| 1 | NAME |
1 | NAME |
| 2 | JSON::XS - JSON serialising/deserialising, done correctly and fast |
2 | EV - perl interface to libevent, monkey.org/~provos/libevent/ |
| 3 | |
|
|
| 4 | JSON::XS - 正しくて高速な JSON |
|
|
| 5 | シリアライザ/デシリアライザ |
|
|
| 6 | (http://fleur.hio.jp/perldoc/mix/lib/JSON/XS.html) |
|
|
| 7 | |
3 | |
| 8 | SYNOPSIS |
4 | SYNOPSIS |
| 9 | use JSON::XS; |
5 | use EV; |
| 10 | |
6 | |
| 11 | # exported functions, they croak on error |
7 | # TIMER |
| 12 | # and expect/generate UTF-8 |
|
|
| 13 | |
8 | |
| 14 | $utf8_encoded_json_text = to_json $perl_hash_or_arrayref; |
9 | my $w = EV::timer 2, 0, sub { |
| 15 | $perl_hash_or_arrayref = from_json $utf8_encoded_json_text; |
10 | warn "is called after 2s"; |
|
|
11 | }; |
| 16 | |
12 | |
| 17 | # OO-interface |
13 | my $w = EV::timer 2, 1, sub { |
|
|
14 | warn "is called roughly every 2s (repeat = 1)"; |
|
|
15 | }; |
| 18 | |
16 | |
| 19 | $coder = JSON::XS->new->ascii->pretty->allow_nonref; |
17 | undef $w; # destroy event watcher again |
| 20 | $pretty_printed_unencoded = $coder->encode ($perl_scalar); |
18 | |
| 21 | $perl_scalar = $coder->decode ($unicode_json_text); |
19 | # IO |
|
|
20 | |
|
|
21 | my $w = EV::timer_abs 0, 60, sub { |
|
|
22 | warn "is called every minute, on the minute, exactly"; |
|
|
23 | }; |
|
|
24 | |
|
|
25 | my $w = EV::io \*STDIN, EV::READ | EV::PERSIST, sub { |
|
|
26 | my ($w, $events) = @_; # all callbacks get the watcher object and event mask |
|
|
27 | if ($events & EV::TIMEOUT) { |
|
|
28 | warn "nothign received on stdin for 10 seconds, retrying"; |
|
|
29 | } else { |
|
|
30 | warn "stdin is readable, you entered: ", <STDIN>; |
|
|
31 | } |
|
|
32 | }; |
|
|
33 | $w->timeout (10); |
|
|
34 | |
|
|
35 | # MAINLOOP |
|
|
36 | EV::dispatch; # loop as long as watchers are active |
|
|
37 | EV::loop; # the same thing |
|
|
38 | EV::loop EV::LOOP_ONCE; |
|
|
39 | EV::loop EV::LOOP_ONSHOT; |
| 22 | |
40 | |
| 23 | DESCRIPTION |
41 | DESCRIPTION |
| 24 | This module converts Perl data structures to JSON and vice versa. Its |
42 | This module provides an interface to libevent |
| 25 | primary goal is to be *correct* and its secondary goal is to be *fast*. |
43 | (<http://monkey.org/~provos/libevent/>). You probably should acquaint |
| 26 | To reach the latter goal it was written in C. |
44 | yourself with its documentation and source code to be able to use this |
|
|
45 | module fully. |
| 27 | |
46 | |
| 28 | As this is the n-th-something JSON module on CPAN, what was the reason |
47 | Please note thta this module disables the libevent EPOLL method by |
| 29 | to write yet another JSON module? While it seems there are many JSON |
48 | default, see BUGS, below, if you need to enable it. |
| 30 | modules, none of them correctly handle all corner cases, and in most |
|
|
| 31 | cases their maintainers are unresponsive, gone missing, or not listening |
|
|
| 32 | to bug reports for other reasons. |
|
|
| 33 | |
|
|
| 34 | See COMPARISON, below, for a comparison to some other JSON modules. |
|
|
| 35 | |
|
|
| 36 | See MAPPING, below, on how JSON::XS maps perl values to JSON values and |
|
|
| 37 | vice versa. |
|
|
| 38 | |
|
|
| 39 | FEATURES |
|
|
| 40 | * correct unicode handling |
|
|
| 41 | This module knows how to handle Unicode, and even documents how and |
|
|
| 42 | when it does so. |
|
|
| 43 | |
|
|
| 44 | * round-trip integrity |
|
|
| 45 | When you serialise a perl data structure using only datatypes |
|
|
| 46 | supported by JSON, the deserialised data structure is identical on |
|
|
| 47 | the Perl level. (e.g. the string "2.0" doesn't suddenly become "2" |
|
|
| 48 | just because it looks like a number). |
|
|
| 49 | |
|
|
| 50 | * strict checking of JSON correctness |
|
|
| 51 | There is no guessing, no generating of illegal JSON texts by |
|
|
| 52 | default, and only JSON is accepted as input by default (the latter |
|
|
| 53 | is a security feature). |
|
|
| 54 | |
|
|
| 55 | * fast |
|
|
| 56 | Compared to other JSON modules, this module compares favourably in |
|
|
| 57 | terms of speed, too. |
|
|
| 58 | |
|
|
| 59 | * simple to use |
|
|
| 60 | This module has both a simple functional interface as well as an OO |
|
|
| 61 | interface. |
|
|
| 62 | |
|
|
| 63 | * reasonably versatile output formats |
|
|
| 64 | You can choose between the most compact guarenteed single-line |
|
|
| 65 | format possible (nice for simple line-based protocols), a pure-ascii |
|
|
| 66 | format (for when your transport is not 8-bit clean, still supports |
|
|
| 67 | the whole unicode range), or a pretty-printed format (for when you |
|
|
| 68 | want to read that stuff). Or you can combine those features in |
|
|
| 69 | whatever way you like. |
|
|
| 70 | |
49 | |
| 71 | FUNCTIONAL INTERFACE |
50 | FUNCTIONAL INTERFACE |
| 72 | The following convinience methods are provided by this module. They are |
51 | $EV::NPRI |
| 73 | exported by default: |
52 | How many priority levels are available. |
| 74 | |
53 | |
| 75 | $json_text = to_json $perl_scalar |
54 | $time = EV::now |
| 76 | Converts the given Perl data structure to a UTF-8 encoded, binary |
55 | Returns the time in (fractional) seconds since the epoch. |
| 77 | string (that is, the string contains octets only). Croaks on error. |
|
|
| 78 | |
56 | |
| 79 | This function call is functionally identical to: |
57 | $version = EV::version |
|
|
58 | $method = EV::method |
|
|
59 | Return version string and event polling method used. |
| 80 | |
60 | |
| 81 | $json_text = JSON::XS->new->utf8->encode ($perl_scalar) |
61 | EV::loop $flags # EV::LOOP_ONCE, EV::LOOP_ONESHOT |
|
|
62 | EV::loopexit $after |
|
|
63 | Exit any active loop or dispatch after $after seconds or immediately |
|
|
64 | if $after is missing or zero. |
| 82 | |
65 | |
| 83 | except being faster. |
66 | EV::dispatch |
|
|
67 | Same as "EV::loop 0". |
| 84 | |
68 | |
| 85 | $perl_scalar = from_json $json_text |
69 | EV::event $callback |
| 86 | The opposite of "to_json": expects an UTF-8 (binary) string and |
70 | Creates a new event watcher waiting for nothing, calling the given |
| 87 | tries to parse that as an UTF-8 encoded JSON text, returning the |
71 | callback. |
| 88 | resulting reference. Croaks on error. |
|
|
| 89 | |
72 | |
| 90 | This function call is functionally identical to: |
73 | my $w = EV::io $fileno_or_fh, $eventmask, $callback |
|
|
74 | my $w = EV::io_ns $fileno_or_fh, $eventmask, $callback |
|
|
75 | As long as the returned watcher object is alive, call the $callback |
|
|
76 | when the events specified in $eventmask happen. Initially, the |
|
|
77 | timeout is disabled. |
| 91 | |
78 | |
| 92 | $perl_scalar = JSON::XS->new->utf8->decode ($json_text) |
79 | Youc an additionall set a timeout to occur on the watcher, but note |
|
|
80 | that this timeout will not be reset when you get an I/O event in the |
|
|
81 | EV::PERSIST case, and reaching a timeout will always stop the |
|
|
82 | watcher even in the EV::PERSIST case. |
| 93 | |
83 | |
| 94 | except being faster. |
84 | If you want a timeout to occur only after a specific time of |
|
|
85 | inactivity, set a repeating timeout and do NOT use EV::PERSIST. |
| 95 | |
86 | |
| 96 | $is_boolean = JSON::XS::is_bool $scalar |
87 | Eventmask can be one or more of these constants ORed together: |
| 97 | Returns true if the passed scalar represents either JSON::XS::true |
|
|
| 98 | or JSON::XS::false, two constants that act like 1 and 0, |
|
|
| 99 | respectively and are used to represent JSON "true" and "false" |
|
|
| 100 | values in Perl. |
|
|
| 101 | |
88 | |
| 102 | See MAPPING, below, for more information on how JSON values are |
89 | EV::READ wait until read() wouldn't block anymore |
| 103 | mapped to Perl. |
90 | EV::WRITE wait until write() wouldn't block anymore |
|
|
91 | EV::PERSIST stay active after a (non-timeout) event occured |
| 104 | |
92 | |
| 105 | A FEW NOTES ON UNICODE AND PERL |
93 | The "io_ns" variant doesn't add/start the newly created watcher. |
| 106 | Since this often leads to confusion, here are a few very clear words on |
|
|
| 107 | how Unicode works in Perl, modulo bugs. |
|
|
| 108 | |
94 | |
| 109 | 1. Perl strings can store characters with ordinal values > 255. |
95 | my $w = EV::timer $after, $repeat, $callback |
| 110 | This enables you to store unicode characters as single characters in |
96 | my $w = EV::timer_ns $after, $repeat, $callback |
| 111 | a Perl string - very natural. |
97 | Calls the callback after $after seconds. If $repeat is true, the |
|
|
98 | timer will be restarted after the callback returns. This means that |
|
|
99 | the callback would be called roughly every $after seconds, prolonged |
|
|
100 | by the time the callback takes. |
| 112 | |
101 | |
| 113 | 2. Perl does *not* associate an encoding with your strings. |
102 | The "timer_ns" variant doesn't add/start the newly created watcher. |
| 114 | Unless you force it to, e.g. when matching it against a regex, or |
|
|
| 115 | printing the scalar to a file, in which case Perl either interprets |
|
|
| 116 | your string as locale-encoded text, octets/binary, or as Unicode, |
|
|
| 117 | depending on various settings. In no case is an encoding stored |
|
|
| 118 | together with your data, it is *use* that decides encoding, not any |
|
|
| 119 | magical metadata. |
|
|
| 120 | |
103 | |
| 121 | 3. The internal utf-8 flag has no meaning with regards to the encoding |
104 | my $w = EV::timer_abs $at, $interval, $callback |
| 122 | of your string. |
105 | my $w = EV::timer_abs_ns $at, $interval, $callback |
| 123 | Just ignore that flag unless you debug a Perl bug, a module written |
106 | Similar to EV::timer, but the time is given as an absolute point in |
| 124 | in XS or want to dive into the internals of perl. Otherwise it will |
107 | time ($at), plus an optional $interval. |
| 125 | only confuse you, as, despite the name, it says nothing about how |
|
|
| 126 | your string is encoded. You can have unicode strings with that flag |
|
|
| 127 | set, with that flag clear, and you can have binary data with that |
|
|
| 128 | flag set and that flag clear. Other possibilities exist, too. |
|
|
| 129 | |
108 | |
| 130 | If you didn't know about that flag, just the better, pretend it |
109 | If the $interval is zero, then the callback will be called at the |
| 131 | doesn't exist. |
110 | time $at if that is in the future, or as soon as possible if its in |
|
|
111 | the past. It will not automatically repeat. |
| 132 | |
112 | |
| 133 | 4. A "Unicode String" is simply a string where each character can be |
113 | If the $interval is nonzero, then the watcher will always be |
| 134 | validly interpreted as a Unicode codepoint. |
114 | scheduled to time out at the next "$at + integer * $interval" time. |
| 135 | If you have UTF-8 encoded data, it is no longer a Unicode string, |
|
|
| 136 | but a Unicode string encoded in UTF-8, giving you a binary string. |
|
|
| 137 | |
115 | |
| 138 | 5. A string containing "high" (> 255) character values is *not* a UTF-8 |
116 | This can be used to schedule a callback to run at very regular |
| 139 | string. |
117 | intervals, as long as the processing time is less then the interval |
| 140 | Its a fact. Learn to live with it. |
118 | (otherwise obviously events will be skipped). |
| 141 | |
119 | |
| 142 | I hope this helps :) |
120 | Another way to think about it (for the mathematically inclined) is |
|
|
121 | that "timer_abs" will try to tun the callback at the next possible |
|
|
122 | time where "$time = $at (mod $interval)", regardless of any time |
|
|
123 | jumps. |
| 143 | |
124 | |
| 144 | OBJECT-ORIENTED INTERFACE |
125 | The "timer_abs_ns" variant doesn't add/start the newly created |
| 145 | The object oriented interface lets you configure your own encoding or |
126 | watcher. |
| 146 | decoding style, within the limits of supported formats. |
|
|
| 147 | |
127 | |
| 148 | $json = new JSON::XS |
128 | my $w = EV::signal $signum, $callback |
| 149 | Creates a new JSON::XS object that can be used to de/encode JSON |
129 | my $w = EV::signal_ns $signum, $callback |
| 150 | strings. All boolean flags described below are by default |
130 | Call the callback when signal $signum is received. |
| 151 | *disabled*. |
|
|
| 152 | |
131 | |
| 153 | The mutators for flags all return the JSON object again and thus |
132 | The "signal_ns" variant doesn't add/start the newly created watcher. |
| 154 | calls can be chained: |
|
|
| 155 | |
133 | |
| 156 | my $json = JSON::XS->new->utf8->space_after->encode ({a => [1,2]}) |
134 | THE EV::Event CLASS |
| 157 | => {"a": [1, 2]} |
135 | All EV functions creating an event watcher (designated by "my $w =" |
|
|
136 | above) support the following methods on the returned watcher object: |
| 158 | |
137 | |
| 159 | $json = $json->ascii ([$enable]) |
138 | $w->add ($timeout) |
| 160 | If $enable is true (or missing), then the "encode" method will not |
139 | Stops and (re-)starts the event watcher, setting the optional |
| 161 | generate characters outside the code range 0..127 (which is ASCII). |
140 | timeout to the given value, or clearing the timeout if none is |
| 162 | Any unicode characters outside that range will be escaped using |
141 | given. |
| 163 | either a single \uXXXX (BMP characters) or a double \uHHHH\uLLLLL |
|
|
| 164 | escape sequence, as per RFC4627. The resulting encoded JSON text can |
|
|
| 165 | be treated as a native unicode string, an ascii-encoded, |
|
|
| 166 | latin1-encoded or UTF-8 encoded string, or any other superset of |
|
|
| 167 | ASCII. |
|
|
| 168 | |
142 | |
| 169 | If $enable is false, then the "encode" method will not escape |
143 | $w->start |
| 170 | Unicode characters unless required by the JSON syntax or other |
144 | Stops and (re-)starts the event watcher without touching the |
| 171 | flags. This results in a faster and more compact format. |
145 | timeout. |
| 172 | |
146 | |
| 173 | The main use for this flag is to produce JSON texts that can be |
147 | $w->del |
| 174 | transmitted over a 7-bit channel, as the encoded JSON texts will not |
148 | $w->stop |
| 175 | contain any 8 bit characters. |
149 | Stop the event watcher if it was started. |
| 176 | |
150 | |
| 177 | JSON::XS->new->ascii (1)->encode ([chr 0x10401]) |
151 | $current_callback = $w->cb |
| 178 | => ["\ud801\udc01"] |
152 | $old_callback = $w->cb ($new_callback) |
|
|
153 | Return the previously set callback and optionally set a new one. |
| 179 | |
154 | |
| 180 | $json = $json->latin1 ([$enable]) |
155 | $current_fh = $w->fh |
| 181 | If $enable is true (or missing), then the "encode" method will |
156 | $old_fh = $w->fh ($new_fh) |
| 182 | encode the resulting JSON text as latin1 (or iso-8859-1), escaping |
157 | Returns the previously set filehandle and optionally set a new one. |
| 183 | any characters outside the code range 0..255. The resulting string |
|
|
| 184 | can be treated as a latin1-encoded JSON text or a native unicode |
|
|
| 185 | string. The "decode" method will not be affected in any way by this |
|
|
| 186 | flag, as "decode" by default expects unicode, which is a strict |
|
|
| 187 | superset of latin1. |
|
|
| 188 | |
158 | |
| 189 | If $enable is false, then the "encode" method will not escape |
159 | $current_eventmask = $w->events |
| 190 | Unicode characters unless required by the JSON syntax or other |
160 | $old_eventmask = $w->events ($new_eventmask) |
| 191 | flags. |
161 | Returns the previously set event mask and optionally set a new one. |
| 192 | |
162 | |
| 193 | The main use for this flag is efficiently encoding binary data as |
163 | $w->timeout ($after, $repeat) |
| 194 | JSON text, as most octets will not be escaped, resulting in a |
164 | Resets the timeout (see "EV::timer" for details). |
| 195 | smaller encoded size. The disadvantage is that the resulting JSON |
|
|
| 196 | text is encoded in latin1 (and must correctly be treated as such |
|
|
| 197 | when storing and transfering), a rare encoding for JSON. It is |
|
|
| 198 | therefore most useful when you want to store data structures known |
|
|
| 199 | to contain binary data efficiently in files or databases, not when |
|
|
| 200 | talking to other JSON encoders/decoders. |
|
|
| 201 | |
165 | |
| 202 | JSON::XS->new->latin1->encode (["\x{89}\x{abc}"] |
166 | $w->timeout_abs ($at, $interval) |
| 203 | => ["\x{89}\\u0abc"] # (perl syntax, U+abc escaped, U+89 not) |
167 | Resets the timeout (see "EV::timer_abs" for details). |
| 204 | |
168 | |
| 205 | $json = $json->utf8 ([$enable]) |
169 | $w->priority_set ($priority) |
| 206 | If $enable is true (or missing), then the "encode" method will |
170 | Set the priority of the watcher to $priority (0 <= $priority < |
| 207 | encode the JSON result into UTF-8, as required by many protocols, |
171 | $EV::NPRI). |
| 208 | while the "decode" method expects to be handled an UTF-8-encoded |
|
|
| 209 | string. Please note that UTF-8-encoded strings do not contain any |
|
|
| 210 | characters outside the range 0..255, they are thus useful for |
|
|
| 211 | bytewise/binary I/O. In future versions, enabling this option might |
|
|
| 212 | enable autodetection of the UTF-16 and UTF-32 encoding families, as |
|
|
| 213 | described in RFC4627. |
|
|
| 214 | |
|
|
| 215 | If $enable is false, then the "encode" method will return the JSON |
|
|
| 216 | string as a (non-encoded) unicode string, while "decode" expects |
|
|
| 217 | thus a unicode string. Any decoding or encoding (e.g. to UTF-8 or |
|
|
| 218 | UTF-16) needs to be done yourself, e.g. using the Encode module. |
|
|
| 219 | |
|
|
| 220 | Example, output UTF-16BE-encoded JSON: |
|
|
| 221 | |
|
|
| 222 | use Encode; |
|
|
| 223 | $jsontext = encode "UTF-16BE", JSON::XS->new->encode ($object); |
|
|
| 224 | |
|
|
| 225 | Example, decode UTF-32LE-encoded JSON: |
|
|
| 226 | |
|
|
| 227 | use Encode; |
|
|
| 228 | $object = JSON::XS->new->decode (decode "UTF-32LE", $jsontext); |
|
|
| 229 | |
|
|
| 230 | $json = $json->pretty ([$enable]) |
|
|
| 231 | This enables (or disables) all of the "indent", "space_before" and |
|
|
| 232 | "space_after" (and in the future possibly more) flags in one call to |
|
|
| 233 | generate the most readable (or most compact) form possible. |
|
|
| 234 | |
|
|
| 235 | Example, pretty-print some simple structure: |
|
|
| 236 | |
|
|
| 237 | my $json = JSON::XS->new->pretty(1)->encode ({a => [1,2]}) |
|
|
| 238 | => |
|
|
| 239 | { |
|
|
| 240 | "a" : [ |
|
|
| 241 | 1, |
|
|
| 242 | 2 |
|
|
| 243 | ] |
|
|
| 244 | } |
|
|
| 245 | |
|
|
| 246 | $json = $json->indent ([$enable]) |
|
|
| 247 | If $enable is true (or missing), then the "encode" method will use a |
|
|
| 248 | multiline format as output, putting every array member or |
|
|
| 249 | object/hash key-value pair into its own line, identing them |
|
|
| 250 | properly. |
|
|
| 251 | |
|
|
| 252 | If $enable is false, no newlines or indenting will be produced, and |
|
|
| 253 | the resulting JSON text is guarenteed not to contain any "newlines". |
|
|
| 254 | |
|
|
| 255 | This setting has no effect when decoding JSON texts. |
|
|
| 256 | |
|
|
| 257 | $json = $json->space_before ([$enable]) |
|
|
| 258 | If $enable is true (or missing), then the "encode" method will add |
|
|
| 259 | an extra optional space before the ":" separating keys from values |
|
|
| 260 | in JSON objects. |
|
|
| 261 | |
|
|
| 262 | If $enable is false, then the "encode" method will not add any extra |
|
|
| 263 | space at those places. |
|
|
| 264 | |
|
|
| 265 | This setting has no effect when decoding JSON texts. You will also |
|
|
| 266 | most likely combine this setting with "space_after". |
|
|
| 267 | |
|
|
| 268 | Example, space_before enabled, space_after and indent disabled: |
|
|
| 269 | |
|
|
| 270 | {"key" :"value"} |
|
|
| 271 | |
|
|
| 272 | $json = $json->space_after ([$enable]) |
|
|
| 273 | If $enable is true (or missing), then the "encode" method will add |
|
|
| 274 | an extra optional space after the ":" separating keys from values in |
|
|
| 275 | JSON objects and extra whitespace after the "," separating key-value |
|
|
| 276 | pairs and array members. |
|
|
| 277 | |
|
|
| 278 | If $enable is false, then the "encode" method will not add any extra |
|
|
| 279 | space at those places. |
|
|
| 280 | |
|
|
| 281 | This setting has no effect when decoding JSON texts. |
|
|
| 282 | |
|
|
| 283 | Example, space_before and indent disabled, space_after enabled: |
|
|
| 284 | |
|
|
| 285 | {"key": "value"} |
|
|
| 286 | |
|
|
| 287 | $json = $json->relaxed ([$enable]) |
|
|
| 288 | If $enable is true (or missing), then "decode" will accept some |
|
|
| 289 | extensions to normal JSON syntax (see below). "encode" will not be |
|
|
| 290 | affected in anyway. *Be aware that this option makes you accept |
|
|
| 291 | invalid JSON texts as if they were valid!*. I suggest only to use |
|
|
| 292 | this option to parse application-specific files written by humans |
|
|
| 293 | (configuration files, resource files etc.) |
|
|
| 294 | |
|
|
| 295 | If $enable is false (the default), then "decode" will only accept |
|
|
| 296 | valid JSON texts. |
|
|
| 297 | |
|
|
| 298 | Currently accepted extensions are: |
|
|
| 299 | |
|
|
| 300 | * list items can have an end-comma |
|
|
| 301 | JSON *separates* array elements and key-value pairs with commas. |
|
|
| 302 | This can be annoying if you write JSON texts manually and want |
|
|
| 303 | to be able to quickly append elements, so this extension accepts |
|
|
| 304 | comma at the end of such items not just between them: |
|
|
| 305 | |
|
|
| 306 | [ |
|
|
| 307 | 1, |
|
|
| 308 | 2, <- this comma not normally allowed |
|
|
| 309 | ] |
|
|
| 310 | { |
|
|
| 311 | "k1": "v1", |
|
|
| 312 | "k2": "v2", <- this comma not normally allowed |
|
|
| 313 | } |
|
|
| 314 | |
|
|
| 315 | * shell-style '#'-comments |
|
|
| 316 | Whenever JSON allows whitespace, shell-style comments are |
|
|
| 317 | additionally allowed. They are terminated by the first |
|
|
| 318 | carriage-return or line-feed character, after which more |
|
|
| 319 | white-space and comments are allowed. |
|
|
| 320 | |
|
|
| 321 | [ |
|
|
| 322 | 1, # this comment not allowed in JSON |
|
|
| 323 | # neither this one... |
|
|
| 324 | ] |
|
|
| 325 | |
|
|
| 326 | $json = $json->canonical ([$enable]) |
|
|
| 327 | If $enable is true (or missing), then the "encode" method will |
|
|
| 328 | output JSON objects by sorting their keys. This is adding a |
|
|
| 329 | comparatively high overhead. |
|
|
| 330 | |
|
|
| 331 | If $enable is false, then the "encode" method will output key-value |
|
|
| 332 | pairs in the order Perl stores them (which will likely change |
|
|
| 333 | between runs of the same script). |
|
|
| 334 | |
|
|
| 335 | This option is useful if you want the same data structure to be |
|
|
| 336 | encoded as the same JSON text (given the same overall settings). If |
|
|
| 337 | it is disabled, the same hash migh be encoded differently even if |
|
|
| 338 | contains the same data, as key-value pairs have no inherent ordering |
|
|
| 339 | in Perl. |
|
|
| 340 | |
|
|
| 341 | This setting has no effect when decoding JSON texts. |
|
|
| 342 | |
|
|
| 343 | $json = $json->allow_nonref ([$enable]) |
|
|
| 344 | If $enable is true (or missing), then the "encode" method can |
|
|
| 345 | convert a non-reference into its corresponding string, number or |
|
|
| 346 | null JSON value, which is an extension to RFC4627. Likewise, |
|
|
| 347 | "decode" will accept those JSON values instead of croaking. |
|
|
| 348 | |
|
|
| 349 | If $enable is false, then the "encode" method will croak if it isn't |
|
|
| 350 | passed an arrayref or hashref, as JSON texts must either be an |
|
|
| 351 | object or array. Likewise, "decode" will croak if given something |
|
|
| 352 | that is not a JSON object or array. |
|
|
| 353 | |
|
|
| 354 | Example, encode a Perl scalar as JSON value with enabled |
|
|
| 355 | "allow_nonref", resulting in an invalid JSON text: |
|
|
| 356 | |
|
|
| 357 | JSON::XS->new->allow_nonref->encode ("Hello, World!") |
|
|
| 358 | => "Hello, World!" |
|
|
| 359 | |
|
|
| 360 | $json = $json->allow_blessed ([$enable]) |
|
|
| 361 | If $enable is true (or missing), then the "encode" method will not |
|
|
| 362 | barf when it encounters a blessed reference. Instead, the value of |
|
|
| 363 | the convert_blessed option will decide wether "null" |
|
|
| 364 | ("convert_blessed" disabled or no "to_json" method found) or a |
|
|
| 365 | representation of the object ("convert_blessed" enabled and |
|
|
| 366 | "to_json" method found) is being encoded. Has no effect on "decode". |
|
|
| 367 | |
|
|
| 368 | If $enable is false (the default), then "encode" will throw an |
|
|
| 369 | exception when it encounters a blessed object. |
|
|
| 370 | |
|
|
| 371 | $json = $json->convert_blessed ([$enable]) |
|
|
| 372 | If $enable is true (or missing), then "encode", upon encountering a |
|
|
| 373 | blessed object, will check for the availability of the "TO_JSON" |
|
|
| 374 | method on the object's class. If found, it will be called in scalar |
|
|
| 375 | context and the resulting scalar will be encoded instead of the |
|
|
| 376 | object. If no "TO_JSON" method is found, the value of |
|
|
| 377 | "allow_blessed" will decide what to do. |
|
|
| 378 | |
|
|
| 379 | The "TO_JSON" method may safely call die if it wants. If "TO_JSON" |
|
|
| 380 | returns other blessed objects, those will be handled in the same |
|
|
| 381 | way. "TO_JSON" must take care of not causing an endless recursion |
|
|
| 382 | cycle (== crash) in this case. The name of "TO_JSON" was chosen |
|
|
| 383 | because other methods called by the Perl core (== not by the user of |
|
|
| 384 | the object) are usually in upper case letters and to avoid |
|
|
| 385 | collisions with the "to_json" function. |
|
|
| 386 | |
|
|
| 387 | This setting does not yet influence "decode" in any way, but in the |
|
|
| 388 | future, global hooks might get installed that influence "decode" and |
|
|
| 389 | are enabled by this setting. |
|
|
| 390 | |
|
|
| 391 | If $enable is false, then the "allow_blessed" setting will decide |
|
|
| 392 | what to do when a blessed object is found. |
|
|
| 393 | |
|
|
| 394 | $json = $json->filter_json_object ([$coderef->($hashref)]) |
|
|
| 395 | When $coderef is specified, it will be called from "decode" each |
|
|
| 396 | time it decodes a JSON object. The only argument is a reference to |
|
|
| 397 | the newly-created hash. If the code references returns a single |
|
|
| 398 | scalar (which need not be a reference), this value (i.e. a copy of |
|
|
| 399 | that scalar to avoid aliasing) is inserted into the deserialised |
|
|
| 400 | data structure. If it returns an empty list (NOTE: *not* "undef", |
|
|
| 401 | which is a valid scalar), the original deserialised hash will be |
|
|
| 402 | inserted. This setting can slow down decoding considerably. |
|
|
| 403 | |
|
|
| 404 | When $coderef is omitted or undefined, any existing callback will be |
|
|
| 405 | removed and "decode" will not change the deserialised hash in any |
|
|
| 406 | way. |
|
|
| 407 | |
|
|
| 408 | Example, convert all JSON objects into the integer 5: |
|
|
| 409 | |
|
|
| 410 | my $js = JSON::XS->new->filter_json_object (sub { 5 }); |
|
|
| 411 | # returns [5] |
|
|
| 412 | $js->decode ('[{}]') |
|
|
| 413 | # throw an exception because allow_nonref is not enabled |
|
|
| 414 | # so a lone 5 is not allowed. |
|
|
| 415 | $js->decode ('{"a":1, "b":2}'); |
|
|
| 416 | |
|
|
| 417 | $json = $json->filter_json_single_key_object ($key [=> |
|
|
| 418 | $coderef->($value)]) |
|
|
| 419 | Works remotely similar to "filter_json_object", but is only called |
|
|
| 420 | for JSON objects having a single key named $key. |
|
|
| 421 | |
|
|
| 422 | This $coderef is called before the one specified via |
|
|
| 423 | "filter_json_object", if any. It gets passed the single value in the |
|
|
| 424 | JSON object. If it returns a single value, it will be inserted into |
|
|
| 425 | the data structure. If it returns nothing (not even "undef" but the |
|
|
| 426 | empty list), the callback from "filter_json_object" will be called |
|
|
| 427 | next, as if no single-key callback were specified. |
|
|
| 428 | |
|
|
| 429 | If $coderef is omitted or undefined, the corresponding callback will |
|
|
| 430 | be disabled. There can only ever be one callback for a given key. |
|
|
| 431 | |
|
|
| 432 | As this callback gets called less often then the |
|
|
| 433 | "filter_json_object" one, decoding speed will not usually suffer as |
|
|
| 434 | much. Therefore, single-key objects make excellent targets to |
|
|
| 435 | serialise Perl objects into, especially as single-key JSON objects |
|
|
| 436 | are as close to the type-tagged value concept as JSON gets (its |
|
|
| 437 | basically an ID/VALUE tuple). Of course, JSON does not support this |
|
|
| 438 | in any way, so you need to make sure your data never looks like a |
|
|
| 439 | serialised Perl hash. |
|
|
| 440 | |
|
|
| 441 | Typical names for the single object key are "__class_whatever__", or |
|
|
| 442 | "$__dollars_are_rarely_used__$" or "}ugly_brace_placement", or even |
|
|
| 443 | things like "__class_md5sum(classname)__", to reduce the risk of |
|
|
| 444 | clashing with real hashes. |
|
|
| 445 | |
|
|
| 446 | Example, decode JSON objects of the form "{ "__widget__" => <id> }" |
|
|
| 447 | into the corresponding $WIDGET{<id>} object: |
|
|
| 448 | |
|
|
| 449 | # return whatever is in $WIDGET{5}: |
|
|
| 450 | JSON::XS |
|
|
| 451 | ->new |
|
|
| 452 | ->filter_json_single_key_object (__widget__ => sub { |
|
|
| 453 | $WIDGET{ $_[0] } |
|
|
| 454 | }) |
|
|
| 455 | ->decode ('{"__widget__": 5') |
|
|
| 456 | |
|
|
| 457 | # this can be used with a TO_JSON method in some "widget" class |
|
|
| 458 | # for serialisation to json: |
|
|
| 459 | sub WidgetBase::TO_JSON { |
|
|
| 460 | my ($self) = @_; |
|
|
| 461 | |
|
|
| 462 | unless ($self->{id}) { |
|
|
| 463 | $self->{id} = ..get..some..id..; |
|
|
| 464 | $WIDGET{$self->{id}} = $self; |
|
|
| 465 | } |
|
|
| 466 | |
|
|
| 467 | { __widget__ => $self->{id} } |
|
|
| 468 | } |
|
|
| 469 | |
|
|
| 470 | $json = $json->shrink ([$enable]) |
|
|
| 471 | Perl usually over-allocates memory a bit when allocating space for |
|
|
| 472 | strings. This flag optionally resizes strings generated by either |
|
|
| 473 | "encode" or "decode" to their minimum size possible. This can save |
|
|
| 474 | memory when your JSON texts are either very very long or you have |
|
|
| 475 | many short strings. It will also try to downgrade any strings to |
|
|
| 476 | octet-form if possible: perl stores strings internally either in an |
|
|
| 477 | encoding called UTF-X or in octet-form. The latter cannot store |
|
|
| 478 | everything but uses less space in general (and some buggy Perl or C |
|
|
| 479 | code might even rely on that internal representation being used). |
|
|
| 480 | |
|
|
| 481 | The actual definition of what shrink does might change in future |
|
|
| 482 | versions, but it will always try to save space at the expense of |
|
|
| 483 | time. |
|
|
| 484 | |
|
|
| 485 | If $enable is true (or missing), the string returned by "encode" |
|
|
| 486 | will be shrunk-to-fit, while all strings generated by "decode" will |
|
|
| 487 | also be shrunk-to-fit. |
|
|
| 488 | |
|
|
| 489 | If $enable is false, then the normal perl allocation algorithms are |
|
|
| 490 | used. If you work with your data, then this is likely to be faster. |
|
|
| 491 | |
|
|
| 492 | In the future, this setting might control other things, such as |
|
|
| 493 | converting strings that look like integers or floats into integers |
|
|
| 494 | or floats internally (there is no difference on the Perl level), |
|
|
| 495 | saving space. |
|
|
| 496 | |
|
|
| 497 | $json = $json->max_depth ([$maximum_nesting_depth]) |
|
|
| 498 | Sets the maximum nesting level (default 512) accepted while encoding |
|
|
| 499 | or decoding. If the JSON text or Perl data structure has an equal or |
|
|
| 500 | higher nesting level then this limit, then the encoder and decoder |
|
|
| 501 | will stop and croak at that point. |
|
|
| 502 | |
|
|
| 503 | Nesting level is defined by number of hash- or arrayrefs that the |
|
|
| 504 | encoder needs to traverse to reach a given point or the number of |
|
|
| 505 | "{" or "[" characters without their matching closing parenthesis |
|
|
| 506 | crossed to reach a given character in a string. |
|
|
| 507 | |
|
|
| 508 | Setting the maximum depth to one disallows any nesting, so that |
|
|
| 509 | ensures that the object is only a single hash/object or array. |
|
|
| 510 | |
|
|
| 511 | The argument to "max_depth" will be rounded up to the next highest |
|
|
| 512 | power of two. If no argument is given, the highest possible setting |
|
|
| 513 | will be used, which is rarely useful. |
|
|
| 514 | |
|
|
| 515 | See SECURITY CONSIDERATIONS, below, for more info on why this is |
|
|
| 516 | useful. |
|
|
| 517 | |
|
|
| 518 | $json = $json->max_size ([$maximum_string_size]) |
|
|
| 519 | Set the maximum length a JSON text may have (in bytes) where |
|
|
| 520 | decoding is being attempted. The default is 0, meaning no limit. |
|
|
| 521 | When "decode" is called on a string longer then this number of |
|
|
| 522 | characters it will not attempt to decode the string but throw an |
|
|
| 523 | exception. This setting has no effect on "encode" (yet). |
|
|
| 524 | |
|
|
| 525 | The argument to "max_size" will be rounded up to the next highest |
|
|
| 526 | power of two (so may be more than requested). If no argument is |
|
|
| 527 | given, the limit check will be deactivated (same as when 0 is |
|
|
| 528 | specified). |
|
|
| 529 | |
|
|
| 530 | See SECURITY CONSIDERATIONS, below, for more info on why this is |
|
|
| 531 | useful. |
|
|
| 532 | |
|
|
| 533 | $json_text = $json->encode ($perl_scalar) |
|
|
| 534 | Converts the given Perl data structure (a simple scalar or a |
|
|
| 535 | reference to a hash or array) to its JSON representation. Simple |
|
|
| 536 | scalars will be converted into JSON string or number sequences, |
|
|
| 537 | while references to arrays become JSON arrays and references to |
|
|
| 538 | hashes become JSON objects. Undefined Perl values (e.g. "undef") |
|
|
| 539 | become JSON "null" values. Neither "true" nor "false" values will be |
|
|
| 540 | generated. |
|
|
| 541 | |
|
|
| 542 | $perl_scalar = $json->decode ($json_text) |
|
|
| 543 | The opposite of "encode": expects a JSON text and tries to parse it, |
|
|
| 544 | returning the resulting simple scalar or reference. Croaks on error. |
|
|
| 545 | |
|
|
| 546 | JSON numbers and strings become simple Perl scalars. JSON arrays |
|
|
| 547 | become Perl arrayrefs and JSON objects become Perl hashrefs. "true" |
|
|
| 548 | becomes 1, "false" becomes 0 and "null" becomes "undef". |
|
|
| 549 | |
|
|
| 550 | ($perl_scalar, $characters) = $json->decode_prefix ($json_text) |
|
|
| 551 | This works like the "decode" method, but instead of raising an |
|
|
| 552 | exception when there is trailing garbage after the first JSON |
|
|
| 553 | object, it will silently stop parsing there and return the number of |
|
|
| 554 | characters consumed so far. |
|
|
| 555 | |
|
|
| 556 | This is useful if your JSON texts are not delimited by an outer |
|
|
| 557 | protocol (which is not the brightest thing to do in the first place) |
|
|
| 558 | and you need to know where the JSON text ends. |
|
|
| 559 | |
|
|
| 560 | JSON::XS->new->decode_prefix ("[1] the tail") |
|
|
| 561 | => ([], 3) |
|
|
| 562 | |
|
|
| 563 | MAPPING |
|
|
| 564 | This section describes how JSON::XS maps Perl values to JSON values and |
|
|
| 565 | vice versa. These mappings are designed to "do the right thing" in most |
|
|
| 566 | circumstances automatically, preserving round-tripping characteristics |
|
|
| 567 | (what you put in comes out as something equivalent). |
|
|
| 568 | |
|
|
| 569 | For the more enlightened: note that in the following descriptions, |
|
|
| 570 | lowercase *perl* refers to the Perl interpreter, while uppcercase *Perl* |
|
|
| 571 | refers to the abstract Perl language itself. |
|
|
| 572 | |
|
|
| 573 | JSON -> PERL |
|
|
| 574 | object |
|
|
| 575 | A JSON object becomes a reference to a hash in Perl. No ordering of |
|
|
| 576 | object keys is preserved (JSON does not preserver object key |
|
|
| 577 | ordering itself). |
|
|
| 578 | |
|
|
| 579 | array |
|
|
| 580 | A JSON array becomes a reference to an array in Perl. |
|
|
| 581 | |
|
|
| 582 | string |
|
|
| 583 | A JSON string becomes a string scalar in Perl - Unicode codepoints |
|
|
| 584 | in JSON are represented by the same codepoints in the Perl string, |
|
|
| 585 | so no manual decoding is necessary. |
|
|
| 586 | |
|
|
| 587 | number |
|
|
| 588 | A JSON number becomes either an integer, numeric (floating point) or |
|
|
| 589 | string scalar in perl, depending on its range and any fractional |
|
|
| 590 | parts. On the Perl level, there is no difference between those as |
|
|
| 591 | Perl handles all the conversion details, but an integer may take |
|
|
| 592 | slightly less memory and might represent more values exactly than |
|
|
| 593 | (floating point) numbers. |
|
|
| 594 | |
|
|
| 595 | If the number consists of digits only, JSON::XS will try to |
|
|
| 596 | represent it as an integer value. If that fails, it will try to |
|
|
| 597 | represent it as a numeric (floating point) value if that is possible |
|
|
| 598 | without loss of precision. Otherwise it will preserve the number as |
|
|
| 599 | a string value. |
|
|
| 600 | |
|
|
| 601 | Numbers containing a fractional or exponential part will always be |
|
|
| 602 | represented as numeric (floating point) values, possibly at a loss |
|
|
| 603 | of precision. |
|
|
| 604 | |
|
|
| 605 | This might create round-tripping problems as numbers might become |
|
|
| 606 | strings, but as Perl is typeless there is no other way to do it. |
|
|
| 607 | |
|
|
| 608 | true, false |
|
|
| 609 | These JSON atoms become "JSON::XS::true" and "JSON::XS::false", |
|
|
| 610 | respectively. They are overloaded to act almost exactly like the |
|
|
| 611 | numbers 1 and 0. You can check wether a scalar is a JSON boolean by |
|
|
| 612 | using the "JSON::XS::is_bool" function. |
|
|
| 613 | |
|
|
| 614 | null |
|
|
| 615 | A JSON null atom becomes "undef" in Perl. |
|
|
| 616 | |
|
|
| 617 | PERL -> JSON |
|
|
| 618 | The mapping from Perl to JSON is slightly more difficult, as Perl is a |
|
|
| 619 | truly typeless language, so we can only guess which JSON type is meant |
|
|
| 620 | by a Perl value. |
|
|
| 621 | |
|
|
| 622 | hash references |
|
|
| 623 | Perl hash references become JSON objects. As there is no inherent |
|
|
| 624 | ordering in hash keys (or JSON objects), they will usually be |
|
|
| 625 | encoded in a pseudo-random order that can change between runs of the |
|
|
| 626 | same program but stays generally the same within a single run of a |
|
|
| 627 | program. JSON::XS can optionally sort the hash keys (determined by |
|
|
| 628 | the *canonical* flag), so the same datastructure will serialise to |
|
|
| 629 | the same JSON text (given same settings and version of JSON::XS), |
|
|
| 630 | but this incurs a runtime overhead and is only rarely useful, e.g. |
|
|
| 631 | when you want to compare some JSON text against another for |
|
|
| 632 | equality. |
|
|
| 633 | |
|
|
| 634 | array references |
|
|
| 635 | Perl array references become JSON arrays. |
|
|
| 636 | |
|
|
| 637 | other references |
|
|
| 638 | Other unblessed references are generally not allowed and will cause |
|
|
| 639 | an exception to be thrown, except for references to the integers 0 |
|
|
| 640 | and 1, which get turned into "false" and "true" atoms in JSON. You |
|
|
| 641 | can also use "JSON::XS::false" and "JSON::XS::true" to improve |
|
|
| 642 | readability. |
|
|
| 643 | |
|
|
| 644 | to_json [\0,JSON::XS::true] # yields [false,true] |
|
|
| 645 | |
|
|
| 646 | JSON::XS::true, JSON::XS::false |
|
|
| 647 | These special values become JSON true and JSON false values, |
|
|
| 648 | respectively. You can also use "\1" and "\0" directly if you want. |
|
|
| 649 | |
|
|
| 650 | blessed objects |
|
|
| 651 | Blessed objects are not allowed. JSON::XS currently tries to encode |
|
|
| 652 | their underlying representation (hash- or arrayref), but this |
|
|
| 653 | behaviour might change in future versions. |
|
|
| 654 | |
|
|
| 655 | simple scalars |
|
|
| 656 | Simple Perl scalars (any scalar that is not a reference) are the |
|
|
| 657 | most difficult objects to encode: JSON::XS will encode undefined |
|
|
| 658 | scalars as JSON null value, scalars that have last been used in a |
|
|
| 659 | string context before encoding as JSON strings and anything else as |
|
|
| 660 | number value: |
|
|
| 661 | |
|
|
| 662 | # dump as number |
|
|
| 663 | to_json [2] # yields [2] |
|
|
| 664 | to_json [-3.0e17] # yields [-3e+17] |
|
|
| 665 | my $value = 5; to_json [$value] # yields [5] |
|
|
| 666 | |
|
|
| 667 | # used as string, so dump as string |
|
|
| 668 | print $value; |
|
|
| 669 | to_json [$value] # yields ["5"] |
|
|
| 670 | |
|
|
| 671 | # undef becomes null |
|
|
| 672 | to_json [undef] # yields [null] |
|
|
| 673 | |
|
|
| 674 | You can force the type to be a string by stringifying it: |
|
|
| 675 | |
|
|
| 676 | my $x = 3.1; # some variable containing a number |
|
|
| 677 | "$x"; # stringified |
|
|
| 678 | $x .= ""; # another, more awkward way to stringify |
|
|
| 679 | print $x; # perl does it for you, too, quite often |
|
|
| 680 | |
|
|
| 681 | You can force the type to be a number by numifying it: |
|
|
| 682 | |
|
|
| 683 | my $x = "3"; # some variable containing a string |
|
|
| 684 | $x += 0; # numify it, ensuring it will be dumped as a number |
|
|
| 685 | $x *= 1; # same thing, the choise is yours. |
|
|
| 686 | |
|
|
| 687 | You can not currently output JSON booleans or force the type in |
|
|
| 688 | other, less obscure, ways. Tell me if you need this capability. |
|
|
| 689 | |
|
|
| 690 | COMPARISON |
|
|
| 691 | As already mentioned, this module was created because none of the |
|
|
| 692 | existing JSON modules could be made to work correctly. First I will |
|
|
| 693 | describe the problems (or pleasures) I encountered with various existing |
|
|
| 694 | JSON modules, followed by some benchmark values. JSON::XS was designed |
|
|
| 695 | not to suffer from any of these problems or limitations. |
|
|
| 696 | |
|
|
| 697 | JSON 1.07 |
|
|
| 698 | Slow (but very portable, as it is written in pure Perl). |
|
|
| 699 | |
|
|
| 700 | Undocumented/buggy Unicode handling (how JSON handles unicode values |
|
|
| 701 | is undocumented. One can get far by feeding it unicode strings and |
|
|
| 702 | doing en-/decoding oneself, but unicode escapes are not working |
|
|
| 703 | properly). |
|
|
| 704 | |
|
|
| 705 | No roundtripping (strings get clobbered if they look like numbers, |
|
|
| 706 | e.g. the string 2.0 will encode to 2.0 instead of "2.0", and that |
|
|
| 707 | will decode into the number 2. |
|
|
| 708 | |
|
|
| 709 | JSON::PC 0.01 |
|
|
| 710 | Very fast. |
|
|
| 711 | |
|
|
| 712 | Undocumented/buggy Unicode handling. |
|
|
| 713 | |
|
|
| 714 | No roundtripping. |
|
|
| 715 | |
|
|
| 716 | Has problems handling many Perl values (e.g. regex results and other |
|
|
| 717 | magic values will make it croak). |
|
|
| 718 | |
|
|
| 719 | Does not even generate valid JSON ("{1,2}" gets converted to "{1:2}" |
|
|
| 720 | which is not a valid JSON text. |
|
|
| 721 | |
|
|
| 722 | Unmaintained (maintainer unresponsive for many months, bugs are not |
|
|
| 723 | getting fixed). |
|
|
| 724 | |
|
|
| 725 | JSON::Syck 0.21 |
|
|
| 726 | Very buggy (often crashes). |
|
|
| 727 | |
|
|
| 728 | Very inflexible (no human-readable format supported, format pretty |
|
|
| 729 | much undocumented. I need at least a format for easy reading by |
|
|
| 730 | humans and a single-line compact format for use in a protocol, and |
|
|
| 731 | preferably a way to generate ASCII-only JSON texts). |
|
|
| 732 | |
|
|
| 733 | Completely broken (and confusingly documented) Unicode handling |
|
|
| 734 | (unicode escapes are not working properly, you need to set |
|
|
| 735 | ImplicitUnicode to *different* values on en- and decoding to get |
|
|
| 736 | symmetric behaviour). |
|
|
| 737 | |
|
|
| 738 | No roundtripping (simple cases work, but this depends on wether the |
|
|
| 739 | scalar value was used in a numeric context or not). |
|
|
| 740 | |
|
|
| 741 | Dumping hashes may skip hash values depending on iterator state. |
|
|
| 742 | |
|
|
| 743 | Unmaintained (maintainer unresponsive for many months, bugs are not |
|
|
| 744 | getting fixed). |
|
|
| 745 | |
|
|
| 746 | Does not check input for validity (i.e. will accept non-JSON input |
|
|
| 747 | and return "something" instead of raising an exception. This is a |
|
|
| 748 | security issue: imagine two banks transfering money between each |
|
|
| 749 | other using JSON. One bank might parse a given non-JSON request and |
|
|
| 750 | deduct money, while the other might reject the transaction with a |
|
|
| 751 | syntax error. While a good protocol will at least recover, that is |
|
|
| 752 | extra unnecessary work and the transaction will still not succeed). |
|
|
| 753 | |
|
|
| 754 | JSON::DWIW 0.04 |
|
|
| 755 | Very fast. Very natural. Very nice. |
|
|
| 756 | |
|
|
| 757 | Undocumented unicode handling (but the best of the pack. Unicode |
|
|
| 758 | escapes still don't get parsed properly). |
|
|
| 759 | |
|
|
| 760 | Very inflexible. |
|
|
| 761 | |
|
|
| 762 | No roundtripping. |
|
|
| 763 | |
|
|
| 764 | Does not generate valid JSON texts (key strings are often unquoted, |
|
|
| 765 | empty keys result in nothing being output) |
|
|
| 766 | |
|
|
| 767 | Does not check input for validity. |
|
|
| 768 | |
|
|
| 769 | JSON and YAML |
|
|
| 770 | You often hear that JSON is a subset (or a close subset) of YAML. This |
|
|
| 771 | is, however, a mass hysteria and very far from the truth. In general, |
|
|
| 772 | there is no way to configure JSON::XS to output a data structure as |
|
|
| 773 | valid YAML. |
|
|
| 774 | |
|
|
| 775 | If you really must use JSON::XS to generate YAML, you should use this |
|
|
| 776 | algorithm (subject to change in future versions): |
|
|
| 777 | |
|
|
| 778 | my $to_yaml = JSON::XS->new->utf8->space_after (1); |
|
|
| 779 | my $yaml = $to_yaml->encode ($ref) . "\n"; |
|
|
| 780 | |
|
|
| 781 | This will usually generate JSON texts that also parse as valid YAML. |
|
|
| 782 | Please note that YAML has hardcoded limits on (simple) object key |
|
|
| 783 | lengths that JSON doesn't have, so you should make sure that your hash |
|
|
| 784 | keys are noticably shorter than the 1024 characters YAML allows. |
|
|
| 785 | |
|
|
| 786 | There might be other incompatibilities that I am not aware of. In |
|
|
| 787 | general you should not try to generate YAML with a JSON generator or |
|
|
| 788 | vice versa, or try to parse JSON with a YAML parser or vice versa: |
|
|
| 789 | chances are high that you will run into severe interoperability |
|
|
| 790 | problems. |
|
|
| 791 | |
|
|
| 792 | SPEED |
|
|
| 793 | It seems that JSON::XS is surprisingly fast, as shown in the following |
|
|
| 794 | tables. They have been generated with the help of the "eg/bench" program |
|
|
| 795 | in the JSON::XS distribution, to make it easy to compare on your own |
|
|
| 796 | system. |
|
|
| 797 | |
|
|
| 798 | First comes a comparison between various modules using a very short |
|
|
| 799 | single-line JSON string: |
|
|
| 800 | |
|
|
| 801 | {"method": "handleMessage", "params": ["user1", "we were just talking"], \ |
|
|
| 802 | "id": null, "array":[1,11,234,-5,1e5,1e7, true, false]} |
|
|
| 803 | |
|
|
| 804 | It shows the number of encodes/decodes per second (JSON::XS uses the |
|
|
| 805 | functional interface, while JSON::XS/2 uses the OO interface with |
|
|
| 806 | pretty-printing and hashkey sorting enabled, JSON::XS/3 enables shrink). |
|
|
| 807 | Higher is better: |
|
|
| 808 | |
|
|
| 809 | Storable | 15779.925 | 14169.946 | |
|
|
| 810 | -----------+------------+------------+ |
|
|
| 811 | module | encode | decode | |
|
|
| 812 | -----------|------------|------------| |
|
|
| 813 | JSON | 4990.842 | 4088.813 | |
|
|
| 814 | JSON::DWIW | 51653.990 | 71575.154 | |
|
|
| 815 | JSON::PC | 65948.176 | 74631.744 | |
|
|
| 816 | JSON::PP | 8931.652 | 3817.168 | |
|
|
| 817 | JSON::Syck | 24877.248 | 27776.848 | |
|
|
| 818 | JSON::XS | 388361.481 | 227951.304 | |
|
|
| 819 | JSON::XS/2 | 227951.304 | 218453.333 | |
|
|
| 820 | JSON::XS/3 | 338250.323 | 218453.333 | |
|
|
| 821 | Storable | 16500.016 | 135300.129 | |
|
|
| 822 | -----------+------------+------------+ |
|
|
| 823 | |
|
|
| 824 | That is, JSON::XS is about five times faster than JSON::DWIW on |
|
|
| 825 | encoding, about three times faster on decoding, and over fourty times |
|
|
| 826 | faster than JSON, even with pretty-printing and key sorting. It also |
|
|
| 827 | compares favourably to Storable for small amounts of data. |
|
|
| 828 | |
|
|
| 829 | Using a longer test string (roughly 18KB, generated from Yahoo! Locals |
|
|
| 830 | search API (http://nanoref.com/yahooapis/mgPdGg): |
|
|
| 831 | |
|
|
| 832 | module | encode | decode | |
|
|
| 833 | -----------|------------|------------| |
|
|
| 834 | JSON | 55.260 | 34.971 | |
|
|
| 835 | JSON::DWIW | 825.228 | 1082.513 | |
|
|
| 836 | JSON::PC | 3571.444 | 2394.829 | |
|
|
| 837 | JSON::PP | 210.987 | 32.574 | |
|
|
| 838 | JSON::Syck | 552.551 | 787.544 | |
|
|
| 839 | JSON::XS | 5780.463 | 4854.519 | |
|
|
| 840 | JSON::XS/2 | 3869.998 | 4798.975 | |
|
|
| 841 | JSON::XS/3 | 5862.880 | 4798.975 | |
|
|
| 842 | Storable | 4445.002 | 5235.027 | |
|
|
| 843 | -----------+------------+------------+ |
|
|
| 844 | |
|
|
| 845 | Again, JSON::XS leads by far (except for Storable which non-surprisingly |
|
|
| 846 | decodes faster). |
|
|
| 847 | |
|
|
| 848 | On large strings containing lots of high unicode characters, some |
|
|
| 849 | modules (such as JSON::PC) seem to decode faster than JSON::XS, but the |
|
|
| 850 | result will be broken due to missing (or wrong) unicode handling. Others |
|
|
| 851 | refuse to decode or encode properly, so it was impossible to prepare a |
|
|
| 852 | fair comparison table for that case. |
|
|
| 853 | |
|
|
| 854 | SECURITY CONSIDERATIONS |
|
|
| 855 | When you are using JSON in a protocol, talking to untrusted potentially |
|
|
| 856 | hostile creatures requires relatively few measures. |
|
|
| 857 | |
|
|
| 858 | First of all, your JSON decoder should be secure, that is, should not |
|
|
| 859 | have any buffer overflows. Obviously, this module should ensure that and |
|
|
| 860 | I am trying hard on making that true, but you never know. |
|
|
| 861 | |
|
|
| 862 | Second, you need to avoid resource-starving attacks. That means you |
|
|
| 863 | should limit the size of JSON texts you accept, or make sure then when |
|
|
| 864 | your resources run out, thats just fine (e.g. by using a separate |
|
|
| 865 | process that can crash safely). The size of a JSON text in octets or |
|
|
| 866 | characters is usually a good indication of the size of the resources |
|
|
| 867 | required to decode it into a Perl structure. While JSON::XS can check |
|
|
| 868 | the size of the JSON text, it might be too late when you already have it |
|
|
| 869 | in memory, so you might want to check the size before you accept the |
|
|
| 870 | string. |
|
|
| 871 | |
|
|
| 872 | Third, JSON::XS recurses using the C stack when decoding objects and |
|
|
| 873 | arrays. The C stack is a limited resource: for instance, on my amd64 |
|
|
| 874 | machine with 8MB of stack size I can decode around 180k nested arrays |
|
|
| 875 | but only 14k nested JSON objects (due to perl itself recursing deeply on |
|
|
| 876 | croak to free the temporary). If that is exceeded, the program crashes. |
|
|
| 877 | to be conservative, the default nesting limit is set to 512. If your |
|
|
| 878 | process has a smaller stack, you should adjust this setting accordingly |
|
|
| 879 | with the "max_depth" method. |
|
|
| 880 | |
|
|
| 881 | And last but least, something else could bomb you that I forgot to think |
|
|
| 882 | of. In that case, you get to keep the pieces. I am always open for |
|
|
| 883 | hints, though... |
|
|
| 884 | |
|
|
| 885 | If you are using JSON::XS to return packets to consumption by javascript |
|
|
| 886 | scripts in a browser you should have a look at |
|
|
| 887 | <http://jpsykes.com/47/practical-csrf-and-json-security> to see wether |
|
|
| 888 | you are vulnerable to some common attack vectors (which really are |
|
|
| 889 | browser design bugs, but it is still you who will have to deal with it, |
|
|
| 890 | as major browser developers care only for features, not about doing |
|
|
| 891 | security right). |
|
|
| 892 | |
|
|
| 893 | THREADS |
|
|
| 894 | This module is *not* guarenteed to be thread safe and there are no plans |
|
|
| 895 | to change this until Perl gets thread support (as opposed to the |
|
|
| 896 | horribly slow so-called "threads" which are simply slow and bloated |
|
|
| 897 | process simulations - use fork, its *much* faster, cheaper, better). |
|
|
| 898 | |
|
|
| 899 | (It might actually work, but you ahve ben warned). |
|
|
| 900 | |
172 | |
| 901 | BUGS |
173 | BUGS |
| 902 | While the goal of this module is to be correct, that unfortunately does |
174 | Lots. Libevent itself isn't well tested and rather buggy, and this |
| 903 | not mean its bug-free, only that I think its design is bug-free. It is |
175 | module is quite new at the moment. |
| 904 | still relatively early in its development. If you keep reporting bugs |
|
|
| 905 | they will be fixed swiftly, though. |
|
|
| 906 | |
176 | |
| 907 | Please refrain from using rt.cpan.org or any other bug reporting |
177 | Please note that the epoll method is not, in general, reliable in |
| 908 | service. I put the contact address into my modules for a reason. |
178 | programs that use fork (even if no libveent calls are being made in the |
|
|
179 | forked process). If your program behaves erratically, try setting the |
|
|
180 | environment variable "EVENT_NOEPOLL" first when running the program. |
|
|
181 | |
|
|
182 | In general, if you fork, then you can only use the EV module in one of |
|
|
183 | the children. |
|
|
184 | |
|
|
185 | SEE ALSO |
|
|
186 | L<EV::DNS>, L<event(3)>, L<event.h>, L<evdns.h>. |
|
|
187 | L<EV::AnyEvent>. |
| 909 | |
188 | |
| 910 | AUTHOR |
189 | AUTHOR |
| 911 | Marc Lehmann <schmorp@schmorp.de> |
190 | Marc Lehmann <schmorp@schmorp.de> |
| 912 | http://home.schmorp.de/ |
191 | http://home.schmorp.de/ |
| 913 | |
192 | |
