[ViewVC] Diff of: cvs/JSON-XS/XS.pm

Comparing JSON-XS/XS.pm (file contents):
Revision 1.82 by root, Sun Dec 30 21:24:57 2007 UTC vs.
Revision 1.139 by root, Thu May 23 09:31:32 2013 UTC

…		…
1	=head1 NAME	1	=head1 NAME
2		2
3	JSON::XS - JSON serialising/deserialising, done correctly and fast	3	JSON::XS - JSON serialising/deserialising, done correctly and fast
		4
		5	=encoding utf-8
4		6
5	JSON::XS - 正しくて高速な JSON シリアライザ/デシリアライザ	7	JSON::XS - 正しくて高速な JSON シリアライザ/デシリアライザ
6	(http://fleur.hio.jp/perldoc/mix/lib/JSON/XS.html)	8	(http://fleur.hio.jp/perldoc/mix/lib/JSON/XS.html)
7		9
8	=head1 SYNOPSIS	10	=head1 SYNOPSIS
35	primary goal is to be I<correct> and its secondary goal is to be	37	primary goal is to be I<correct> and its secondary goal is to be
36	I<fast>. To reach the latter goal it was written in C.	38	I<fast>. To reach the latter goal it was written in C.
37		39
38	Beginning with version 2.0 of the JSON module, when both JSON and	40	Beginning with version 2.0 of the JSON module, when both JSON and
39	JSON::XS are installed, then JSON will fall back on JSON::XS (this can be	41	JSON::XS are installed, then JSON will fall back on JSON::XS (this can be
40	overriden) with no overhead due to emulation (by inheritign constructor	42	overridden) with no overhead due to emulation (by inheriting constructor
41	and methods). If JSON::XS is not available, it will fall back to the	43	and methods). If JSON::XS is not available, it will fall back to the
42	compatible JSON::PP module as backend, so using JSON instead of JSON::XS	44	compatible JSON::PP module as backend, so using JSON instead of JSON::XS
43	gives you a portable JSON API that can be fast when you need and doesn't	45	gives you a portable JSON API that can be fast when you need and doesn't
44	require a C compiler when that is a problem.	46	require a C compiler when that is a problem.
45		47
…		…
47	to write yet another JSON module? While it seems there are many JSON	49	to write yet another JSON module? While it seems there are many JSON
48	modules, none of them correctly handle all corner cases, and in most cases	50	modules, none of them correctly handle all corner cases, and in most cases
49	their maintainers are unresponsive, gone missing, or not listening to bug	51	their maintainers are unresponsive, gone missing, or not listening to bug
50	reports for other reasons.	52	reports for other reasons.
51		53
52	See COMPARISON, below, for a comparison to some other JSON modules.
53
54	See MAPPING, below, on how JSON::XS maps perl values to JSON values and	54	See MAPPING, below, on how JSON::XS maps perl values to JSON values and
55	vice versa.	55	vice versa.
56		56
57	=head2 FEATURES	57	=head2 FEATURES
58		58
59	=over 4	59	=over 4
60		60
61	=item * correct Unicode handling	61	=item * correct Unicode handling
62		62
63	This module knows how to handle Unicode, and even documents how and when	63	This module knows how to handle Unicode, documents how and when it does
64	it does so.	64	so, and even documents what "correct" means.
65		65
66	=item * round-trip integrity	66	=item * round-trip integrity
67		67
68	When you serialise a perl data structure using only datatypes supported	68	When you serialise a perl data structure using only data types supported
69	by JSON, the deserialised data structure is identical on the Perl level.	69	by JSON and Perl, the deserialised data structure is identical on the Perl
70	(e.g. the string "2.0" doesn't suddenly become "2" just because it looks	70	level. (e.g. the string "2.0" doesn't suddenly become "2" just because
71	like a number).	71	it looks like a number). There I<are> minor exceptions to this, read the
		72	MAPPING section below to learn about those.
72		73
73	=item * strict checking of JSON correctness	74	=item * strict checking of JSON correctness
74		75
75	There is no guessing, no generating of illegal JSON texts by default,	76	There is no guessing, no generating of illegal JSON texts by default,
76	and only JSON is accepted as input by default (the latter is a security	77	and only JSON is accepted as input by default (the latter is a security
77	feature).	78	feature).
78		79
79	=item * fast	80	=item * fast
80		81
81	Compared to other JSON modules, this module compares favourably in terms	82	Compared to other JSON modules and other serialisers such as Storable,
82	of speed, too.	83	this module usually compares favourably in terms of speed, too.
83		84
84	=item * simple to use	85	=item * simple to use
85		86
86	This module has both a simple functional interface as well as an OO	87	This module has both a simple functional interface as well as an object
87	interface.	88	oriented interface interface.
88		89
89	=item * reasonably versatile output formats	90	=item * reasonably versatile output formats
90		91
91	You can choose between the most compact guaranteed single-line format	92	You can choose between the most compact guaranteed-single-line format
92	possible (nice for simple line-based protocols), a pure-ascii format	93	possible (nice for simple line-based protocols), a pure-ASCII format
93	(for when your transport is not 8-bit clean, still supports the whole	94	(for when your transport is not 8-bit clean, still supports the whole
94	Unicode range), or a pretty-printed format (for when you want to read that	95	Unicode range), or a pretty-printed format (for when you want to read that
95	stuff). Or you can combine those features in whatever way you like.	96	stuff). Or you can combine those features in whatever way you like.
96		97
97	=back	98	=back
98		99
99	=cut	100	=cut
100		101
101	package JSON::XS;	102	package JSON::XS;
102		103
103	use strict;	104	use common::sense;
104		105
105	our $VERSION = '2.01';	106	our $VERSION = 2.34;
106	our @ISA = qw(Exporter);	107	our @ISA = qw(Exporter);
107		108
108	our @EXPORT = qw(encode_json decode_json to_json from_json);	109	our @EXPORT = qw(encode_json decode_json to_json from_json);
109		110
110	sub to_json($) {	111	sub to_json($) {
…		…
134		135
135	This function call is functionally identical to:	136	This function call is functionally identical to:
136		137
137	$json_text = JSON::XS->new->utf8->encode ($perl_scalar)	138	$json_text = JSON::XS->new->utf8->encode ($perl_scalar)
138		139
139	except being faster.	140	Except being faster.
140		141
141	=item $perl_scalar = decode_json $json_text	142	=item $perl_scalar = decode_json $json_text
142		143
143	The opposite of C<encode_json>: expects an UTF-8 (binary) string and tries	144	The opposite of C<encode_json>: expects an UTF-8 (binary) string and tries
144	to parse that as an UTF-8 encoded JSON text, returning the resulting	145	to parse that as an UTF-8 encoded JSON text, returning the resulting
…		…
146		147
147	This function call is functionally identical to:	148	This function call is functionally identical to:
148		149
149	$perl_scalar = JSON::XS->new->utf8->decode ($json_text)	150	$perl_scalar = JSON::XS->new->utf8->decode ($json_text)
150		151
151	except being faster.	152	Except being faster.
152		153
153	=item $is_boolean = JSON::XS::is_bool $scalar	154	=item $is_boolean = JSON::XS::is_bool $scalar
154		155
155	Returns true if the passed scalar represents either JSON::XS::true or	156	Returns true if the passed scalar represents either JSON::XS::true or
156	JSON::XS::false, two constants that act like C<1> and C<0>, respectively	157	JSON::XS::false, two constants that act like C<1> and C<0>, respectively
…		…
174	This enables you to store Unicode characters as single characters in a	175	This enables you to store Unicode characters as single characters in a
175	Perl string - very natural.	176	Perl string - very natural.
176		177
177	=item 2. Perl does I<not> associate an encoding with your strings.	178	=item 2. Perl does I<not> associate an encoding with your strings.
178		179
179	Unless you force it to, e.g. when matching it against a regex, or printing	180	... until you force it to, e.g. when matching it against a regex, or
180	the scalar to a file, in which case Perl either interprets your string as	181	printing the scalar to a file, in which case Perl either interprets your
181	locale-encoded text, octets/binary, or as Unicode, depending on various	182	string as locale-encoded text, octets/binary, or as Unicode, depending
182	settings. In no case is an encoding stored together with your data, it is	183	on various settings. In no case is an encoding stored together with your
183	I<use> that decides encoding, not any magical metadata.	184	data, it is I<use> that decides encoding, not any magical meta data.
184		185
185	=item 3. The internal utf-8 flag has no meaning with regards to the	186	=item 3. The internal utf-8 flag has no meaning with regards to the
186	encoding of your string.	187	encoding of your string.
187		188
188	Just ignore that flag unless you debug a Perl bug, a module written in	189	Just ignore that flag unless you debug a Perl bug, a module written in
…		…
194		195
195	If you didn't know about that flag, just the better, pretend it doesn't	196	If you didn't know about that flag, just the better, pretend it doesn't
196	exist.	197	exist.
197		198
198	=item 4. A "Unicode String" is simply a string where each character can be	199	=item 4. A "Unicode String" is simply a string where each character can be
199	validly interpreted as a Unicode codepoint.	200	validly interpreted as a Unicode code point.
200		201
201	If you have UTF-8 encoded data, it is no longer a Unicode string, but a	202	If you have UTF-8 encoded data, it is no longer a Unicode string, but a
202	Unicode string encoded in UTF-8, giving you a binary string.	203	Unicode string encoded in UTF-8, giving you a binary string.
203		204
204	=item 5. A string containing "high" (> 255) character values is I<not> a UTF-8 string.	205	=item 5. A string containing "high" (> 255) character values is I<not> a UTF-8 string.
…		…
242		243
243	If C<$enable> is false, then the C<encode> method will not escape Unicode	244	If C<$enable> is false, then the C<encode> method will not escape Unicode
244	characters unless required by the JSON syntax or other flags. This results	245	characters unless required by the JSON syntax or other flags. This results
245	in a faster and more compact format.	246	in a faster and more compact format.
246		247
		248	See also the section I<ENCODING/CODESET FLAG NOTES> later in this
		249	document.
		250
247	The main use for this flag is to produce JSON texts that can be	251	The main use for this flag is to produce JSON texts that can be
248	transmitted over a 7-bit channel, as the encoded JSON texts will not	252	transmitted over a 7-bit channel, as the encoded JSON texts will not
249	contain any 8 bit characters.	253	contain any 8 bit characters.
250		254
251	JSON::XS->new->ascii (1)->encode ([chr 0x10401])	255	JSON::XS->new->ascii (1)->encode ([chr 0x10401])
…		…
262	will not be affected in any way by this flag, as C<decode> by default	266	will not be affected in any way by this flag, as C<decode> by default
263	expects Unicode, which is a strict superset of latin1.	267	expects Unicode, which is a strict superset of latin1.
264		268
265	If C<$enable> is false, then the C<encode> method will not escape Unicode	269	If C<$enable> is false, then the C<encode> method will not escape Unicode
266	characters unless required by the JSON syntax or other flags.	270	characters unless required by the JSON syntax or other flags.
		271
		272	See also the section I<ENCODING/CODESET FLAG NOTES> later in this
		273	document.
267		274
268	The main use for this flag is efficiently encoding binary data as JSON	275	The main use for this flag is efficiently encoding binary data as JSON
269	text, as most octets will not be escaped, resulting in a smaller encoded	276	text, as most octets will not be escaped, resulting in a smaller encoded
270	size. The disadvantage is that the resulting JSON text is encoded	277	size. The disadvantage is that the resulting JSON text is encoded
271	in latin1 (and must correctly be treated as such when storing and	278	in latin1 (and must correctly be treated as such when storing and
…		…
290		297
291	If C<$enable> is false, then the C<encode> method will return the JSON	298	If C<$enable> is false, then the C<encode> method will return the JSON
292	string as a (non-encoded) Unicode string, while C<decode> expects thus a	299	string as a (non-encoded) Unicode string, while C<decode> expects thus a
293	Unicode string. Any decoding or encoding (e.g. to UTF-8 or UTF-16) needs	300	Unicode string. Any decoding or encoding (e.g. to UTF-8 or UTF-16) needs
294	to be done yourself, e.g. using the Encode module.	301	to be done yourself, e.g. using the Encode module.
		302
		303	See also the section I<ENCODING/CODESET FLAG NOTES> later in this
		304	document.
295		305
296	Example, output UTF-16BE-encoded JSON:	306	Example, output UTF-16BE-encoded JSON:
297		307
298	use Encode;	308	use Encode;
299	$jsontext = encode "UTF-16BE", JSON::XS->new->encode ($object);	309	$jsontext = encode "UTF-16BE", JSON::XS->new->encode ($object);
…		…
422	If C<$enable> is true (or missing), then the C<encode> method will output JSON objects	432	If C<$enable> is true (or missing), then the C<encode> method will output JSON objects
423	by sorting their keys. This is adding a comparatively high overhead.	433	by sorting their keys. This is adding a comparatively high overhead.
424		434
425	If C<$enable> is false, then the C<encode> method will output key-value	435	If C<$enable> is false, then the C<encode> method will output key-value
426	pairs in the order Perl stores them (which will likely change between runs	436	pairs in the order Perl stores them (which will likely change between runs
427	of the same script).	437	of the same script, and can change even within the same run from 5.18
		438	onwards).
428		439
429	This option is useful if you want the same data structure to be encoded as	440	This option is useful if you want the same data structure to be encoded as
430	the same JSON text (given the same overall settings). If it is disabled,	441	the same JSON text (given the same overall settings). If it is disabled,
431	the same hash might be encoded differently even if contains the same data,	442	the same hash might be encoded differently even if contains the same data,
432	as key-value pairs have no inherent ordering in Perl.	443	as key-value pairs have no inherent ordering in Perl.
433		444
434	This setting has no effect when decoding JSON texts.	445	This setting has no effect when decoding JSON texts.
435		446
		447	This setting has currently no effect on tied hashes.
		448
436	=item $json = $json->allow_nonref ([$enable])	449	=item $json = $json->allow_nonref ([$enable])
437		450
438	=item $enabled = $json->get_allow_nonref	451	=item $enabled = $json->get_allow_nonref
439		452
440	If C<$enable> is true (or missing), then the C<encode> method can convert a	453	If C<$enable> is true (or missing), then the C<encode> method can convert a
…		…
450	Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>,	463	Example, encode a Perl scalar as JSON value with enabled C<allow_nonref>,
451	resulting in an invalid JSON text:	464	resulting in an invalid JSON text:
452		465
453	JSON::XS->new->allow_nonref->encode ("Hello, World!")	466	JSON::XS->new->allow_nonref->encode ("Hello, World!")
454	=> "Hello, World!"	467	=> "Hello, World!"
		468
		469	=item $json = $json->allow_unknown ([$enable])
		470
		471	=item $enabled = $json->get_allow_unknown
		472
		473	If C<$enable> is true (or missing), then C<encode> will I<not> throw an
		474	exception when it encounters values it cannot represent in JSON (for
		475	example, filehandles) but instead will encode a JSON C<null> value. Note
		476	that blessed objects are not included here and are handled separately by
		477	c<allow_nonref>.
		478
		479	If C<$enable> is false (the default), then C<encode> will throw an
		480	exception when it encounters anything it cannot encode as JSON.
		481
		482	This option does not affect C<decode> in any way, and it is recommended to
		483	leave it off unless you know your communications partner.
455		484
456	=item $json = $json->allow_blessed ([$enable])	485	=item $json = $json->allow_blessed ([$enable])
457		486
458	=item $enabled = $json->get_allow_blessed	487	=item $enabled = $json->get_allow_blessed
459		488
…		…
600	=item $json = $json->max_depth ([$maximum_nesting_depth])	629	=item $json = $json->max_depth ([$maximum_nesting_depth])
601		630
602	=item $max_depth = $json->get_max_depth	631	=item $max_depth = $json->get_max_depth
603		632
604	Sets the maximum nesting level (default C<512>) accepted while encoding	633	Sets the maximum nesting level (default C<512>) accepted while encoding
605	or decoding. If the JSON text or Perl data structure has an equal or	634	or decoding. If a higher nesting level is detected in JSON text or a Perl
606	higher nesting level then this limit, then the encoder and decoder will	635	data structure, then the encoder and decoder will stop and croak at that
607	stop and croak at that point.	636	point.
608		637
609	Nesting level is defined by number of hash- or arrayrefs that the encoder	638	Nesting level is defined by number of hash- or arrayrefs that the encoder
610	needs to traverse to reach a given point or the number of C<{> or C<[>	639	needs to traverse to reach a given point or the number of C<{> or C<[>
611	characters without their matching closing parenthesis crossed to reach a	640	characters without their matching closing parenthesis crossed to reach a
612	given character in a string.	641	given character in a string.
613		642
614	Setting the maximum depth to one disallows any nesting, so that ensures	643	Setting the maximum depth to one disallows any nesting, so that ensures
615	that the object is only a single hash/object or array.	644	that the object is only a single hash/object or array.
616		645
617	The argument to C<max_depth> will be rounded up to the next highest power
618	of two. If no argument is given, the highest possible setting will be	646	If no argument is given, the highest possible setting will be used, which
619	used, which is rarely useful.	647	is rarely useful.
		648
		649	Note that nesting is implemented by recursion in C. The default value has
		650	been chosen to be as large as typical operating systems allow without
		651	crashing.
620		652
621	See SECURITY CONSIDERATIONS, below, for more info on why this is useful.	653	See SECURITY CONSIDERATIONS, below, for more info on why this is useful.
622		654
623	=item $json = $json->max_size ([$maximum_string_size])	655	=item $json = $json->max_size ([$maximum_string_size])
624		656
625	=item $max_size = $json->get_max_size	657	=item $max_size = $json->get_max_size
626		658
627	Set the maximum length a JSON text may have (in bytes) where decoding is	659	Set the maximum length a JSON text may have (in bytes) where decoding is
628	being attempted. The default is C<0>, meaning no limit. When C<decode>	660	being attempted. The default is C<0>, meaning no limit. When C<decode>
629	is called on a string longer then this number of characters it will not	661	is called on a string that is longer then this many bytes, it will not
630	attempt to decode the string but throw an exception. This setting has no	662	attempt to decode the string but throw an exception. This setting has no
631	effect on C<encode> (yet).	663	effect on C<encode> (yet).
632		664
633	The argument to C<max_size> will be rounded up to the next B<highest>	665	If no argument is given, the limit check will be deactivated (same as when
634	power of two (so may be more than requested). If no argument is given, the	666	C<0> is specified).
635	limit check will be deactivated (same as when C<0> is specified).
636		667
637	See SECURITY CONSIDERATIONS, below, for more info on why this is useful.	668	See SECURITY CONSIDERATIONS, below, for more info on why this is useful.
638		669
639	=item $json_text = $json->encode ($perl_scalar)	670	=item $json_text = $json->encode ($perl_scalar)
640		671
…		…
669	=> ([], 3)	700	=> ([], 3)
670		701
671	=back	702	=back
672		703
673		704
		705	=head1 INCREMENTAL PARSING
		706
		707	In some cases, there is the need for incremental parsing of JSON
		708	texts. While this module always has to keep both JSON text and resulting
		709	Perl data structure in memory at one time, it does allow you to parse a
		710	JSON stream incrementally. It does so by accumulating text until it has
		711	a full JSON object, which it then can decode. This process is similar to
		712	using C<decode_prefix> to see if a full JSON object is available, but
		713	is much more efficient (and can be implemented with a minimum of method
		714	calls).
		715
		716	JSON::XS will only attempt to parse the JSON text once it is sure it
		717	has enough text to get a decisive result, using a very simple but
		718	truly incremental parser. This means that it sometimes won't stop as
		719	early as the full parser, for example, it doesn't detect mismatched
		720	parentheses. The only thing it guarantees is that it starts decoding as
		721	soon as a syntactically valid JSON text has been seen. This means you need
		722	to set resource limits (e.g. C<max_size>) to ensure the parser will stop
		723	parsing in the presence if syntax errors.
		724
		725	The following methods implement this incremental parser.
		726
		727	=over 4
		728
		729	=item [void, scalar or list context] = $json->incr_parse ([$string])
		730
		731	This is the central parsing function. It can both append new text and
		732	extract objects from the stream accumulated so far (both of these
		733	functions are optional).
		734
		735	If C<$string> is given, then this string is appended to the already
		736	existing JSON fragment stored in the C<$json> object.
		737
		738	After that, if the function is called in void context, it will simply
		739	return without doing anything further. This can be used to add more text
		740	in as many chunks as you want.
		741
		742	If the method is called in scalar context, then it will try to extract
		743	exactly I<one> JSON object. If that is successful, it will return this
		744	object, otherwise it will return C<undef>. If there is a parse error,
		745	this method will croak just as C<decode> would do (one can then use
		746	C<incr_skip> to skip the errornous part). This is the most common way of
		747	using the method.
		748
		749	And finally, in list context, it will try to extract as many objects
		750	from the stream as it can find and return them, or the empty list
		751	otherwise. For this to work, there must be no separators between the JSON
		752	objects or arrays, instead they must be concatenated back-to-back. If
		753	an error occurs, an exception will be raised as in the scalar context
		754	case. Note that in this case, any previously-parsed JSON texts will be
		755	lost.
		756
		757	Example: Parse some JSON arrays/objects in a given string and return
		758	them.
		759
		760	my @objs = JSON::XS->new->incr_parse ("[5][7][1,2]");
		761
		762	=item $lvalue_string = $json->incr_text
		763
		764	This method returns the currently stored JSON fragment as an lvalue, that
		765	is, you can manipulate it. This I<only> works when a preceding call to
		766	C<incr_parse> in I<scalar context> successfully returned an object. Under
		767	all other circumstances you must not call this function (I mean it.
		768	although in simple tests it might actually work, it I<will> fail under
		769	real world conditions). As a special exception, you can also call this
		770	method before having parsed anything.
		771
		772	This function is useful in two cases: a) finding the trailing text after a
		773	JSON object or b) parsing multiple JSON objects separated by non-JSON text
		774	(such as commas).
		775
		776	=item $json->incr_skip
		777
		778	This will reset the state of the incremental parser and will remove
		779	the parsed text from the input buffer so far. This is useful after
		780	C<incr_parse> died, in which case the input buffer and incremental parser
		781	state is left unchanged, to skip the text parsed so far and to reset the
		782	parse state.
		783
		784	The difference to C<incr_reset> is that only text until the parse error
		785	occured is removed.
		786
		787	=item $json->incr_reset
		788
		789	This completely resets the incremental parser, that is, after this call,
		790	it will be as if the parser had never parsed anything.
		791
		792	This is useful if you want to repeatedly parse JSON objects and want to
		793	ignore any trailing data, which means you have to reset the parser after
		794	each successful decode.
		795
		796	=back
		797
		798	=head2 LIMITATIONS
		799
		800	All options that affect decoding are supported, except
		801	C<allow_nonref>. The reason for this is that it cannot be made to
		802	work sensibly: JSON objects and arrays are self-delimited, i.e. you can concatenate
		803	them back to back and still decode them perfectly. This does not hold true
		804	for JSON numbers, however.
		805
		806	For example, is the string C<1> a single JSON number, or is it simply the
		807	start of C<12>? Or is C<12> a single JSON number, or the concatenation
		808	of C<1> and C<2>? In neither case you can tell, and this is why JSON::XS
		809	takes the conservative route and disallows this case.
		810
		811	=head2 EXAMPLES
		812
		813	Some examples will make all this clearer. First, a simple example that
		814	works similarly to C<decode_prefix>: We want to decode the JSON object at
		815	the start of a string and identify the portion after the JSON object:
		816
		817	my $text = "[1,2,3] hello";
		818
		819	my $json = new JSON::XS;
		820
		821	my $obj = $json->incr_parse ($text)
		822	or die "expected JSON object or array at beginning of string";
		823
		824	my $tail = $json->incr_text;
		825	# $tail now contains " hello"
		826
		827	Easy, isn't it?
		828
		829	Now for a more complicated example: Imagine a hypothetical protocol where
		830	you read some requests from a TCP stream, and each request is a JSON
		831	array, without any separation between them (in fact, it is often useful to
		832	use newlines as "separators", as these get interpreted as whitespace at
		833	the start of the JSON text, which makes it possible to test said protocol
		834	with C<telnet>...).
		835
		836	Here is how you'd do it (it is trivial to write this in an event-based
		837	manner):
		838
		839	my $json = new JSON::XS;
		840
		841	# read some data from the socket
		842	while (sysread $socket, my $buf, 4096) {
		843
		844	# split and decode as many requests as possible
		845	for my $request ($json->incr_parse ($buf)) {
		846	# act on the $request
		847	}
		848	}
		849
		850	Another complicated example: Assume you have a string with JSON objects
		851	or arrays, all separated by (optional) comma characters (e.g. C<[1],[2],
		852	[3]>). To parse them, we have to skip the commas between the JSON texts,
		853	and here is where the lvalue-ness of C<incr_text> comes in useful:
		854
		855	my $text = "[1],[2], [3]";
		856	my $json = new JSON::XS;
		857
		858	# void context, so no parsing done
		859	$json->incr_parse ($text);
		860
		861	# now extract as many objects as possible. note the
		862	# use of scalar context so incr_text can be called.
		863	while (my $obj = $json->incr_parse) {
		864	# do something with $obj
		865
		866	# now skip the optional comma
		867	$json->incr_text =~ s/^ \s* , //x;
		868	}
		869
		870	Now lets go for a very complex example: Assume that you have a gigantic
		871	JSON array-of-objects, many gigabytes in size, and you want to parse it,
		872	but you cannot load it into memory fully (this has actually happened in
		873	the real world :).
		874
		875	Well, you lost, you have to implement your own JSON parser. But JSON::XS
		876	can still help you: You implement a (very simple) array parser and let
		877	JSON decode the array elements, which are all full JSON objects on their
		878	own (this wouldn't work if the array elements could be JSON numbers, for
		879	example):
		880
		881	my $json = new JSON::XS;
		882
		883	# open the monster
		884	open my $fh, "<bigfile.json"
		885	or die "bigfile: $!";
		886
		887	# first parse the initial "["
		888	for (;;) {
		889	sysread $fh, my $buf, 65536
		890	or die "read error: $!";
		891	$json->incr_parse ($buf); # void context, so no parsing
		892
		893	# Exit the loop once we found and removed(!) the initial "[".
		894	# In essence, we are (ab-)using the $json object as a simple scalar
		895	# we append data to.
		896	last if $json->incr_text =~ s/^ \s* \[ //x;
		897	}
		898
		899	# now we have the skipped the initial "[", so continue
		900	# parsing all the elements.
		901	for (;;) {
		902	# in this loop we read data until we got a single JSON object
		903	for (;;) {
		904	if (my $obj = $json->incr_parse) {
		905	# do something with $obj
		906	last;
		907	}
		908
		909	# add more data
		910	sysread $fh, my $buf, 65536
		911	or die "read error: $!";
		912	$json->incr_parse ($buf); # void context, so no parsing
		913	}
		914
		915	# in this loop we read data until we either found and parsed the
		916	# separating "," between elements, or the final "]"
		917	for (;;) {
		918	# first skip whitespace
		919	$json->incr_text =~ s/^\s*//;
		920
		921	# if we find "]", we are done
		922	if ($json->incr_text =~ s/^\]//) {
		923	print "finished.\n";
		924	exit;
		925	}
		926
		927	# if we find ",", we can continue with the next element
		928	if ($json->incr_text =~ s/^,//) {
		929	last;
		930	}
		931
		932	# if we find anything else, we have a parse error!
		933	if (length $json->incr_text) {
		934	die "parse error near ", $json->incr_text;
		935	}
		936
		937	# else add more data
		938	sysread $fh, my $buf, 65536
		939	or die "read error: $!";
		940	$json->incr_parse ($buf); # void context, so no parsing
		941	}
		942
		943	This is a complex example, but most of the complexity comes from the fact
		944	that we are trying to be correct (bear with me if I am wrong, I never ran
		945	the above example :).
		946
		947
		948
674	=head1 MAPPING	949	=head1 MAPPING
675		950
676	This section describes how JSON::XS maps Perl values to JSON values and	951	This section describes how JSON::XS maps Perl values to JSON values and
677	vice versa. These mappings are designed to "do the right thing" in most	952	vice versa. These mappings are designed to "do the right thing" in most
678	circumstances automatically, preserving round-tripping characteristics	953	circumstances automatically, preserving round-tripping characteristics
…		…
706		981
707	A JSON number becomes either an integer, numeric (floating point) or	982	A JSON number becomes either an integer, numeric (floating point) or
708	string scalar in perl, depending on its range and any fractional parts. On	983	string scalar in perl, depending on its range and any fractional parts. On
709	the Perl level, there is no difference between those as Perl handles all	984	the Perl level, there is no difference between those as Perl handles all
710	the conversion details, but an integer may take slightly less memory and	985	the conversion details, but an integer may take slightly less memory and
711	might represent more values exactly than (floating point) numbers.	986	might represent more values exactly than floating point numbers.
712		987
713	If the number consists of digits only, JSON::XS will try to represent	988	If the number consists of digits only, JSON::XS will try to represent
714	it as an integer value. If that fails, it will try to represent it as	989	it as an integer value. If that fails, it will try to represent it as
715	a numeric (floating point) value if that is possible without loss of	990	a numeric (floating point) value if that is possible without loss of
716	precision. Otherwise it will preserve the number as a string value.	991	precision. Otherwise it will preserve the number as a string value (in
		992	which case you lose roundtripping ability, as the JSON number will be
		993	re-encoded toa JSON string).
717		994
718	Numbers containing a fractional or exponential part will always be	995	Numbers containing a fractional or exponential part will always be
719	represented as numeric (floating point) values, possibly at a loss of	996	represented as numeric (floating point) values, possibly at a loss of
720	precision.	997	precision (in which case you might lose perfect roundtripping ability, but
		998	the JSON number will still be re-encoded as a JSON number).
721		999
722	This might create round-tripping problems as numbers might become strings,	1000	Note that precision is not accuracy - binary floating point values cannot
723	but as Perl is typeless there is no other way to do it.	1001	represent most decimal fractions exactly, and when converting from and to
		1002	floating point, JSON::XS only guarantees precision up to but not including
		1003	the leats significant bit.
724		1004
725	=item true, false	1005	=item true, false
726		1006
727	These JSON atoms become C<JSON::XS::true> and C<JSON::XS::false>,	1007	These JSON atoms become C<JSON::XS::true> and C<JSON::XS::false>,
728	respectively. They are overloaded to act almost exactly like the numbers	1008	respectively. They are overloaded to act almost exactly like the numbers
…		…
765	Other unblessed references are generally not allowed and will cause an	1045	Other unblessed references are generally not allowed and will cause an
766	exception to be thrown, except for references to the integers C<0> and	1046	exception to be thrown, except for references to the integers C<0> and
767	C<1>, which get turned into C<false> and C<true> atoms in JSON. You can	1047	C<1>, which get turned into C<false> and C<true> atoms in JSON. You can
768	also use C<JSON::XS::false> and C<JSON::XS::true> to improve readability.	1048	also use C<JSON::XS::false> and C<JSON::XS::true> to improve readability.
769		1049
770	encode_json [\0,JSON::XS::true] # yields [false,true]	1050	encode_json [\0, JSON::XS::true] # yields [false,true]
771		1051
772	=item JSON::XS::true, JSON::XS::false	1052	=item JSON::XS::true, JSON::XS::false
773		1053
774	These special values become JSON true and JSON false values,	1054	These special values become JSON true and JSON false values,
775	respectively. You can also use C<\1> and C<\0> directly if you want.	1055	respectively. You can also use C<\1> and C<\0> directly if you want.
776		1056
777	=item blessed objects	1057	=item blessed objects
778		1058
779	Blessed objects are not allowed. JSON::XS currently tries to encode their	1059	Blessed objects are not directly representable in JSON. See the
780	underlying representation (hash- or arrayref), but this behaviour might	1060	C<allow_blessed> and C<convert_blessed> methods on various options on
781	change in future versions.	1061	how to deal with this: basically, you can choose between throwing an
		1062	exception, encoding the reference as if it weren't blessed, or provide
		1063	your own serialiser method.
782		1064
783	=item simple scalars	1065	=item simple scalars
784		1066
785	Simple Perl scalars (any scalar that is not a reference) are the most	1067	Simple Perl scalars (any scalar that is not a reference) are the most
786	difficult objects to encode: JSON::XS will encode undefined scalars as	1068	difficult objects to encode: JSON::XS will encode undefined scalars as
787	JSON null value, scalars that have last been used in a string context	1069	JSON C<null> values, scalars that have last been used in a string context
788	before encoding as JSON strings and anything else as number value:	1070	before encoding as JSON strings, and anything else as number value:
789		1071
790	# dump as number	1072	# dump as number
791	encode_json [2] # yields [2]	1073	encode_json [2] # yields [2]
792	encode_json [-3.0e17] # yields [-3e+17]	1074	encode_json [-3.0e17] # yields [-3e+17]
793	my $value = 5; encode_json [$value] # yields [5]	1075	my $value = 5; encode_json [$value] # yields [5]
…		…
811	my $x = "3"; # some variable containing a string	1093	my $x = "3"; # some variable containing a string
812	$x += 0; # numify it, ensuring it will be dumped as a number	1094	$x += 0; # numify it, ensuring it will be dumped as a number
813	$x *= 1; # same thing, the choice is yours.	1095	$x *= 1; # same thing, the choice is yours.
814		1096
815	You can not currently force the type in other, less obscure, ways. Tell me	1097	You can not currently force the type in other, less obscure, ways. Tell me
816	if you need this capability.	1098	if you need this capability (but don't forget to explain why it's needed
		1099	:).
		1100
		1101	Note that numerical precision has the same meaning as under Perl (so
		1102	binary to decimal conversion follows the same rules as in Perl, which
		1103	can differ to other languages). Also, your perl interpreter might expose
		1104	extensions to the floating point numbers of your platform, such as
		1105	infinities or NaN's - these cannot be represented in JSON, and it is an
		1106	error to pass those in.
817		1107
818	=back	1108	=back
819		1109
820		1110
821	=head1 COMPARISON	1111	=head1 ENCODING/CODESET FLAG NOTES
822		1112
823	As already mentioned, this module was created because none of the existing	1113	The interested reader might have seen a number of flags that signify
824	JSON modules could be made to work correctly. First I will describe the	1114	encodings or codesets - C<utf8>, C<latin1> and C<ascii>. There seems to be
825	problems (or pleasures) I encountered with various existing JSON modules,	1115	some confusion on what these do, so here is a short comparison:
826	followed by some benchmark values. JSON::XS was designed not to suffer	1116
827	from any of these problems or limitations.	1117	C<utf8> controls whether the JSON text created by C<encode> (and expected
		1118	by C<decode>) is UTF-8 encoded or not, while C<latin1> and C<ascii> only
		1119	control whether C<encode> escapes character values outside their respective
		1120	codeset range. Neither of these flags conflict with each other, although
		1121	some combinations make less sense than others.
		1122
		1123	Care has been taken to make all flags symmetrical with respect to
		1124	C<encode> and C<decode>, that is, texts encoded with any combination of
		1125	these flag values will be correctly decoded when the same flags are used
		1126	- in general, if you use different flag settings while encoding vs. when
		1127	decoding you likely have a bug somewhere.
		1128
		1129	Below comes a verbose discussion of these flags. Note that a "codeset" is
		1130	simply an abstract set of character-codepoint pairs, while an encoding
		1131	takes those codepoint numbers and I<encodes> them, in our case into
		1132	octets. Unicode is (among other things) a codeset, UTF-8 is an encoding,
		1133	and ISO-8859-1 (= latin 1) and ASCII are both codesets I<and> encodings at
		1134	the same time, which can be confusing.
828		1135
829	=over 4	1136	=over 4
830		1137
831	=item JSON 1.07	1138	=item C<utf8> flag disabled
832		1139
833	Slow (but very portable, as it is written in pure Perl).	1140	When C<utf8> is disabled (the default), then C<encode>/C<decode> generate
		1141	and expect Unicode strings, that is, characters with high ordinal Unicode
		1142	values (> 255) will be encoded as such characters, and likewise such
		1143	characters are decoded as-is, no canges to them will be done, except
		1144	"(re-)interpreting" them as Unicode codepoints or Unicode characters,
		1145	respectively (to Perl, these are the same thing in strings unless you do
		1146	funny/weird/dumb stuff).
834		1147
835	Undocumented/buggy Unicode handling (how JSON handles Unicode values is	1148	This is useful when you want to do the encoding yourself (e.g. when you
836	undocumented. One can get far by feeding it Unicode strings and doing	1149	want to have UTF-16 encoded JSON texts) or when some other layer does
837	en-/decoding oneself, but Unicode escapes are not working properly).	1150	the encoding for you (for example, when printing to a terminal using a
		1151	filehandle that transparently encodes to UTF-8 you certainly do NOT want
		1152	to UTF-8 encode your data first and have Perl encode it another time).
838		1153
839	No round-tripping (strings get clobbered if they look like numbers, e.g.	1154	=item C<utf8> flag enabled
840	the string C<2.0> will encode to C<2.0> instead of C<"2.0">, and that will
841	decode into the number 2.
842		1155
843	=item JSON::PC 0.01	1156	If the C<utf8>-flag is enabled, C<encode>/C<decode> will encode all
		1157	characters using the corresponding UTF-8 multi-byte sequence, and will
		1158	expect your input strings to be encoded as UTF-8, that is, no "character"
		1159	of the input string must have any value > 255, as UTF-8 does not allow
		1160	that.
844		1161
845	Very fast.	1162	The C<utf8> flag therefore switches between two modes: disabled means you
		1163	will get a Unicode string in Perl, enabled means you get an UTF-8 encoded
		1164	octet/binary string in Perl.
846		1165
847	Undocumented/buggy Unicode handling.	1166	=item C<latin1> or C<ascii> flags enabled
848		1167
849	No round-tripping.	1168	With C<latin1> (or C<ascii>) enabled, C<encode> will escape characters
		1169	with ordinal values > 255 (> 127 with C<ascii>) and encode the remaining
		1170	characters as specified by the C<utf8> flag.
850		1171
851	Has problems handling many Perl values (e.g. regex results and other magic	1172	If C<utf8> is disabled, then the result is also correctly encoded in those
852	values will make it croak).	1173	character sets (as both are proper subsets of Unicode, meaning that a
		1174	Unicode string with all character values < 256 is the same thing as a
		1175	ISO-8859-1 string, and a Unicode string with all character values < 128 is
		1176	the same thing as an ASCII string in Perl).
853		1177
854	Does not even generate valid JSON (C<{1,2}> gets converted to C<{1:2}>	1178	If C<utf8> is enabled, you still get a correct UTF-8-encoded string,
855	which is not a valid JSON text.	1179	regardless of these flags, just some more characters will be escaped using
		1180	C<\uXXXX> then before.
856		1181
857	Unmaintained (maintainer unresponsive for many months, bugs are not	1182	Note that ISO-8859-1-I<encoded> strings are not compatible with UTF-8
858	getting fixed).	1183	encoding, while ASCII-encoded strings are. That is because the ISO-8859-1
		1184	encoding is NOT a subset of UTF-8 (despite the ISO-8859-1 I<codeset> being
		1185	a subset of Unicode), while ASCII is.
859		1186
860	=item JSON::Syck 0.21	1187	Surprisingly, C<decode> will ignore these flags and so treat all input
		1188	values as governed by the C<utf8> flag. If it is disabled, this allows you
		1189	to decode ISO-8859-1- and ASCII-encoded strings, as both strict subsets of
		1190	Unicode. If it is enabled, you can correctly decode UTF-8 encoded strings.
861		1191
862	Very buggy (often crashes).	1192	So neither C<latin1> nor C<ascii> are incompatible with the C<utf8> flag -
		1193	they only govern when the JSON output engine escapes a character or not.
863		1194
864	Very inflexible (no human-readable format supported, format pretty much	1195	The main use for C<latin1> is to relatively efficiently store binary data
865	undocumented. I need at least a format for easy reading by humans and a	1196	as JSON, at the expense of breaking compatibility with most JSON decoders.
866	single-line compact format for use in a protocol, and preferably a way to
867	generate ASCII-only JSON texts).
868		1197
869	Completely broken (and confusingly documented) Unicode handling (Unicode	1198	The main use for C<ascii> is to force the output to not contain characters
870	escapes are not working properly, you need to set ImplicitUnicode to	1199	with values > 127, which means you can interpret the resulting string
871	I<different> values on en- and decoding to get symmetric behaviour).	1200	as UTF-8, ISO-8859-1, ASCII, KOI8-R or most about any character set and
872		1201	8-bit-encoding, and still get the same data structure back. This is useful
873	No round-tripping (simple cases work, but this depends on whether the scalar	1202	when your channel for JSON transfer is not 8-bit clean or the encoding
874	value was used in a numeric context or not).	1203	might be mangled in between (e.g. in mail), and works because ASCII is a
875		1204	proper subset of most 8-bit and multibyte encodings in use in the world.
876	Dumping hashes may skip hash values depending on iterator state.
877
878	Unmaintained (maintainer unresponsive for many months, bugs are not
879	getting fixed).
880
881	Does not check input for validity (i.e. will accept non-JSON input and
882	return "something" instead of raising an exception. This is a security
883	issue: imagine two banks transferring money between each other using
884	JSON. One bank might parse a given non-JSON request and deduct money,
885	while the other might reject the transaction with a syntax error. While a
886	good protocol will at least recover, that is extra unnecessary work and
887	the transaction will still not succeed).
888
889	=item JSON::DWIW 0.04
890
891	Very fast. Very natural. Very nice.
892
893	Undocumented Unicode handling (but the best of the pack. Unicode escapes
894	still don't get parsed properly).
895
896	Very inflexible.
897
898	No round-tripping.
899
900	Does not generate valid JSON texts (key strings are often unquoted, empty keys
901	result in nothing being output)
902
903	Does not check input for validity.
904		1205
905	=back	1206	=back
906		1207
907		1208
		1209	=head2 JSON and ECMAscript
		1210
		1211	JSON syntax is based on how literals are represented in javascript (the
		1212	not-standardised predecessor of ECMAscript) which is presumably why it is
		1213	called "JavaScript Object Notation".
		1214
		1215	However, JSON is not a subset (and also not a superset of course) of
		1216	ECMAscript (the standard) or javascript (whatever browsers actually
		1217	implement).
		1218
		1219	If you want to use javascript's C<eval> function to "parse" JSON, you
		1220	might run into parse errors for valid JSON texts, or the resulting data
		1221	structure might not be queryable:
		1222
		1223	One of the problems is that U+2028 and U+2029 are valid characters inside
		1224	JSON strings, but are not allowed in ECMAscript string literals, so the
		1225	following Perl fragment will not output something that can be guaranteed
		1226	to be parsable by javascript's C<eval>:
		1227
		1228	use JSON::XS;
		1229
		1230	print encode_json [chr 0x2028];
		1231
		1232	The right fix for this is to use a proper JSON parser in your javascript
		1233	programs, and not rely on C<eval> (see for example Douglas Crockford's
		1234	F<json2.js> parser).
		1235
		1236	If this is not an option, you can, as a stop-gap measure, simply encode to
		1237	ASCII-only JSON:
		1238
		1239	use JSON::XS;
		1240
		1241	print JSON::XS->new->ascii->encode ([chr 0x2028]);
		1242
		1243	Note that this will enlarge the resulting JSON text quite a bit if you
		1244	have many non-ASCII characters. You might be tempted to run some regexes
		1245	to only escape U+2028 and U+2029, e.g.:
		1246
		1247	# DO NOT USE THIS!
		1248	my $json = JSON::XS->new->utf8->encode ([chr 0x2028]);
		1249	$json =~ s/\xe2\x80\xa8/\\u2028/g; # escape U+2028
		1250	$json =~ s/\xe2\x80\xa9/\\u2029/g; # escape U+2029
		1251	print $json;
		1252
		1253	Note that I<this is a bad idea>: the above only works for U+2028 and
		1254	U+2029 and thus only for fully ECMAscript-compliant parsers. Many existing
		1255	javascript implementations, however, have issues with other characters as
		1256	well - using C<eval> naively simply I<will> cause problems.
		1257
		1258	Another problem is that some javascript implementations reserve
		1259	some property names for their own purposes (which probably makes
		1260	them non-ECMAscript-compliant). For example, Iceweasel reserves the
		1261	C<__proto__> property name for its own purposes.
		1262
		1263	If that is a problem, you could parse try to filter the resulting JSON
		1264	output for these property strings, e.g.:
		1265
		1266	$json =~ s/"__proto__"\s*:/"__proto__renamed":/g;
		1267
		1268	This works because C<__proto__> is not valid outside of strings, so every
		1269	occurence of C<"__proto__"\s*:> must be a string used as property name.
		1270
		1271	If you know of other incompatibilities, please let me know.
		1272
		1273
908	=head2 JSON and YAML	1274	=head2 JSON and YAML
909		1275
910	You often hear that JSON is a subset of YAML. This is, however, a mass	1276	You often hear that JSON is a subset of YAML. This is, however, a mass
911	hysteria(*) and very far from the truth. In general, there is no way to	1277	hysteria(*) and very far from the truth (as of the time of this writing),
		1278	so let me state it clearly: I<in general, there is no way to configure
912	configure JSON::XS to output a data structure as valid YAML that works for	1279	JSON::XS to output a data structure as valid YAML> that works in all
913	all cases.	1280	cases.
914		1281
915	If you really must use JSON::XS to generate YAML, you should use this	1282	If you really must use JSON::XS to generate YAML, you should use this
916	algorithm (subject to change in future versions):	1283	algorithm (subject to change in future versions):
917		1284
918	my $to_yaml = JSON::XS->new->utf8->space_after (1);	1285	my $to_yaml = JSON::XS->new->utf8->space_after (1);
919	my $yaml = $to_yaml->encode ($ref) . "\n";	1286	my $yaml = $to_yaml->encode ($ref) . "\n";
920		1287
921	This will usually generate JSON texts that also parse as valid	1288	This will I<usually> generate JSON texts that also parse as valid
922	YAML. Please note that YAML has hardcoded limits on (simple) object key	1289	YAML. Please note that YAML has hardcoded limits on (simple) object key
923	lengths that JSON doesn't have and also has different and incompatible	1290	lengths that JSON doesn't have and also has different and incompatible
924	unicode handling, so you should make sure that your hash keys are	1291	unicode character escape syntax, so you should make sure that your hash
925	noticeably shorter than the 1024 "stream characters" YAML allows and that	1292	keys are noticeably shorter than the 1024 "stream characters" YAML allows
926	you do not have codepoints with values outside the Unicode BMP (basic	1293	and that you do not have characters with codepoint values outside the
927	multilingual page). YAML also does not allow C<\/> sequences in strings	1294	Unicode BMP (basic multilingual page). YAML also does not allow C<\/>
928	(which JSON::XS does not I<currently> generate).	1295	sequences in strings (which JSON::XS does not I<currently> generate, but
		1296	other JSON generators might).
929		1297
930	There might be other incompatibilities that I am not aware of. In general	1298	There might be other incompatibilities that I am not aware of (or the YAML
		1299	specification has been changed yet again - it does so quite often). In
931	you should not try to generate YAML with a JSON generator or vice versa,	1300	general you should not try to generate YAML with a JSON generator or vice
932	or try to parse JSON with a YAML parser or vice versa: chances are high	1301	versa, or try to parse JSON with a YAML parser or vice versa: chances are
933	that you will run into severe interoperability problems when you least	1302	high that you will run into severe interoperability problems when you
934	expect it.	1303	least expect it.
935		1304
936	=over 4	1305	=over 4
937		1306
938	=item (*)	1307	=item (*)
939		1308
940	This is spread actively by the YAML team, however. For many years now they	1309	I have been pressured multiple times by Brian Ingerson (one of the
941	claim YAML were a superset of JSON, even when proven otherwise.	1310	authors of the YAML specification) to remove this paragraph, despite him
		1311	acknowledging that the actual incompatibilities exist. As I was personally
		1312	bitten by this "JSON is YAML" lie, I refused and said I will continue to
		1313	educate people about these issues, so others do not run into the same
		1314	problem again and again. After this, Brian called me a (quote)I<complete
		1315	and worthless idiot>(unquote).
942		1316
943	Even the author of this manpage was at some point accused of providing	1317	In my opinion, instead of pressuring and insulting people who actually
944	"incorrect" information, despite the evidence presented (claims ranged	1318	clarify issues with YAML and the wrong statements of some of its
945	from "your documentation contains inaccurate and negative statements about	1319	proponents, I would kindly suggest reading the JSON spec (which is not
946	YAML" (the only negative comment is this footnote, and it didn't exist	1320	that difficult or long) and finally make YAML compatible to it, and
947	back then; the question on which claims were inaccurate was never answered	1321	educating users about the changes, instead of spreading lies about the
948	etc.) to "the YAML spec is not up-to-date" (the real and suppsedly	1322	real compatibility for many I<years> and trying to silence people who
949	JSON-compatible spec is apparently not currently publicly available)	1323	point out that it isn't true.
950	to actual requests to replace this section by incorrect information,
951	suppressing information about the real problem).
952		1324
953	So whenever you are told that YAML was a superset of JSON, first check	1325	Addendum/2009: the YAML 1.2 spec is still incompatible with JSON, even
954	wether it is really true (it might be when you check it, but it cetrainly	1326	though the incompatibilities have been documented (and are known to Brian)
955	is not true when this was written). I would much prefer if the YAML team	1327	for many years and the spec makes explicit claims that YAML is a superset
956	would spent their time on actually making JSON compatibility a truth	1328	of JSON. It would be so easy to fix, but apparently, bullying people and
957	(JSON, after all, has a very small and simple specification) instead of	1329	corrupting userdata is so much easier.
958	trying to lobby/force people into reporting untruths.
959		1330
960	=back	1331	=back
961		1332
962		1333
963	=head2 SPEED	1334	=head2 SPEED
…		…
965	It seems that JSON::XS is surprisingly fast, as shown in the following	1336	It seems that JSON::XS is surprisingly fast, as shown in the following
966	tables. They have been generated with the help of the C<eg/bench> program	1337	tables. They have been generated with the help of the C<eg/bench> program
967	in the JSON::XS distribution, to make it easy to compare on your own	1338	in the JSON::XS distribution, to make it easy to compare on your own
968	system.	1339	system.
969		1340
970	First comes a comparison between various modules using a very short	1341	First comes a comparison between various modules using
971	single-line JSON string:	1342	a very short single-line JSON string (also available at
		1343	L<http://dist.schmorp.de/misc/json/short.json>).
972		1344
973	{"method": "handleMessage", "params": ["user1", "we were just talking"], \	1345	{"method": "handleMessage", "params": ["user1",
974	"id": null, "array":[1,11,234,-5,1e5,1e7, true, false]}	1346	"we were just talking"], "id": null, "array":[1,11,234,-5,1e5,1e7,
		1347	1, 0]}
975		1348
976	It shows the number of encodes/decodes per second (JSON::XS uses	1349	It shows the number of encodes/decodes per second (JSON::XS uses
977	the functional interface, while JSON::XS/2 uses the OO interface	1350	the functional interface, while JSON::XS/2 uses the OO interface
978	with pretty-printing and hashkey sorting enabled, JSON::XS/3 enables	1351	with pretty-printing and hashkey sorting enabled, JSON::XS/3 enables
979	shrink). Higher is better:	1352	shrink. JSON::DWIW/DS uses the deserialise function, while JSON::DWIW::FJ
		1353	uses the from_json method). Higher is better:
980		1354
981	module \| encode \| decode \|	1355	module \| encode \| decode \|
982	-----------\|------------\|------------\|	1356	--------------\|------------\|------------\|
983	JSON 1.x \| 4990.842 \| 4088.813 \|	1357	JSON::DWIW/DS \| 86302.551 \| 102300.098 \|
984	JSON::DWIW \| 51653.990 \| 71575.154 \|	1358	JSON::DWIW/FJ \| 86302.551 \| 75983.768 \|
985	JSON::PC \| 65948.176 \| 74631.744 \|	1359	JSON::PP \| 15827.562 \| 6638.658 \|
986	JSON::PP \| 8931.652 \| 3817.168 \|	1360	JSON::Syck \| 63358.066 \| 47662.545 \|
987	JSON::Syck \| 24877.248 \| 27776.848 \|	1361	JSON::XS \| 511500.488 \| 511500.488 \|
988	JSON::XS \| 388361.481 \| 227951.304 \|	1362	JSON::XS/2 \| 291271.111 \| 388361.481 \|
989	JSON::XS/2 \| 227951.304 \| 218453.333 \|	1363	JSON::XS/3 \| 361577.931 \| 361577.931 \|
990	JSON::XS/3 \| 338250.323 \| 218453.333 \|	1364	Storable \| 66788.280 \| 265462.278 \|
991	Storable \| 16500.016 \| 135300.129 \|
992	-----------+------------+------------+	1365	--------------+------------+------------+
993		1366
994	That is, JSON::XS is about five times faster than JSON::DWIW on encoding,	1367	That is, JSON::XS is almost six times faster than JSON::DWIW on encoding,
995	about three times faster on decoding, and over forty times faster	1368	about five times faster on decoding, and over thirty to seventy times
996	than JSON, even with pretty-printing and key sorting. It also compares	1369	faster than JSON's pure perl implementation. It also compares favourably
997	favourably to Storable for small amounts of data.	1370	to Storable for small amounts of data.
998		1371
999	Using a longer test string (roughly 18KB, generated from Yahoo! Locals	1372	Using a longer test string (roughly 18KB, generated from Yahoo! Locals
1000	search API (http://nanoref.com/yahooapis/mgPdGg):	1373	search API (L<http://dist.schmorp.de/misc/json/long.json>).
1001		1374
1002	module \| encode \| decode \|	1375	module \| encode \| decode \|
1003	-----------\|------------\|------------\|	1376	--------------\|------------\|------------\|
1004	JSON 1.x \| 55.260 \| 34.971 \|	1377	JSON::DWIW/DS \| 1647.927 \| 2673.916 \|
1005	JSON::DWIW \| 825.228 \| 1082.513 \|	1378	JSON::DWIW/FJ \| 1630.249 \| 2596.128 \|
1006	JSON::PC \| 3571.444 \| 2394.829 \|
1007	JSON::PP \| 210.987 \| 32.574 \|	1379	JSON::PP \| 400.640 \| 62.311 \|
1008	JSON::Syck \| 552.551 \| 787.544 \|	1380	JSON::Syck \| 1481.040 \| 1524.869 \|
1009	JSON::XS \| 5780.463 \| 4854.519 \|	1381	JSON::XS \| 20661.596 \| 9541.183 \|
1010	JSON::XS/2 \| 3869.998 \| 4798.975 \|	1382	JSON::XS/2 \| 10683.403 \| 9416.938 \|
1011	JSON::XS/3 \| 5862.880 \| 4798.975 \|	1383	JSON::XS/3 \| 20661.596 \| 9400.054 \|
1012	Storable \| 4445.002 \| 5235.027 \|	1384	Storable \| 19765.806 \| 10000.725 \|
1013	-----------+------------+------------+	1385	--------------+------------+------------+
1014		1386
1015	Again, JSON::XS leads by far (except for Storable which non-surprisingly	1387	Again, JSON::XS leads by far (except for Storable which non-surprisingly
1016	decodes faster).	1388	decodes a bit faster).
1017		1389
1018	On large strings containing lots of high Unicode characters, some modules	1390	On large strings containing lots of high Unicode characters, some modules
1019	(such as JSON::PC) seem to decode faster than JSON::XS, but the result	1391	(such as JSON::PC) seem to decode faster than JSON::XS, but the result
1020	will be broken due to missing (or wrong) Unicode handling. Others refuse	1392	will be broken due to missing (or wrong) Unicode handling. Others refuse
1021	to decode or encode properly, so it was impossible to prepare a fair	1393	to decode or encode properly, so it was impossible to prepare a fair
…		…
1047	to free the temporary). If that is exceeded, the program crashes. To be	1419	to free the temporary). If that is exceeded, the program crashes. To be
1048	conservative, the default nesting limit is set to 512. If your process	1420	conservative, the default nesting limit is set to 512. If your process
1049	has a smaller stack, you should adjust this setting accordingly with the	1421	has a smaller stack, you should adjust this setting accordingly with the
1050	C<max_depth> method.	1422	C<max_depth> method.
1051		1423
1052	And last but least, something else could bomb you that I forgot to think	1424	Something else could bomb you, too, that I forgot to think of. In that
1053	of. In that case, you get to keep the pieces. I am always open for hints,	1425	case, you get to keep the pieces. I am always open for hints, though...
1054	though...	1426
		1427	Also keep in mind that JSON::XS might leak contents of your Perl data
		1428	structures in its error messages, so when you serialise sensitive
		1429	information you might want to make sure that exceptions thrown by JSON::XS
		1430	will not end up in front of untrusted eyes.
1055		1431
1056	If you are using JSON::XS to return packets to consumption	1432	If you are using JSON::XS to return packets to consumption
1057	by JavaScript scripts in a browser you should have a look at	1433	by JavaScript scripts in a browser you should have a look at
1058	L<http://jpsykes.com/47/practical-csrf-and-json-security> to see whether	1434	L<http://blog.archive.jpsykes.com/47/practical-csrf-and-json-security/> to
1059	you are vulnerable to some common attack vectors (which really are browser	1435	see whether you are vulnerable to some common attack vectors (which really
1060	design bugs, but it is still you who will have to deal with it, as major	1436	are browser design bugs, but it is still you who will have to deal with
1061	browser developers care only for features, not about getting security	1437	it, as major browser developers care only for features, not about getting
1062	right).	1438	security right).
1063		1439
1064		1440
1065	=head1 THREADS	1441	=head1 THREADS
1066		1442
1067	This module is I<not> guaranteed to be thread safe and there are no	1443	This module is I<not> guaranteed to be thread safe and there are no
1068	plans to change this until Perl gets thread support (as opposed to the	1444	plans to change this until Perl gets thread support (as opposed to the
1069	horribly slow so-called "threads" which are simply slow and bloated	1445	horribly slow so-called "threads" which are simply slow and bloated
1070	process simulations - use fork, its I<much> faster, cheaper, better).	1446	process simulations - use fork, it's I<much> faster, cheaper, better).
1071		1447
1072	(It might actually work, but you have been warned).	1448	(It might actually work, but you have been warned).
1073		1449
1074		1450
		1451	=head1 THE PERILS OF SETLOCALE
		1452
		1453	Sometimes people avoid the Perl locale support and directly call the
		1454	system's setlocale function with C<LC_ALL>.
		1455
		1456	This breaks both perl and modules such as JSON::XS, as stringification of
		1457	numbers no longer works correcly (e.g. C<$x = 0.1; print "$x"+1> might
		1458	print C<1>, and JSON::XS might output illegal JSON as JSON::XS relies on
		1459	perl to stringify numbers).
		1460
		1461	The solution is simple: don't call C<setlocale>, or use it for only those
		1462	categories you need, such as C<LC_MESSAGES> or C<LC_CTYPE>.
		1463
		1464	If you need C<LC_NUMERIC>, you should enable it only around the code that
		1465	actually needs it (avoiding stringification of numbers), and restore it
		1466	afterwards.
		1467
		1468
1075	=head1 BUGS	1469	=head1 BUGS
1076		1470
1077	While the goal of this module is to be correct, that unfortunately does	1471	While the goal of this module is to be correct, that unfortunately does
1078	not mean its bug-free, only that I think its design is bug-free. It is	1472	not mean it's bug-free, only that I think its design is bug-free. If you
1079	still relatively early in its development. If you keep reporting bugs they	1473	keep reporting bugs they will be fixed swiftly, though.
1080	will be fixed swiftly, though.
1081		1474
1082	Please refrain from using rt.cpan.org or any other bug reporting	1475	Please refrain from using rt.cpan.org or any other bug reporting
1083	service. I put the contact address into my modules for a reason.	1476	service. I put the contact address into my modules for a reason.
1084		1477
1085	=cut	1478	=cut
…		…
1105	"--" => sub { $_[0] = ${$_[0]} - 1 },	1498	"--" => sub { $_[0] = ${$_[0]} - 1 },
1106	fallback => 1;	1499	fallback => 1;
1107		1500
1108	1;	1501	1;
1109		1502
		1503	=head1 SEE ALSO
		1504
		1505	The F<json_xs> command line utility for quick experiments.
		1506
1110	=head1 AUTHOR	1507	=head1 AUTHOR
1111		1508
1112	Marc Lehmann <schmorp@schmorp.de>	1509	Marc Lehmann <schmorp@schmorp.de>
1113	http://home.schmorp.de/	1510	http://home.schmorp.de/
1114		1511

Diff Legend

-–
+Removed lines
-+
+Added lines
-<
+Changed lines
->
+Changed lines

Comparing JSON-XS/XS.pm (file contents): Revision 1.82 by root, Sun Dec 30 21:24:57 2007 UTC vs. Revision 1.139 by root, Thu May 23 09:31:32 2013 UTC

Diff Legend

Comparing JSON-XS/XS.pm (file contents):
Revision 1.82 by root, Sun Dec 30 21:24:57 2007 UTC vs.
Revision 1.139 by root, Thu May 23 09:31:32 2013 UTC