1 |
=head1 NAME |
2 |
|
3 |
JSON::XS - JSON serialising/deserialising, done correctly and fast |
4 |
|
5 |
=head1 SYNOPSIS |
6 |
|
7 |
use JSON::XS; |
8 |
|
9 |
# exported functions, they croak on error |
10 |
# and expect/generate UTF-8 |
11 |
|
12 |
$utf8_encoded_json_text = to_json $perl_hash_or_arrayref; |
13 |
$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 |
# OO-interface |
20 |
|
21 |
$coder = JSON::XS->new->ascii->pretty->allow_nonref; |
22 |
$pretty_printed_unencoded = $coder->encode ($perl_scalar); |
23 |
$perl_scalar = $coder->decode ($unicode_json_text); |
24 |
|
25 |
=head1 DESCRIPTION |
26 |
|
27 |
This module converts Perl data structures to JSON and vice versa. Its |
28 |
primary goal is to be I<correct> and its secondary goal is to be |
29 |
I<fast>. To reach the latter goal it was written in C. |
30 |
|
31 |
As this is the n-th-something JSON module on CPAN, what was the reason |
32 |
to write yet another JSON module? While it seems there are many JSON |
33 |
modules, none of them correctly handle all corner cases, and in most cases |
34 |
their maintainers are unresponsive, gone missing, or not listening to bug |
35 |
reports for other reasons. |
36 |
|
37 |
See COMPARISON, below, for a comparison to some other JSON modules. |
38 |
|
39 |
See MAPPING, below, on how JSON::XS maps perl values to JSON values and |
40 |
vice versa. |
41 |
|
42 |
=head2 FEATURES |
43 |
|
44 |
=over 4 |
45 |
|
46 |
=item * correct unicode handling |
47 |
|
48 |
This module knows how to handle Unicode, and even documents how and when |
49 |
it does so. |
50 |
|
51 |
=item * round-trip integrity |
52 |
|
53 |
When you serialise a perl data structure using only datatypes supported |
54 |
by JSON, the deserialised data structure is identical on the Perl level. |
55 |
(e.g. the string "2.0" doesn't suddenly become "2" just because it looks |
56 |
like a number). |
57 |
|
58 |
=item * strict checking of JSON correctness |
59 |
|
60 |
There is no guessing, no generating of illegal JSON texts by default, |
61 |
and only JSON is accepted as input by default (the latter is a security |
62 |
feature). |
63 |
|
64 |
=item * fast |
65 |
|
66 |
Compared to other JSON modules, this module compares favourably in terms |
67 |
of speed, too. |
68 |
|
69 |
=item * simple to use |
70 |
|
71 |
This module has both a simple functional interface as well as an OO |
72 |
interface. |
73 |
|
74 |
=item * reasonably versatile output formats |
75 |
|
76 |
You can choose between the most compact guarenteed single-line format |
77 |
possible (nice for simple line-based protocols), a pure-ascii format |
78 |
(for when your transport is not 8-bit clean, still supports the whole |
79 |
unicode range), or a pretty-printed format (for when you want to read that |
80 |
stuff). Or you can combine those features in whatever way you like. |
81 |
|
82 |
=back |
83 |
|
84 |
=cut |
85 |
|
86 |
package JSON::XS; |
87 |
|
88 |
use strict; |
89 |
|
90 |
our $VERSION = '1.4'; |
91 |
our @ISA = qw(Exporter); |
92 |
|
93 |
our @EXPORT = qw(to_json from_json objToJson jsonToObj); |
94 |
|
95 |
use Exporter; |
96 |
use XSLoader; |
97 |
|
98 |
=head1 FUNCTIONAL INTERFACE |
99 |
|
100 |
The following convinience methods are provided by this module. They are |
101 |
exported by default: |
102 |
|
103 |
=over 4 |
104 |
|
105 |
=item $json_text = to_json $perl_scalar |
106 |
|
107 |
Converts the given Perl data structure (a simple scalar or a reference to |
108 |
a hash or array) to a UTF-8 encoded, binary string (that is, the string contains |
109 |
octets only). Croaks on error. |
110 |
|
111 |
This function call is functionally identical to: |
112 |
|
113 |
$json_text = JSON::XS->new->utf8->encode ($perl_scalar) |
114 |
|
115 |
except being faster. |
116 |
|
117 |
=item $perl_scalar = from_json $json_text |
118 |
|
119 |
The opposite of C<to_json>: expects an UTF-8 (binary) string and tries to |
120 |
parse that as an UTF-8 encoded JSON text, returning the resulting simple |
121 |
scalar or reference. Croaks on error. |
122 |
|
123 |
This function call is functionally identical to: |
124 |
|
125 |
$perl_scalar = JSON::XS->new->utf8->decode ($json_text) |
126 |
|
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. |
137 |
|
138 |
=back |
139 |
|
140 |
|
141 |
=head1 OBJECT-ORIENTED INTERFACE |
142 |
|
143 |
The object oriented interface lets you configure your own encoding or |
144 |
decoding style, within the limits of supported formats. |
145 |
|
146 |
=over 4 |
147 |
|
148 |
=item $json = new JSON::XS |
149 |
|
150 |
Creates a new JSON::XS object that can be used to de/encode JSON |
151 |
strings. All boolean flags described below are by default I<disabled>. |
152 |
|
153 |
The mutators for flags all return the JSON object again and thus calls can |
154 |
be chained: |
155 |
|
156 |
my $json = JSON::XS->new->utf8->space_after->encode ({a => [1,2]}) |
157 |
=> {"a": [1, 2]} |
158 |
|
159 |
=item $json = $json->ascii ([$enable]) |
160 |
|
161 |
If C<$enable> is true (or missing), then the C<encode> method will not |
162 |
generate characters outside the code range C<0..127> (which is ASCII). Any |
163 |
unicode characters outside that range will be escaped using either a |
164 |
single \uXXXX (BMP characters) or a double \uHHHH\uLLLLL escape sequence, |
165 |
as per RFC4627. The resulting encoded JSON text can be treated as a native |
166 |
unicode string, an ascii-encoded, latin1-encoded or UTF-8 encoded string, |
167 |
or any other superset of ASCII. |
168 |
|
169 |
If C<$enable> is false, then the C<encode> method will not escape Unicode |
170 |
characters unless required by the JSON syntax or other flags. This results |
171 |
in a faster and more compact format. |
172 |
|
173 |
The main use for this flag is to produce JSON texts that can be |
174 |
transmitted over a 7-bit channel, as the encoded JSON texts will not |
175 |
contain any 8 bit characters. |
176 |
|
177 |
JSON::XS->new->ascii (1)->encode ([chr 0x10401]) |
178 |
=> ["\ud801\udc01"] |
179 |
|
180 |
=item $json = $json->latin1 ([$enable]) |
181 |
|
182 |
If C<$enable> is true (or missing), then the C<encode> method will encode |
183 |
the resulting JSON text as latin1 (or iso-8859-1), escaping any characters |
184 |
outside the code range C<0..255>. The resulting string can be treated as a |
185 |
latin1-encoded JSON text or a native unicode string. The C<decode> method |
186 |
will not be affected in any way by this flag, as C<decode> by default |
187 |
expects unicode, which is a strict superset of latin1. |
188 |
|
189 |
If C<$enable> is false, then the C<encode> method will not escape Unicode |
190 |
characters unless required by the JSON syntax or other flags. |
191 |
|
192 |
The main use for this flag is efficiently encoding binary data as JSON |
193 |
text, as most octets will not be escaped, resulting in a smaller encoded |
194 |
size. The disadvantage is that the resulting JSON text is encoded |
195 |
in latin1 (and must correctly be treated as such when storing and |
196 |
transfering), a rare encoding for JSON. It is therefore most useful when |
197 |
you want to store data structures known to contain binary data efficiently |
198 |
in files or databases, not when talking to other JSON encoders/decoders. |
199 |
|
200 |
JSON::XS->new->latin1->encode (["\x{89}\x{abc}"] |
201 |
=> ["\x{89}\\u0abc"] # (perl syntax, U+abc escaped, U+89 not) |
202 |
|
203 |
=item $json = $json->utf8 ([$enable]) |
204 |
|
205 |
If C<$enable> is true (or missing), then the C<encode> method will encode |
206 |
the JSON result into UTF-8, as required by many protocols, while the |
207 |
C<decode> method expects to be handled an UTF-8-encoded string. Please |
208 |
note that UTF-8-encoded strings do not contain any characters outside the |
209 |
range C<0..255>, they are thus useful for bytewise/binary I/O. In future |
210 |
versions, enabling this option might enable autodetection of the UTF-16 |
211 |
and UTF-32 encoding families, as described in RFC4627. |
212 |
|
213 |
If C<$enable> is false, then the C<encode> method will return the JSON |
214 |
string as a (non-encoded) unicode string, while C<decode> expects thus a |
215 |
unicode string. Any decoding or encoding (e.g. to UTF-8 or UTF-16) needs |
216 |
to be done yourself, e.g. using the Encode module. |
217 |
|
218 |
Example, output UTF-16BE-encoded JSON: |
219 |
|
220 |
use Encode; |
221 |
$jsontext = encode "UTF-16BE", JSON::XS->new->encode ($object); |
222 |
|
223 |
Example, decode UTF-32LE-encoded JSON: |
224 |
|
225 |
use Encode; |
226 |
$object = JSON::XS->new->decode (decode "UTF-32LE", $jsontext); |
227 |
|
228 |
=item $json = $json->pretty ([$enable]) |
229 |
|
230 |
This enables (or disables) all of the C<indent>, C<space_before> and |
231 |
C<space_after> (and in the future possibly more) flags in one call to |
232 |
generate the most readable (or most compact) form possible. |
233 |
|
234 |
Example, pretty-print some simple structure: |
235 |
|
236 |
my $json = JSON::XS->new->pretty(1)->encode ({a => [1,2]}) |
237 |
=> |
238 |
{ |
239 |
"a" : [ |
240 |
1, |
241 |
2 |
242 |
] |
243 |
} |
244 |
|
245 |
=item $json = $json->indent ([$enable]) |
246 |
|
247 |
If C<$enable> is true (or missing), then the C<encode> method will use a multiline |
248 |
format as output, putting every array member or object/hash key-value pair |
249 |
into its own line, identing them properly. |
250 |
|
251 |
If C<$enable> is false, no newlines or indenting will be produced, and the |
252 |
resulting JSON text is guarenteed not to contain any C<newlines>. |
253 |
|
254 |
This setting has no effect when decoding JSON texts. |
255 |
|
256 |
=item $json = $json->space_before ([$enable]) |
257 |
|
258 |
If C<$enable> is true (or missing), then the C<encode> method will add an extra |
259 |
optional space before the C<:> separating keys from values in JSON objects. |
260 |
|
261 |
If C<$enable> is false, then the C<encode> method will not add any extra |
262 |
space at those places. |
263 |
|
264 |
This setting has no effect when decoding JSON texts. You will also |
265 |
most likely combine this setting with C<space_after>. |
266 |
|
267 |
Example, space_before enabled, space_after and indent disabled: |
268 |
|
269 |
{"key" :"value"} |
270 |
|
271 |
=item $json = $json->space_after ([$enable]) |
272 |
|
273 |
If C<$enable> is true (or missing), then the C<encode> method will add an extra |
274 |
optional space after the C<:> separating keys from values in JSON objects |
275 |
and extra whitespace after the C<,> separating key-value pairs and array |
276 |
members. |
277 |
|
278 |
If C<$enable> is false, then the C<encode> method will not add any extra |
279 |
space at those places. |
280 |
|
281 |
This setting has no effect when decoding JSON texts. |
282 |
|
283 |
Example, space_before and indent disabled, space_after enabled: |
284 |
|
285 |
{"key": "value"} |
286 |
|
287 |
=item $json = $json->canonical ([$enable]) |
288 |
|
289 |
If C<$enable> is true (or missing), then the C<encode> method will output JSON objects |
290 |
by sorting their keys. This is adding a comparatively high overhead. |
291 |
|
292 |
If C<$enable> is false, then the C<encode> method will output key-value |
293 |
pairs in the order Perl stores them (which will likely change between runs |
294 |
of the same script). |
295 |
|
296 |
This option is useful if you want the same data structure to be encoded as |
297 |
the same JSON text (given the same overall settings). If it is disabled, |
298 |
the same hash migh be encoded differently even if contains the same data, |
299 |
as key-value pairs have no inherent ordering in Perl. |
300 |
|
301 |
This setting has no effect when decoding JSON texts. |
302 |
|
303 |
=item $json = $json->allow_nonref ([$enable]) |
304 |
|
305 |
If C<$enable> is true (or missing), then the C<encode> method can convert a |
306 |
non-reference into its corresponding string, number or null JSON value, |
307 |
which is an extension to RFC4627. Likewise, C<decode> will accept those JSON |
308 |
values instead of croaking. |
309 |
|
310 |
If C<$enable> is false, then the C<encode> method will croak if it isn't |
311 |
passed an arrayref or hashref, as JSON texts must either be an object |
312 |
or array. Likewise, C<decode> will croak if given something that is not a |
313 |
JSON object or array. |
314 |
|
315 |
Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>, |
316 |
resulting in an invalid JSON text: |
317 |
|
318 |
JSON::XS->new->allow_nonref->encode ("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. |
356 |
|
357 |
=item $json = $json->shrink ([$enable]) |
358 |
|
359 |
Perl usually over-allocates memory a bit when allocating space for |
360 |
strings. This flag optionally resizes strings generated by either |
361 |
C<encode> or C<decode> to their minimum size possible. This can save |
362 |
memory when your JSON texts are either very very long or you have many |
363 |
short strings. It will also try to downgrade any strings to octet-form |
364 |
if possible: perl stores strings internally either in an encoding called |
365 |
UTF-X or in octet-form. The latter cannot store everything but uses less |
366 |
space in general (and some buggy Perl or C code might even rely on that |
367 |
internal representation being used). |
368 |
|
369 |
The actual definition of what shrink does might change in future versions, |
370 |
but it will always try to save space at the expense of time. |
371 |
|
372 |
If C<$enable> is true (or missing), the string returned by C<encode> will |
373 |
be shrunk-to-fit, while all strings generated by C<decode> will also be |
374 |
shrunk-to-fit. |
375 |
|
376 |
If C<$enable> is false, then the normal perl allocation algorithms are used. |
377 |
If you work with your data, then this is likely to be faster. |
378 |
|
379 |
In the future, this setting might control other things, such as converting |
380 |
strings that look like integers or floats into integers or floats |
381 |
internally (there is no difference on the Perl level), saving space. |
382 |
|
383 |
=item $json = $json->max_depth ([$maximum_nesting_depth]) |
384 |
|
385 |
Sets the maximum nesting level (default C<512>) accepted while encoding |
386 |
or decoding. If the JSON text or Perl data structure has an equal or |
387 |
higher nesting level then this limit, then the encoder and decoder will |
388 |
stop and croak at that point. |
389 |
|
390 |
Nesting level is defined by number of hash- or arrayrefs that the encoder |
391 |
needs to traverse to reach a given point or the number of C<{> or C<[> |
392 |
characters without their matching closing parenthesis crossed to reach a |
393 |
given character in a string. |
394 |
|
395 |
Setting the maximum depth to one disallows any nesting, so that ensures |
396 |
that the object is only a single hash/object or array. |
397 |
|
398 |
The argument to C<max_depth> will be rounded up to the next nearest power |
399 |
of two. |
400 |
|
401 |
See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
402 |
|
403 |
=item $json_text = $json->encode ($perl_scalar) |
404 |
|
405 |
Converts the given Perl data structure (a simple scalar or a reference |
406 |
to a hash or array) to its JSON representation. Simple scalars will be |
407 |
converted into JSON string or number sequences, while references to arrays |
408 |
become JSON arrays and references to hashes become JSON objects. Undefined |
409 |
Perl values (e.g. C<undef>) become JSON C<null> values. Neither C<true> |
410 |
nor C<false> values will be generated. |
411 |
|
412 |
=item $perl_scalar = $json->decode ($json_text) |
413 |
|
414 |
The opposite of C<encode>: expects a JSON text and tries to parse it, |
415 |
returning the resulting simple scalar or reference. Croaks on error. |
416 |
|
417 |
JSON numbers and strings become simple Perl scalars. JSON arrays become |
418 |
Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes |
419 |
C<1>, C<false> becomes C<0> and C<null> becomes C<undef>. |
420 |
|
421 |
=item ($perl_scalar, $characters) = $json->decode_prefix ($json_text) |
422 |
|
423 |
This works like the C<decode> method, but instead of raising an exception |
424 |
when there is trailing garbage after the first JSON object, it will |
425 |
silently stop parsing there and return the number of characters consumed |
426 |
so far. |
427 |
|
428 |
This is useful if your JSON texts are not delimited by an outer protocol |
429 |
(which is not the brightest thing to do in the first place) and you need |
430 |
to know where the JSON text ends. |
431 |
|
432 |
JSON::XS->new->decode_prefix ("[1] the tail") |
433 |
=> ([], 3) |
434 |
|
435 |
=back |
436 |
|
437 |
|
438 |
=head1 MAPPING |
439 |
|
440 |
This section describes how JSON::XS maps Perl values to JSON values and |
441 |
vice versa. These mappings are designed to "do the right thing" in most |
442 |
circumstances automatically, preserving round-tripping characteristics |
443 |
(what you put in comes out as something equivalent). |
444 |
|
445 |
For the more enlightened: note that in the following descriptions, |
446 |
lowercase I<perl> refers to the Perl interpreter, while uppcercase I<Perl> |
447 |
refers to the abstract Perl language itself. |
448 |
|
449 |
|
450 |
=head2 JSON -> PERL |
451 |
|
452 |
=over 4 |
453 |
|
454 |
=item object |
455 |
|
456 |
A JSON object becomes a reference to a hash in Perl. No ordering of object |
457 |
keys is preserved (JSON does not preserver object key ordering itself). |
458 |
|
459 |
=item array |
460 |
|
461 |
A JSON array becomes a reference to an array in Perl. |
462 |
|
463 |
=item string |
464 |
|
465 |
A JSON string becomes a string scalar in Perl - Unicode codepoints in JSON |
466 |
are represented by the same codepoints in the Perl string, so no manual |
467 |
decoding is necessary. |
468 |
|
469 |
=item number |
470 |
|
471 |
A JSON number becomes either an integer or numeric (floating point) |
472 |
scalar in perl, depending on its range and any fractional parts. On the |
473 |
Perl level, there is no difference between those as Perl handles all the |
474 |
conversion details, but an integer may take slightly less memory and might |
475 |
represent more values exactly than (floating point) numbers. |
476 |
|
477 |
=item true, false |
478 |
|
479 |
These JSON atoms become C<JSON::XS::true> and C<JSON::XS::false>, |
480 |
respectively. They are overloaded to act almost exactly like the numbers |
481 |
C<1> and C<0>. You can check wether a scalar is a JSON boolean by using |
482 |
the C<JSON::XS::is_bool> function. |
483 |
|
484 |
=item null |
485 |
|
486 |
A JSON null atom becomes C<undef> in Perl. |
487 |
|
488 |
=back |
489 |
|
490 |
|
491 |
=head2 PERL -> JSON |
492 |
|
493 |
The mapping from Perl to JSON is slightly more difficult, as Perl is a |
494 |
truly typeless language, so we can only guess which JSON type is meant by |
495 |
a Perl value. |
496 |
|
497 |
=over 4 |
498 |
|
499 |
=item hash references |
500 |
|
501 |
Perl hash references become JSON objects. As there is no inherent ordering |
502 |
in hash keys (or JSON objects), they will usually be encoded in a |
503 |
pseudo-random order that can change between runs of the same program but |
504 |
stays generally the same within a single run of a program. JSON::XS can |
505 |
optionally sort the hash keys (determined by the I<canonical> flag), so |
506 |
the same datastructure will serialise to the same JSON text (given same |
507 |
settings and version of JSON::XS), but this incurs a runtime overhead |
508 |
and is only rarely useful, e.g. when you want to compare some JSON text |
509 |
against another for equality. |
510 |
|
511 |
=item array references |
512 |
|
513 |
Perl array references become JSON arrays. |
514 |
|
515 |
=item other references |
516 |
|
517 |
Other unblessed references are generally not allowed and will cause an |
518 |
exception to be thrown, except for references to the integers C<0> and |
519 |
C<1>, which get turned into C<false> and C<true> atoms in JSON. You can |
520 |
also use C<JSON::XS::false> and C<JSON::XS::true> to improve readability. |
521 |
|
522 |
to_json [\0,JSON::XS::true] # yields [false,true] |
523 |
|
524 |
=item JSON::XS::true, JSON::XS::false |
525 |
|
526 |
These special values become JSON true and JSON false values, |
527 |
respectively. You cna alos use C<\1> and C<\0> directly if you want. |
528 |
|
529 |
=item blessed objects |
530 |
|
531 |
Blessed objects are not allowed. JSON::XS currently tries to encode their |
532 |
underlying representation (hash- or arrayref), but this behaviour might |
533 |
change in future versions. |
534 |
|
535 |
=item simple scalars |
536 |
|
537 |
Simple Perl scalars (any scalar that is not a reference) are the most |
538 |
difficult objects to encode: JSON::XS will encode undefined scalars as |
539 |
JSON null value, scalars that have last been used in a string context |
540 |
before encoding as JSON strings and anything else as number value: |
541 |
|
542 |
# dump as number |
543 |
to_json [2] # yields [2] |
544 |
to_json [-3.0e17] # yields [-3e+17] |
545 |
my $value = 5; to_json [$value] # yields [5] |
546 |
|
547 |
# used as string, so dump as string |
548 |
print $value; |
549 |
to_json [$value] # yields ["5"] |
550 |
|
551 |
# undef becomes null |
552 |
to_json [undef] # yields [null] |
553 |
|
554 |
You can force the type to be a string by stringifying it: |
555 |
|
556 |
my $x = 3.1; # some variable containing a number |
557 |
"$x"; # stringified |
558 |
$x .= ""; # another, more awkward way to stringify |
559 |
print $x; # perl does it for you, too, quite often |
560 |
|
561 |
You can force the type to be a number by numifying it: |
562 |
|
563 |
my $x = "3"; # some variable containing a string |
564 |
$x += 0; # numify it, ensuring it will be dumped as a number |
565 |
$x *= 1; # same thing, the choise is yours. |
566 |
|
567 |
You can not currently output JSON booleans or force the type in other, |
568 |
less obscure, ways. Tell me if you need this capability. |
569 |
|
570 |
=back |
571 |
|
572 |
|
573 |
=head1 COMPARISON |
574 |
|
575 |
As already mentioned, this module was created because none of the existing |
576 |
JSON modules could be made to work correctly. First I will describe the |
577 |
problems (or pleasures) I encountered with various existing JSON modules, |
578 |
followed by some benchmark values. JSON::XS was designed not to suffer |
579 |
from any of these problems or limitations. |
580 |
|
581 |
=over 4 |
582 |
|
583 |
=item JSON 1.07 |
584 |
|
585 |
Slow (but very portable, as it is written in pure Perl). |
586 |
|
587 |
Undocumented/buggy Unicode handling (how JSON handles unicode values is |
588 |
undocumented. One can get far by feeding it unicode strings and doing |
589 |
en-/decoding oneself, but unicode escapes are not working properly). |
590 |
|
591 |
No roundtripping (strings get clobbered if they look like numbers, e.g. |
592 |
the string C<2.0> will encode to C<2.0> instead of C<"2.0">, and that will |
593 |
decode into the number 2. |
594 |
|
595 |
=item JSON::PC 0.01 |
596 |
|
597 |
Very fast. |
598 |
|
599 |
Undocumented/buggy Unicode handling. |
600 |
|
601 |
No roundtripping. |
602 |
|
603 |
Has problems handling many Perl values (e.g. regex results and other magic |
604 |
values will make it croak). |
605 |
|
606 |
Does not even generate valid JSON (C<{1,2}> gets converted to C<{1:2}> |
607 |
which is not a valid JSON text. |
608 |
|
609 |
Unmaintained (maintainer unresponsive for many months, bugs are not |
610 |
getting fixed). |
611 |
|
612 |
=item JSON::Syck 0.21 |
613 |
|
614 |
Very buggy (often crashes). |
615 |
|
616 |
Very inflexible (no human-readable format supported, format pretty much |
617 |
undocumented. I need at least a format for easy reading by humans and a |
618 |
single-line compact format for use in a protocol, and preferably a way to |
619 |
generate ASCII-only JSON texts). |
620 |
|
621 |
Completely broken (and confusingly documented) Unicode handling (unicode |
622 |
escapes are not working properly, you need to set ImplicitUnicode to |
623 |
I<different> values on en- and decoding to get symmetric behaviour). |
624 |
|
625 |
No roundtripping (simple cases work, but this depends on wether the scalar |
626 |
value was used in a numeric context or not). |
627 |
|
628 |
Dumping hashes may skip hash values depending on iterator state. |
629 |
|
630 |
Unmaintained (maintainer unresponsive for many months, bugs are not |
631 |
getting fixed). |
632 |
|
633 |
Does not check input for validity (i.e. will accept non-JSON input and |
634 |
return "something" instead of raising an exception. This is a security |
635 |
issue: imagine two banks transfering money between each other using |
636 |
JSON. One bank might parse a given non-JSON request and deduct money, |
637 |
while the other might reject the transaction with a syntax error. While a |
638 |
good protocol will at least recover, that is extra unnecessary work and |
639 |
the transaction will still not succeed). |
640 |
|
641 |
=item JSON::DWIW 0.04 |
642 |
|
643 |
Very fast. Very natural. Very nice. |
644 |
|
645 |
Undocumented unicode handling (but the best of the pack. Unicode escapes |
646 |
still don't get parsed properly). |
647 |
|
648 |
Very inflexible. |
649 |
|
650 |
No roundtripping. |
651 |
|
652 |
Does not generate valid JSON texts (key strings are often unquoted, empty keys |
653 |
result in nothing being output) |
654 |
|
655 |
Does not check input for validity. |
656 |
|
657 |
=back |
658 |
|
659 |
|
660 |
=head2 JSON and YAML |
661 |
|
662 |
You often hear that JSON is a subset (or a close subset) of YAML. This is, |
663 |
however, a mass hysteria and very far from the truth. In general, there is |
664 |
no way to configure JSON::XS to output a data structure as valid YAML. |
665 |
|
666 |
If you really must use JSON::XS to generate YAML, you should use this |
667 |
algorithm (subject to change in future versions): |
668 |
|
669 |
my $to_yaml = JSON::XS->new->utf8->space_after (1); |
670 |
my $yaml = $to_yaml->encode ($ref) . "\n"; |
671 |
|
672 |
This will usually generate JSON texts that also parse as valid |
673 |
YAML. Please note that YAML has hardcoded limits on (simple) object key |
674 |
lengths that JSON doesn't have, so you should make sure that your hash |
675 |
keys are noticably shorter than the 1024 characters YAML allows. |
676 |
|
677 |
There might be other incompatibilities that I am not aware of. In general |
678 |
you should not try to generate YAML with a JSON generator or vice versa, |
679 |
or try to parse JSON with a YAML parser or vice versa: chances are high |
680 |
that you will run into severe interoperability problems. |
681 |
|
682 |
|
683 |
=head2 SPEED |
684 |
|
685 |
It seems that JSON::XS is surprisingly fast, as shown in the following |
686 |
tables. They have been generated with the help of the C<eg/bench> program |
687 |
in the JSON::XS distribution, to make it easy to compare on your own |
688 |
system. |
689 |
|
690 |
First comes a comparison between various modules using a very short |
691 |
single-line JSON string: |
692 |
|
693 |
{"method": "handleMessage", "params": ["user1", "we were just talking"], \ |
694 |
"id": null, "array":[1,11,234,-5,1e5,1e7, true, false]} |
695 |
|
696 |
It shows the number of encodes/decodes per second (JSON::XS uses |
697 |
the functional interface, while JSON::XS/2 uses the OO interface |
698 |
with pretty-printing and hashkey sorting enabled, JSON::XS/3 enables |
699 |
shrink). Higher is better: |
700 |
|
701 |
module | encode | decode | |
702 |
-----------|------------|------------| |
703 |
JSON | 7645.468 | 4208.613 | |
704 |
JSON::DWIW | 40721.398 | 77101.176 | |
705 |
JSON::PC | 65948.176 | 78251.940 | |
706 |
JSON::Syck | 22844.793 | 26479.192 | |
707 |
JSON::XS | 388361.481 | 199728.762 | |
708 |
JSON::XS/2 | 218453.333 | 192399.266 | |
709 |
JSON::XS/3 | 338250.323 | 192399.266 | |
710 |
Storable | 15779.925 | 14169.946 | |
711 |
-----------+------------+------------+ |
712 |
|
713 |
That is, JSON::XS is about five times faster than JSON::DWIW on encoding, |
714 |
about three times faster on decoding, and over fourty times faster |
715 |
than JSON, even with pretty-printing and key sorting. It also compares |
716 |
favourably to Storable for small amounts of data. |
717 |
|
718 |
Using a longer test string (roughly 18KB, generated from Yahoo! Locals |
719 |
search API (http://nanoref.com/yahooapis/mgPdGg): |
720 |
|
721 |
module | encode | decode | |
722 |
-----------|------------|------------| |
723 |
JSON | 254.685 | 37.665 | |
724 |
JSON::DWIW | 843.343 | 1049.731 | |
725 |
JSON::PC | 3602.116 | 2307.352 | |
726 |
JSON::Syck | 505.107 | 787.899 | |
727 |
JSON::XS | 5747.196 | 3690.220 | |
728 |
JSON::XS/2 | 3968.121 | 3676.634 | |
729 |
JSON::XS/3 | 6105.246 | 3662.508 | |
730 |
Storable | 4417.337 | 5285.161 | |
731 |
-----------+------------+------------+ |
732 |
|
733 |
Again, JSON::XS leads by far (except for Storable which non-surprisingly |
734 |
decodes faster). |
735 |
|
736 |
On large strings containing lots of high unicode characters, some modules |
737 |
(such as JSON::PC) seem to decode faster than JSON::XS, but the result |
738 |
will be broken due to missing (or wrong) unicode handling. Others refuse |
739 |
to decode or encode properly, so it was impossible to prepare a fair |
740 |
comparison table for that case. |
741 |
|
742 |
|
743 |
=head1 SECURITY CONSIDERATIONS |
744 |
|
745 |
When you are using JSON in a protocol, talking to untrusted potentially |
746 |
hostile creatures requires relatively few measures. |
747 |
|
748 |
First of all, your JSON decoder should be secure, that is, should not have |
749 |
any buffer overflows. Obviously, this module should ensure that and I am |
750 |
trying hard on making that true, but you never know. |
751 |
|
752 |
Second, you need to avoid resource-starving attacks. That means you should |
753 |
limit the size of JSON texts you accept, or make sure then when your |
754 |
resources run out, thats just fine (e.g. by using a separate process that |
755 |
can crash safely). The size of a JSON text in octets or characters is |
756 |
usually a good indication of the size of the resources required to decode |
757 |
it into a Perl structure. |
758 |
|
759 |
Third, JSON::XS recurses using the C stack when decoding objects and |
760 |
arrays. The C stack is a limited resource: for instance, on my amd64 |
761 |
machine with 8MB of stack size I can decode around 180k nested arrays but |
762 |
only 14k nested JSON objects (due to perl itself recursing deeply on croak |
763 |
to free the temporary). If that is exceeded, the program crashes. to be |
764 |
conservative, the default nesting limit is set to 512. If your process |
765 |
has a smaller stack, you should adjust this setting accordingly with the |
766 |
C<max_depth> method. |
767 |
|
768 |
And last but least, something else could bomb you that I forgot to think |
769 |
of. In that case, you get to keep the pieces. I am always open for hints, |
770 |
though... |
771 |
|
772 |
If you are using JSON::XS to return packets to consumption |
773 |
by javascript scripts in a browser you should have a look at |
774 |
L<http://jpsykes.com/47/practical-csrf-and-json-security> to see wether |
775 |
you are vulnerable to some common attack vectors (which really are browser |
776 |
design bugs, but it is still you who will have to deal with it, as major |
777 |
browser developers care only for features, not about doing security |
778 |
right). |
779 |
|
780 |
|
781 |
=head1 BUGS |
782 |
|
783 |
While the goal of this module is to be correct, that unfortunately does |
784 |
not mean its bug-free, only that I think its design is bug-free. It is |
785 |
still relatively early in its development. If you keep reporting bugs they |
786 |
will be fixed swiftly, though. |
787 |
|
788 |
=cut |
789 |
|
790 |
our $true = do { bless \(my $dummy = 1), "JSON::XS::Boolean" }; |
791 |
our $false = do { bless \(my $dummy = 0), "JSON::XS::Boolean" }; |
792 |
|
793 |
sub true() { $true } |
794 |
sub false() { $false } |
795 |
|
796 |
sub is_bool($) { |
797 |
UNIVERSAL::isa $_[0], "JSON::XS::Boolean" |
798 |
# or UNIVERSAL::isa $_[0], "JSON::Literal" |
799 |
} |
800 |
|
801 |
XSLoader::load "JSON::XS", $VERSION; |
802 |
|
803 |
package JSON::XS::Boolean; |
804 |
|
805 |
use overload |
806 |
"0+" => sub { ${$_[0]} }, |
807 |
"++" => sub { $_[0] = ${$_[0]} + 1 }, |
808 |
"--" => sub { $_[0] = ${$_[0]} - 1 }, |
809 |
fallback => 1; |
810 |
|
811 |
1; |
812 |
|
813 |
=head1 AUTHOR |
814 |
|
815 |
Marc Lehmann <schmorp@schmorp.de> |
816 |
http://home.schmorp.de/ |
817 |
|
818 |
=cut |
819 |
|