1 | =head1 NAME |
1 | =head1 NAME |
2 | |
2 | |
|
|
3 | =encoding utf-8 |
|
|
4 | |
3 | JSON::XS - JSON serialising/deserialising, done correctly and fast |
5 | JSON::XS - JSON serialising/deserialising, done correctly and fast |
4 | |
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) |
|
|
9 | |
5 | =head1 SYNOPSIS |
10 | =head1 SYNOPSIS |
6 | |
11 | |
7 | use JSON::XS; |
12 | use JSON::XS; |
8 | |
13 | |
|
|
14 | # exported functions, they croak on error |
|
|
15 | # and expect/generate UTF-8 |
|
|
16 | |
|
|
17 | $utf8_encoded_json_text = encode_json $perl_hash_or_arrayref; |
|
|
18 | $perl_hash_or_arrayref = decode_json $utf8_encoded_json_text; |
|
|
19 | |
|
|
20 | # OO-interface |
|
|
21 | |
|
|
22 | $coder = JSON::XS->new->ascii->pretty->allow_nonref; |
|
|
23 | $pretty_printed_unencoded = $coder->encode ($perl_scalar); |
|
|
24 | $perl_scalar = $coder->decode ($unicode_json_text); |
|
|
25 | |
|
|
26 | # Note that JSON version 2.0 and above will automatically use JSON::XS |
|
|
27 | # if available, at virtually no speed overhead either, so you should |
|
|
28 | # be able to just: |
|
|
29 | |
|
|
30 | use JSON; |
|
|
31 | |
|
|
32 | # and do the same things, except that you have a pure-perl fallback now. |
|
|
33 | |
9 | =head1 DESCRIPTION |
34 | =head1 DESCRIPTION |
10 | |
35 | |
|
|
36 | This module converts Perl data structures to JSON and vice versa. Its |
|
|
37 | primary goal is to be I<correct> and its secondary goal is to be |
|
|
38 | I<fast>. To reach the latter goal it was written in C. |
|
|
39 | |
|
|
40 | Beginning with version 2.0 of the JSON module, when both JSON and |
|
|
41 | JSON::XS are installed, then JSON will fall back on JSON::XS (this can be |
|
|
42 | overriden) with no overhead due to emulation (by inheritign constructor |
|
|
43 | and methods). If JSON::XS is not available, it will fall back to the |
|
|
44 | compatible JSON::PP module as backend, so using JSON instead of JSON::XS |
|
|
45 | gives you a portable JSON API that can be fast when you need and doesn't |
|
|
46 | require a C compiler when that is a problem. |
|
|
47 | |
|
|
48 | As this is the n-th-something JSON module on CPAN, what was the reason |
|
|
49 | to write yet another JSON module? While it seems there are many JSON |
|
|
50 | modules, none of them correctly handle all corner cases, and in most cases |
|
|
51 | their maintainers are unresponsive, gone missing, or not listening to bug |
|
|
52 | reports for other reasons. |
|
|
53 | |
|
|
54 | See COMPARISON, below, for a comparison to some other JSON modules. |
|
|
55 | |
|
|
56 | See MAPPING, below, on how JSON::XS maps perl values to JSON values and |
|
|
57 | vice versa. |
|
|
58 | |
|
|
59 | =head2 FEATURES |
|
|
60 | |
11 | =over 4 |
61 | =over 4 |
12 | |
62 | |
|
|
63 | =item * correct Unicode handling |
|
|
64 | |
|
|
65 | This module knows how to handle Unicode, documents how and when it does |
|
|
66 | so, and even documents what "correct" means. |
|
|
67 | |
|
|
68 | =item * round-trip integrity |
|
|
69 | |
|
|
70 | When you serialise a perl data structure using only datatypes supported |
|
|
71 | by JSON, the deserialised data structure is identical on the Perl level. |
|
|
72 | (e.g. the string "2.0" doesn't suddenly become "2" just because it looks |
|
|
73 | like a number). There minor I<are> exceptions to this, read the MAPPING |
|
|
74 | section below to learn about those. |
|
|
75 | |
|
|
76 | =item * strict checking of JSON correctness |
|
|
77 | |
|
|
78 | There is no guessing, no generating of illegal JSON texts by default, |
|
|
79 | and only JSON is accepted as input by default (the latter is a security |
|
|
80 | feature). |
|
|
81 | |
|
|
82 | =item * fast |
|
|
83 | |
|
|
84 | Compared to other JSON modules and other serialisers such as Storable, |
|
|
85 | this module usually compares favourably in terms of speed, too. |
|
|
86 | |
|
|
87 | =item * simple to use |
|
|
88 | |
|
|
89 | This module has both a simple functional interface as well as an objetc |
|
|
90 | oriented interface interface. |
|
|
91 | |
|
|
92 | =item * reasonably versatile output formats |
|
|
93 | |
|
|
94 | You can choose between the most compact guaranteed-single-line format |
|
|
95 | possible (nice for simple line-based protocols), a pure-ascii format |
|
|
96 | (for when your transport is not 8-bit clean, still supports the whole |
|
|
97 | Unicode range), or a pretty-printed format (for when you want to read that |
|
|
98 | stuff). Or you can combine those features in whatever way you like. |
|
|
99 | |
|
|
100 | =back |
|
|
101 | |
13 | =cut |
102 | =cut |
14 | |
103 | |
15 | package JSON::XS; |
104 | package JSON::XS; |
16 | |
105 | |
17 | BEGIN { |
106 | use strict; |
|
|
107 | |
18 | $VERSION = '0.1'; |
108 | our $VERSION = '2.1'; |
19 | @ISA = qw(Exporter); |
109 | our @ISA = qw(Exporter); |
20 | |
110 | |
21 | require Exporter; |
111 | our @EXPORT = qw(encode_json decode_json to_json from_json); |
22 | |
112 | |
|
|
113 | sub to_json($) { |
23 | require XSLoader; |
114 | require Carp; |
24 | XSLoader::load JSON::XS::, $VERSION; |
115 | Carp::croak ("JSON::XS::to_json has been renamed to encode_json, either downgrade to pre-2.0 versions of JSON::XS or rename the call"); |
25 | } |
116 | } |
26 | |
117 | |
27 | =item |
118 | sub from_json($) { |
|
|
119 | require Carp; |
|
|
120 | Carp::croak ("JSON::XS::from_json has been renamed to decode_json, either downgrade to pre-2.0 versions of JSON::XS or rename the call"); |
|
|
121 | } |
|
|
122 | |
|
|
123 | use Exporter; |
|
|
124 | use XSLoader; |
|
|
125 | |
|
|
126 | =head1 FUNCTIONAL INTERFACE |
|
|
127 | |
|
|
128 | The following convenience methods are provided by this module. They are |
|
|
129 | exported by default: |
|
|
130 | |
|
|
131 | =over 4 |
|
|
132 | |
|
|
133 | =item $json_text = encode_json $perl_scalar |
|
|
134 | |
|
|
135 | Converts the given Perl data structure to a UTF-8 encoded, binary string |
|
|
136 | (that is, the string contains octets only). Croaks on error. |
|
|
137 | |
|
|
138 | This function call is functionally identical to: |
|
|
139 | |
|
|
140 | $json_text = JSON::XS->new->utf8->encode ($perl_scalar) |
|
|
141 | |
|
|
142 | except being faster. |
|
|
143 | |
|
|
144 | =item $perl_scalar = decode_json $json_text |
|
|
145 | |
|
|
146 | The opposite of C<encode_json>: expects an UTF-8 (binary) string and tries |
|
|
147 | to parse that as an UTF-8 encoded JSON text, returning the resulting |
|
|
148 | reference. Croaks on error. |
|
|
149 | |
|
|
150 | This function call is functionally identical to: |
|
|
151 | |
|
|
152 | $perl_scalar = JSON::XS->new->utf8->decode ($json_text) |
|
|
153 | |
|
|
154 | except being faster. |
|
|
155 | |
|
|
156 | =item $is_boolean = JSON::XS::is_bool $scalar |
|
|
157 | |
|
|
158 | Returns true if the passed scalar represents either JSON::XS::true or |
|
|
159 | JSON::XS::false, two constants that act like C<1> and C<0>, respectively |
|
|
160 | and are used to represent JSON C<true> and C<false> values in Perl. |
|
|
161 | |
|
|
162 | See MAPPING, below, for more information on how JSON values are mapped to |
|
|
163 | Perl. |
|
|
164 | |
|
|
165 | =back |
|
|
166 | |
|
|
167 | |
|
|
168 | =head1 A FEW NOTES ON UNICODE AND PERL |
|
|
169 | |
|
|
170 | Since this often leads to confusion, here are a few very clear words on |
|
|
171 | how Unicode works in Perl, modulo bugs. |
|
|
172 | |
|
|
173 | =over 4 |
|
|
174 | |
|
|
175 | =item 1. Perl strings can store characters with ordinal values > 255. |
|
|
176 | |
|
|
177 | This enables you to store Unicode characters as single characters in a |
|
|
178 | Perl string - very natural. |
|
|
179 | |
|
|
180 | =item 2. Perl does I<not> associate an encoding with your strings. |
|
|
181 | |
|
|
182 | ... until you force it to, e.g. when matching it against a regex, or |
|
|
183 | printing the scalar to a file, in which case Perl either interprets your |
|
|
184 | string as locale-encoded text, octets/binary, or as Unicode, depending |
|
|
185 | on various settings. In no case is an encoding stored together with your |
|
|
186 | data, it is I<use> that decides encoding, not any magical meta data. |
|
|
187 | |
|
|
188 | =item 3. The internal utf-8 flag has no meaning with regards to the |
|
|
189 | encoding of your string. |
|
|
190 | |
|
|
191 | Just ignore that flag unless you debug a Perl bug, a module written in |
|
|
192 | XS or want to dive into the internals of perl. Otherwise it will only |
|
|
193 | confuse you, as, despite the name, it says nothing about how your string |
|
|
194 | is encoded. You can have Unicode strings with that flag set, with that |
|
|
195 | flag clear, and you can have binary data with that flag set and that flag |
|
|
196 | clear. Other possibilities exist, too. |
|
|
197 | |
|
|
198 | If you didn't know about that flag, just the better, pretend it doesn't |
|
|
199 | exist. |
|
|
200 | |
|
|
201 | =item 4. A "Unicode String" is simply a string where each character can be |
|
|
202 | validly interpreted as a Unicode codepoint. |
|
|
203 | |
|
|
204 | If you have UTF-8 encoded data, it is no longer a Unicode string, but a |
|
|
205 | Unicode string encoded in UTF-8, giving you a binary string. |
|
|
206 | |
|
|
207 | =item 5. A string containing "high" (> 255) character values is I<not> a UTF-8 string. |
|
|
208 | |
|
|
209 | It's a fact. Learn to live with it. |
|
|
210 | |
|
|
211 | =back |
|
|
212 | |
|
|
213 | I hope this helps :) |
|
|
214 | |
|
|
215 | |
|
|
216 | =head1 OBJECT-ORIENTED INTERFACE |
|
|
217 | |
|
|
218 | The object oriented interface lets you configure your own encoding or |
|
|
219 | decoding style, within the limits of supported formats. |
|
|
220 | |
|
|
221 | =over 4 |
|
|
222 | |
|
|
223 | =item $json = new JSON::XS |
|
|
224 | |
|
|
225 | Creates a new JSON::XS object that can be used to de/encode JSON |
|
|
226 | strings. All boolean flags described below are by default I<disabled>. |
|
|
227 | |
|
|
228 | The mutators for flags all return the JSON object again and thus calls can |
|
|
229 | be chained: |
|
|
230 | |
|
|
231 | my $json = JSON::XS->new->utf8->space_after->encode ({a => [1,2]}) |
|
|
232 | => {"a": [1, 2]} |
|
|
233 | |
|
|
234 | =item $json = $json->ascii ([$enable]) |
|
|
235 | |
|
|
236 | =item $enabled = $json->get_ascii |
|
|
237 | |
|
|
238 | If C<$enable> is true (or missing), then the C<encode> method will not |
|
|
239 | generate characters outside the code range C<0..127> (which is ASCII). Any |
|
|
240 | Unicode characters outside that range will be escaped using either a |
|
|
241 | single \uXXXX (BMP characters) or a double \uHHHH\uLLLLL escape sequence, |
|
|
242 | as per RFC4627. The resulting encoded JSON text can be treated as a native |
|
|
243 | Unicode string, an ascii-encoded, latin1-encoded or UTF-8 encoded string, |
|
|
244 | or any other superset of ASCII. |
|
|
245 | |
|
|
246 | If C<$enable> is false, then the C<encode> method will not escape Unicode |
|
|
247 | characters unless required by the JSON syntax or other flags. This results |
|
|
248 | in a faster and more compact format. |
|
|
249 | |
|
|
250 | See also the section I<ENCODING/CODESET FLAG NOTES> later in this |
|
|
251 | document. |
|
|
252 | |
|
|
253 | The main use for this flag is to produce JSON texts that can be |
|
|
254 | transmitted over a 7-bit channel, as the encoded JSON texts will not |
|
|
255 | contain any 8 bit characters. |
|
|
256 | |
|
|
257 | JSON::XS->new->ascii (1)->encode ([chr 0x10401]) |
|
|
258 | => ["\ud801\udc01"] |
|
|
259 | |
|
|
260 | =item $json = $json->latin1 ([$enable]) |
|
|
261 | |
|
|
262 | =item $enabled = $json->get_latin1 |
|
|
263 | |
|
|
264 | If C<$enable> is true (or missing), then the C<encode> method will encode |
|
|
265 | the resulting JSON text as latin1 (or iso-8859-1), escaping any characters |
|
|
266 | outside the code range C<0..255>. The resulting string can be treated as a |
|
|
267 | latin1-encoded JSON text or a native Unicode string. The C<decode> method |
|
|
268 | will not be affected in any way by this flag, as C<decode> by default |
|
|
269 | expects Unicode, which is a strict superset of latin1. |
|
|
270 | |
|
|
271 | If C<$enable> is false, then the C<encode> method will not escape Unicode |
|
|
272 | characters unless required by the JSON syntax or other flags. |
|
|
273 | |
|
|
274 | See also the section I<ENCODING/CODESET FLAG NOTES> later in this |
|
|
275 | document. |
|
|
276 | |
|
|
277 | The main use for this flag is efficiently encoding binary data as JSON |
|
|
278 | text, as most octets will not be escaped, resulting in a smaller encoded |
|
|
279 | size. The disadvantage is that the resulting JSON text is encoded |
|
|
280 | in latin1 (and must correctly be treated as such when storing and |
|
|
281 | transferring), a rare encoding for JSON. It is therefore most useful when |
|
|
282 | you want to store data structures known to contain binary data efficiently |
|
|
283 | in files or databases, not when talking to other JSON encoders/decoders. |
|
|
284 | |
|
|
285 | JSON::XS->new->latin1->encode (["\x{89}\x{abc}"] |
|
|
286 | => ["\x{89}\\u0abc"] # (perl syntax, U+abc escaped, U+89 not) |
|
|
287 | |
|
|
288 | =item $json = $json->utf8 ([$enable]) |
|
|
289 | |
|
|
290 | =item $enabled = $json->get_utf8 |
|
|
291 | |
|
|
292 | If C<$enable> is true (or missing), then the C<encode> method will encode |
|
|
293 | the JSON result into UTF-8, as required by many protocols, while the |
|
|
294 | C<decode> method expects to be handled an UTF-8-encoded string. Please |
|
|
295 | note that UTF-8-encoded strings do not contain any characters outside the |
|
|
296 | range C<0..255>, they are thus useful for bytewise/binary I/O. In future |
|
|
297 | versions, enabling this option might enable autodetection of the UTF-16 |
|
|
298 | and UTF-32 encoding families, as described in RFC4627. |
|
|
299 | |
|
|
300 | If C<$enable> is false, then the C<encode> method will return the JSON |
|
|
301 | string as a (non-encoded) Unicode string, while C<decode> expects thus a |
|
|
302 | Unicode string. Any decoding or encoding (e.g. to UTF-8 or UTF-16) needs |
|
|
303 | to be done yourself, e.g. using the Encode module. |
|
|
304 | |
|
|
305 | See also the section I<ENCODING/CODESET FLAG NOTES> later in this |
|
|
306 | document. |
|
|
307 | |
|
|
308 | Example, output UTF-16BE-encoded JSON: |
|
|
309 | |
|
|
310 | use Encode; |
|
|
311 | $jsontext = encode "UTF-16BE", JSON::XS->new->encode ($object); |
|
|
312 | |
|
|
313 | Example, decode UTF-32LE-encoded JSON: |
|
|
314 | |
|
|
315 | use Encode; |
|
|
316 | $object = JSON::XS->new->decode (decode "UTF-32LE", $jsontext); |
|
|
317 | |
|
|
318 | =item $json = $json->pretty ([$enable]) |
|
|
319 | |
|
|
320 | This enables (or disables) all of the C<indent>, C<space_before> and |
|
|
321 | C<space_after> (and in the future possibly more) flags in one call to |
|
|
322 | generate the most readable (or most compact) form possible. |
|
|
323 | |
|
|
324 | Example, pretty-print some simple structure: |
|
|
325 | |
|
|
326 | my $json = JSON::XS->new->pretty(1)->encode ({a => [1,2]}) |
|
|
327 | => |
|
|
328 | { |
|
|
329 | "a" : [ |
|
|
330 | 1, |
|
|
331 | 2 |
|
|
332 | ] |
|
|
333 | } |
|
|
334 | |
|
|
335 | =item $json = $json->indent ([$enable]) |
|
|
336 | |
|
|
337 | =item $enabled = $json->get_indent |
|
|
338 | |
|
|
339 | If C<$enable> is true (or missing), then the C<encode> method will use a multiline |
|
|
340 | format as output, putting every array member or object/hash key-value pair |
|
|
341 | into its own line, indenting them properly. |
|
|
342 | |
|
|
343 | If C<$enable> is false, no newlines or indenting will be produced, and the |
|
|
344 | resulting JSON text is guaranteed not to contain any C<newlines>. |
|
|
345 | |
|
|
346 | This setting has no effect when decoding JSON texts. |
|
|
347 | |
|
|
348 | =item $json = $json->space_before ([$enable]) |
|
|
349 | |
|
|
350 | =item $enabled = $json->get_space_before |
|
|
351 | |
|
|
352 | If C<$enable> is true (or missing), then the C<encode> method will add an extra |
|
|
353 | optional space before the C<:> separating keys from values in JSON objects. |
|
|
354 | |
|
|
355 | If C<$enable> is false, then the C<encode> method will not add any extra |
|
|
356 | space at those places. |
|
|
357 | |
|
|
358 | This setting has no effect when decoding JSON texts. You will also |
|
|
359 | most likely combine this setting with C<space_after>. |
|
|
360 | |
|
|
361 | Example, space_before enabled, space_after and indent disabled: |
|
|
362 | |
|
|
363 | {"key" :"value"} |
|
|
364 | |
|
|
365 | =item $json = $json->space_after ([$enable]) |
|
|
366 | |
|
|
367 | =item $enabled = $json->get_space_after |
|
|
368 | |
|
|
369 | If C<$enable> is true (or missing), then the C<encode> method will add an extra |
|
|
370 | optional space after the C<:> separating keys from values in JSON objects |
|
|
371 | and extra whitespace after the C<,> separating key-value pairs and array |
|
|
372 | members. |
|
|
373 | |
|
|
374 | If C<$enable> is false, then the C<encode> method will not add any extra |
|
|
375 | space at those places. |
|
|
376 | |
|
|
377 | This setting has no effect when decoding JSON texts. |
|
|
378 | |
|
|
379 | Example, space_before and indent disabled, space_after enabled: |
|
|
380 | |
|
|
381 | {"key": "value"} |
|
|
382 | |
|
|
383 | =item $json = $json->relaxed ([$enable]) |
|
|
384 | |
|
|
385 | =item $enabled = $json->get_relaxed |
|
|
386 | |
|
|
387 | If C<$enable> is true (or missing), then C<decode> will accept some |
|
|
388 | extensions to normal JSON syntax (see below). C<encode> will not be |
|
|
389 | affected in anyway. I<Be aware that this option makes you accept invalid |
|
|
390 | JSON texts as if they were valid!>. I suggest only to use this option to |
|
|
391 | parse application-specific files written by humans (configuration files, |
|
|
392 | resource files etc.) |
|
|
393 | |
|
|
394 | If C<$enable> is false (the default), then C<decode> will only accept |
|
|
395 | valid JSON texts. |
|
|
396 | |
|
|
397 | Currently accepted extensions are: |
|
|
398 | |
|
|
399 | =over 4 |
|
|
400 | |
|
|
401 | =item * list items can have an end-comma |
|
|
402 | |
|
|
403 | JSON I<separates> array elements and key-value pairs with commas. This |
|
|
404 | can be annoying if you write JSON texts manually and want to be able to |
|
|
405 | quickly append elements, so this extension accepts comma at the end of |
|
|
406 | such items not just between them: |
|
|
407 | |
|
|
408 | [ |
|
|
409 | 1, |
|
|
410 | 2, <- this comma not normally allowed |
|
|
411 | ] |
|
|
412 | { |
|
|
413 | "k1": "v1", |
|
|
414 | "k2": "v2", <- this comma not normally allowed |
|
|
415 | } |
|
|
416 | |
|
|
417 | =item * shell-style '#'-comments |
|
|
418 | |
|
|
419 | Whenever JSON allows whitespace, shell-style comments are additionally |
|
|
420 | allowed. They are terminated by the first carriage-return or line-feed |
|
|
421 | character, after which more white-space and comments are allowed. |
|
|
422 | |
|
|
423 | [ |
|
|
424 | 1, # this comment not allowed in JSON |
|
|
425 | # neither this one... |
|
|
426 | ] |
|
|
427 | |
|
|
428 | =back |
|
|
429 | |
|
|
430 | =item $json = $json->canonical ([$enable]) |
|
|
431 | |
|
|
432 | =item $enabled = $json->get_canonical |
|
|
433 | |
|
|
434 | If C<$enable> is true (or missing), then the C<encode> method will output JSON objects |
|
|
435 | by sorting their keys. This is adding a comparatively high overhead. |
|
|
436 | |
|
|
437 | If C<$enable> is false, then the C<encode> method will output key-value |
|
|
438 | pairs in the order Perl stores them (which will likely change between runs |
|
|
439 | of the same script). |
|
|
440 | |
|
|
441 | This option is useful if you want the same data structure to be encoded as |
|
|
442 | the same JSON text (given the same overall settings). If it is disabled, |
|
|
443 | the same hash might be encoded differently even if contains the same data, |
|
|
444 | as key-value pairs have no inherent ordering in Perl. |
|
|
445 | |
|
|
446 | This setting has no effect when decoding JSON texts. |
|
|
447 | |
|
|
448 | =item $json = $json->allow_nonref ([$enable]) |
|
|
449 | |
|
|
450 | =item $enabled = $json->get_allow_nonref |
|
|
451 | |
|
|
452 | If C<$enable> is true (or missing), then the C<encode> method can convert a |
|
|
453 | non-reference into its corresponding string, number or null JSON value, |
|
|
454 | which is an extension to RFC4627. Likewise, C<decode> will accept those JSON |
|
|
455 | values instead of croaking. |
|
|
456 | |
|
|
457 | If C<$enable> is false, then the C<encode> method will croak if it isn't |
|
|
458 | passed an arrayref or hashref, as JSON texts must either be an object |
|
|
459 | or array. Likewise, C<decode> will croak if given something that is not a |
|
|
460 | JSON object or array. |
|
|
461 | |
|
|
462 | Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>, |
|
|
463 | resulting in an invalid JSON text: |
|
|
464 | |
|
|
465 | JSON::XS->new->allow_nonref->encode ("Hello, World!") |
|
|
466 | => "Hello, World!" |
|
|
467 | |
|
|
468 | =item $json = $json->allow_blessed ([$enable]) |
|
|
469 | |
|
|
470 | =item $enabled = $json->get_allow_blessed |
|
|
471 | |
|
|
472 | If C<$enable> is true (or missing), then the C<encode> method will not |
|
|
473 | barf when it encounters a blessed reference. Instead, the value of the |
|
|
474 | B<convert_blessed> option will decide whether C<null> (C<convert_blessed> |
|
|
475 | disabled or no C<TO_JSON> method found) or a representation of the |
|
|
476 | object (C<convert_blessed> enabled and C<TO_JSON> method found) is being |
|
|
477 | encoded. Has no effect on C<decode>. |
|
|
478 | |
|
|
479 | If C<$enable> is false (the default), then C<encode> will throw an |
|
|
480 | exception when it encounters a blessed object. |
|
|
481 | |
|
|
482 | =item $json = $json->convert_blessed ([$enable]) |
|
|
483 | |
|
|
484 | =item $enabled = $json->get_convert_blessed |
|
|
485 | |
|
|
486 | If C<$enable> is true (or missing), then C<encode>, upon encountering a |
|
|
487 | blessed object, will check for the availability of the C<TO_JSON> method |
|
|
488 | on the object's class. If found, it will be called in scalar context |
|
|
489 | and the resulting scalar will be encoded instead of the object. If no |
|
|
490 | C<TO_JSON> method is found, the value of C<allow_blessed> will decide what |
|
|
491 | to do. |
|
|
492 | |
|
|
493 | The C<TO_JSON> method may safely call die if it wants. If C<TO_JSON> |
|
|
494 | returns other blessed objects, those will be handled in the same |
|
|
495 | way. C<TO_JSON> must take care of not causing an endless recursion cycle |
|
|
496 | (== crash) in this case. The name of C<TO_JSON> was chosen because other |
|
|
497 | methods called by the Perl core (== not by the user of the object) are |
|
|
498 | usually in upper case letters and to avoid collisions with any C<to_json> |
|
|
499 | function or method. |
|
|
500 | |
|
|
501 | This setting does not yet influence C<decode> in any way, but in the |
|
|
502 | future, global hooks might get installed that influence C<decode> and are |
|
|
503 | enabled by this setting. |
|
|
504 | |
|
|
505 | If C<$enable> is false, then the C<allow_blessed> setting will decide what |
|
|
506 | to do when a blessed object is found. |
|
|
507 | |
|
|
508 | =item $json = $json->filter_json_object ([$coderef->($hashref)]) |
|
|
509 | |
|
|
510 | When C<$coderef> is specified, it will be called from C<decode> each |
|
|
511 | time it decodes a JSON object. The only argument is a reference to the |
|
|
512 | newly-created hash. If the code references returns a single scalar (which |
|
|
513 | need not be a reference), this value (i.e. a copy of that scalar to avoid |
|
|
514 | aliasing) is inserted into the deserialised data structure. If it returns |
|
|
515 | an empty list (NOTE: I<not> C<undef>, which is a valid scalar), the |
|
|
516 | original deserialised hash will be inserted. This setting can slow down |
|
|
517 | decoding considerably. |
|
|
518 | |
|
|
519 | When C<$coderef> is omitted or undefined, any existing callback will |
|
|
520 | be removed and C<decode> will not change the deserialised hash in any |
|
|
521 | way. |
|
|
522 | |
|
|
523 | Example, convert all JSON objects into the integer 5: |
|
|
524 | |
|
|
525 | my $js = JSON::XS->new->filter_json_object (sub { 5 }); |
|
|
526 | # returns [5] |
|
|
527 | $js->decode ('[{}]') |
|
|
528 | # throw an exception because allow_nonref is not enabled |
|
|
529 | # so a lone 5 is not allowed. |
|
|
530 | $js->decode ('{"a":1, "b":2}'); |
|
|
531 | |
|
|
532 | =item $json = $json->filter_json_single_key_object ($key [=> $coderef->($value)]) |
|
|
533 | |
|
|
534 | Works remotely similar to C<filter_json_object>, but is only called for |
|
|
535 | JSON objects having a single key named C<$key>. |
|
|
536 | |
|
|
537 | This C<$coderef> is called before the one specified via |
|
|
538 | C<filter_json_object>, if any. It gets passed the single value in the JSON |
|
|
539 | object. If it returns a single value, it will be inserted into the data |
|
|
540 | structure. If it returns nothing (not even C<undef> but the empty list), |
|
|
541 | the callback from C<filter_json_object> will be called next, as if no |
|
|
542 | single-key callback were specified. |
|
|
543 | |
|
|
544 | If C<$coderef> is omitted or undefined, the corresponding callback will be |
|
|
545 | disabled. There can only ever be one callback for a given key. |
|
|
546 | |
|
|
547 | As this callback gets called less often then the C<filter_json_object> |
|
|
548 | one, decoding speed will not usually suffer as much. Therefore, single-key |
|
|
549 | objects make excellent targets to serialise Perl objects into, especially |
|
|
550 | as single-key JSON objects are as close to the type-tagged value concept |
|
|
551 | as JSON gets (it's basically an ID/VALUE tuple). Of course, JSON does not |
|
|
552 | support this in any way, so you need to make sure your data never looks |
|
|
553 | like a serialised Perl hash. |
|
|
554 | |
|
|
555 | Typical names for the single object key are C<__class_whatever__>, or |
|
|
556 | C<$__dollars_are_rarely_used__$> or C<}ugly_brace_placement>, or even |
|
|
557 | things like C<__class_md5sum(classname)__>, to reduce the risk of clashing |
|
|
558 | with real hashes. |
|
|
559 | |
|
|
560 | Example, decode JSON objects of the form C<< { "__widget__" => <id> } >> |
|
|
561 | into the corresponding C<< $WIDGET{<id>} >> object: |
|
|
562 | |
|
|
563 | # return whatever is in $WIDGET{5}: |
|
|
564 | JSON::XS |
|
|
565 | ->new |
|
|
566 | ->filter_json_single_key_object (__widget__ => sub { |
|
|
567 | $WIDGET{ $_[0] } |
|
|
568 | }) |
|
|
569 | ->decode ('{"__widget__": 5') |
|
|
570 | |
|
|
571 | # this can be used with a TO_JSON method in some "widget" class |
|
|
572 | # for serialisation to json: |
|
|
573 | sub WidgetBase::TO_JSON { |
|
|
574 | my ($self) = @_; |
|
|
575 | |
|
|
576 | unless ($self->{id}) { |
|
|
577 | $self->{id} = ..get..some..id..; |
|
|
578 | $WIDGET{$self->{id}} = $self; |
|
|
579 | } |
|
|
580 | |
|
|
581 | { __widget__ => $self->{id} } |
|
|
582 | } |
|
|
583 | |
|
|
584 | =item $json = $json->shrink ([$enable]) |
|
|
585 | |
|
|
586 | =item $enabled = $json->get_shrink |
|
|
587 | |
|
|
588 | Perl usually over-allocates memory a bit when allocating space for |
|
|
589 | strings. This flag optionally resizes strings generated by either |
|
|
590 | C<encode> or C<decode> to their minimum size possible. This can save |
|
|
591 | memory when your JSON texts are either very very long or you have many |
|
|
592 | short strings. It will also try to downgrade any strings to octet-form |
|
|
593 | if possible: perl stores strings internally either in an encoding called |
|
|
594 | UTF-X or in octet-form. The latter cannot store everything but uses less |
|
|
595 | space in general (and some buggy Perl or C code might even rely on that |
|
|
596 | internal representation being used). |
|
|
597 | |
|
|
598 | The actual definition of what shrink does might change in future versions, |
|
|
599 | but it will always try to save space at the expense of time. |
|
|
600 | |
|
|
601 | If C<$enable> is true (or missing), the string returned by C<encode> will |
|
|
602 | be shrunk-to-fit, while all strings generated by C<decode> will also be |
|
|
603 | shrunk-to-fit. |
|
|
604 | |
|
|
605 | If C<$enable> is false, then the normal perl allocation algorithms are used. |
|
|
606 | If you work with your data, then this is likely to be faster. |
|
|
607 | |
|
|
608 | In the future, this setting might control other things, such as converting |
|
|
609 | strings that look like integers or floats into integers or floats |
|
|
610 | internally (there is no difference on the Perl level), saving space. |
|
|
611 | |
|
|
612 | =item $json = $json->max_depth ([$maximum_nesting_depth]) |
|
|
613 | |
|
|
614 | =item $max_depth = $json->get_max_depth |
|
|
615 | |
|
|
616 | Sets the maximum nesting level (default C<512>) accepted while encoding |
|
|
617 | or decoding. If the JSON text or Perl data structure has an equal or |
|
|
618 | higher nesting level then this limit, then the encoder and decoder will |
|
|
619 | stop and croak at that point. |
|
|
620 | |
|
|
621 | Nesting level is defined by number of hash- or arrayrefs that the encoder |
|
|
622 | needs to traverse to reach a given point or the number of C<{> or C<[> |
|
|
623 | characters without their matching closing parenthesis crossed to reach a |
|
|
624 | given character in a string. |
|
|
625 | |
|
|
626 | Setting the maximum depth to one disallows any nesting, so that ensures |
|
|
627 | that the object is only a single hash/object or array. |
|
|
628 | |
|
|
629 | The argument to C<max_depth> will be rounded up to the next highest power |
|
|
630 | of two. If no argument is given, the highest possible setting will be |
|
|
631 | used, which is rarely useful. |
|
|
632 | |
|
|
633 | See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
|
|
634 | |
|
|
635 | =item $json = $json->max_size ([$maximum_string_size]) |
|
|
636 | |
|
|
637 | =item $max_size = $json->get_max_size |
|
|
638 | |
|
|
639 | Set the maximum length a JSON text may have (in bytes) where decoding is |
|
|
640 | being attempted. The default is C<0>, meaning no limit. When C<decode> |
|
|
641 | is called on a string longer then this number of characters it will not |
|
|
642 | attempt to decode the string but throw an exception. This setting has no |
|
|
643 | effect on C<encode> (yet). |
|
|
644 | |
|
|
645 | The argument to C<max_size> will be rounded up to the next B<highest> |
|
|
646 | power of two (so may be more than requested). If no argument is given, the |
|
|
647 | limit check will be deactivated (same as when C<0> is specified). |
|
|
648 | |
|
|
649 | See SECURITY CONSIDERATIONS, below, for more info on why this is useful. |
|
|
650 | |
|
|
651 | =item $json_text = $json->encode ($perl_scalar) |
|
|
652 | |
|
|
653 | Converts the given Perl data structure (a simple scalar or a reference |
|
|
654 | to a hash or array) to its JSON representation. Simple scalars will be |
|
|
655 | converted into JSON string or number sequences, while references to arrays |
|
|
656 | become JSON arrays and references to hashes become JSON objects. Undefined |
|
|
657 | Perl values (e.g. C<undef>) become JSON C<null> values. Neither C<true> |
|
|
658 | nor C<false> values will be generated. |
|
|
659 | |
|
|
660 | =item $perl_scalar = $json->decode ($json_text) |
|
|
661 | |
|
|
662 | The opposite of C<encode>: expects a JSON text and tries to parse it, |
|
|
663 | returning the resulting simple scalar or reference. Croaks on error. |
|
|
664 | |
|
|
665 | JSON numbers and strings become simple Perl scalars. JSON arrays become |
|
|
666 | Perl arrayrefs and JSON objects become Perl hashrefs. C<true> becomes |
|
|
667 | C<1>, C<false> becomes C<0> and C<null> becomes C<undef>. |
|
|
668 | |
|
|
669 | =item ($perl_scalar, $characters) = $json->decode_prefix ($json_text) |
|
|
670 | |
|
|
671 | This works like the C<decode> method, but instead of raising an exception |
|
|
672 | when there is trailing garbage after the first JSON object, it will |
|
|
673 | silently stop parsing there and return the number of characters consumed |
|
|
674 | so far. |
|
|
675 | |
|
|
676 | This is useful if your JSON texts are not delimited by an outer protocol |
|
|
677 | (which is not the brightest thing to do in the first place) and you need |
|
|
678 | to know where the JSON text ends. |
|
|
679 | |
|
|
680 | JSON::XS->new->decode_prefix ("[1] the tail") |
|
|
681 | => ([], 3) |
|
|
682 | |
|
|
683 | =back |
|
|
684 | |
|
|
685 | |
|
|
686 | =head1 MAPPING |
|
|
687 | |
|
|
688 | This section describes how JSON::XS maps Perl values to JSON values and |
|
|
689 | vice versa. These mappings are designed to "do the right thing" in most |
|
|
690 | circumstances automatically, preserving round-tripping characteristics |
|
|
691 | (what you put in comes out as something equivalent). |
|
|
692 | |
|
|
693 | For the more enlightened: note that in the following descriptions, |
|
|
694 | lowercase I<perl> refers to the Perl interpreter, while uppercase I<Perl> |
|
|
695 | refers to the abstract Perl language itself. |
|
|
696 | |
|
|
697 | |
|
|
698 | =head2 JSON -> PERL |
|
|
699 | |
|
|
700 | =over 4 |
|
|
701 | |
|
|
702 | =item object |
|
|
703 | |
|
|
704 | A JSON object becomes a reference to a hash in Perl. No ordering of object |
|
|
705 | keys is preserved (JSON does not preserve object key ordering itself). |
|
|
706 | |
|
|
707 | =item array |
|
|
708 | |
|
|
709 | A JSON array becomes a reference to an array in Perl. |
|
|
710 | |
|
|
711 | =item string |
|
|
712 | |
|
|
713 | A JSON string becomes a string scalar in Perl - Unicode codepoints in JSON |
|
|
714 | are represented by the same codepoints in the Perl string, so no manual |
|
|
715 | decoding is necessary. |
|
|
716 | |
|
|
717 | =item number |
|
|
718 | |
|
|
719 | A JSON number becomes either an integer, numeric (floating point) or |
|
|
720 | string scalar in perl, depending on its range and any fractional parts. On |
|
|
721 | the Perl level, there is no difference between those as Perl handles all |
|
|
722 | the conversion details, but an integer may take slightly less memory and |
|
|
723 | might represent more values exactly than floating point numbers. |
|
|
724 | |
|
|
725 | If the number consists of digits only, JSON::XS will try to represent |
|
|
726 | it as an integer value. If that fails, it will try to represent it as |
|
|
727 | a numeric (floating point) value if that is possible without loss of |
|
|
728 | precision. Otherwise it will preserve the number as a string value (in |
|
|
729 | which case you lose roundtripping ability, as the JSON number will be |
|
|
730 | re-encoded toa JSON string). |
|
|
731 | |
|
|
732 | Numbers containing a fractional or exponential part will always be |
|
|
733 | represented as numeric (floating point) values, possibly at a loss of |
|
|
734 | precision (in which case you might lose perfect roundtripping ability, but |
|
|
735 | the JSON number will still be re-encoded as a JSON number). |
|
|
736 | |
|
|
737 | =item true, false |
|
|
738 | |
|
|
739 | These JSON atoms become C<JSON::XS::true> and C<JSON::XS::false>, |
|
|
740 | respectively. They are overloaded to act almost exactly like the numbers |
|
|
741 | C<1> and C<0>. You can check whether a scalar is a JSON boolean by using |
|
|
742 | the C<JSON::XS::is_bool> function. |
|
|
743 | |
|
|
744 | =item null |
|
|
745 | |
|
|
746 | A JSON null atom becomes C<undef> in Perl. |
|
|
747 | |
|
|
748 | =back |
|
|
749 | |
|
|
750 | |
|
|
751 | =head2 PERL -> JSON |
|
|
752 | |
|
|
753 | The mapping from Perl to JSON is slightly more difficult, as Perl is a |
|
|
754 | truly typeless language, so we can only guess which JSON type is meant by |
|
|
755 | a Perl value. |
|
|
756 | |
|
|
757 | =over 4 |
|
|
758 | |
|
|
759 | =item hash references |
|
|
760 | |
|
|
761 | Perl hash references become JSON objects. As there is no inherent ordering |
|
|
762 | in hash keys (or JSON objects), they will usually be encoded in a |
|
|
763 | pseudo-random order that can change between runs of the same program but |
|
|
764 | stays generally the same within a single run of a program. JSON::XS can |
|
|
765 | optionally sort the hash keys (determined by the I<canonical> flag), so |
|
|
766 | the same datastructure will serialise to the same JSON text (given same |
|
|
767 | settings and version of JSON::XS), but this incurs a runtime overhead |
|
|
768 | and is only rarely useful, e.g. when you want to compare some JSON text |
|
|
769 | against another for equality. |
|
|
770 | |
|
|
771 | =item array references |
|
|
772 | |
|
|
773 | Perl array references become JSON arrays. |
|
|
774 | |
|
|
775 | =item other references |
|
|
776 | |
|
|
777 | Other unblessed references are generally not allowed and will cause an |
|
|
778 | exception to be thrown, except for references to the integers C<0> and |
|
|
779 | C<1>, which get turned into C<false> and C<true> atoms in JSON. You can |
|
|
780 | also use C<JSON::XS::false> and C<JSON::XS::true> to improve readability. |
|
|
781 | |
|
|
782 | encode_json [\0,JSON::XS::true] # yields [false,true] |
|
|
783 | |
|
|
784 | =item JSON::XS::true, JSON::XS::false |
|
|
785 | |
|
|
786 | These special values become JSON true and JSON false values, |
|
|
787 | respectively. You can also use C<\1> and C<\0> directly if you want. |
|
|
788 | |
|
|
789 | =item blessed objects |
|
|
790 | |
|
|
791 | Blessed objects are not directly representable in JSON. See the |
|
|
792 | C<allow_blessed> and C<convert_blessed> methods on various options on |
|
|
793 | how to deal with this: basically, you can choose between throwing an |
|
|
794 | exception, encoding the reference as if it weren't blessed, or provide |
|
|
795 | your own serialiser method. |
|
|
796 | |
|
|
797 | =item simple scalars |
|
|
798 | |
|
|
799 | Simple Perl scalars (any scalar that is not a reference) are the most |
|
|
800 | difficult objects to encode: JSON::XS will encode undefined scalars as |
|
|
801 | JSON C<null> values, scalars that have last been used in a string context |
|
|
802 | before encoding as JSON strings, and anything else as number value: |
|
|
803 | |
|
|
804 | # dump as number |
|
|
805 | encode_json [2] # yields [2] |
|
|
806 | encode_json [-3.0e17] # yields [-3e+17] |
|
|
807 | my $value = 5; encode_json [$value] # yields [5] |
|
|
808 | |
|
|
809 | # used as string, so dump as string |
|
|
810 | print $value; |
|
|
811 | encode_json [$value] # yields ["5"] |
|
|
812 | |
|
|
813 | # undef becomes null |
|
|
814 | encode_json [undef] # yields [null] |
|
|
815 | |
|
|
816 | You can force the type to be a JSON string by stringifying it: |
|
|
817 | |
|
|
818 | my $x = 3.1; # some variable containing a number |
|
|
819 | "$x"; # stringified |
|
|
820 | $x .= ""; # another, more awkward way to stringify |
|
|
821 | print $x; # perl does it for you, too, quite often |
|
|
822 | |
|
|
823 | You can force the type to be a JSON number by numifying it: |
|
|
824 | |
|
|
825 | my $x = "3"; # some variable containing a string |
|
|
826 | $x += 0; # numify it, ensuring it will be dumped as a number |
|
|
827 | $x *= 1; # same thing, the choice is yours. |
|
|
828 | |
|
|
829 | You can not currently force the type in other, less obscure, ways. Tell me |
|
|
830 | if you need this capability (but don't forget to explain why it's needed |
|
|
831 | :). |
|
|
832 | |
|
|
833 | =back |
|
|
834 | |
|
|
835 | |
|
|
836 | =head1 ENCODING/CODESET FLAG NOTES |
|
|
837 | |
|
|
838 | The interested reader might have seen a number of flags that signify |
|
|
839 | encodings or codesets - C<utf8>, C<latin1> and C<ascii>. There seems to be |
|
|
840 | some confusion on what these do, so here is a short comparison: |
|
|
841 | |
|
|
842 | C<utf8> controls whether the JSON text created by C<encode> (and expected |
|
|
843 | by C<decode>) is UTF-8 encoded or not, while C<latin1> and C<ascii> only |
|
|
844 | control whether C<encode> escapes character values outside their respective |
|
|
845 | codeset range. Neither of these flags conflict with each other, although |
|
|
846 | some combinations make less sense than others. |
|
|
847 | |
|
|
848 | Care has been taken to make all flags symmetrical with respect to |
|
|
849 | C<encode> and C<decode>, that is, texts encoded with any combination of |
|
|
850 | these flag values will be correctly decoded when the same flags are used |
|
|
851 | - in general, if you use different flag settings while encoding vs. when |
|
|
852 | decoding you likely have a bug somewhere. |
|
|
853 | |
|
|
854 | Below comes a verbose discussion of these flags. Note that a "codeset" is |
|
|
855 | simply an abstract set of character-codepoint pairs, while an encoding |
|
|
856 | takes those codepoint numbers and I<encodes> them, in our case into |
|
|
857 | octets. Unicode is (among other things) a codeset, UTF-8 is an encoding, |
|
|
858 | and ISO-8859-1 (= latin 1) and ASCII are both codesets I<and> encodings at |
|
|
859 | the same time, which can be confusing. |
|
|
860 | |
|
|
861 | =over 4 |
|
|
862 | |
|
|
863 | =item C<utf8> flag disabled |
|
|
864 | |
|
|
865 | When C<utf8> is disabled (the default), then C<encode>/C<decode> generate |
|
|
866 | and expect Unicode strings, that is, characters with high ordinal Unicode |
|
|
867 | values (> 255) will be encoded as such characters, and likewise such |
|
|
868 | characters are decoded as-is, no canges to them will be done, except |
|
|
869 | "(re-)interpreting" them as Unicode codepoints or Unicode characters, |
|
|
870 | respectively (to Perl, these are the same thing in strings unless you do |
|
|
871 | funny/weird/dumb stuff). |
|
|
872 | |
|
|
873 | This is useful when you want to do the encoding yourself (e.g. when you |
|
|
874 | want to have UTF-16 encoded JSON texts) or when some other layer does |
|
|
875 | the encoding for you (for example, when printing to a terminal using a |
|
|
876 | filehandle that transparently encodes to UTF-8 you certainly do NOT want |
|
|
877 | to UTF-8 encode your data first and have Perl encode it another time). |
|
|
878 | |
|
|
879 | =item C<utf8> flag enabled |
|
|
880 | |
|
|
881 | If the C<utf8>-flag is enabled, C<encode>/C<decode> will encode all |
|
|
882 | characters using the corresponding UTF-8 multi-byte sequence, and will |
|
|
883 | expect your input strings to be encoded as UTF-8, that is, no "character" |
|
|
884 | of the input string must have any value > 255, as UTF-8 does not allow |
|
|
885 | that. |
|
|
886 | |
|
|
887 | The C<utf8> flag therefore switches between two modes: disabled means you |
|
|
888 | will get a Unicode string in Perl, enabled means you get an UTF-8 encoded |
|
|
889 | octet/binary string in Perl. |
|
|
890 | |
|
|
891 | =item C<latin1> or C<ascii> flags enabled |
|
|
892 | |
|
|
893 | With C<latin1> (or C<ascii>) enabled, C<encode> will escape characters |
|
|
894 | with ordinal values > 255 (> 127 with C<ascii>) and encode the remaining |
|
|
895 | characters as specified by the C<utf8> flag. |
|
|
896 | |
|
|
897 | If C<utf8> is disabled, then the result is also correctly encoded in those |
|
|
898 | character sets (as both are proper subsets of Unicode, meaning that a |
|
|
899 | Unicode string with all character values < 256 is the same thing as a |
|
|
900 | ISO-8859-1 string, and a Unicode string with all character values < 128 is |
|
|
901 | the same thing as an ASCII string in Perl). |
|
|
902 | |
|
|
903 | If C<utf8> is enabled, you still get a correct UTF-8-encoded string, |
|
|
904 | regardless of these flags, just some more characters will be escaped using |
|
|
905 | C<\uXXXX> then before. |
|
|
906 | |
|
|
907 | Note that ISO-8859-1-I<encoded> strings are not compatible with UTF-8 |
|
|
908 | encoding, while ASCII-encoded strings are. That is because the ISO-8859-1 |
|
|
909 | encoding is NOT a subset of UTF-8 (despite the ISO-8859-1 I<codeset> being |
|
|
910 | a subset of Unicode), while ASCII is. |
|
|
911 | |
|
|
912 | Surprisingly, C<decode> will ignore these flags and so treat all input |
|
|
913 | values as governed by the C<utf8> flag. If it is disabled, this allows you |
|
|
914 | to decode ISO-8859-1- and ASCII-encoded strings, as both strict subsets of |
|
|
915 | Unicode. If it is enabled, you can correctly decode UTF-8 encoded strings. |
|
|
916 | |
|
|
917 | So neither C<latin1> nor C<ascii> are incompatible with the C<utf8> flag - |
|
|
918 | they only govern when the JSON output engine escapes a character or not. |
|
|
919 | |
|
|
920 | The main use for C<latin1> is to relatively efficiently store binary data |
|
|
921 | as JSON, at the expense of breaking compatibility with most JSON decoders. |
|
|
922 | |
|
|
923 | The main use for C<ascii> is to force the output to not contain characters |
|
|
924 | with values > 127, which means you can interpret the resulting string |
|
|
925 | as UTF-8, ISO-8859-1, ASCII, KOI8-R or most about any character set and |
|
|
926 | 8-bit-encoding, and still get the same data structure back. This is useful |
|
|
927 | when your channel for JSON transfer is not 8-bit clean or the encoding |
|
|
928 | might be mangled in between (e.g. in mail), and works because ASCII is a |
|
|
929 | proper subset of most 8-bit and multibyte encodings in use in the world. |
|
|
930 | |
|
|
931 | =back |
|
|
932 | |
|
|
933 | |
|
|
934 | =head1 COMPARISON |
|
|
935 | |
|
|
936 | As already mentioned, this module was created because none of the existing |
|
|
937 | JSON modules could be made to work correctly. First I will describe the |
|
|
938 | problems (or pleasures) I encountered with various existing JSON modules, |
|
|
939 | followed by some benchmark values. JSON::XS was designed not to suffer |
|
|
940 | from any of these problems or limitations. |
|
|
941 | |
|
|
942 | =over 4 |
|
|
943 | |
|
|
944 | =item JSON 2.xx |
|
|
945 | |
|
|
946 | A marvellous piece of engineering, this module either uses JSON::XS |
|
|
947 | directly when available (so will be 100% compatible with it, including |
|
|
948 | speed), or it uses JSON::PP, which is basically JSON::XS translated to |
|
|
949 | Pure Perl, which should be 100% compatible with JSON::XS, just a bit |
|
|
950 | slower. |
|
|
951 | |
|
|
952 | You cannot really lose by using this module, especially as it tries very |
|
|
953 | hard to work even with ancient Perl versions, while JSON::XS does not. |
|
|
954 | |
|
|
955 | =item JSON 1.07 |
|
|
956 | |
|
|
957 | Slow (but very portable, as it is written in pure Perl). |
|
|
958 | |
|
|
959 | Undocumented/buggy Unicode handling (how JSON handles Unicode values is |
|
|
960 | undocumented. One can get far by feeding it Unicode strings and doing |
|
|
961 | en-/decoding oneself, but Unicode escapes are not working properly). |
|
|
962 | |
|
|
963 | No round-tripping (strings get clobbered if they look like numbers, e.g. |
|
|
964 | the string C<2.0> will encode to C<2.0> instead of C<"2.0">, and that will |
|
|
965 | decode into the number 2. |
|
|
966 | |
|
|
967 | =item JSON::PC 0.01 |
|
|
968 | |
|
|
969 | Very fast. |
|
|
970 | |
|
|
971 | Undocumented/buggy Unicode handling. |
|
|
972 | |
|
|
973 | No round-tripping. |
|
|
974 | |
|
|
975 | Has problems handling many Perl values (e.g. regex results and other magic |
|
|
976 | values will make it croak). |
|
|
977 | |
|
|
978 | Does not even generate valid JSON (C<{1,2}> gets converted to C<{1:2}> |
|
|
979 | which is not a valid JSON text. |
|
|
980 | |
|
|
981 | Unmaintained (maintainer unresponsive for many months, bugs are not |
|
|
982 | getting fixed). |
|
|
983 | |
|
|
984 | =item JSON::Syck 0.21 |
|
|
985 | |
|
|
986 | Very buggy (often crashes). |
|
|
987 | |
|
|
988 | Very inflexible (no human-readable format supported, format pretty much |
|
|
989 | undocumented. I need at least a format for easy reading by humans and a |
|
|
990 | single-line compact format for use in a protocol, and preferably a way to |
|
|
991 | generate ASCII-only JSON texts). |
|
|
992 | |
|
|
993 | Completely broken (and confusingly documented) Unicode handling (Unicode |
|
|
994 | escapes are not working properly, you need to set ImplicitUnicode to |
|
|
995 | I<different> values on en- and decoding to get symmetric behaviour). |
|
|
996 | |
|
|
997 | No round-tripping (simple cases work, but this depends on whether the scalar |
|
|
998 | value was used in a numeric context or not). |
|
|
999 | |
|
|
1000 | Dumping hashes may skip hash values depending on iterator state. |
|
|
1001 | |
|
|
1002 | Unmaintained (maintainer unresponsive for many months, bugs are not |
|
|
1003 | getting fixed). |
|
|
1004 | |
|
|
1005 | Does not check input for validity (i.e. will accept non-JSON input and |
|
|
1006 | return "something" instead of raising an exception. This is a security |
|
|
1007 | issue: imagine two banks transferring money between each other using |
|
|
1008 | JSON. One bank might parse a given non-JSON request and deduct money, |
|
|
1009 | while the other might reject the transaction with a syntax error. While a |
|
|
1010 | good protocol will at least recover, that is extra unnecessary work and |
|
|
1011 | the transaction will still not succeed). |
|
|
1012 | |
|
|
1013 | =item JSON::DWIW 0.04 |
|
|
1014 | |
|
|
1015 | Very fast. Very natural. Very nice. |
|
|
1016 | |
|
|
1017 | Undocumented Unicode handling (but the best of the pack. Unicode escapes |
|
|
1018 | still don't get parsed properly). |
|
|
1019 | |
|
|
1020 | Very inflexible. |
|
|
1021 | |
|
|
1022 | No round-tripping. |
|
|
1023 | |
|
|
1024 | Does not generate valid JSON texts (key strings are often unquoted, empty keys |
|
|
1025 | result in nothing being output) |
|
|
1026 | |
|
|
1027 | Does not check input for validity. |
|
|
1028 | |
|
|
1029 | =back |
|
|
1030 | |
|
|
1031 | |
|
|
1032 | =head2 JSON and YAML |
|
|
1033 | |
|
|
1034 | You often hear that JSON is a subset of YAML. This is, however, a mass |
|
|
1035 | hysteria(*) and very far from the truth (as of the time of this writing), |
|
|
1036 | so let me state it clearly: I<in general, there is no way to configure |
|
|
1037 | JSON::XS to output a data structure as valid YAML> that works in all |
|
|
1038 | cases. |
|
|
1039 | |
|
|
1040 | If you really must use JSON::XS to generate YAML, you should use this |
|
|
1041 | algorithm (subject to change in future versions): |
|
|
1042 | |
|
|
1043 | my $to_yaml = JSON::XS->new->utf8->space_after (1); |
|
|
1044 | my $yaml = $to_yaml->encode ($ref) . "\n"; |
|
|
1045 | |
|
|
1046 | This will I<usually> generate JSON texts that also parse as valid |
|
|
1047 | YAML. Please note that YAML has hardcoded limits on (simple) object key |
|
|
1048 | lengths that JSON doesn't have and also has different and incompatible |
|
|
1049 | unicode handling, so you should make sure that your hash keys are |
|
|
1050 | noticeably shorter than the 1024 "stream characters" YAML allows and that |
|
|
1051 | you do not have characters with codepoint values outside the Unicode BMP |
|
|
1052 | (basic multilingual page). YAML also does not allow C<\/> sequences in |
|
|
1053 | strings (which JSON::XS does not I<currently> generate, but other JSON |
|
|
1054 | generators might). |
|
|
1055 | |
|
|
1056 | There might be other incompatibilities that I am not aware of (or the YAML |
|
|
1057 | specification has been changed yet again - it does so quite often). In |
|
|
1058 | general you should not try to generate YAML with a JSON generator or vice |
|
|
1059 | versa, or try to parse JSON with a YAML parser or vice versa: chances are |
|
|
1060 | high that you will run into severe interoperability problems when you |
|
|
1061 | least expect it. |
|
|
1062 | |
|
|
1063 | =over 4 |
|
|
1064 | |
|
|
1065 | =item (*) |
|
|
1066 | |
|
|
1067 | I have been pressured multiple times by Brian Ingerson (one of the |
|
|
1068 | authors of the YAML specification) to remove this paragraph, despite him |
|
|
1069 | acknowledging that the actual incompatibilities exist. As I was personally |
|
|
1070 | bitten by this "JSON is YAML" lie, I refused and said I will continue to |
|
|
1071 | educate people about these issues, so others do not run into the same |
|
|
1072 | problem again and again. After this, Brian called me a (quote)I<complete |
|
|
1073 | and worthless idiot>(unquote). |
|
|
1074 | |
|
|
1075 | In my opinion, instead of pressuring and insulting people who actually |
|
|
1076 | clarify issues with YAML and the wrong statements of some of its |
|
|
1077 | proponents, I would kindly suggest reading the JSON spec (which is not |
|
|
1078 | that difficult or long) and finally make YAML compatible to it, and |
|
|
1079 | educating users about the changes, instead of spreading lies about the |
|
|
1080 | real compatibility for many I<years> and trying to silence people who |
|
|
1081 | point out that it isn't true. |
|
|
1082 | |
|
|
1083 | =back |
|
|
1084 | |
|
|
1085 | |
|
|
1086 | =head2 SPEED |
|
|
1087 | |
|
|
1088 | It seems that JSON::XS is surprisingly fast, as shown in the following |
|
|
1089 | tables. They have been generated with the help of the C<eg/bench> program |
|
|
1090 | in the JSON::XS distribution, to make it easy to compare on your own |
|
|
1091 | system. |
|
|
1092 | |
|
|
1093 | First comes a comparison between various modules using |
|
|
1094 | a very short single-line JSON string (also available at |
|
|
1095 | L<http://dist.schmorp.de/misc/json/short.json>). |
|
|
1096 | |
|
|
1097 | {"method": "handleMessage", "params": ["user1", "we were just talking"], \ |
|
|
1098 | "id": null, "array":[1,11,234,-5,1e5,1e7, true, false]} |
|
|
1099 | |
|
|
1100 | It shows the number of encodes/decodes per second (JSON::XS uses |
|
|
1101 | the functional interface, while JSON::XS/2 uses the OO interface |
|
|
1102 | with pretty-printing and hashkey sorting enabled, JSON::XS/3 enables |
|
|
1103 | shrink). Higher is better: |
|
|
1104 | |
|
|
1105 | module | encode | decode | |
|
|
1106 | -----------|------------|------------| |
|
|
1107 | JSON 1.x | 4990.842 | 4088.813 | |
|
|
1108 | JSON::DWIW | 51653.990 | 71575.154 | |
|
|
1109 | JSON::PC | 65948.176 | 74631.744 | |
|
|
1110 | JSON::PP | 8931.652 | 3817.168 | |
|
|
1111 | JSON::Syck | 24877.248 | 27776.848 | |
|
|
1112 | JSON::XS | 388361.481 | 227951.304 | |
|
|
1113 | JSON::XS/2 | 227951.304 | 218453.333 | |
|
|
1114 | JSON::XS/3 | 338250.323 | 218453.333 | |
|
|
1115 | Storable | 16500.016 | 135300.129 | |
|
|
1116 | -----------+------------+------------+ |
|
|
1117 | |
|
|
1118 | That is, JSON::XS is about five times faster than JSON::DWIW on encoding, |
|
|
1119 | about three times faster on decoding, and over forty times faster |
|
|
1120 | than JSON, even with pretty-printing and key sorting. It also compares |
|
|
1121 | favourably to Storable for small amounts of data. |
|
|
1122 | |
|
|
1123 | Using a longer test string (roughly 18KB, generated from Yahoo! Locals |
|
|
1124 | search API (L<http://dist.schmorp.de/misc/json/long.json>). |
|
|
1125 | |
|
|
1126 | module | encode | decode | |
|
|
1127 | -----------|------------|------------| |
|
|
1128 | JSON 1.x | 55.260 | 34.971 | |
|
|
1129 | JSON::DWIW | 825.228 | 1082.513 | |
|
|
1130 | JSON::PC | 3571.444 | 2394.829 | |
|
|
1131 | JSON::PP | 210.987 | 32.574 | |
|
|
1132 | JSON::Syck | 552.551 | 787.544 | |
|
|
1133 | JSON::XS | 5780.463 | 4854.519 | |
|
|
1134 | JSON::XS/2 | 3869.998 | 4798.975 | |
|
|
1135 | JSON::XS/3 | 5862.880 | 4798.975 | |
|
|
1136 | Storable | 4445.002 | 5235.027 | |
|
|
1137 | -----------+------------+------------+ |
|
|
1138 | |
|
|
1139 | Again, JSON::XS leads by far (except for Storable which non-surprisingly |
|
|
1140 | decodes faster). |
|
|
1141 | |
|
|
1142 | On large strings containing lots of high Unicode characters, some modules |
|
|
1143 | (such as JSON::PC) seem to decode faster than JSON::XS, but the result |
|
|
1144 | will be broken due to missing (or wrong) Unicode handling. Others refuse |
|
|
1145 | to decode or encode properly, so it was impossible to prepare a fair |
|
|
1146 | comparison table for that case. |
|
|
1147 | |
|
|
1148 | |
|
|
1149 | =head1 SECURITY CONSIDERATIONS |
|
|
1150 | |
|
|
1151 | When you are using JSON in a protocol, talking to untrusted potentially |
|
|
1152 | hostile creatures requires relatively few measures. |
|
|
1153 | |
|
|
1154 | First of all, your JSON decoder should be secure, that is, should not have |
|
|
1155 | any buffer overflows. Obviously, this module should ensure that and I am |
|
|
1156 | trying hard on making that true, but you never know. |
|
|
1157 | |
|
|
1158 | Second, you need to avoid resource-starving attacks. That means you should |
|
|
1159 | limit the size of JSON texts you accept, or make sure then when your |
|
|
1160 | resources run out, that's just fine (e.g. by using a separate process that |
|
|
1161 | can crash safely). The size of a JSON text in octets or characters is |
|
|
1162 | usually a good indication of the size of the resources required to decode |
|
|
1163 | it into a Perl structure. While JSON::XS can check the size of the JSON |
|
|
1164 | text, it might be too late when you already have it in memory, so you |
|
|
1165 | might want to check the size before you accept the string. |
|
|
1166 | |
|
|
1167 | Third, JSON::XS recurses using the C stack when decoding objects and |
|
|
1168 | arrays. The C stack is a limited resource: for instance, on my amd64 |
|
|
1169 | machine with 8MB of stack size I can decode around 180k nested arrays but |
|
|
1170 | only 14k nested JSON objects (due to perl itself recursing deeply on croak |
|
|
1171 | to free the temporary). If that is exceeded, the program crashes. To be |
|
|
1172 | conservative, the default nesting limit is set to 512. If your process |
|
|
1173 | has a smaller stack, you should adjust this setting accordingly with the |
|
|
1174 | C<max_depth> method. |
|
|
1175 | |
|
|
1176 | Something else could bomb you, too, that I forgot to think of. In that |
|
|
1177 | case, you get to keep the pieces. I am always open for hints, though... |
|
|
1178 | |
|
|
1179 | Also keep in mind that JSON::XS might leak contents of your Perl data |
|
|
1180 | structures in its error messages, so when you serialise sensitive |
|
|
1181 | information you might want to make sure that exceptions thrown by JSON::XS |
|
|
1182 | will not end up in front of untrusted eyes. |
|
|
1183 | |
|
|
1184 | If you are using JSON::XS to return packets to consumption |
|
|
1185 | by JavaScript scripts in a browser you should have a look at |
|
|
1186 | L<http://jpsykes.com/47/practical-csrf-and-json-security> to see whether |
|
|
1187 | you are vulnerable to some common attack vectors (which really are browser |
|
|
1188 | design bugs, but it is still you who will have to deal with it, as major |
|
|
1189 | browser developers care only for features, not about getting security |
|
|
1190 | right). |
|
|
1191 | |
|
|
1192 | |
|
|
1193 | =head1 THREADS |
|
|
1194 | |
|
|
1195 | This module is I<not> guaranteed to be thread safe and there are no |
|
|
1196 | plans to change this until Perl gets thread support (as opposed to the |
|
|
1197 | horribly slow so-called "threads" which are simply slow and bloated |
|
|
1198 | process simulations - use fork, it's I<much> faster, cheaper, better). |
|
|
1199 | |
|
|
1200 | (It might actually work, but you have been warned). |
|
|
1201 | |
|
|
1202 | |
|
|
1203 | =head1 BUGS |
|
|
1204 | |
|
|
1205 | While the goal of this module is to be correct, that unfortunately does |
|
|
1206 | not mean it's bug-free, only that I think its design is bug-free. It is |
|
|
1207 | still relatively early in its development. If you keep reporting bugs they |
|
|
1208 | will be fixed swiftly, though. |
|
|
1209 | |
|
|
1210 | Please refrain from using rt.cpan.org or any other bug reporting |
|
|
1211 | service. I put the contact address into my modules for a reason. |
28 | |
1212 | |
29 | =cut |
1213 | =cut |
30 | |
1214 | |
31 | use JSON::DWIW; |
1215 | our $true = do { bless \(my $dummy = 1), "JSON::XS::Boolean" }; |
32 | use Benchmark; |
1216 | our $false = do { bless \(my $dummy = 0), "JSON::XS::Boolean" }; |
33 | |
1217 | |
34 | use utf8; |
1218 | sub true() { $true } |
35 | #my $json = '{"ü":1,"a":[1,{"3":4},2],"b":5,"üü":2}'; |
1219 | sub false() { $false } |
36 | my $json = '{"test":9555555555555555555,"hu" : -1e+5, "arr" : [ 1,2,3,4,5]}'; |
|
|
37 | |
1220 | |
38 | my $js = JSON::XS->new; |
1221 | sub is_bool($) { |
39 | warn $js->indent (0); |
1222 | UNIVERSAL::isa $_[0], "JSON::XS::Boolean" |
40 | warn $js->canonical (0); |
1223 | # or UNIVERSAL::isa $_[0], "JSON::Literal" |
41 | warn $js->ascii (0); |
1224 | } |
42 | warn $js->space_after (0); |
|
|
43 | use Data::Dumper; |
|
|
44 | warn Dumper $js->decode ($json); |
|
|
45 | warn Dumper $js->encode ($js->decode ($json)); |
|
|
46 | #my $x = {"üü" => 2, "ü" => 1, "a" => [1,{3,4},2], b => 5}; |
|
|
47 | |
1225 | |
48 | #my $js2 = JSON::DWIW->new; |
1226 | XSLoader::load "JSON::XS", $VERSION; |
49 | # |
1227 | |
50 | #timethese 200000, { |
1228 | package JSON::XS::Boolean; |
51 | # a => sub { $js->encode ($x) }, |
1229 | |
52 | # b => sub { $js2->to_json ($x) }, |
1230 | use overload |
53 | #}; |
1231 | "0+" => sub { ${$_[0]} }, |
|
|
1232 | "++" => sub { $_[0] = ${$_[0]} + 1 }, |
|
|
1233 | "--" => sub { $_[0] = ${$_[0]} - 1 }, |
|
|
1234 | fallback => 1; |
54 | |
1235 | |
55 | 1; |
1236 | 1; |
56 | |
1237 | |
57 | =back |
1238 | =head1 SEE ALSO |
|
|
1239 | |
|
|
1240 | The F<json_xs> command line utility for quick experiments. |
58 | |
1241 | |
59 | =head1 AUTHOR |
1242 | =head1 AUTHOR |
60 | |
1243 | |
61 | Marc Lehmann <schmorp@schmorp.de> |
1244 | Marc Lehmann <schmorp@schmorp.de> |
62 | http://home.schmorp.de/ |
1245 | http://home.schmorp.de/ |