 | not mean its bug-free, only that I think its design is bug-free. It is |
986 | not mean its bug-free, only that I think its design is bug-free. It is |
655 | still relatively early in its development. If you keep reporting bugs they |
987 | still relatively early in its development. If you keep reporting bugs they |
656 | will be fixed swiftly, though. |
988 | will be fixed swiftly, though. |
657 | |
989 | |
658 | =cut |
990 | =cut |
659 | |
991 | |
|
|
992 | our $true = do { bless \(my $dummy = 1), "JSON::XS::Boolean" }; |
|
|
993 | our $false = do { bless \(my $dummy = 0), "JSON::XS::Boolean" }; |
|
|
994 | |
660 | sub true() { \1 } |
995 | sub true() { $true } |
661 | sub false() { \0 } |
996 | sub false() { $false } |
|
|
997 | |
|
|
998 | sub is_bool($) { |
|
|
999 | UNIVERSAL::isa $_[0], "JSON::XS::Boolean" |
|
|
1000 | # or UNIVERSAL::isa $_[0], "JSON::Literal" |
|
|
1001 | } |
|
|
1002 | |
|
|
1003 | XSLoader::load "JSON::XS", $VERSION; |
|
|
1004 | |
|
|
1005 | package JSON::XS::Boolean; |
|
|
1006 | |
|
|
1007 | use overload |
|
|
1008 | "0+" => sub { ${$_[0]} }, |
|
|
1009 | "++" => sub { $_[0] = ${$_[0]} + 1 }, |
|
|
1010 | "--" => sub { $_[0] = ${$_[0]} - 1 }, |
|
|
1011 | fallback => 1; |
662 | |
1012 | |
663 | 1; |
1013 | 1; |
664 | |
1014 | |
665 | =head1 AUTHOR |
1015 | =head1 AUTHOR |
666 | |
1016 | |