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

Comparing CBOR-XS/XS.pm (file contents):
Revision 1.23 by root, Fri Nov 22 16:00:30 2013 UTC vs.
Revision 1.89 by root, Fri Sep 8 20:03:06 2023 UTC

…		…
26	substr $many_cbor_strings, 0, $length, ""; # remove decoded cbor string	26	substr $many_cbor_strings, 0, $length, ""; # remove decoded cbor string
27	}	27	}
28		28
29	=head1 DESCRIPTION	29	=head1 DESCRIPTION
30		30
31	WARNING! This module is very new, and not very well tested (that's up
32	to you to do). Furthermore, details of the implementation might change
33	freely before version 1.0. And lastly, most extensions depend on an IANA
34	assignment, and until that assignment is official, this implementation is
35	not interoperable with other implementations (even future versions of this
36	module) until the assignment is done.
37
38	You are still invited to try out CBOR, and this module.
39
40	This module converts Perl data structures to the Concise Binary Object	31	This module converts Perl data structures to the Concise Binary Object
41	Representation (CBOR) and vice versa. CBOR is a fast binary serialisation	32	Representation (CBOR) and vice versa. CBOR is a fast binary serialisation
42	format that aims to use a superset of the JSON data model, i.e. when you	33	format that aims to use an (almost) superset of the JSON data model, i.e.
43	can represent something in JSON, you should be able to represent it in	34	when you can represent something useful in JSON, you should be able to
44	CBOR.	35	represent it in CBOR.
45		36
46	In short, CBOR is a faster and very compact binary alternative to JSON,	37	In short, CBOR is a faster and quite compact binary alternative to JSON,
47	with the added ability of supporting serialisation of Perl objects. (JSON	38	with the added ability of supporting serialisation of Perl objects. (JSON
48	often compresses better than CBOR though, so if you plan to compress the	39	often compresses better than CBOR though, so if you plan to compress the
49	data later you might want to compare both formats first).	40	data later and speed is less important you might want to compare both
		41	formats first).
		42
		43	The primary goal of this module is to be I<correct> and the secondary goal
		44	is to be I<fast>. To reach the latter goal it was written in C.
50		45
51	To give you a general idea about speed, with texts in the megabyte range,	46	To give you a general idea about speed, with texts in the megabyte range,
52	C<CBOR::XS> usually encodes roughly twice as fast as L<Storable> or	47	C<CBOR::XS> usually encodes roughly twice as fast as L<Storable> or
53	L<JSON::XS> and decodes about 15%-30% faster than those. The shorter the	48	L<JSON::XS> and decodes about 15%-30% faster than those. The shorter the
54	data, the worse L<Storable> performs in comparison.	49	data, the worse L<Storable> performs in comparison.
55		50
56	As for compactness, C<CBOR::XS> encoded data structures are usually about	51	Regarding compactness, C<CBOR::XS>-encoded data structures are usually
57	20% smaller than the same data encoded as (compact) JSON or L<Storable>.	52	about 20% smaller than the same data encoded as (compact) JSON or
		53	L<Storable>.
58		54
59	In addition to the core CBOR data format, this module implements a number	55	In addition to the core CBOR data format, this module implements a
60	of extensions, to support cyclic and self-referencing data structures	56	number of extensions, to support cyclic and shared data structures
61	(see C<allow_sharing>), string deduplication (see C<allow_stringref>) and	57	(see C<allow_sharing> and C<allow_cycles>), string deduplication (see
62	scalar references (always enabled).	58	C<pack_strings>) and scalar references (always enabled).
63
64	The primary goal of this module is to be I<correct> and the secondary goal
65	is to be I<fast>. To reach the latter goal it was written in C.
66		59
67	See MAPPING, below, on how CBOR::XS maps perl values to CBOR values and	60	See MAPPING, below, on how CBOR::XS maps perl values to CBOR values and
68	vice versa.	61	vice versa.
69		62
70	=cut	63	=cut
71		64
72	package CBOR::XS;	65	package CBOR::XS;
73		66
74	use common::sense;	67	use common::sense;
75		68
76	our $VERSION = 0.08;	69	our $VERSION = 1.87;
77	our @ISA = qw(Exporter);	70	our @ISA = qw(Exporter);
78		71
79	our @EXPORT = qw(encode_cbor decode_cbor);	72	our @EXPORT = qw(encode_cbor decode_cbor);
80		73
81	use Exporter;	74	use Exporter;
…		…
119		112
120	The mutators for flags all return the CBOR object again and thus calls can	113	The mutators for flags all return the CBOR object again and thus calls can
121	be chained:	114	be chained:
122		115
123	my $cbor = CBOR::XS->new->encode ({a => [1,2]});	116	my $cbor = CBOR::XS->new->encode ({a => [1,2]});
		117
		118	=item $cbor = new_safe CBOR::XS
		119
		120	Create a new, safe/secure CBOR::XS object. This is similar to C<new>,
		121	but configures the coder object to be safe to use with untrusted
		122	data. Currently, this is equivalent to:
		123
		124	my $cbor = CBOR::XS
		125	->new
		126	->validate_utf8
		127	->forbid_objects
		128	->filter (\&CBOR::XS::safe_filter)
		129	->max_size (1e8);
		130
		131	But is more future proof (it is better to crash because of a change than
		132	to be exploited in other ways).
		133
		134	=cut
		135
		136	sub new_safe {
		137	CBOR::XS
		138	->new
		139	->validate_utf8
		140	->forbid_objects
		141	->filter (\&CBOR::XS::safe_filter)
		142	->max_size (1e8)
		143	}
124		144
125	=item $cbor = $cbor->max_depth ([$maximum_nesting_depth])	145	=item $cbor = $cbor->max_depth ([$maximum_nesting_depth])
126		146
127	=item $max_depth = $cbor->get_max_depth	147	=item $max_depth = $cbor->get_max_depth
128		148
…		…
144		164
145	Note that nesting is implemented by recursion in C. The default value has	165	Note that nesting is implemented by recursion in C. The default value has
146	been chosen to be as large as typical operating systems allow without	166	been chosen to be as large as typical operating systems allow without
147	crashing.	167	crashing.
148		168
149	See SECURITY CONSIDERATIONS, below, for more info on why this is useful.	169	See L<SECURITY CONSIDERATIONS>, below, for more info on why this is useful.
150		170
151	=item $cbor = $cbor->max_size ([$maximum_string_size])	171	=item $cbor = $cbor->max_size ([$maximum_string_size])
152		172
153	=item $max_size = $cbor->get_max_size	173	=item $max_size = $cbor->get_max_size
154		174
…		…
159	effect on C<encode> (yet).	179	effect on C<encode> (yet).
160		180
161	If no argument is given, the limit check will be deactivated (same as when	181	If no argument is given, the limit check will be deactivated (same as when
162	C<0> is specified).	182	C<0> is specified).
163		183
164	See SECURITY CONSIDERATIONS, below, for more info on why this is useful.	184	See L<SECURITY CONSIDERATIONS>, below, for more info on why this is useful.
165		185
166	=item $cbor = $cbor->allow_unknown ([$enable])	186	=item $cbor = $cbor->allow_unknown ([$enable])
167		187
168	=item $enabled = $cbor->get_allow_unknown	188	=item $enabled = $cbor->get_allow_unknown
169		189
…		…
186	as an array, is referenced multiple times), but instead will emit a	206	as an array, is referenced multiple times), but instead will emit a
187	reference to the earlier value.	207	reference to the earlier value.
188		208
189	This means that such values will only be encoded once, and will not result	209	This means that such values will only be encoded once, and will not result
190	in a deep cloning of the value on decode, in decoders supporting the value	210	in a deep cloning of the value on decode, in decoders supporting the value
191	sharing extension.	211	sharing extension. This also makes it possible to encode cyclic data
		212	structures (which need C<allow_cycles> to be enabled to be decoded by this
		213	module).
192		214
193	It is recommended to leave it off unless you know your	215	It is recommended to leave it off unless you know your
194	communication partner supports the value sharing extensions to CBOR	216	communication partner supports the value sharing extensions to CBOR
195	(http://cbor.schmorp.de/value-sharing).	217	(L<http://cbor.schmorp.de/value-sharing>), as without decoder support, the
		218	resulting data structure might be unusable.
196		219
197	Detecting shared values incurs a runtime overhead when values are encoded	220	Detecting shared values incurs a runtime overhead when values are encoded
198	that have a reference counter large than one, and might unnecessarily	221	that have a reference counter larger than one, and might unnecessarily
199	increase the encoded size, as potentially shared values are encode as	222	increase the encoded size, as potentially shared values are encoded as
200	sharable whether or not they are actually shared.	223	shareable whether or not they are actually shared.
201		224
202	At the moment, only targets of references can be shared (e.g. scalars,	225	At the moment, only targets of references can be shared (e.g. scalars,
203	arrays or hashes pointed to by a reference). Weirder constructs, such as	226	arrays or hashes pointed to by a reference). Weirder constructs, such as
204	an array with multiple "copies" of the I<same> string, which are hard but	227	an array with multiple "copies" of the I<same> string, which are hard but
205	not impossible to create in Perl, are not supported (this is the same as	228	not impossible to create in Perl, are not supported (this is the same as
206	for L<Storable>).	229	with L<Storable>).
207		230
208	If C<$enable> is false (the default), then C<encode> will encode	231	If C<$enable> is false (the default), then C<encode> will encode shared
209	exception when it encounters anything it cannot encode as CBOR.	232	data structures repeatedly, unsharing them in the process. Cyclic data
		233	structures cannot be encoded in this mode.
210		234
211	This option does not affect C<decode> in any way - shared values and	235	This option does not affect C<decode> in any way - shared values and
212	references will always be decoded properly if present.	236	references will always be decoded properly if present.
213		237
		238	=item $cbor = $cbor->allow_cycles ([$enable])
		239
		240	=item $enabled = $cbor->get_allow_cycles
		241
		242	If C<$enable> is true (or missing), then C<decode> will happily decode
		243	self-referential (cyclic) data structures. By default these will not be
		244	decoded, as they need manual cleanup to avoid memory leaks, so code that
		245	isn't prepared for this will not leak memory.
		246
		247	If C<$enable> is false (the default), then C<decode> will throw an error
		248	when it encounters a self-referential/cyclic data structure.
		249
		250	This option does not affect C<encode> in any way - shared values and
		251	references will always be encoded properly if present.
		252
		253	=item $cbor = $cbor->allow_weak_cycles ([$enable])
		254
		255	=item $enabled = $cbor->get_allow_weak_cycles
		256
		257	This works like C<allow_cycles> in that it allows the resulting data
		258	structures to contain cycles, but unlike C<allow_cycles>, those cyclic
		259	rreferences will be weak. That means that code that recurrsively walks
		260	the data structure must be prepared with cycles, but at least not special
		261	precautions must be implemented to free these data structures.
		262
		263	Only those references leading to actual cycles will be weakened - other
		264	references, e.g. when the same hash or arrray is referenced multiple times
		265	in an arrray, will be normal references.
		266
		267	This option does not affect C<encode> in any way - shared values and
		268	references will always be encoded properly if present.
		269
		270	=item $cbor = $cbor->forbid_objects ([$enable])
		271
		272	=item $enabled = $cbor->get_forbid_objects
		273
		274	Disables the use of the object serialiser protocol.
		275
		276	If C<$enable> is true (or missing), then C<encode> will will throw an
		277	exception when it encounters perl objects that would be encoded using the
		278	perl-object tag (26). When C<decode> encounters such tags, it will fall
		279	back to the general filter/tagged logic as if this were an unknown tag (by
		280	default resulting in a C<CBOR::XC::Tagged> object).
		281
		282	If C<$enable> is false (the default), then C<encode> will use the
		283	L<Types::Serialiser> object serialisation protocol to serialise objects
		284	into perl-object tags, and C<decode> will do the same to decode such tags.
		285
		286	See L<SECURITY CONSIDERATIONS>, below, for more info on why forbidding this
		287	protocol can be useful.
		288
214	=item $cbor = $cbor->allow_stringref ([$enable])	289	=item $cbor = $cbor->pack_strings ([$enable])
215		290
216	=item $enabled = $cbor->get_allow_stringref	291	=item $enabled = $cbor->get_pack_strings
217		292
218	If C<$enable> is true (or missing), then C<encode> will try not to encode	293	If C<$enable> is true (or missing), then C<encode> will try not to encode
219	the same string twice, but will instead encode a reference to the string	294	the same string twice, but will instead encode a reference to the string
220	instead. Depending on your data format. this can save a lot of space, but	295	instead. Depending on your data format, this can save a lot of space, but
221	also results in a very large runtime overhead (expect encoding times to be	296	also results in a very large runtime overhead (expect encoding times to be
222	2-4 times as high as without).	297	2-4 times as high as without).
223		298
224	It is recommended to leave it off unless you know your	299	It is recommended to leave it off unless you know your
225	communications partner supports the stringref extension to CBOR	300	communications partner supports the stringref extension to CBOR
226	(http://cbor.schmorp.de/stringref).	301	(L<http://cbor.schmorp.de/stringref>), as without decoder support, the
		302	resulting data structure might not be usable.
227		303
228	If C<$enable> is false (the default), then C<encode> will encode	304	If C<$enable> is false (the default), then C<encode> will encode strings
229	exception when it encounters anything it cannot encode as CBOR.	305	the standard CBOR way.
230		306
231	This option does not affect C<decode> in any way - string references will	307	This option does not affect C<decode> in any way - string references will
232	always be decoded properly if present.	308	always be decoded properly if present.
233		309
		310	=item $cbor = $cbor->text_keys ([$enable])
		311
		312	=item $enabled = $cbor->get_text_keys
		313
		314	If C<$enabled> is true (or missing), then C<encode> will encode all
		315	perl hash keys as CBOR text strings/UTF-8 string, upgrading them as needed.
		316
		317	If C<$enable> is false (the default), then C<encode> will encode hash keys
		318	normally - upgraded perl strings (strings internally encoded as UTF-8) as
		319	CBOR text strings, and downgraded perl strings as CBOR byte strings.
		320
		321	This option does not affect C<decode> in any way.
		322
		323	This option is useful for interoperability with CBOR decoders that don't
		324	treat byte strings as a form of text. It is especially useful as Perl
		325	gives very little control over hash keys.
		326
		327	Enabling this option can be slow, as all downgraded hash keys that are
		328	encoded need to be scanned and converted to UTF-8.
		329
		330	=item $cbor = $cbor->text_strings ([$enable])
		331
		332	=item $enabled = $cbor->get_text_strings
		333
		334	This option works similar to C<text_keys>, above, but works on all strings
		335	(including hash keys), so C<text_keys> has no further effect after
		336	enabling C<text_strings>.
		337
		338	If C<$enabled> is true (or missing), then C<encode> will encode all perl
		339	strings as CBOR text strings/UTF-8 strings, upgrading them as needed.
		340
		341	If C<$enable> is false (the default), then C<encode> will encode strings
		342	normally (but see C<text_keys>) - upgraded perl strings (strings
		343	internally encoded as UTF-8) as CBOR text strings, and downgraded perl
		344	strings as CBOR byte strings.
		345
		346	This option does not affect C<decode> in any way.
		347
		348	This option has similar advantages and disadvantages as C<text_keys>. In
		349	addition, this option effectively removes the ability to automatically
		350	encode byte strings, which might break some C<FREEZE> and C<TO_CBOR>
		351	methods that rely on this.
		352
		353	A workaround is to use explicit type casts, which are unaffected by this option.
		354
		355	=item $cbor = $cbor->validate_utf8 ([$enable])
		356
		357	=item $enabled = $cbor->get_validate_utf8
		358
		359	If C<$enable> is true (or missing), then C<decode> will validate that
		360	elements (text strings) containing UTF-8 data in fact contain valid UTF-8
		361	data (instead of blindly accepting it). This validation obviously takes
		362	extra time during decoding.
		363
		364	The concept of "valid UTF-8" used is perl's concept, which is a superset
		365	of the official UTF-8.
		366
		367	If C<$enable> is false (the default), then C<decode> will blindly accept
		368	UTF-8 data, marking them as valid UTF-8 in the resulting data structure
		369	regardless of whether that's true or not.
		370
		371	Perl isn't too happy about corrupted UTF-8 in strings, but should
		372	generally not crash or do similarly evil things. Extensions might be not
		373	so forgiving, so it's recommended to turn on this setting if you receive
		374	untrusted CBOR.
		375
		376	This option does not affect C<encode> in any way - strings that are
		377	supposedly valid UTF-8 will simply be dumped into the resulting CBOR
		378	string without checking whether that is, in fact, true or not.
		379
234	=item $cbor = $cbor->filter ([$cb->($tag, $value)])	380	=item $cbor = $cbor->filter ([$cb->($tag, $value)])
235		381
236	=item $cb_or_undef = $cbor->get_filter	382	=item $cb_or_undef = $cbor->get_filter
237		383
238	TODO	384	Sets or replaces the tagged value decoding filter (when C<$cb> is
		385	specified) or clears the filter (if no argument or C<undef> is provided).
		386
		387	The filter callback is called only during decoding, when a non-enforced
		388	tagged value has been decoded (see L<TAG HANDLING AND EXTENSIONS> for a
		389	list of enforced tags). For specific tags, it's often better to provide a
		390	default converter using the C<%CBOR::XS::FILTER> hash (see below).
		391
		392	The first argument is the numerical tag, the second is the (decoded) value
		393	that has been tagged.
		394
		395	The filter function should return either exactly one value, which will
		396	replace the tagged value in the decoded data structure, or no values,
		397	which will result in default handling, which currently means the decoder
		398	creates a C<CBOR::XS::Tagged> object to hold the tag and the value.
		399
		400	When the filter is cleared (the default state), the default filter
		401	function, C<CBOR::XS::default_filter>, is used. This function simply
		402	looks up the tag in the C<%CBOR::XS::FILTER> hash. If an entry exists
		403	it must be a code reference that is called with tag and value, and is
		404	responsible for decoding the value. If no entry exists, it returns no
		405	values. C<CBOR::XS> provides a number of default filter functions already,
		406	the the C<%CBOR::XS::FILTER> hash can be freely extended with more.
		407
		408	C<CBOR::XS> additionally provides an alternative filter function that is
		409	supposed to be safe to use with untrusted data (which the default filter
		410	might not), called C<CBOR::XS::safe_filter>, which works the same as
		411	the C<default_filter> but uses the C<%CBOR::XS::SAFE_FILTER> variable
		412	instead. It is prepopulated with the tag decoding functions that are
		413	deemed safe (basically the same as C<%CBOR::XS::FILTER> without all
		414	the bignum tags), and can be extended by user code as wlel, although,
		415	obviously, one should be very careful about adding decoding functions
		416	here, since the expectation is that they are safe to use on untrusted
		417	data, after all.
		418
		419	Example: decode all tags not handled internally into C<CBOR::XS::Tagged>
		420	objects, with no other special handling (useful when working with
		421	potentially "unsafe" CBOR data).
		422
		423	CBOR::XS->new->filter (sub { })->decode ($cbor_data);
		424
		425	Example: provide a global filter for tag 1347375694, converting the value
		426	into some string form.
		427
		428	$CBOR::XS::FILTER{1347375694} = sub {
		429	my ($tag, $value);
		430
		431	"tag 1347375694 value $value"
		432	};
		433
		434	Example: provide your own filter function that looks up tags in your own
		435	hash:
		436
		437	my %my_filter = (
		438	998347484 => sub {
		439	my ($tag, $value);
		440
		441	"tag 998347484 value $value"
		442	};
		443	);
		444
		445	my $coder = CBOR::XS->new->filter (sub {
		446	&{ $my_filter{$_[0]} or return }
		447	});
		448
		449
		450	Example: use the safe filter function (see L<SECURITY CONSIDERATIONS> for
		451	more considerations on security).
		452
		453	CBOR::XS->new->filter (\&CBOR::XS::safe_filter)->decode ($cbor_data);
239		454
240	=item $cbor_data = $cbor->encode ($perl_scalar)	455	=item $cbor_data = $cbor->encode ($perl_scalar)
241		456
242	Converts the given Perl data structure (a scalar value) to its CBOR	457	Converts the given Perl data structure (a scalar value) to its CBOR
243	representation.	458	representation.
…		…
253	when there is trailing garbage after the CBOR string, it will silently	468	when there is trailing garbage after the CBOR string, it will silently
254	stop parsing there and return the number of characters consumed so far.	469	stop parsing there and return the number of characters consumed so far.
255		470
256	This is useful if your CBOR texts are not delimited by an outer protocol	471	This is useful if your CBOR texts are not delimited by an outer protocol
257	and you need to know where the first CBOR string ends amd the next one	472	and you need to know where the first CBOR string ends amd the next one
258	starts.	473	starts - CBOR strings are self-delimited, so it is possible to concatenate
		474	CBOR strings without any delimiters or size fields and recover their data.
259		475
260	CBOR::XS->new->decode_prefix ("......")	476	CBOR::XS->new->decode_prefix ("......")
261	=> ("...", 3)	477	=> ("...", 3)
		478
		479	=back
		480
		481	=head2 INCREMENTAL PARSING
		482
		483	In some cases, there is the need for incremental parsing of JSON
		484	texts. While this module always has to keep both CBOR text and resulting
		485	Perl data structure in memory at one time, it does allow you to parse a
		486	CBOR stream incrementally, using a similar to using "decode_prefix" to see
		487	if a full CBOR object is available, but is much more efficient.
		488
		489	It basically works by parsing as much of a CBOR string as possible - if
		490	the CBOR data is not complete yet, the parser will remember where it was,
		491	to be able to restart when more data has been accumulated. Once enough
		492	data is available to either decode a complete CBOR value or raise an
		493	error, a real decode will be attempted.
		494
		495	A typical use case would be a network protocol that consists of sending
		496	and receiving CBOR-encoded messages. The solution that works with CBOR and
		497	about anything else is by prepending a length to every CBOR value, so the
		498	receiver knows how many octets to read. More compact (and slightly slower)
		499	would be to just send CBOR values back-to-back, as C<CBOR::XS> knows where
		500	a CBOR value ends, and doesn't need an explicit length.
		501
		502	The following methods help with this:
		503
		504	=over 4
		505
		506	=item @decoded = $cbor->incr_parse ($buffer)
		507
		508	This method attempts to decode exactly one CBOR value from the beginning
		509	of the given C<$buffer>. The value is removed from the C<$buffer> on
		510	success. When C<$buffer> doesn't contain a complete value yet, it returns
		511	nothing. Finally, when the C<$buffer> doesn't start with something
		512	that could ever be a valid CBOR value, it raises an exception, just as
		513	C<decode> would. In the latter case the decoder state is undefined and
		514	must be reset before being able to parse further.
		515
		516	This method modifies the C<$buffer> in place. When no CBOR value can be
		517	decoded, the decoder stores the current string offset. On the next call,
		518	continues decoding at the place where it stopped before. For this to make
		519	sense, the C<$buffer> must begin with the same octets as on previous
		520	unsuccessful calls.
		521
		522	You can call this method in scalar context, in which case it either
		523	returns a decoded value or C<undef>. This makes it impossible to
		524	distinguish between CBOR null values (which decode to C<undef>) and an
		525	unsuccessful decode, which is often acceptable.
		526
		527	=item @decoded = $cbor->incr_parse_multiple ($buffer)
		528
		529	Same as C<incr_parse>, but attempts to decode as many CBOR values as
		530	possible in one go, instead of at most one. Calls to C<incr_parse> and
		531	C<incr_parse_multiple> can be interleaved.
		532
		533	=item $cbor->incr_reset
		534
		535	Resets the incremental decoder. This throws away any saved state, so that
		536	subsequent calls to C<incr_parse> or C<incr_parse_multiple> start to parse
		537	a new CBOR value from the beginning of the C<$buffer> again.
		538
		539	This method can be called at any time, but it I<must> be called if you want
		540	to change your C<$buffer> or there was a decoding error and you want to
		541	reuse the C<$cbor> object for future incremental parsings.
262		542
263	=back	543	=back
264		544
265		545
266	=head1 MAPPING	546	=head1 MAPPING
…		…
284	CBOR integers become (numeric) perl scalars. On perls without 64 bit	564	CBOR integers become (numeric) perl scalars. On perls without 64 bit
285	support, 64 bit integers will be truncated or otherwise corrupted.	565	support, 64 bit integers will be truncated or otherwise corrupted.
286		566
287	=item byte strings	567	=item byte strings
288		568
289	Byte strings will become octet strings in Perl (the byte values 0..255	569	Byte strings will become octet strings in Perl (the Byte values 0..255
290	will simply become characters of the same value in Perl).	570	will simply become characters of the same value in Perl).
291		571
292	=item UTF-8 strings	572	=item UTF-8 strings
293		573
294	UTF-8 strings in CBOR will be decoded, i.e. the UTF-8 octets will be	574	UTF-8 strings in CBOR will be decoded, i.e. the UTF-8 octets will be
…		…
317	=item tagged values	597	=item tagged values
318		598
319	Tagged items consists of a numeric tag and another CBOR value.	599	Tagged items consists of a numeric tag and another CBOR value.
320		600
321	See L<TAG HANDLING AND EXTENSIONS> and the description of C<< ->filter >>	601	See L<TAG HANDLING AND EXTENSIONS> and the description of C<< ->filter >>
322	for details.	602	for details on which tags are handled how.
323		603
324	=item anything else	604	=item anything else
325		605
326	Anything else (e.g. unsupported simple values) will raise a decoding	606	Anything else (e.g. unsupported simple values) will raise a decoding
327	error.	607	error.
…		…
330		610
331		611
332	=head2 PERL -> CBOR	612	=head2 PERL -> CBOR
333		613
334	The mapping from Perl to CBOR is slightly more difficult, as Perl is a	614	The mapping from Perl to CBOR is slightly more difficult, as Perl is a
335	truly typeless language, so we can only guess which CBOR type is meant by	615	typeless language. That means this module can only guess which CBOR type
336	a Perl value.	616	is meant by a perl value.
337		617
338	=over 4	618	=over 4
339		619
340	=item hash references	620	=item hash references
341		621
342	Perl hash references become CBOR maps. As there is no inherent ordering in	622	Perl hash references become CBOR maps. As there is no inherent ordering in
343	hash keys (or CBOR maps), they will usually be encoded in a pseudo-random	623	hash keys (or CBOR maps), they will usually be encoded in a pseudo-random
344	order.	624	order. This order can be different each time a hash is encoded.
345		625
346	Currently, tied hashes will use the indefinite-length format, while normal	626	Currently, tied hashes will use the indefinite-length format, while normal
347	hashes will use the fixed-length format.	627	hashes will use the fixed-length format.
348		628
349	=item array references	629	=item array references
350		630
351	Perl array references become fixed-length CBOR arrays.	631	Perl array references become fixed-length CBOR arrays.
352		632
353	=item other references	633	=item other references
354		634
355	Other unblessed references are generally not allowed and will cause an	635	Other unblessed references will be represented using
356	exception to be thrown, except for references to the integers C<0> and	636	the indirection tag extension (tag value C<22098>,
357	C<1>, which get turned into false and true in CBOR.	637	L<http://cbor.schmorp.de/indirection>). CBOR decoders are guaranteed
		638	to be able to decode these values somehow, by either "doing the right
		639	thing", decoding into a generic tagged object, simply ignoring the tag, or
		640	something else.
358		641
359	=item CBOR::XS::Tagged objects	642	=item CBOR::XS::Tagged objects
360		643
361	Objects of this type must be arrays consisting of a single C<[tag, value]>	644	Objects of this type must be arrays consisting of a single C<[tag, value]>
362	pair. The (numerical) tag will be encoded as a CBOR tag, the value will	645	pair. The (numerical) tag will be encoded as a CBOR tag, the value will
363	be encoded as appropriate for the value. You cna use C<CBOR::XS::tag> to	646	be encoded as appropriate for the value. You must use C<CBOR::XS::tag> to
364	create such objects.	647	create such objects.
365		648
366	=item Types::Serialiser::true, Types::Serialiser::false, Types::Serialiser::error	649	=item Types::Serialiser::true, Types::Serialiser::false, Types::Serialiser::error
367		650
368	These special values become CBOR true, CBOR false and CBOR undefined	651	These special values become CBOR true, CBOR false and CBOR undefined
369	values, respectively. You can also use C<\1>, C<\0> and C<\undef> directly	652	values, respectively.
370	if you want.
371		653
372	=item other blessed objects	654	=item other blessed objects
373		655
374	Other blessed objects are serialised via C<TO_CBOR> or C<FREEZE>. See	656	Other blessed objects are serialised via C<TO_CBOR> or C<FREEZE>. See
375	L<TAG HANDLING AND EXTENSIONS> for specific classes handled by this	657	L<TAG HANDLING AND EXTENSIONS> for specific classes handled by this
…		…
385	# dump as number	667	# dump as number
386	encode_cbor [2] # yields [2]	668	encode_cbor [2] # yields [2]
387	encode_cbor [-3.0e17] # yields [-3e+17]	669	encode_cbor [-3.0e17] # yields [-3e+17]
388	my $value = 5; encode_cbor [$value] # yields [5]	670	my $value = 5; encode_cbor [$value] # yields [5]
389		671
390	# used as string, so dump as string	672	# used as string, so dump as string (either byte or text)
391	print $value;	673	print $value;
392	encode_cbor [$value] # yields ["5"]	674	encode_cbor [$value] # yields ["5"]
393		675
394	# undef becomes null	676	# undef becomes null
395	encode_cbor [undef] # yields [null]	677	encode_cbor [undef] # yields [null]
…		…
398		680
399	my $x = 3.1; # some variable containing a number	681	my $x = 3.1; # some variable containing a number
400	"$x"; # stringified	682	"$x"; # stringified
401	$x .= ""; # another, more awkward way to stringify	683	$x .= ""; # another, more awkward way to stringify
402	print $x; # perl does it for you, too, quite often	684	print $x; # perl does it for you, too, quite often
		685
		686	You can force whether a string is encoded as byte or text string by using
		687	C<utf8::upgrade> and C<utf8::downgrade> (if C<text_strings> is disabled).
		688
		689	utf8::upgrade $x; # encode $x as text string
		690	utf8::downgrade $x; # encode $x as byte string
		691
		692	More options are available, see L<TYPE CASTS>, below, and the C<text_keys>
		693	and C<text_strings> options.
		694
		695	Perl doesn't define what operations up- and downgrade strings, so if the
		696	difference between byte and text is important, you should up- or downgrade
		697	your string as late as possible before encoding. You can also force the
		698	use of CBOR text strings by using C<text_keys> or C<text_strings>.
403		699
404	You can force the type to be a CBOR number by numifying it:	700	You can force the type to be a CBOR number by numifying it:
405		701
406	my $x = "3"; # some variable containing a string	702	my $x = "3"; # some variable containing a string
407	$x += 0; # numify it, ensuring it will be dumped as a number	703	$x += 0; # numify it, ensuring it will be dumped as a number
…		…
418	represent numerical values are supported, but might suffer loss of	714	represent numerical values are supported, but might suffer loss of
419	precision.	715	precision.
420		716
421	=back	717	=back
422		718
		719	=head2 TYPE CASTS
		720
		721	B<EXPERIMENTAL>: As an experimental extension, C<CBOR::XS> allows you to
		722	force specific CBOR types to be used when encoding. That allows you to
		723	encode types not normally accessible (e.g. half floats) as well as force
		724	string types even when C<text_strings> is in effect.
		725
		726	Type forcing is done by calling a special "cast" function which keeps a
		727	copy of the value and returns a new value that can be handed over to any
		728	CBOR encoder function.
		729
		730	The following casts are currently available (all of which are unary
		731	operators, that is, have a prototype of C<$>):
		732
		733	=over
		734
		735	=item CBOR::XS::as_int $value
		736
		737	Forces the value to be encoded as some form of (basic, not bignum) integer
		738	type.
		739
		740	=item CBOR::XS::as_text $value
		741
		742	Forces the value to be encoded as (UTF-8) text values.
		743
		744	=item CBOR::XS::as_bytes $value
		745
		746	Forces the value to be encoded as a (binary) string value.
		747
		748	Example: encode a perl string as binary even though C<text_strings> is in
		749	effect.
		750
		751	CBOR::XS->new->text_strings->encode ([4, "text", CBOR::XS::bytes "bytevalue"]);
		752
		753	=item CBOR::XS::as_bool $value
		754
		755	Converts a Perl boolean (which can be any kind of scalar) into a CBOR
		756	boolean. Strictly the same, but shorter to write, than:
		757
		758	$value ? Types::Serialiser::true : Types::Serialiser::false
		759
		760	=item CBOR::XS::as_float16 $value
		761
		762	Forces half-float (IEEE 754 binary16) encoding of the given value.
		763
		764	=item CBOR::XS::as_float32 $value
		765
		766	Forces single-float (IEEE 754 binary32) encoding of the given value.
		767
		768	=item CBOR::XS::as_float64 $value
		769
		770	Forces double-float (IEEE 754 binary64) encoding of the given value.
		771
		772	=item CBOR::XS::as_cbor $cbor_text
		773
		774	Not a type cast per-se, this type cast forces the argument to be encoded
		775	as-is. This can be used to embed pre-encoded CBOR data.
		776
		777	Note that no checking on the validity of the C<$cbor_text> is done - it's
		778	the callers responsibility to correctly encode values.
		779
		780	=item CBOR::XS::as_map [key => value...]
		781
		782	Treat the array reference as key value pairs and output a CBOR map. This
		783	allows you to generate CBOR maps with arbitrary key types (or, if you
		784	don't care about semantics, duplicate keys or pairs in a custom order),
		785	which is otherwise hard to do with Perl.
		786
		787	The single argument must be an array reference with an even number of
		788	elements.
		789
		790	Note that only the reference to the array is copied, the array itself is
		791	not. Modifications done to the array before calling an encoding function
		792	will be reflected in the encoded output.
		793
		794	Example: encode a CBOR map with a string and an integer as keys.
		795
		796	encode_cbor CBOR::XS::as_map [string => "value", 5 => "value"]
		797
		798	=back
		799
		800	=cut
		801
		802	sub CBOR::XS::as_cbor ($) { bless [$_[0], 0, undef], CBOR::XS::Tagged:: }
		803	sub CBOR::XS::as_int ($) { bless [$_[0], 1, undef], CBOR::XS::Tagged:: }
		804	sub CBOR::XS::as_bytes ($) { bless [$_[0], 2, undef], CBOR::XS::Tagged:: }
		805	sub CBOR::XS::as_text ($) { bless [$_[0], 3, undef], CBOR::XS::Tagged:: }
		806	sub CBOR::XS::as_float16 ($) { bless [$_[0], 4, undef], CBOR::XS::Tagged:: }
		807	sub CBOR::XS::as_float32 ($) { bless [$_[0], 5, undef], CBOR::XS::Tagged:: }
		808	sub CBOR::XS::as_float64 ($) { bless [$_[0], 6, undef], CBOR::XS::Tagged:: }
		809
		810	sub CBOR::XS::as_bool ($) { $_[0] ? $Types::Serialiser::true : $Types::Serialiser::false }
		811
		812	sub CBOR::XS::as_map ($) {
		813	ARRAY:: eq ref $_[0]
		814	and $#{ $_[0] } & 1
		815	or do { require Carp; Carp::croak ("CBOR::XS::as_map only acepts array references with an even number of elements, caught") };
		816
		817	bless [$_[0], 7, undef], CBOR::XS::Tagged::
		818	}
		819
423	=head2 OBJECT SERIALISATION	820	=head2 OBJECT SERIALISATION
		821
		822	This module implements both a CBOR-specific and the generic
		823	L<Types::Serialier> object serialisation protocol. The following
		824	subsections explain both methods.
		825
		826	=head3 ENCODING
424		827
425	This module knows two way to serialise a Perl object: The CBOR-specific	828	This module knows two way to serialise a Perl object: The CBOR-specific
426	way, and the generic way.	829	way, and the generic way.
427		830
428	Whenever the encoder encounters a Perl object that it cnanot serialise	831	Whenever the encoder encounters a Perl object that it cannot serialise
429	directly (most of them), it will first look up the C<TO_CBOR> method on	832	directly (most of them), it will first look up the C<TO_CBOR> method on
430	it.	833	it.
431		834
432	If it has a C<TO_CBOR> method, it will call it with the object as only	835	If it has a C<TO_CBOR> method, it will call it with the object as only
433	argument, and expects exactly one return value, which it will then	836	argument, and expects exactly one return value, which it will then
…		…
439		842
440	The C<FREEZE> method can return any number of values (i.e. zero or	843	The C<FREEZE> method can return any number of values (i.e. zero or
441	more). These will be encoded as CBOR perl object, together with the	844	more). These will be encoded as CBOR perl object, together with the
442	classname.	845	classname.
443		846
		847	These methods I<MUST NOT> change the data structure that is being
		848	serialised. Failure to comply to this can result in memory corruption -
		849	and worse.
		850
444	If an object supports neither C<TO_CBOR> nor C<FREEZE>, encoding will fail	851	If an object supports neither C<TO_CBOR> nor C<FREEZE>, encoding will fail
445	with an error.	852	with an error.
446		853
		854	=head3 DECODING
		855
447	Objects encoded via C<TO_CBOR> cannot be automatically decoded, but	856	Objects encoded via C<TO_CBOR> cannot (normally) be automatically decoded,
448	objects encoded via C<FREEZE> can be decoded using the following protocol:	857	but objects encoded via C<FREEZE> can be decoded using the following
		858	protocol:
449		859
450	When an encoded CBOR perl object is encountered by the decoder, it will	860	When an encoded CBOR perl object is encountered by the decoder, it will
451	look up the C<THAW> method, by using the stored classname, and will fail	861	look up the C<THAW> method, by using the stored classname, and will fail
452	if the method cannot be found.	862	if the method cannot be found.
453		863
454	After the lookup it will call the C<THAW> method with the stored classname	864	After the lookup it will call the C<THAW> method with the stored classname
455	as first argument, the constant string C<CBOR> as second argument, and all	865	as first argument, the constant string C<CBOR> as second argument, and all
456	values returned by C<FREEZE> as remaining arguments.	866	values returned by C<FREEZE> as remaining arguments.
457		867
458	=head4 EXAMPLES	868	=head3 EXAMPLES
459		869
460	Here is an example C<TO_CBOR> method:	870	Here is an example C<TO_CBOR> method:
461		871
462	sub My::Object::TO_CBOR {	872	sub My::Object::TO_CBOR {
463	my ($obj) = @_;	873	my ($obj) = @_;
…		…
474		884
475	sub URI::TO_CBOR {	885	sub URI::TO_CBOR {
476	my ($self) = @_;	886	my ($self) = @_;
477	my $uri = "$self"; # stringify uri	887	my $uri = "$self"; # stringify uri
478	utf8::upgrade $uri; # make sure it will be encoded as UTF-8 string	888	utf8::upgrade $uri; # make sure it will be encoded as UTF-8 string
479	CBOR::XS::tagged 32, "$_[0]"	889	CBOR::XS::tag 32, "$_[0]"
480	}	890	}
481		891
482	This will encode URIs as a UTF-8 string with tag 32, which indicates an	892	This will encode URIs as a UTF-8 string with tag 32, which indicates an
483	URI.	893	URI.
484		894
…		…
495	"$self" # encode url string	905	"$self" # encode url string
496	}	906	}
497		907
498	sub URI::THAW {	908	sub URI::THAW {
499	my ($class, $serialiser, $uri) = @_;	909	my ($class, $serialiser, $uri) = @_;
500
501	$class->new ($uri)	910	$class->new ($uri)
502	}	911	}
503		912
504	Unlike C<TO_CBOR>, multiple values can be returned by C<FREEZE>. For	913	Unlike C<TO_CBOR>, multiple values can be returned by C<FREEZE>. For
505	example, a C<FREEZE> method that returns "type", "id" and "variant" values	914	example, a C<FREEZE> method that returns "type", "id" and "variant" values
…		…
636	additional tags (such as base64url).	1045	additional tags (such as base64url).
637		1046
638	=head2 ENFORCED TAGS	1047	=head2 ENFORCED TAGS
639		1048
640	These tags are always handled when decoding, and their handling cannot be	1049	These tags are always handled when decoding, and their handling cannot be
641	overriden by the user.	1050	overridden by the user.
642		1051
643	=over 4	1052	=over 4
644		1053
645	=item <unassigned> (perl-object, L<http://cbor.schmorp.de/perl-object>)	1054	=item 26 (perl-object, L<http://cbor.schmorp.de/perl-object>)
646		1055
647	These tags are automatically created (and decoded) for serialisable	1056	These tags are automatically created (and decoded) for serialisable
648	objects using the C<FREEZE/THAW> methods (the L<Types::Serialier> object	1057	objects using the C<FREEZE/THAW> methods (the L<Types::Serialier> object
649	serialisation protocol). See L<OBJECT SERIALISATION> for details.	1058	serialisation protocol). See L<OBJECT SERIALISATION> for details.
650		1059
651	=item <unassigned>, <unassigned> (sharable, sharedref, L <http://cbor.schmorp.de/value-sharing>)	1060	=item 28, 29 (shareable, sharedref, L<http://cbor.schmorp.de/value-sharing>)
652		1061
653	These tags are automatically decoded when encountered, resulting in	1062	These tags are automatically decoded when encountered (and they do not
		1063	result in a cyclic data structure, see C<allow_cycles>), resulting in
654	shared values in the decoded object. They are only encoded, however, when	1064	shared values in the decoded object. They are only encoded, however, when
655	C<allow_sharable> is enabled.	1065	C<allow_sharing> is enabled.
656		1066
		1067	Not all shared values can be successfully decoded: values that reference
		1068	themselves will I<currently> decode as C<undef> (this is not the same
		1069	as a reference pointing to itself, which will be represented as a value
		1070	that contains an indirect reference to itself - these will be decoded
		1071	properly).
		1072
		1073	Note that considerably more shared value data structures can be decoded
		1074	than will be encoded - currently, only values pointed to by references
		1075	will be shared, others will not. While non-reference shared values can be
		1076	generated in Perl with some effort, they were considered too unimportant
		1077	to be supported in the encoder. The decoder, however, will decode these
		1078	values as shared values.
		1079
657	=item <unassigned>, <unassigned> (stringref-namespace, stringref, L <http://cbor.schmorp.de/stringref>)	1080	=item 256, 25 (stringref-namespace, stringref, L<http://cbor.schmorp.de/stringref>)
658		1081
659	These tags are automatically decoded when encountered. They are only	1082	These tags are automatically decoded when encountered. They are only
660	encoded, however, when C<allow_stringref> is enabled.	1083	encoded, however, when C<pack_strings> is enabled.
661		1084
662	=item 22098 (indirection, L<http://cbor.schmorp.de/indirection>)	1085	=item 22098 (indirection, L<http://cbor.schmorp.de/indirection>)
663		1086
664	This tag is automatically generated when a reference are encountered (with	1087	This tag is automatically generated when a reference are encountered (with
665	the exception of hash and array refernces). It is converted to a reference	1088	the exception of hash and array references). It is converted to a reference
666	when decoding.	1089	when decoding.
667		1090
668	=item 55799 (self-describe CBOR, RFC 7049)	1091	=item 55799 (self-describe CBOR, RFC 7049)
669		1092
670	This value is not generated on encoding (unless explicitly requested by	1093	This value is not generated on encoding (unless explicitly requested by
671	the user), and is simply ignored when decoding.	1094	the user), and is simply ignored when decoding.
672		1095
673	=back	1096	=back
674		1097
675	=head2 OPTIONAL TAGS	1098	=head2 NON-ENFORCED TAGS
676		1099
677	These tags have default filters provided when decoding. Their handling can	1100	These tags have default filters provided when decoding. Their handling can
678	be overriden by changing the C<%CBOR::XS::FILTER> entry for the tag, or by	1101	be overridden by changing the C<%CBOR::XS::FILTER> entry for the tag, or by
679	providing a custom C<filter> function when decoding.	1102	providing a custom C<filter> callback when decoding.
680		1103
681	When they result in decoding into a specific Perl class, the module	1104	When they result in decoding into a specific Perl class, the module
682	usually provides a corresponding C<TO_CBOR> method as well.	1105	usually provides a corresponding C<TO_CBOR> method as well.
683		1106
684	When any of these need to load additional modules that are not part of the	1107	When any of these need to load additional modules that are not part of the
…		…
686	provide these modules. The decoding usually fails with an exception if the	1109	provide these modules. The decoding usually fails with an exception if the
687	required module cannot be loaded.	1110	required module cannot be loaded.
688		1111
689	=over 4	1112	=over 4
690		1113
		1114	=item 0, 1 (date/time string, seconds since the epoch)
		1115
		1116	These tags are decoded into L<Time::Piece> objects. The corresponding
		1117	C<Time::Piece::TO_CBOR> method always encodes into tag 1 values currently.
		1118
		1119	The L<Time::Piece> API is generally surprisingly bad, and fractional
		1120	seconds are only accidentally kept intact, so watch out. On the plus side,
		1121	the module comes with perl since 5.10, which has to count for something.
		1122
691	=item 2, 3 (positive/negative bignum)	1123	=item 2, 3 (positive/negative bignum)
692		1124
693	These tags are decoded into L<Math::BigInt> objects. The corresponding	1125	These tags are decoded into L<Math::BigInt> objects. The corresponding
694	C<Math::BigInt::TO_CBOR> method encodes "small" bigints into normal CBOR	1126	C<Math::BigInt::TO_CBOR> method encodes "small" bigints into normal CBOR
695	integers, and others into positive/negative CBOR bignums.	1127	integers, and others into positive/negative CBOR bignums.
696		1128
697	=item 4, 5 (decimal fraction/bigfloat)	1129	=item 4, 5, 264, 265 (decimal fraction/bigfloat)
698		1130
699	Both decimal fractions and bigfloats are decoded into L<Math::BigFloat>	1131	Both decimal fractions and bigfloats are decoded into L<Math::BigFloat>
700	objects. The corresponding C<Math::BigFloat::TO_CBOR> method I<always>	1132	objects. The corresponding C<Math::BigFloat::TO_CBOR> method I<always>
701	encodes into a decimal fraction.	1133	encodes into a decimal fraction (either tag 4 or 264).
702		1134
703	CBOR cannot represent bigfloats with I<very> large exponents - conversion	1135	NaN and infinities are not encoded properly, as they cannot be represented
704	of such big float objects is undefined.	1136	in CBOR.
705		1137
706	Also, NaN and infinities are not encoded properly.	1138	See L<BIGNUM SECURITY CONSIDERATIONS> for more info.
		1139
		1140	=item 30 (rational numbers)
		1141
		1142	These tags are decoded into L<Math::BigRat> objects. The corresponding
		1143	C<Math::BigRat::TO_CBOR> method encodes rational numbers with denominator
		1144	C<1> via their numerator only, i.e., they become normal integers or
		1145	C<bignums>.
		1146
		1147	See L<BIGNUM SECURITY CONSIDERATIONS> for more info.
707		1148
708	=item 21, 22, 23 (expected later JSON conversion)	1149	=item 21, 22, 23 (expected later JSON conversion)
709		1150
710	CBOR::XS is not a CBOR-to-JSON converter, and will simply ignore these	1151	CBOR::XS is not a CBOR-to-JSON converter, and will simply ignore these
711	tags.	1152	tags.
…		…
716	C<URI::TO_CBOR> method again results in a CBOR URI value.	1157	C<URI::TO_CBOR> method again results in a CBOR URI value.
717		1158
718	=back	1159	=back
719		1160
720	=cut	1161	=cut
721
722	our %FILTER = (
723	# 0 # rfc4287 datetime, utf-8
724	# 1 # unix timestamp, any
725
726	2 => sub { # pos bigint
727	require Math::BigInt;
728	Math::BigInt->new ("0x" . unpack "H*", pop)
729	},
730
731	3 => sub { # neg bigint
732	require Math::BigInt;
733	-Math::BigInt->new ("0x" . unpack "H*", pop)
734	},
735
736	4 => sub { # decimal fraction, array
737	require Math::BigFloat;
738	Math::BigFloat->new ($_[1][1] . "E" . $_[1][0])
739	},
740
741	5 => sub { # bigfloat, array
742	require Math::BigFloat;
743	scalar Math::BigFloat->new ($_[1][1])->blsft ($_[1][0], 2)
744	},
745
746	21 => sub { pop }, # expected conversion to base64url encoding
747	22 => sub { pop }, # expected conversion to base64 encoding
748	23 => sub { pop }, # expected conversion to base16 encoding
749
750	# 24 # embedded cbor, byte string
751
752	32 => sub {
753	require URI;
754	URI->new (pop)
755	},
756
757	# 33 # base64url rfc4648, utf-8
758	# 34 # base64 rfc46484, utf-8
759	# 35 # regex pcre/ecma262, utf-8
760	# 36 # mime message rfc2045, utf-8
761	);
762
763		1162
764	=head1 CBOR and JSON	1163	=head1 CBOR and JSON
765		1164
766	CBOR is supposed to implement a superset of the JSON data model, and is,	1165	CBOR is supposed to implement a superset of the JSON data model, and is,
767	with some coercion, able to represent all JSON texts (something that other	1166	with some coercion, able to represent all JSON texts (something that other
…		…
776	CBOR intact.	1175	CBOR intact.
777		1176
778		1177
779	=head1 SECURITY CONSIDERATIONS	1178	=head1 SECURITY CONSIDERATIONS
780		1179
781	When you are using CBOR in a protocol, talking to untrusted potentially	1180	Tl;dr... if you want to decode or encode CBOR from untrusted sources, you
782	hostile creatures requires relatively few measures.	1181	should start with a coder object created via C<new_safe> (which implements
		1182	the mitigations explained below):
783		1183
		1184	my $coder = CBOR::XS->new_safe;
		1185
		1186	my $data = $coder->decode ($cbor_text);
		1187	my $cbor = $coder->encode ($data);
		1188
		1189	Longer version: When you are using CBOR in a protocol, talking to
		1190	untrusted potentially hostile creatures requires some thought:
		1191
		1192	=over 4
		1193
		1194	=item Security of the CBOR decoder itself
		1195
784	First of all, your CBOR decoder should be secure, that is, should not have	1196	First and foremost, your CBOR decoder should be secure, that is, should
		1197	not have any buffer overflows or similar bugs that could potentially be
785	any buffer overflows. Obviously, this module should ensure that and I am	1198	exploited. Obviously, this module should ensure that and I am trying hard
786	trying hard on making that true, but you never know.	1199	on making that true, but you never know.
787		1200
		1201	=item CBOR::XS can invoke almost arbitrary callbacks during decoding
		1202
		1203	CBOR::XS supports object serialisation - decoding CBOR can cause calls
		1204	to I<any> C<THAW> method in I<any> package that exists in your process
		1205	(that is, CBOR::XS will not try to load modules, but any existing C<THAW>
		1206	method or function can be called, so they all have to be secure).
		1207
		1208	Less obviously, it will also invoke C<TO_CBOR> and C<FREEZE> methods -
		1209	even if all your C<THAW> methods are secure, encoding data structures from
		1210	untrusted sources can invoke those and trigger bugs in those.
		1211
		1212	So, if you are not sure about the security of all the modules you
		1213	have loaded (you shouldn't), you should disable this part using
		1214	C<forbid_objects> or using C<new_safe>.
		1215
		1216	=item CBOR can be extended with tags that call library code
		1217
		1218	CBOR can be extended with tags, and C<CBOR::XS> has a registry of
		1219	conversion functions for many existing tags that can be extended via
		1220	third-party modules (see the C<filter> method).
		1221
		1222	If you don't trust these, you should configure the "safe" filter function,
		1223	C<CBOR::XS::safe_filter> (C<new_safe> does this), which by default only
		1224	includes conversion functions that are considered "safe" by the author
		1225	(but again, they can be extended by third party modules).
		1226
		1227	Depending on your level of paranoia, you can use the "safe" filter:
		1228
		1229	$cbor->filter (\&CBOR::XS::safe_filter);
		1230
		1231	... your own filter...
		1232
		1233	$cbor->filter (sub { ... do your stuffs here ... });
		1234
		1235	... or even no filter at all, disabling all tag decoding:
		1236
		1237	$cbor->filter (sub { });
		1238
		1239	This is never a problem for encoding, as the tag mechanism only exists in
		1240	CBOR texts.
		1241
		1242	=item Resource-starving attacks: object memory usage
		1243
788	Second, you need to avoid resource-starving attacks. That means you should	1244	You need to avoid resource-starving attacks. That means you should limit
789	limit the size of CBOR data you accept, or make sure then when your	1245	the size of CBOR data you accept, or make sure then when your resources
790	resources run out, that's just fine (e.g. by using a separate process that	1246	run out, that's just fine (e.g. by using a separate process that can
791	can crash safely). The size of a CBOR string in octets is usually a good	1247	crash safely). The size of a CBOR string in octets is usually a good
792	indication of the size of the resources required to decode it into a Perl	1248	indication of the size of the resources required to decode it into a Perl
793	structure. While CBOR::XS can check the size of the CBOR text, it might be	1249	structure. While CBOR::XS can check the size of the CBOR text (using
794	too late when you already have it in memory, so you might want to check	1250	C<max_size> - done by C<new_safe>), it might be too late when you already
795	the size before you accept the string.	1251	have it in memory, so you might want to check the size before you accept
		1252	the string.
796		1253
		1254	As for encoding, it is possible to construct data structures that are
		1255	relatively small but result in large CBOR texts (for example by having an
		1256	array full of references to the same big data structure, which will all be
		1257	deep-cloned during encoding by default). This is rarely an actual issue
		1258	(and the worst case is still just running out of memory), but you can
		1259	reduce this risk by using C<allow_sharing>.
		1260
		1261	=item Resource-starving attacks: stack overflows
		1262
797	Third, CBOR::XS recurses using the C stack when decoding objects and	1263	CBOR::XS recurses using the C stack when decoding objects and arrays. The
798	arrays. The C stack is a limited resource: for instance, on my amd64	1264	C stack is a limited resource: for instance, on my amd64 machine with 8MB
799	machine with 8MB of stack size I can decode around 180k nested arrays but	1265	of stack size I can decode around 180k nested arrays but only 14k nested
800	only 14k nested CBOR objects (due to perl itself recursing deeply on croak	1266	CBOR objects (due to perl itself recursing deeply on croak to free the
801	to free the temporary). If that is exceeded, the program crashes. To be	1267	temporary). If that is exceeded, the program crashes. To be conservative,
802	conservative, the default nesting limit is set to 512. If your process	1268	the default nesting limit is set to 512. If your process has a smaller
803	has a smaller stack, you should adjust this setting accordingly with the	1269	stack, you should adjust this setting accordingly with the C<max_depth>
804	C<max_depth> method.	1270	method.
		1271
		1272	=item Resource-starving attacks: CPU en-/decoding complexity
		1273
		1274	CBOR::XS will use the L<Math::BigInt>, L<Math::BigFloat> and
		1275	L<Math::BigRat> libraries to represent encode/decode bignums. These can be
		1276	very slow (as in, centuries of CPU time) and can even crash your program
		1277	(and are generally not very trustworthy). See the next section on bignum
		1278	security for details.
		1279
		1280	=item Data breaches: leaking information in error messages
		1281
		1282	CBOR::XS might leak contents of your Perl data structures in its error
		1283	messages, so when you serialise sensitive information you might want to
		1284	make sure that exceptions thrown by CBOR::XS will not end up in front of
		1285	untrusted eyes.
		1286
		1287	=item Something else...
805		1288
806	Something else could bomb you, too, that I forgot to think of. In that	1289	Something else could bomb you, too, that I forgot to think of. In that
807	case, you get to keep the pieces. I am always open for hints, though...	1290	case, you get to keep the pieces. I am always open for hints, though...
808		1291
809	Also keep in mind that CBOR::XS might leak contents of your Perl data	1292	=back
810	structures in its error messages, so when you serialise sensitive	1293
811	information you might want to make sure that exceptions thrown by CBOR::XS	1294
812	will not end up in front of untrusted eyes.	1295	=head1 BIGNUM SECURITY CONSIDERATIONS
		1296
		1297	CBOR::XS provides a C<TO_CBOR> method for both L<Math::BigInt> and
		1298	L<Math::BigFloat> that tries to encode the number in the simplest possible
		1299	way, that is, either a CBOR integer, a CBOR bigint/decimal fraction (tag
		1300	4) or an arbitrary-exponent decimal fraction (tag 264). Rational numbers
		1301	(L<Math::BigRat>, tag 30) can also contain bignums as members.
		1302
		1303	CBOR::XS will also understand base-2 bigfloat or arbitrary-exponent
		1304	bigfloats (tags 5 and 265), but it will never generate these on its own.
		1305
		1306	Using the built-in L<Math::BigInt::Calc> support, encoding and decoding
		1307	decimal fractions is generally fast. Decoding bigints can be slow for very
		1308	big numbers (tens of thousands of digits, something that could potentially
		1309	be caught by limiting the size of CBOR texts), and decoding bigfloats or
		1310	arbitrary-exponent bigfloats can be I<extremely> slow (minutes, decades)
		1311	for large exponents (roughly 40 bit and longer).
		1312
		1313	Additionally, L<Math::BigInt> can take advantage of other bignum
		1314	libraries, such as L<Math::GMP>, which cannot handle big floats with large
		1315	exponents, and might simply abort or crash your program, due to their code
		1316	quality.
		1317
		1318	This can be a concern if you want to parse untrusted CBOR. If it is, you
		1319	might want to disable decoding of tag 2 (bigint) and 3 (negative bigint)
		1320	types. You should also disable types 5 and 265, as these can be slow even
		1321	without bigints.
		1322
		1323	Disabling bigints will also partially or fully disable types that rely on
		1324	them, e.g. rational numbers that use bignums.
		1325
813		1326
814	=head1 CBOR IMPLEMENTATION NOTES	1327	=head1 CBOR IMPLEMENTATION NOTES
815		1328
816	This section contains some random implementation notes. They do not	1329	This section contains some random implementation notes. They do not
817	describe guaranteed behaviour, but merely behaviour as-is implemented	1330	describe guaranteed behaviour, but merely behaviour as-is implemented
…		…
826	Only the double data type is supported for NV data types - when Perl uses	1339	Only the double data type is supported for NV data types - when Perl uses
827	long double to represent floating point values, they might not be encoded	1340	long double to represent floating point values, they might not be encoded
828	properly. Half precision types are accepted, but not encoded.	1341	properly. Half precision types are accepted, but not encoded.
829		1342
830	Strict mode and canonical mode are not implemented.	1343	Strict mode and canonical mode are not implemented.
		1344
		1345
		1346	=head1 LIMITATIONS ON PERLS WITHOUT 64-BIT INTEGER SUPPORT
		1347
		1348	On perls that were built without 64 bit integer support (these are rare
		1349	nowadays, even on 32 bit architectures, as all major Perl distributions
		1350	are built with 64 bit integer support), support for any kind of 64 bit
		1351	value in CBOR is very limited - most likely, these 64 bit values will
		1352	be truncated, corrupted, or otherwise not decoded correctly. This also
		1353	includes string, float, array and map sizes that are stored as 64 bit
		1354	integers.
831		1355
832		1356
833	=head1 THREADS	1357	=head1 THREADS
834		1358
835	This module is I<not> guaranteed to be thread safe and there are no	1359	This module is I<not> guaranteed to be thread safe and there are no
…		…
849	Please refrain from using rt.cpan.org or any other bug reporting	1373	Please refrain from using rt.cpan.org or any other bug reporting
850	service. I put the contact address into my modules for a reason.	1374	service. I put the contact address into my modules for a reason.
851		1375
852	=cut	1376	=cut
853		1377
		1378	# clumsy and slow hv_store-in-hash helper function
		1379	sub _hv_store {
		1380	$_[0]{$_[1]} = $_[2];
		1381	}
		1382
854	our %FILTER = (	1383	our %FILTER = (
855	# 0 # rfc4287 datetime, utf-8	1384	0 => sub { # rfc4287 datetime, utf-8
856	# 1 # unix timestamp, any	1385	require Time::Piece;
		1386	# Time::Piece::Strptime uses the "incredibly flexible date parsing routine"
		1387	# from FreeBSD, which can't parse ISO 8601, RFC3339, RFC4287 or much of anything
		1388	# else either. Whats incredibe over standard strptime totally escapes me.
		1389	# doesn't do fractional times, either. sigh.
		1390	# In fact, it's all a lie, it uses whatever strptime it wants, and of course,
		1391	# they are all incompatible. The openbsd one simply ignores %z (but according to the
		1392	# docs, it would be much more incredibly flexible indeed. If it worked, that is.).
		1393	scalar eval {
		1394	my $s = $_[1];
		1395
		1396	$s =~ s/Z$/+00:00/;
		1397	$s =~ s/(\.[0-9]+)?([+-][0-9][0-9]):([0-9][0-9])$//
		1398	or die;
		1399
		1400	my $b = $1 - ($2 * 60 + $3) * 60; # fractional part + offset. hopefully
		1401	my $d = Time::Piece->strptime ($s, "%Y-%m-%dT%H:%M:%S");
		1402
		1403	Time::Piece::gmtime ($d->epoch + $b)
		1404	} \|\| die "corrupted CBOR date/time string ($_[0])";
		1405	},
		1406
		1407	1 => sub { # seconds since the epoch, possibly fractional
		1408	require Time::Piece;
		1409	scalar Time::Piece::gmtime (pop)
		1410	},
857		1411
858	2 => sub { # pos bigint	1412	2 => sub { # pos bigint
859	require Math::BigInt;	1413	require Math::BigInt;
860	Math::BigInt->new ("0x" . unpack "H*", pop)	1414	Math::BigInt->new ("0x" . unpack "H*", pop)
861	},	1415	},
…		…
868	4 => sub { # decimal fraction, array	1422	4 => sub { # decimal fraction, array
869	require Math::BigFloat;	1423	require Math::BigFloat;
870	Math::BigFloat->new ($_[1][1] . "E" . $_[1][0])	1424	Math::BigFloat->new ($_[1][1] . "E" . $_[1][0])
871	},	1425	},
872		1426
		1427	264 => sub { # decimal fraction with arbitrary exponent
		1428	require Math::BigFloat;
		1429	Math::BigFloat->new ($_[1][1] . "E" . $_[1][0])
		1430	},
		1431
873	5 => sub { # bigfloat, array	1432	5 => sub { # bigfloat, array
874	require Math::BigFloat;	1433	require Math::BigFloat;
875	scalar Math::BigFloat->new ($_[1][1])->blsft ($_[1][0], 2)	1434	scalar Math::BigFloat->new ($_[1][1]) * Math::BigFloat->new (2)->bpow ($_[1][0])
		1435	},
		1436
		1437	265 => sub { # bigfloat with arbitrary exponent
		1438	require Math::BigFloat;
		1439	scalar Math::BigFloat->new ($_[1][1]) * Math::BigFloat->new (2)->bpow ($_[1][0])
		1440	},
		1441
		1442	30 => sub { # rational number
		1443	require Math::BigRat;
		1444	Math::BigRat->new ("$_[1][0]/$_[1][1]") # separate parameters only work in recent versons
876	},	1445	},
877		1446
878	21 => sub { pop }, # expected conversion to base64url encoding	1447	21 => sub { pop }, # expected conversion to base64url encoding
879	22 => sub { pop }, # expected conversion to base64 encoding	1448	22 => sub { pop }, # expected conversion to base64 encoding
880	23 => sub { pop }, # expected conversion to base16 encoding	1449	23 => sub { pop }, # expected conversion to base16 encoding
…		…
890	# 34 # base64 rfc46484, utf-8	1459	# 34 # base64 rfc46484, utf-8
891	# 35 # regex pcre/ecma262, utf-8	1460	# 35 # regex pcre/ecma262, utf-8
892	# 36 # mime message rfc2045, utf-8	1461	# 36 # mime message rfc2045, utf-8
893	);	1462	);
894		1463
895	sub CBOR::XS::default_filter {	1464	sub default_filter {
896	&{ $FILTER{$_[0]} or return }	1465	&{ $FILTER{$_[0]} or return }
		1466	}
		1467
		1468	our %SAFE_FILTER = map { $_ => $FILTER{$_} } 0, 1, 21, 22, 23, 32;
		1469
		1470	sub safe_filter {
		1471	&{ $SAFE_FILTER{$_[0]} or return }
897	}	1472	}
898		1473
899	sub URI::TO_CBOR {	1474	sub URI::TO_CBOR {
900	my $uri = $_[0]->as_string;	1475	my $uri = $_[0]->as_string;
901	utf8::upgrade $uri;	1476	utf8::upgrade $uri;
902	CBOR::XS::tag 32, $uri	1477	tag 32, $uri
903	}	1478	}
904		1479
905	sub Math::BigInt::TO_CBOR {	1480	sub Math::BigInt::TO_CBOR {
906	if ($_[0] >= -2147483648 && $_[0] <= 2147483647) {	1481	if (-2147483648 <= $_[0] && $_[0] <= 2147483647) {
907	$_[0]->numify	1482	$_[0]->numify
908	} else {	1483	} else {
909	my $hex = substr $_[0]->as_hex, 2;	1484	my $hex = substr $_[0]->as_hex, 2;
910	$hex = "0$hex" if 1 & length $hex; # sigh	1485	$hex = "0$hex" if 1 & length $hex; # sigh
911	CBOR::XS::tag $_[0] >= 0 ? 2 : 3, pack "H*", $hex	1486	tag $_[0] >= 0 ? 2 : 3, pack "H*", $hex
912	}	1487	}
913	}	1488	}
914		1489
915	sub Math::BigFloat::TO_CBOR {	1490	sub Math::BigFloat::TO_CBOR {
916	my ($m, $e) = $_[0]->parts;	1491	my ($m, $e) = $_[0]->parts;
		1492
		1493	-9223372036854775808 <= $e && $e <= 18446744073709551615
917	CBOR::XS::tag 4, [$e->numify, $m]	1494	? tag 4, [$e->numify, $m]
		1495	: tag 264, [$e, $m]
		1496	}
		1497
		1498	sub Math::BigRat::TO_CBOR {
		1499	my ($n, $d) = $_[0]->parts;
		1500
		1501	# older versions of BigRat need *1, as they not always return numbers
		1502
		1503	$d*1 == 1
		1504	? $n*1
		1505	: tag 30, [$n1, $d1]
		1506	}
		1507
		1508	sub Time::Piece::TO_CBOR {
		1509	tag 1, 0 + $_[0]->epoch
918	}	1510	}
919		1511
920	XSLoader::load "CBOR::XS", $VERSION;	1512	XSLoader::load "CBOR::XS", $VERSION;
921		1513
922	=head1 SEE ALSO	1514	=head1 SEE ALSO

Diff Legend

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

Comparing CBOR-XS/XS.pm (file contents): Revision 1.23 by root, Fri Nov 22 16:00:30 2013 UTC vs. Revision 1.89 by root, Fri Sep 8 20:03:06 2023 UTC

Diff Legend

Comparing CBOR-XS/XS.pm (file contents):
Revision 1.23 by root, Fri Nov 22 16:00:30 2013 UTC vs.
Revision 1.89 by root, Fri Sep 8 20:03:06 2023 UTC