ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/JSON-XS/XS.pm
(Generate patch)

Comparing JSON-XS/XS.pm (file contents):
Revision 1.2 by root, Thu Mar 22 17:28:50 2007 UTC vs.
Revision 1.24 by root, Thu Mar 29 01:27:36 2007 UTC

3JSON::XS - JSON serialising/deserialising, done correctly and fast 3JSON::XS - JSON serialising/deserialising, done correctly and fast
4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 use JSON::XS; 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);
8 24
9=head1 DESCRIPTION 25=head1 DESCRIPTION
10 26
11This module converts Perl data structures to JSON and vice versa. Its 27This module converts Perl data structures to JSON and vice versa. Its
12primary goal is to be I<correct> and its secondary goal is to be 28primary goal is to be I<correct> and its secondary goal is to be
18their maintainers are unresponsive, gone missing, or not listening to bug 34their maintainers are unresponsive, gone missing, or not listening to bug
19reports for other reasons. 35reports for other reasons.
20 36
21See COMPARISON, below, for a comparison to some other JSON modules. 37See COMPARISON, below, for a comparison to some other JSON modules.
22 38
39See MAPPING, below, on how JSON::XS maps perl values to JSON values and
40vice versa.
41
23=head2 FEATURES 42=head2 FEATURES
24 43
25=over 4 44=over 4
26 45
27=item * correct handling of unicode issues 46=item * correct unicode handling
28 47
29This module knows how to handle Unicode, and even documents how it does so. 48This module knows how to handle Unicode, and even documents how and when
49it does so.
30 50
31=item * round-trip integrity 51=item * round-trip integrity
32 52
33When you serialise a perl data structure using only datatypes supported 53When you serialise a perl data structure using only datatypes supported
34by JSON, the deserialised data structure is identical on the Perl level. 54by JSON, the deserialised data structure is identical on the Perl level.
35(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
56like a number).
36 57
37=item * strict checking of JSON correctness 58=item * strict checking of JSON correctness
38 59
39There is no guessing, no generating of illegal JSON strings by default, 60There is no guessing, no generating of illegal JSON texts by default,
40and only JSON is accepted as input (the latter is a security feature). 61and only JSON is accepted as input by default (the latter is a security
62feature).
41 63
42=item * fast 64=item * fast
43 65
44compared to other JSON modules, this module compares favourably. 66Compared to other JSON modules, this module compares favourably in terms
67of speed, too.
45 68
46=item * simple to use 69=item * simple to use
47 70
48This module has both a simple functional interface as well as an OO 71This module has both a simple functional interface as well as an OO
49interface. 72interface.
50 73
51=item * reasonably versatile output formats 74=item * reasonably versatile output formats
52 75
53You can choose between the most compact format possible, a pure-ascii 76You can choose between the most compact guarenteed single-line format
54format, or a pretty-printed format. Or you can combine those features in 77possible (nice for simple line-based protocols), a pure-ascii format
55whatever way you like. 78(for when your transport is not 8-bit clean, still supports the whole
79unicode range), or a pretty-printed format (for when you want to read that
80stuff). Or you can combine those features in whatever way you like.
56 81
57=back 82=back
58 83
59=cut 84=cut
60 85
61package JSON::XS; 86package JSON::XS;
62 87
88use strict;
89
63BEGIN { 90BEGIN {
64 $VERSION = '0.1'; 91 our $VERSION = '0.8';
65 @ISA = qw(Exporter); 92 our @ISA = qw(Exporter);
66 93
67 @EXPORT = qw(to_json from_json); 94 our @EXPORT = qw(to_json from_json objToJson jsonToObj);
68 require Exporter; 95 require Exporter;
69 96
70 require XSLoader; 97 require XSLoader;
71 XSLoader::load JSON::XS::, $VERSION; 98 XSLoader::load JSON::XS::, $VERSION;
72} 99}
76The following convinience methods are provided by this module. They are 103The following convinience methods are provided by this module. They are
77exported by default: 104exported by default:
78 105
79=over 4 106=over 4
80 107
81=item $json_string = to_json $perl_scalar 108=item $json_text = to_json $perl_scalar
82 109
83Converts the given Perl data structure (a simple scalar or a reference to 110Converts the given Perl data structure (a simple scalar or a reference to
84a hash or array) to a UTF-8 encoded, binary string (that is, the string contains 111a hash or array) to a UTF-8 encoded, binary string (that is, the string contains
85octets only). Croaks on error. 112octets only). Croaks on error.
86 113
87This function call is functionally identical to C<< JSON::XS->new->utf8 114This function call is functionally identical to:
88(1)->encode ($perl_scalar) >>.
89 115
116 $json_text = JSON::XS->new->utf8->encode ($perl_scalar)
117
118except being faster.
119
90=item $perl_scalar = from_json $json_string 120=item $perl_scalar = from_json $json_text
91 121
92The opposite of C<to_json>: expects an UTF-8 (binary) string and tries to 122The opposite of C<to_json>: expects an UTF-8 (binary) string and tries to
93parse that as an UTF-8 encoded JSON string, returning the resulting simple 123parse that as an UTF-8 encoded JSON text, returning the resulting simple
94scalar or reference. Croaks on error. 124scalar or reference. Croaks on error.
95 125
96This function call is functionally identical to C<< JSON::XS->new->utf8 126This function call is functionally identical to:
97(1)->decode ($json_string) >>. 127
128 $perl_scalar = JSON::XS->new->utf8->decode ($json_text)
129
130except being faster.
98 131
99=back 132=back
133
100 134
101=head1 OBJECT-ORIENTED INTERFACE 135=head1 OBJECT-ORIENTED INTERFACE
102 136
103The object oriented interface lets you configure your own encoding or 137The object oriented interface lets you configure your own encoding or
104decoding style, within the limits of supported formats. 138decoding style, within the limits of supported formats.
111strings. All boolean flags described below are by default I<disabled>. 145strings. All boolean flags described below are by default I<disabled>.
112 146
113The mutators for flags all return the JSON object again and thus calls can 147The mutators for flags all return the JSON object again and thus calls can
114be chained: 148be chained:
115 149
116 my $json = JSON::XS->new->utf8(1)->pretty(1)->encode ({a => [1,2]}) 150 my $json = JSON::XS->new->utf8->space_after->encode ({a => [1,2]})
117 => {"a" : [1, 2]} 151 => {"a": [1, 2]}
118 152
119=item $json = $json->ascii ($enable) 153=item $json = $json->ascii ([$enable])
120 154
121If C<$enable> is true, then the C<encode> method will not generate 155If C<$enable> is true (or missing), then the C<encode> method will not
122characters outside the code range C<0..127>. Any unicode characters 156generate characters outside the code range C<0..127> (which is ASCII). Any
123outside that range will be escaped using either a single \uXXXX (BMP 157unicode characters outside that range will be escaped using either a
124characters) or a double \uHHHH\uLLLLL escape sequence, as per RFC4627. 158single \uXXXX (BMP characters) or a double \uHHHH\uLLLLL escape sequence,
159as per RFC4627.
125 160
126If C<$enable> is false, then the C<encode> method will not escape Unicode 161If C<$enable> is false, then the C<encode> method will not escape Unicode
127characters unless necessary. 162characters unless required by the JSON syntax. This results in a faster
163and more compact format.
128 164
165 JSON::XS->new->ascii (1)->encode ([chr 0x10401])
166 => ["\ud801\udc01"]
167
129=item $json = $json->utf8 ($enable) 168=item $json = $json->utf8 ([$enable])
130 169
131If C<$enable> is true, then the C<encode> method will encode the JSON 170If C<$enable> is true (or missing), then the C<encode> method will encode
132string into UTF-8, as required by many protocols, while the C<decode> 171the JSON result into UTF-8, as required by many protocols, while the
133method expects to be handled an UTF-8-encoded string. Please note that 172C<decode> method expects to be handled an UTF-8-encoded string. Please
134UTF-8-encoded strings do not contain any characters outside the range 173note that UTF-8-encoded strings do not contain any characters outside the
135C<0..255>, they are thus useful for bytewise/binary I/O. 174range C<0..255>, they are thus useful for bytewise/binary I/O. In future
175versions, enabling this option might enable autodetection of the UTF-16
176and UTF-32 encoding families, as described in RFC4627.
136 177
137If C<$enable> is false, then the C<encode> method will return the JSON 178If C<$enable> is false, then the C<encode> method will return the JSON
138string as a (non-encoded) unicode string, while C<decode> expects thus a 179string as a (non-encoded) unicode string, while C<decode> expects thus a
139unicode string. Any decoding or encoding (e.g. to UTF-8 or UTF-16) needs 180unicode string. Any decoding or encoding (e.g. to UTF-8 or UTF-16) needs
140to be done yourself, e.g. using the Encode module. 181to be done yourself, e.g. using the Encode module.
141 182
183Example, output UTF-16BE-encoded JSON:
184
185 use Encode;
186 $jsontext = encode "UTF-16BE", JSON::XS->new->encode ($object);
187
188Example, decode UTF-32LE-encoded JSON:
189
190 use Encode;
191 $object = JSON::XS->new->decode (decode "UTF-32LE", $jsontext);
192
142=item $json = $json->pretty ($enabla) 193=item $json = $json->pretty ([$enable])
143 194
144This enables (or disables) all of the C<indent>, C<space_before> and 195This enables (or disables) all of the C<indent>, C<space_before> and
145C<space_after> (and in the future possibly more) settings in one call to 196C<space_after> (and in the future possibly more) flags in one call to
146generate the most readable (or most compact) form possible. 197generate the most readable (or most compact) form possible.
147 198
199Example, pretty-print some simple structure:
200
201 my $json = JSON::XS->new->pretty(1)->encode ({a => [1,2]})
202 =>
203 {
204 "a" : [
205 1,
206 2
207 ]
208 }
209
148=item $json = $json->indent ($enable) 210=item $json = $json->indent ([$enable])
149 211
150If C<$enable> is true, then the C<encode> method will use a multiline 212If C<$enable> is true (or missing), then the C<encode> method will use a multiline
151format as output, putting every array member or object/hash key-value pair 213format as output, putting every array member or object/hash key-value pair
152into its own line, identing them properly. 214into its own line, identing them properly.
153 215
154If C<$enable> is false, no newlines or indenting will be produced, and the 216If C<$enable> is false, no newlines or indenting will be produced, and the
155resulting JSON strings is guarenteed not to contain any C<newlines>. 217resulting JSON text is guarenteed not to contain any C<newlines>.
156 218
157This setting has no effect when decoding JSON strings. 219This setting has no effect when decoding JSON texts.
158 220
159=item $json = $json->space_before ($enable) 221=item $json = $json->space_before ([$enable])
160 222
161If C<$enable> is true, then the C<encode> method will add an extra 223If C<$enable> is true (or missing), then the C<encode> method will add an extra
162optional space before the C<:> separating keys from values in JSON objects. 224optional space before the C<:> separating keys from values in JSON objects.
163 225
164If C<$enable> is false, then the C<encode> method will not add any extra 226If C<$enable> is false, then the C<encode> method will not add any extra
165space at those places. 227space at those places.
166 228
167This setting has no effect when decoding JSON strings. You will also most 229This setting has no effect when decoding JSON texts. You will also
168likely combine this setting with C<space_after>. 230most likely combine this setting with C<space_after>.
169 231
232Example, space_before enabled, space_after and indent disabled:
233
234 {"key" :"value"}
235
170=item $json = $json->space_after ($enable) 236=item $json = $json->space_after ([$enable])
171 237
172If C<$enable> is true, then the C<encode> method will add an extra 238If C<$enable> is true (or missing), then the C<encode> method will add an extra
173optional space after the C<:> separating keys from values in JSON objects 239optional space after the C<:> separating keys from values in JSON objects
174and extra whitespace after the C<,> separating key-value pairs and array 240and extra whitespace after the C<,> separating key-value pairs and array
175members. 241members.
176 242
177If C<$enable> is false, then the C<encode> method will not add any extra 243If C<$enable> is false, then the C<encode> method will not add any extra
178space at those places. 244space at those places.
179 245
180This setting has no effect when decoding JSON strings. 246This setting has no effect when decoding JSON texts.
181 247
248Example, space_before and indent disabled, space_after enabled:
249
250 {"key": "value"}
251
182=item $json = $json->canonical ($enable) 252=item $json = $json->canonical ([$enable])
183 253
184If C<$enable> is true, then the C<encode> method will output JSON objects 254If C<$enable> is true (or missing), then the C<encode> method will output JSON objects
185by sorting their keys. This is adding a comparatively high overhead. 255by sorting their keys. This is adding a comparatively high overhead.
186 256
187If C<$enable> is false, then the C<encode> method will output key-value 257If C<$enable> is false, then the C<encode> method will output key-value
188pairs in the order Perl stores them (which will likely change between runs 258pairs in the order Perl stores them (which will likely change between runs
189of the same script). 259of the same script).
190 260
191This option is useful if you want the same data structure to be encoded as 261This option is useful if you want the same data structure to be encoded as
192the same JSON string (given the same overall settings). If it is disabled, 262the same JSON text (given the same overall settings). If it is disabled,
193the same hash migh be encoded differently even if contains the same data, 263the same hash migh be encoded differently even if contains the same data,
194as key-value pairs have no inherent ordering in Perl. 264as key-value pairs have no inherent ordering in Perl.
195 265
196This setting has no effect when decoding JSON strings. 266This setting has no effect when decoding JSON texts.
197 267
268=item $json = $json->allow_nonref ([$enable])
269
270If C<$enable> is true (or missing), then the C<encode> method can convert a
271non-reference into its corresponding string, number or null JSON value,
272which is an extension to RFC4627. Likewise, C<decode> will accept those JSON
273values instead of croaking.
274
275If C<$enable> is false, then the C<encode> method will croak if it isn't
276passed an arrayref or hashref, as JSON texts must either be an object
277or array. Likewise, C<decode> will croak if given something that is not a
278JSON object or array.
279
280Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>,
281resulting in an invalid JSON text:
282
283 JSON::XS->new->allow_nonref->encode ("Hello, World!")
284 => "Hello, World!"
285
286=item $json = $json->shrink ([$enable])
287
288Perl usually over-allocates memory a bit when allocating space for
289strings. This flag optionally resizes strings generated by either
290C<encode> or C<decode> to their minimum size possible. This can save
291memory when your JSON texts are either very very long or you have many
292short strings. It will also try to downgrade any strings to octet-form
293if possible: perl stores strings internally either in an encoding called
294UTF-X or in octet-form. The latter cannot store everything but uses less
295space in general (and some buggy Perl or C code might even rely on that
296internal representation being used).
297
298The actual definition of what shrink does might change in future versions,
299but it will always try to save space at the expense of time.
300
301If C<$enable> is true (or missing), the string returned by C<encode> will
302be shrunk-to-fit, while all strings generated by C<decode> will also be
303shrunk-to-fit.
304
305If C<$enable> is false, then the normal perl allocation algorithms are used.
306If you work with your data, then this is likely to be faster.
307
308In the future, this setting might control other things, such as converting
309strings that look like integers or floats into integers or floats
310internally (there is no difference on the Perl level), saving space.
311
312=item $json = $json->max_depth ([$maximum_nesting_depth])
313
314Sets the maximum nesting level (default C<8192>) accepted while encoding
315or decoding. If the JSON text or Perl data structure has an equal or
316higher nesting level then this limit, then the encoder and decoder will
317stop and croak at that point.
318
319Nesting level is defined by number of hash- or arrayrefs that the encoder
320needs to traverse to reach a given point or the number of C<{> or C<[>
321characters without their matching closing parenthesis crossed to reach a
322given character in a string.
323
324Setting the maximum depth to one disallows any nesting, so that ensures
325that the object is only a single hash/object or array.
326
327The argument to C<max_depth> will be rounded up to the next nearest power
328of two.
329
330See SECURITY CONSIDERATIONS, below, for more info on why this is useful.
331
198=item $json_string = $json->encode ($perl_scalar) 332=item $json_text = $json->encode ($perl_scalar)
199 333
200Converts the given Perl data structure (a simple scalar or a reference 334Converts the given Perl data structure (a simple scalar or a reference
201to a hash or array) to its JSON representation. Simple scalars will be 335to a hash or array) to its JSON representation. Simple scalars will be
202converted into JSON string or number sequences, while references to arrays 336converted into JSON string or number sequences, while references to arrays
203become JSON arrays and references to hashes become JSON objects. Undefined 337become JSON arrays and references to hashes become JSON objects. Undefined
204Perl values (e.g. C<undef>) become JSON C<null> values. Neither C<true> 338Perl values (e.g. C<undef>) become JSON C<null> values. Neither C<true>
205nor C<false> values will be generated. 339nor C<false> values will be generated.
206 340
207=item $perl_scalar = $json->decode ($json_string) 341=item $perl_scalar = $json->decode ($json_text)
208 342
209The opposite of C<encode>: expects a JSON string and tries to parse it, 343The opposite of C<encode>: expects a JSON text and tries to parse it,
210returning the resulting simple scalar or reference. Croaks on error. 344returning the resulting simple scalar or reference. Croaks on error.
211 345
212JSON numbers and strings become simple Perl scalars. JSON arrays become 346JSON numbers and strings become simple Perl scalars. JSON arrays become
213Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes 347Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes
214C<1>, C<false> becomes C<0> and C<null> becomes C<undef>. 348C<1>, C<false> becomes C<0> and C<null> becomes C<undef>.
215 349
216=back 350=back
217 351
352
353=head1 MAPPING
354
355This section describes how JSON::XS maps Perl values to JSON values and
356vice versa. These mappings are designed to "do the right thing" in most
357circumstances automatically, preserving round-tripping characteristics
358(what you put in comes out as something equivalent).
359
360For the more enlightened: note that in the following descriptions,
361lowercase I<perl> refers to the Perl interpreter, while uppcercase I<Perl>
362refers to the abstract Perl language itself.
363
364=head2 JSON -> PERL
365
366=over 4
367
368=item object
369
370A JSON object becomes a reference to a hash in Perl. No ordering of object
371keys is preserved (JSON does not preserver object key ordering itself).
372
373=item array
374
375A JSON array becomes a reference to an array in Perl.
376
377=item string
378
379A JSON string becomes a string scalar in Perl - Unicode codepoints in JSON
380are represented by the same codepoints in the Perl string, so no manual
381decoding is necessary.
382
383=item number
384
385A JSON number becomes either an integer or numeric (floating point)
386scalar in perl, depending on its range and any fractional parts. On the
387Perl level, there is no difference between those as Perl handles all the
388conversion details, but an integer may take slightly less memory and might
389represent more values exactly than (floating point) numbers.
390
391=item true, false
392
393These JSON atoms become C<0>, C<1>, respectively. Information is lost in
394this process. Future versions might represent those values differently,
395but they will be guarenteed to act like these integers would normally in
396Perl.
397
398=item null
399
400A JSON null atom becomes C<undef> in Perl.
401
402=back
403
404=head2 PERL -> JSON
405
406The mapping from Perl to JSON is slightly more difficult, as Perl is a
407truly typeless language, so we can only guess which JSON type is meant by
408a Perl value.
409
410=over 4
411
412=item hash references
413
414Perl hash references become JSON objects. As there is no inherent ordering
415in hash keys, they will usually be encoded in a pseudo-random order that
416can change between runs of the same program but stays generally the same
417within a single run of a program. JSON::XS can optionally sort the hash
418keys (determined by the I<canonical> flag), so the same datastructure
419will serialise to the same JSON text (given same settings and version of
420JSON::XS), but this incurs a runtime overhead.
421
422=item array references
423
424Perl array references become JSON arrays.
425
426=item blessed objects
427
428Blessed objects are not allowed. JSON::XS currently tries to encode their
429underlying representation (hash- or arrayref), but this behaviour might
430change in future versions.
431
432=item simple scalars
433
434Simple Perl scalars (any scalar that is not a reference) are the most
435difficult objects to encode: JSON::XS will encode undefined scalars as
436JSON null value, scalars that have last been used in a string context
437before encoding as JSON strings and anything else as number value:
438
439 # dump as number
440 to_json [2] # yields [2]
441 to_json [-3.0e17] # yields [-3e+17]
442 my $value = 5; to_json [$value] # yields [5]
443
444 # used as string, so dump as string
445 print $value;
446 to_json [$value] # yields ["5"]
447
448 # undef becomes null
449 to_json [undef] # yields [null]
450
451You can force the type to be a string by stringifying it:
452
453 my $x = 3.1; # some variable containing a number
454 "$x"; # stringified
455 $x .= ""; # another, more awkward way to stringify
456 print $x; # perl does it for you, too, quite often
457
458You can force the type to be a number by numifying it:
459
460 my $x = "3"; # some variable containing a string
461 $x += 0; # numify it, ensuring it will be dumped as a number
462 $x *= 1; # same thing, the choise is yours.
463
464You can not currently output JSON booleans or force the type in other,
465less obscure, ways. Tell me if you need this capability.
466
467=item circular data structures
468
469Those will be encoded until memory or stackspace runs out.
470
471=back
472
473
474=head1 COMPARISON
475
476As already mentioned, this module was created because none of the existing
477JSON modules could be made to work correctly. First I will describe the
478problems (or pleasures) I encountered with various existing JSON modules,
479followed by some benchmark values. JSON::XS was designed not to suffer
480from any of these problems or limitations.
481
482=over 4
483
484=item JSON 1.07
485
486Slow (but very portable, as it is written in pure Perl).
487
488Undocumented/buggy Unicode handling (how JSON handles unicode values is
489undocumented. One can get far by feeding it unicode strings and doing
490en-/decoding oneself, but unicode escapes are not working properly).
491
492No roundtripping (strings get clobbered if they look like numbers, e.g.
493the string C<2.0> will encode to C<2.0> instead of C<"2.0">, and that will
494decode into the number 2.
495
496=item JSON::PC 0.01
497
498Very fast.
499
500Undocumented/buggy Unicode handling.
501
502No roundtripping.
503
504Has problems handling many Perl values (e.g. regex results and other magic
505values will make it croak).
506
507Does not even generate valid JSON (C<{1,2}> gets converted to C<{1:2}>
508which is not a valid JSON text.
509
510Unmaintained (maintainer unresponsive for many months, bugs are not
511getting fixed).
512
513=item JSON::Syck 0.21
514
515Very buggy (often crashes).
516
517Very inflexible (no human-readable format supported, format pretty much
518undocumented. I need at least a format for easy reading by humans and a
519single-line compact format for use in a protocol, and preferably a way to
520generate ASCII-only JSON texts).
521
522Completely broken (and confusingly documented) Unicode handling (unicode
523escapes are not working properly, you need to set ImplicitUnicode to
524I<different> values on en- and decoding to get symmetric behaviour).
525
526No roundtripping (simple cases work, but this depends on wether the scalar
527value was used in a numeric context or not).
528
529Dumping hashes may skip hash values depending on iterator state.
530
531Unmaintained (maintainer unresponsive for many months, bugs are not
532getting fixed).
533
534Does not check input for validity (i.e. will accept non-JSON input and
535return "something" instead of raising an exception. This is a security
536issue: imagine two banks transfering money between each other using
537JSON. One bank might parse a given non-JSON request and deduct money,
538while the other might reject the transaction with a syntax error. While a
539good protocol will at least recover, that is extra unnecessary work and
540the transaction will still not succeed).
541
542=item JSON::DWIW 0.04
543
544Very fast. Very natural. Very nice.
545
546Undocumented unicode handling (but the best of the pack. Unicode escapes
547still don't get parsed properly).
548
549Very inflexible.
550
551No roundtripping.
552
553Does not generate valid JSON texts (key strings are often unquoted, empty keys
554result in nothing being output)
555
556Does not check input for validity.
557
558=back
559
560=head2 SPEED
561
562It seems that JSON::XS is surprisingly fast, as shown in the following
563tables. They have been generated with the help of the C<eg/bench> program
564in the JSON::XS distribution, to make it easy to compare on your own
565system.
566
567First comes a comparison between various modules using a very short JSON
568string:
569
570 {"method": "handleMessage", "params": ["user1", "we were just talking"], "id": null}
571
572It shows the number of encodes/decodes per second (JSON::XS uses the
573functional interface, while JSON::XS/2 uses the OO interface with
574pretty-printing and hashkey sorting enabled). Higher is better:
575
576 module | encode | decode |
577 -----------|------------|------------|
578 JSON | 11488.516 | 7823.035 |
579 JSON::DWIW | 94708.054 | 129094.260 |
580 JSON::PC | 63884.157 | 128528.212 |
581 JSON::Syck | 34898.677 | 42096.911 |
582 JSON::XS | 654027.064 | 396423.669 |
583 JSON::XS/2 | 371564.190 | 371725.613 |
584 -----------+------------+------------+
585
586That is, JSON::XS is more than six times faster than JSON::DWIW on
587encoding, more than three times faster on decoding, and about thirty times
588faster than JSON, even with pretty-printing and key sorting.
589
590Using a longer test string (roughly 18KB, generated from Yahoo! Locals
591search API (http://nanoref.com/yahooapis/mgPdGg):
592
593 module | encode | decode |
594 -----------|------------|------------|
595 JSON | 273.023 | 44.674 |
596 JSON::DWIW | 1089.383 | 1145.704 |
597 JSON::PC | 3097.419 | 2393.921 |
598 JSON::Syck | 514.060 | 843.053 |
599 JSON::XS | 6479.668 | 3636.364 |
600 JSON::XS/2 | 3774.221 | 3599.124 |
601 -----------+------------+------------+
602
603Again, JSON::XS leads by far.
604
605On large strings containing lots of high unicode characters, some modules
606(such as JSON::PC) seem to decode faster than JSON::XS, but the result
607will be broken due to missing (or wrong) unicode handling. Others refuse
608to decode or encode properly, so it was impossible to prepare a fair
609comparison table for that case.
610
611
612=head1 SECURITY CONSIDERATIONS
613
614When you are using JSON in a protocol, talking to untrusted potentially
615hostile creatures requires relatively few measures.
616
617First of all, your JSON decoder should be secure, that is, should not have
618any buffer overflows. Obviously, this module should ensure that and I am
619trying hard on making that true, but you never know.
620
621Second, you need to avoid resource-starving attacks. That means you should
622limit the size of JSON texts you accept, or make sure then when your
623resources run out, thats just fine (e.g. by using a separate process that
624can crash safely). The size of a JSON text in octets or characters is
625usually a good indication of the size of the resources required to decode
626it into a Perl structure.
627
628Third, JSON::XS recurses using the C stack when decoding objects and
629arrays. The C stack is a limited resource: for instance, on my amd64
630machine with 8MB of stack size I can decode around 180k nested arrays
631but only 14k nested JSON objects. If that is exceeded, the program
632crashes. Thats why the default nesting limit is set to 8192. If your
633process has a smaller stack, you should adjust this setting accordingly
634with the C<max_depth> method.
635
636And last but least, something else could bomb you that I forgot to think
637of. In that case, you get to keep the pieces. I am alway sopen for hints,
638though...
639
640
641=head1 BUGS
642
643While the goal of this module is to be correct, that unfortunately does
644not mean its bug-free, only that I think its design is bug-free. It is
645still relatively early in its development. If you keep reporting bugs they
646will be fixed swiftly, though.
647
218=cut 648=cut
219 649
2201; 6501;
221 651
222=head1 AUTHOR 652=head1 AUTHOR

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines