1XS(3) User Contributed Perl Documentation XS(3)
2
6 CBOR::XS - Concise Binary Object Representation (CBOR, RFC7049)
7
9 use CBOR::XS;
10 $binary_cbor_data = encode_cbor $perl_value;
12 $perl_value = decode_cbor $binary_cbor_data;
13 # OO-interface
15 $coder = CBOR::XS->new;
17 $binary_cbor_data = $coder->encode ($perl_value);
18 $perl_value = $coder->decode ($binary_cbor_data);
19 # prefix decoding
21 my $many_cbor_strings = ...;
23 while (length $many_cbor_strings) {
24 my ($data, $length) = $cbor->decode_prefix ($many_cbor_strings);
25 # data was decoded
26 substr $many_cbor_strings, 0, $length, ""; # remove decoded cbor string
27 }
28
30 This module converts Perl data structures to the Concise Binary Object
31 Representation (CBOR) and vice versa. CBOR is a fast binary
32 serialisation format that aims to use an (almost) superset of the JSON
33 data model, i.e. when you can represent something useful in JSON, you
34 should be able to represent it in CBOR.
35 In short, CBOR is a faster and quite compact binary alternative to
37 JSON, with the added ability of supporting serialisation of Perl
38 objects. (JSON often compresses better than CBOR though, so if you plan
39 to compress the data later and speed is less important you might want
40 to compare both formats first).
41 To give you a general idea about speed, with texts in the megabyte
43 range, "CBOR::XS" usually encodes roughly twice as fast as Storable or
44 JSON::XS and decodes about 15%-30% faster than those. The shorter the
45 data, the worse Storable performs in comparison.
46 Regarding compactness, "CBOR::XS"-encoded data structures are usually
48 about 20% smaller than the same data encoded as (compact) JSON or
49 Storable.
50 In addition to the core CBOR data format, this module implements a
52 number of extensions, to support cyclic and shared data structures (see
53 "allow_sharing" and "allow_cycles"), string deduplication (see
54 "pack_strings") and scalar references (always enabled).
55 The primary goal of this module is to be correct and the secondary goal
57 is to be fast. To reach the latter goal it was written in C.
58 See MAPPING, below, on how CBOR::XS maps perl values to CBOR values and
60 vice versa.
61
63 The following convenience methods are provided by this module. They are
64 exported by default:
65 $cbor_data = encode_cbor $perl_scalar
67 Converts the given Perl data structure to CBOR representation.
68 Croaks on error.
69 $perl_scalar = decode_cbor $cbor_data
71 The opposite of "encode_cbor": expects a valid CBOR string to
72 parse, returning the resulting perl scalar. Croaks on error.
73
75 The object oriented interface lets you configure your own encoding or
76 decoding style, within the limits of supported formats.
77 $cbor = new CBOR::XS
79 Creates a new CBOR::XS object that can be used to de/encode CBOR
80 strings. All boolean flags described below are by default disabled.
81 The mutators for flags all return the CBOR object again and thus
83 calls can be chained:
84 my $cbor = CBOR::XS->new->encode ({a => [1,2]});
86 $cbor = new_safe CBOR::XS
88 Create a new, safe/secure CBOR::XS object. This is similar to
89 "new", but configures the coder object to be safe to use with
90 untrusted data. Currently, this is equivalent to:
91 my $cbor = CBOR::XS
93 ->new
94 ->forbid_objects
95 ->filter (\&CBOR::XS::safe_filter)
96 ->max_size (1e8);
97 But is more future proof (it is better to crash because of a change
99 than to be exploited in other ways).
100 $cbor = $cbor->max_depth ([$maximum_nesting_depth])
102 $max_depth = $cbor->get_max_depth
103 Sets the maximum nesting level (default 512) accepted while
104 encoding or decoding. If a higher nesting level is detected in CBOR
105 data or a Perl data structure, then the encoder and decoder will
106 stop and croak at that point.
107 Nesting level is defined by number of hash- or arrayrefs that the
109 encoder needs to traverse to reach a given point or the number of
110 "{" or "[" characters without their matching closing parenthesis
111 crossed to reach a given character in a string.
112 Setting the maximum depth to one disallows any nesting, so that
114 ensures that the object is only a single hash/object or array.
115 If no argument is given, the highest possible setting will be used,
117 which is rarely useful.
118 Note that nesting is implemented by recursion in C. The default
120 value has been chosen to be as large as typical operating systems
121 allow without crashing.
122 See "SECURITY CONSIDERATIONS", below, for more info on why this is
124 useful.
125 $cbor = $cbor->max_size ([$maximum_string_size])
127 $max_size = $cbor->get_max_size
128 Set the maximum length a CBOR string may have (in bytes) where
129 decoding is being attempted. The default is 0, meaning no limit.
130 When "decode" is called on a string that is longer then this many
131 bytes, it will not attempt to decode the string but throw an
132 exception. This setting has no effect on "encode" (yet).
133 If no argument is given, the limit check will be deactivated (same
135 as when 0 is specified).
136 See "SECURITY CONSIDERATIONS", below, for more info on why this is
138 useful.
139 $cbor = $cbor->allow_unknown ([$enable])
141 $enabled = $cbor->get_allow_unknown
142 If $enable is true (or missing), then "encode" will not throw an
143 exception when it encounters values it cannot represent in CBOR
144 (for example, filehandles) but instead will encode a CBOR "error"
145 value.
146 If $enable is false (the default), then "encode" will throw an
148 exception when it encounters anything it cannot encode as CBOR.
149 This option does not affect "decode" in any way, and it is
151 recommended to leave it off unless you know your communications
152 partner.
153 $cbor = $cbor->allow_sharing ([$enable])
155 $enabled = $cbor->get_allow_sharing
156 If $enable is true (or missing), then "encode" will not double-
157 encode values that have been referenced before (e.g. when the same
158 object, such as an array, is referenced multiple times), but
159 instead will emit a reference to the earlier value.
160 This means that such values will only be encoded once, and will not
162 result in a deep cloning of the value on decode, in decoders
163 supporting the value sharing extension. This also makes it possible
164 to encode cyclic data structures (which need "allow_cycles" to be
165 enabled to be decoded by this module).
166 It is recommended to leave it off unless you know your
168 communication partner supports the value sharing extensions to CBOR
169 (<http://cbor.schmorp.de/value-sharing>), as without decoder
170 support, the resulting data structure might be unusable.
171 Detecting shared values incurs a runtime overhead when values are
173 encoded that have a reference counter large than one, and might
174 unnecessarily increase the encoded size, as potentially shared
175 values are encode as shareable whether or not they are actually
176 shared.
177 At the moment, only targets of references can be shared (e.g.
179 scalars, arrays or hashes pointed to by a reference). Weirder
180 constructs, such as an array with multiple "copies" of the same
181 string, which are hard but not impossible to create in Perl, are
182 not supported (this is the same as with Storable).
183 If $enable is false (the default), then "encode" will encode shared
185 data structures repeatedly, unsharing them in the process. Cyclic
186 data structures cannot be encoded in this mode.
187 This option does not affect "decode" in any way - shared values and
189 references will always be decoded properly if present.
190 $cbor = $cbor->allow_cycles ([$enable])
192 $enabled = $cbor->get_allow_cycles
193 If $enable is true (or missing), then "decode" will happily decode
194 self-referential (cyclic) data structures. By default these will
195 not be decoded, as they need manual cleanup to avoid memory leaks,
196 so code that isn't prepared for this will not leak memory.
197 If $enable is false (the default), then "decode" will throw an
199 error when it encounters a self-referential/cyclic data structure.
200 FUTURE DIRECTION: the motivation behind this option is to avoid
202 real cycles - future versions of this module might chose to decode
203 cyclic data structures using weak references when this option is
204 off, instead of throwing an error.
205 This option does not affect "encode" in any way - shared values and
207 references will always be encoded properly if present.
208 $cbor = $cbor->forbid_objects ([$enable])
210 $enabled = $cbor->get_forbid_objects
211 Disables the use of the object serialiser protocol.
212 If $enable is true (or missing), then "encode" will will throw an
214 exception when it encounters perl objects that would be encoded
215 using the perl-object tag (26). When "decode" encounters such tags,
216 it will fall back to the general filter/tagged logic as if this
217 were an unknown tag (by default resulting in a "CBOR::XC::Tagged"
218 object).
219 If $enable is false (the default), then "encode" will use the
221 Types::Serialiser object serialisation protocol to serialise
222 objects into perl-object tags, and "decode" will do the same to
223 decode such tags.
224 See "SECURITY CONSIDERATIONS", below, for more info on why
226 forbidding this protocol can be useful.
227 $cbor = $cbor->pack_strings ([$enable])
229 $enabled = $cbor->get_pack_strings
230 If $enable is true (or missing), then "encode" will try not to
231 encode the same string twice, but will instead encode a reference
232 to the string instead. Depending on your data format, this can save
233 a lot of space, but also results in a very large runtime overhead
234 (expect encoding times to be 2-4 times as high as without).
235 It is recommended to leave it off unless you know your
237 communications partner supports the stringref extension to CBOR
238 (<http://cbor.schmorp.de/stringref>), as without decoder support,
239 the resulting data structure might not be usable.
240 If $enable is false (the default), then "encode" will encode
242 strings the standard CBOR way.
243 This option does not affect "decode" in any way - string references
245 will always be decoded properly if present.
246 $cbor = $cbor->text_keys ([$enable])
248 $enabled = $cbor->get_text_keys
249 If $enabled is true (or missing), then "encode" will encode all
250 perl hash keys as CBOR text strings/UTF-8 string, upgrading them as
251 needed.
252 If $enable is false (the default), then "encode" will encode hash
254 keys normally - upgraded perl strings (strings internally encoded
255 as UTF-8) as CBOR text strings, and downgraded perl strings as CBOR
256 byte strings.
257 This option does not affect "decode" in any way.
259 This option is useful for interoperability with CBOR decoders that
261 don't treat byte strings as a form of text. It is especially useful
262 as Perl gives very little control over hash keys.
263 Enabling this option can be slow, as all downgraded hash keys that
265 are encoded need to be scanned and converted to UTF-8.
266 $cbor = $cbor->text_strings ([$enable])
268 $enabled = $cbor->get_text_strings
269 This option works similar to "text_keys", above, but works on all
270 strings (including hash keys), so "text_keys" has no further effect
271 after enabling "text_strings".
272 If $enabled is true (or missing), then "encode" will encode all
274 perl strings as CBOR text strings/UTF-8 strings, upgrading them as
275 needed.
276 If $enable is false (the default), then "encode" will encode
278 strings normally (but see "text_keys") - upgraded perl strings
279 (strings internally encoded as UTF-8) as CBOR text strings, and
280 downgraded perl strings as CBOR byte strings.
281 This option does not affect "decode" in any way.
283 This option has similar advantages and disadvantages as
285 "text_keys". In addition, this option effectively removes the
286 ability to encode byte strings, which might break some "FREEZE" and
287 "TO_CBOR" methods that rely on this, such as bignum encoding, so
288 this option is mainly useful for very simple data.
289 $cbor = $cbor->validate_utf8 ([$enable])
291 $enabled = $cbor->get_validate_utf8
292 If $enable is true (or missing), then "decode" will validate that
293 elements (text strings) containing UTF-8 data in fact contain valid
294 UTF-8 data (instead of blindly accepting it). This validation
295 obviously takes extra time during decoding.
296 The concept of "valid UTF-8" used is perl's concept, which is a
298 superset of the official UTF-8.
299 If $enable is false (the default), then "decode" will blindly
301 accept UTF-8 data, marking them as valid UTF-8 in the resulting
302 data structure regardless of whether that's true or not.
303 Perl isn't too happy about corrupted UTF-8 in strings, but should
305 generally not crash or do similarly evil things. Extensions might
306 be not so forgiving, so it's recommended to turn on this setting if
307 you receive untrusted CBOR.
308 This option does not affect "encode" in any way - strings that are
310 supposedly valid UTF-8 will simply be dumped into the resulting
311 CBOR string without checking whether that is, in fact, true or not.
312 $cbor = $cbor->filter ([$cb->($tag, $value)])
314 $cb_or_undef = $cbor->get_filter
315 Sets or replaces the tagged value decoding filter (when $cb is
316 specified) or clears the filter (if no argument or "undef" is
317 provided).
318 The filter callback is called only during decoding, when a non-
320 enforced tagged value has been decoded (see "TAG HANDLING AND
321 EXTENSIONS" for a list of enforced tags). For specific tags, it's
322 often better to provide a default converter using the
323 %CBOR::XS::FILTER hash (see below).
324 The first argument is the numerical tag, the second is the
326 (decoded) value that has been tagged.
327 The filter function should return either exactly one value, which
329 will replace the tagged value in the decoded data structure, or no
330 values, which will result in default handling, which currently
331 means the decoder creates a "CBOR::XS::Tagged" object to hold the
332 tag and the value.
333 When the filter is cleared (the default state), the default filter
335 function, "CBOR::XS::default_filter", is used. This function simply
336 looks up the tag in the %CBOR::XS::FILTER hash. If an entry exists
337 it must be a code reference that is called with tag and value, and
338 is responsible for decoding the value. If no entry exists, it
339 returns no values. "CBOR::XS" provides a number of default filter
340 functions already, the the %CBOR::XS::FILTER hash can be freely
341 extended with more.
342 "CBOR::XS" additionally provides an alternative filter function
344 that is supposed to be safe to use with untrusted data (which the
345 default filter might not), called "CBOR::XS::safe_filter", which
346 works the same as the "default_filter" but uses the
347 %CBOR::XS::SAFE_FILTER variable instead. It is prepopulated with
348 the tag decoding functions that are deemed safe (basically the same
349 as %CBOR::XS::FILTER without all the bignum tags), and can be
350 extended by user code as wlel, although, obviously, one should be
351 very careful about adding decoding functions here, since the
352 expectation is that they are safe to use on untrusted data, after
353 all.
354 Example: decode all tags not handled internally into
356 "CBOR::XS::Tagged" objects, with no other special handling (useful
357 when working with potentially "unsafe" CBOR data).
358 CBOR::XS->new->filter (sub { })->decode ($cbor_data);
360 Example: provide a global filter for tag 1347375694, converting the
362 value into some string form.
363 $CBOR::XS::FILTER{1347375694} = sub {
365 my ($tag, $value);
366 "tag 1347375694 value $value"
368 };
369 Example: provide your own filter function that looks up tags in
371 your own hash:
372 my %my_filter = (
374 998347484 => sub {
375 my ($tag, $value);
376 "tag 998347484 value $value"
378 };
379 );
380 my $coder = CBOR::XS->new->filter (sub {
382 &{ $my_filter{$_[0]} or return }
383 });
384 Example: use the safe filter function (see "SECURITY
386 CONSIDERATIONS" for more considerations on security).
387 CBOR::XS->new->filter (\&CBOR::XS::safe_filter)->decode ($cbor_data);
389 $cbor_data = $cbor->encode ($perl_scalar)
391 Converts the given Perl data structure (a scalar value) to its CBOR
392 representation.
393 $perl_scalar = $cbor->decode ($cbor_data)
395 The opposite of "encode": expects CBOR data and tries to parse it,
396 returning the resulting simple scalar or reference. Croaks on
397 error.
398 ($perl_scalar, $octets) = $cbor->decode_prefix ($cbor_data)
400 This works like the "decode" method, but instead of raising an
401 exception when there is trailing garbage after the CBOR string, it
402 will silently stop parsing there and return the number of
403 characters consumed so far.
404 This is useful if your CBOR texts are not delimited by an outer
406 protocol and you need to know where the first CBOR string ends amd
407 the next one starts.
408 CBOR::XS->new->decode_prefix ("......")
410 => ("...", 3)
411 INCREMENTAL PARSING
413 In some cases, there is the need for incremental parsing of JSON texts.
414 While this module always has to keep both CBOR text and resulting Perl
415 data structure in memory at one time, it does allow you to parse a CBOR
416 stream incrementally, using a similar to using "decode_prefix" to see
417 if a full CBOR object is available, but is much more efficient.
418 It basically works by parsing as much of a CBOR string as possible - if
420 the CBOR data is not complete yet, the pasrer will remember where it
421 was, to be able to restart when more data has been accumulated. Once
422 enough data is available to either decode a complete CBOR value or
423 raise an error, a real decode will be attempted.
424 A typical use case would be a network protocol that consists of sending
426 and receiving CBOR-encoded messages. The solution that works with CBOR
427 and about anything else is by prepending a length to every CBOR value,
428 so the receiver knows how many octets to read. More compact (and
429 slightly slower) would be to just send CBOR values back-to-back, as
430 "CBOR::XS" knows where a CBOR value ends, and doesn't need an explicit
431 length.
432 The following methods help with this:
434 @decoded = $cbor->incr_parse ($buffer)
436 This method attempts to decode exactly one CBOR value from the
437 beginning of the given $buffer. The value is removed from the
438 $buffer on success. When $buffer doesn't contain a complete value
439 yet, it returns nothing. Finally, when the $buffer doesn't start
440 with something that could ever be a valid CBOR value, it raises an
441 exception, just as "decode" would. In the latter case the decoder
442 state is undefined and must be reset before being able to parse
443 further.
444 This method modifies the $buffer in place. When no CBOR value can
446 be decoded, the decoder stores the current string offset. On the
447 next call, continues decoding at the place where it stopped before.
448 For this to make sense, the $buffer must begin with the same octets
449 as on previous unsuccessful calls.
450 You can call this method in scalar context, in which case it either
452 returns a decoded value or "undef". This makes it impossible to
453 distinguish between CBOR null values (which decode to "undef") and
454 an unsuccessful decode, which is often acceptable.
455 @decoded = $cbor->incr_parse_multiple ($buffer)
457 Same as "incr_parse", but attempts to decode as many CBOR values as
458 possible in one go, instead of at most one. Calls to "incr_parse"
459 and "incr_parse_multiple" can be interleaved.
460 $cbor->incr_reset
462 Resets the incremental decoder. This throws away any saved state,
463 so that subsequent calls to "incr_parse" or "incr_parse_multiple"
464 start to parse a new CBOR value from the beginning of the $buffer
465 again.
466 This method can be called at any time, but it must be called if you
468 want to change your $buffer or there was a decoding error and you
469 want to reuse the $cbor object for future incremental parsings.
470
472 This section describes how CBOR::XS maps Perl values to CBOR values and
473 vice versa. These mappings are designed to "do the right thing" in most
474 circumstances automatically, preserving round-tripping characteristics
475 (what you put in comes out as something equivalent).
476 For the more enlightened: note that in the following descriptions,
478 lowercase perl refers to the Perl interpreter, while uppercase Perl
479 refers to the abstract Perl language itself.
480 CBOR -> PERL
482 integers
483 CBOR integers become (numeric) perl scalars. On perls without 64
484 bit support, 64 bit integers will be truncated or otherwise
485 corrupted.
486 byte strings
488 Byte strings will become octet strings in Perl (the Byte values
489 0..255 will simply become characters of the same value in Perl).
490 UTF-8 strings
492 UTF-8 strings in CBOR will be decoded, i.e. the UTF-8 octets will
493 be decoded into proper Unicode code points. At the moment, the
494 validity of the UTF-8 octets will not be validated - corrupt input
495 will result in corrupted Perl strings.
496 arrays, maps
498 CBOR arrays and CBOR maps will be converted into references to a
499 Perl array or hash, respectively. The keys of the map will be
500 stringified during this process.
501 null
503 CBOR null becomes "undef" in Perl.
504 true, false, undefined
506 These CBOR values become "Types:Serialiser::true",
507 "Types:Serialiser::false" and "Types::Serialiser::error",
508 respectively. They are overloaded to act almost exactly like the
509 numbers 1 and 0 (for true and false) or to throw an exception on
510 access (for error). See the Types::Serialiser manpage for details.
511 tagged values
513 Tagged items consists of a numeric tag and another CBOR value.
514 See "TAG HANDLING AND EXTENSIONS" and the description of "->filter"
516 for details on which tags are handled how.
517 anything else
519 Anything else (e.g. unsupported simple values) will raise a
520 decoding error.
521 PERL -> CBOR
523 The mapping from Perl to CBOR is slightly more difficult, as Perl is a
524 typeless language. That means this module can only guess which CBOR
525 type is meant by a perl value.
526 hash references
528 Perl hash references become CBOR maps. As there is no inherent
529 ordering in hash keys (or CBOR maps), they will usually be encoded
530 in a pseudo-random order. This order can be different each time a
531 hash is encoded.
532 Currently, tied hashes will use the indefinite-length format, while
534 normal hashes will use the fixed-length format.
535 array references
537 Perl array references become fixed-length CBOR arrays.
538 other references
540 Other unblessed references will be represented using the
541 indirection tag extension (tag value 22098,
542 <http://cbor.schmorp.de/indirection>). CBOR decoders are guaranteed
543 to be able to decode these values somehow, by either "doing the
544 right thing", decoding into a generic tagged object, simply
545 ignoring the tag, or something else.
546 CBOR::XS::Tagged objects
548 Objects of this type must be arrays consisting of a single "[tag,
549 value]" pair. The (numerical) tag will be encoded as a CBOR tag,
550 the value will be encoded as appropriate for the value. You must
551 use "CBOR::XS::tag" to create such objects.
552 Types::Serialiser::true, Types::Serialiser::false,
554 Types::Serialiser::error
555 These special values become CBOR true, CBOR false and CBOR
556 undefined values, respectively. You can also use "\1", "\0" and
557 "\undef" directly if you want.
558 other blessed objects
560 Other blessed objects are serialised via "TO_CBOR" or "FREEZE". See
561 "TAG HANDLING AND EXTENSIONS" for specific classes handled by this
562 module, and "OBJECT SERIALISATION" for generic object
563 serialisation.
564 simple scalars
566 Simple Perl scalars (any scalar that is not a reference) are the
567 most difficult objects to encode: CBOR::XS will encode undefined
568 scalars as CBOR null values, scalars that have last been used in a
569 string context before encoding as CBOR strings, and anything else
570 as number value:
571 # dump as number
573 encode_cbor [2] # yields [2]
574 encode_cbor [-3.0e17] # yields [-3e+17]
575 my $value = 5; encode_cbor [$value] # yields [5]
576 # used as string, so dump as string (either byte or text)
578 print $value;
579 encode_cbor [$value] # yields ["5"]
580 # undef becomes null
582 encode_cbor [undef] # yields [null]
583 You can force the type to be a CBOR string by stringifying it:
585 my $x = 3.1; # some variable containing a number
587 "$x"; # stringified
588 $x .= ""; # another, more awkward way to stringify
589 print $x; # perl does it for you, too, quite often
590 You can force whether a string is encoded as byte or text string by
592 using "utf8::upgrade" and "utf8::downgrade" (if "text_strings" is
593 disabled):
594 utf8::upgrade $x; # encode $x as text string
596 utf8::downgrade $x; # encode $x as byte string
597 Perl doesn't define what operations up- and downgrade strings, so
599 if the difference between byte and text is important, you should
600 up- or downgrade your string as late as possible before encoding.
601 You can also force the use of CBOR text strings by using
602 "text_keys" or "text_strings".
603 You can force the type to be a CBOR number by numifying it:
605 my $x = "3"; # some variable containing a string
607 $x += 0; # numify it, ensuring it will be dumped as a number
608 $x *= 1; # same thing, the choice is yours.
609 You can not currently force the type in other, less obscure, ways.
611 Tell me if you need this capability (but don't forget to explain
612 why it's needed :).
613 Perl values that seem to be integers generally use the shortest
615 possible representation. Floating-point values will use either the
616 IEEE single format if possible without loss of precision, otherwise
617 the IEEE double format will be used. Perls that use formats other
618 than IEEE double to represent numerical values are supported, but
619 might suffer loss of precision.
620 OBJECT SERIALISATION
622 This module implements both a CBOR-specific and the generic
623 Types::Serialier object serialisation protocol. The following
624 subsections explain both methods.
625 ENCODING
627 This module knows two way to serialise a Perl object: The CBOR-specific
629 way, and the generic way.
630 Whenever the encoder encounters a Perl object that it cannot serialise
632 directly (most of them), it will first look up the "TO_CBOR" method on
633 it.
634 If it has a "TO_CBOR" method, it will call it with the object as only
636 argument, and expects exactly one return value, which it will then
637 substitute and encode it in the place of the object.
638 Otherwise, it will look up the "FREEZE" method. If it exists, it will
640 call it with the object as first argument, and the constant string
641 "CBOR" as the second argument, to distinguish it from other
642 serialisers.
643 The "FREEZE" method can return any number of values (i.e. zero or
645 more). These will be encoded as CBOR perl object, together with the
646 classname.
647 These methods MUST NOT change the data structure that is being
649 serialised. Failure to comply to this can result in memory corruption -
650 and worse.
651 If an object supports neither "TO_CBOR" nor "FREEZE", encoding will
653 fail with an error.
654 DECODING
656 Objects encoded via "TO_CBOR" cannot (normally) be automatically
658 decoded, but objects encoded via "FREEZE" can be decoded using the
659 following protocol:
660 When an encoded CBOR perl object is encountered by the decoder, it will
662 look up the "THAW" method, by using the stored classname, and will fail
663 if the method cannot be found.
664 After the lookup it will call the "THAW" method with the stored
666 classname as first argument, the constant string "CBOR" as second
667 argument, and all values returned by "FREEZE" as remaining arguments.
668 EXAMPLES
670 Here is an example "TO_CBOR" method:
672 sub My::Object::TO_CBOR {
674 my ($obj) = @_;
675 ["this is a serialised My::Object object", $obj->{id}]
677 }
678 When a "My::Object" is encoded to CBOR, it will instead encode a simple
680 array with two members: a string, and the "object id". Decoding this
681 CBOR string will yield a normal perl array reference in place of the
682 object.
683 A more useful and practical example would be a serialisation method for
685 the URI module. CBOR has a custom tag value for URIs, namely 32:
686 sub URI::TO_CBOR {
688 my ($self) = @_;
689 my $uri = "$self"; # stringify uri
690 utf8::upgrade $uri; # make sure it will be encoded as UTF-8 string
691 CBOR::XS::tag 32, "$_[0]"
692 }
693 This will encode URIs as a UTF-8 string with tag 32, which indicates an
695 URI.
696 Decoding such an URI will not (currently) give you an URI object, but
698 instead a CBOR::XS::Tagged object with tag number 32 and the string -
699 exactly what was returned by "TO_CBOR".
700 To serialise an object so it can automatically be deserialised, you
702 need to use "FREEZE" and "THAW". To take the URI module as example,
703 this would be a possible implementation:
704 sub URI::FREEZE {
706 my ($self, $serialiser) = @_;
707 "$self" # encode url string
708 }
709 sub URI::THAW {
711 my ($class, $serialiser, $uri) = @_;
712 $class->new ($uri)
713 }
714 Unlike "TO_CBOR", multiple values can be returned by "FREEZE". For
716 example, a "FREEZE" method that returns "type", "id" and "variant"
717 values would cause an invocation of "THAW" with 5 arguments:
718 sub My::Object::FREEZE {
720 my ($self, $serialiser) = @_;
721 ($self->{type}, $self->{id}, $self->{variant})
723 }
724 sub My::Object::THAW {
726 my ($class, $serialiser, $type, $id, $variant) = @_;
727 $class-<new (type => $type, id => $id, variant => $variant)
729 }
730
732 There is no way to distinguish CBOR from other formats
733 programmatically. To make it easier to distinguish CBOR from other
734 formats, the CBOR specification has a special "magic string" that can
735 be prepended to any CBOR string without changing its meaning.
736 This string is available as $CBOR::XS::MAGIC. This module does not
738 prepend this string to the CBOR data it generates, but it will ignore
739 it if present, so users can prepend this string as a "file type"
740 indicator as required.
741
743 CBOR has the concept of tagged values - any CBOR value can be tagged
744 with a numeric 64 bit number, which are centrally administered.
745 "CBOR::XS" handles a few tags internally when en- or decoding. You can
747 also create tags yourself by encoding "CBOR::XS::Tagged" objects, and
748 the decoder will create "CBOR::XS::Tagged" objects itself when it hits
749 an unknown tag.
750 These objects are simply blessed array references - the first member of
752 the array being the numerical tag, the second being the value.
753 You can interact with "CBOR::XS::Tagged" objects in the following ways:
755 $tagged = CBOR::XS::tag $tag, $value
757 This function(!) creates a new "CBOR::XS::Tagged" object using the
758 given $tag (0..2**64-1) to tag the given $value (which can be any
759 Perl value that can be encoded in CBOR, including serialisable Perl
760 objects and "CBOR::XS::Tagged" objects).
761 $tagged->[0]
763 $tagged->[0] = $new_tag
764 $tag = $tagged->tag
765 $new_tag = $tagged->tag ($new_tag)
766 Access/mutate the tag.
767 $tagged->[1]
769 $tagged->[1] = $new_value
770 $value = $tagged->value
771 $new_value = $tagged->value ($new_value)
772 Access/mutate the tagged value.
773 EXAMPLES
775 Here are some examples of "CBOR::XS::Tagged" uses to tag objects.
776 You can look up CBOR tag value and emanings in the IANA registry at
778 <http://www.iana.org/assignments/cbor-tags/cbor-tags.xhtml>.
779 Prepend a magic header ($CBOR::XS::MAGIC):
781 my $cbor = encode_cbor CBOR::XS::tag 55799, $value;
783 # same as:
784 my $cbor = $CBOR::XS::MAGIC . encode_cbor $value;
785 Serialise some URIs and a regex in an array:
787 my $cbor = encode_cbor [
789 (CBOR::XS::tag 32, "http://www.nethype.de/"),
790 (CBOR::XS::tag 32, "http://software.schmorp.de/"),
791 (CBOR::XS::tag 35, "^[Pp][Ee][Rr][lL]\$"),
792 ];
793 Wrap CBOR data in CBOR:
795 my $cbor_cbor = encode_cbor
797 CBOR::XS::tag 24,
798 encode_cbor [1, 2, 3];
799
801 This section describes how this module handles specific tagged values
802 and extensions. If a tag is not mentioned here and no additional
803 filters are provided for it, then the default handling applies
804 (creating a CBOR::XS::Tagged object on decoding, and only encoding the
805 tag when explicitly requested).
806 Tags not handled specifically are currently converted into a
808 CBOR::XS::Tagged object, which is simply a blessed array reference
809 consisting of the numeric tag value followed by the (decoded) CBOR
810 value.
811 Future versions of this module reserve the right to special case
813 additional tags (such as base64url).
814 ENFORCED TAGS
816 These tags are always handled when decoding, and their handling cannot
817 be overridden by the user.
818 26 (perl-object, <http://cbor.schmorp.de/perl-object>)
820 These tags are automatically created (and decoded) for serialisable
821 objects using the "FREEZE/THAW" methods (the Types::Serialier
822 object serialisation protocol). See "OBJECT SERIALISATION" for
823 details.
824 28, 29 (shareable, sharedref, <http://cbor.schmorp.de/value-sharing>)
826 These tags are automatically decoded when encountered (and they do
827 not result in a cyclic data structure, see "allow_cycles"),
828 resulting in shared values in the decoded object. They are only
829 encoded, however, when "allow_sharing" is enabled.
830 Not all shared values can be successfully decoded: values that
832 reference themselves will currently decode as "undef" (this is not
833 the same as a reference pointing to itself, which will be
834 represented as a value that contains an indirect reference to
835 itself - these will be decoded properly).
836 Note that considerably more shared value data structures can be
838 decoded than will be encoded - currently, only values pointed to by
839 references will be shared, others will not. While non-reference
840 shared values can be generated in Perl with some effort, they were
841 considered too unimportant to be supported in the encoder. The
842 decoder, however, will decode these values as shared values.
843 256, 25 (stringref-namespace, stringref,
845 <http://cbor.schmorp.de/stringref>)
846 These tags are automatically decoded when encountered. They are
847 only encoded, however, when "pack_strings" is enabled.
848 22098 (indirection, <http://cbor.schmorp.de/indirection>)
850 This tag is automatically generated when a reference are
851 encountered (with the exception of hash and array references). It
852 is converted to a reference when decoding.
853 55799 (self-describe CBOR, RFC 7049)
855 This value is not generated on encoding (unless explicitly
856 requested by the user), and is simply ignored when decoding.
857 NON-ENFORCED TAGS
859 These tags have default filters provided when decoding. Their handling
860 can be overridden by changing the %CBOR::XS::FILTER entry for the tag,
861 or by providing a custom "filter" callback when decoding.
862 When they result in decoding into a specific Perl class, the module
864 usually provides a corresponding "TO_CBOR" method as well.
865 When any of these need to load additional modules that are not part of
867 the perl core distribution (e.g. URI), it is (currently) up to the user
868 to provide these modules. The decoding usually fails with an exception
869 if the required module cannot be loaded.
870 0, 1 (date/time string, seconds since the epoch)
872 These tags are decoded into Time::Piece objects. The corresponding
873 "Time::Piece::TO_CBOR" method always encodes into tag 1 values
874 currently.
875 The Time::Piece API is generally surprisingly bad, and fractional
877 seconds are only accidentally kept intact, so watch out. On the
878 plus side, the module comes with perl since 5.10, which has to
879 count for something.
880 2, 3 (positive/negative bignum)
882 These tags are decoded into Math::BigInt objects. The corresponding
883 "Math::BigInt::TO_CBOR" method encodes "small" bigints into normal
884 CBOR integers, and others into positive/negative CBOR bignums.
885 4, 5, 264, 265 (decimal fraction/bigfloat)
887 Both decimal fractions and bigfloats are decoded into
888 Math::BigFloat objects. The corresponding "Math::BigFloat::TO_CBOR"
889 method always encodes into a decimal fraction (either tag 4 or
890 264).
891 NaN and infinities are not encoded properly, as they cannot be
893 represented in CBOR.
894 See "BIGNUM SECURITY CONSIDERATIONS" for more info.
896 30 (rational numbers)
898 These tags are decoded into Math::BigRat objects. The corresponding
899 "Math::BigRat::TO_CBOR" method encodes rational numbers with
900 denominator 1 via their numerator only, i.e., they become normal
901 integers or "bignums".
902 See "BIGNUM SECURITY CONSIDERATIONS" for more info.
904 21, 22, 23 (expected later JSON conversion)
906 CBOR::XS is not a CBOR-to-JSON converter, and will simply ignore
907 these tags.
908 32 (URI)
910 These objects decode into URI objects. The corresponding
911 "URI::TO_CBOR" method again results in a CBOR URI value.
912
914 CBOR is supposed to implement a superset of the JSON data model, and
915 is, with some coercion, able to represent all JSON texts (something
916 that other "binary JSON" formats such as BSON generally do not
917 support).
918 CBOR implements some extra hints and support for JSON interoperability,
920 and the spec offers further guidance for conversion between CBOR and
921 JSON. None of this is currently implemented in CBOR, and the guidelines
922 in the spec do not result in correct round-tripping of data. If JSON
923 interoperability is improved in the future, then the goal will be to
924 ensure that decoded JSON data will round-trip encoding and decoding to
925 CBOR intact.
926
928 Tl;dr... if you want to decode or encode CBOR from untrusted sources,
929 you should start with a coder object created via "new_safe":
930 my $coder = CBOR::XS->new_safe;
932 my $data = $coder->decode ($cbor_text);
934 my $cbor = $coder->encode ($data);
935 Longer version: When you are using CBOR in a protocol, talking to
937 untrusted potentially hostile creatures requires some thought:
938 Security of the CBOR decoder itself
940 First and foremost, your CBOR decoder should be secure, that is,
941 should not have any buffer overflows or similar bugs that could
942 potentially be exploited. Obviously, this module should ensure that
943 and I am trying hard on making that true, but you never know.
944 CBOR::XS can invoke almost arbitrary callbacks during decoding
946 CBOR::XS supports object serialisation - decoding CBOR can cause
947 calls to any "THAW" method in any package that exists in your
948 process (that is, CBOR::XS will not try to load modules, but any
949 existing "THAW" method or function can be called, so they all have
950 to be secure).
951 Less obviously, it will also invoke "TO_CBOR" and "FREEZE" methods
953 - even if all your "THAW" methods are secure, encoding data
954 structures from untrusted sources can invoke those and trigger bugs
955 in those.
956 So, if you are not sure about the security of all the modules you
958 have loaded (you shouldn't), you should disable this part using
959 "forbid_objects".
960 CBOR can be extended with tags that call library code
962 CBOR can be extended with tags, and "CBOR::XS" has a registry of
963 conversion functions for many existing tags that can be extended
964 via third-party modules (see the "filter" method).
965 If you don't trust these, you should configure the "safe" filter
967 function, "CBOR::XS::safe_filter", which by default only includes
968 conversion functions that are considered "safe" by the author (but
969 again, they can be extended by third party modules).
970 Depending on your level of paranoia, you can use the "safe" filter:
972 $cbor->filter (\&CBOR::XS::safe_filter);
974 ... your own filter...
976 $cbor->filter (sub { ... do your stuffs here ... });
978 ... or even no filter at all, disabling all tag decoding:
980 $cbor->filter (sub { });
982 This is never a problem for encoding, as the tag mechanism only
984 exists in CBOR texts.
985 Resource-starving attacks: object memory usage
987 You need to avoid resource-starving attacks. That means you should
988 limit the size of CBOR data you accept, or make sure then when your
989 resources run out, that's just fine (e.g. by using a separate
990 process that can crash safely). The size of a CBOR string in octets
991 is usually a good indication of the size of the resources required
992 to decode it into a Perl structure. While CBOR::XS can check the
993 size of the CBOR text (using "max_size"), it might be too late when
994 you already have it in memory, so you might want to check the size
995 before you accept the string.
996 As for encoding, it is possible to construct data structures that
998 are relatively small but result in large CBOR texts (for example by
999 having an array full of references to the same big data structure,
1000 which will all be deep-cloned during encoding by default). This is
1001 rarely an actual issue (and the worst case is still just running
1002 out of memory), but you can reduce this risk by using
1003 "allow_sharing".
1004 Resource-starving attacks: stack overflows
1006 CBOR::XS recurses using the C stack when decoding objects and
1007 arrays. The C stack is a limited resource: for instance, on my
1008 amd64 machine with 8MB of stack size I can decode around 180k
1009 nested arrays but only 14k nested CBOR objects (due to perl itself
1010 recursing deeply on croak to free the temporary). If that is
1011 exceeded, the program crashes. To be conservative, the default
1012 nesting limit is set to 512. If your process has a smaller stack,
1013 you should adjust this setting accordingly with the "max_depth"
1014 method.
1015 Resource-starving attacks: CPU en-/decoding complexity
1017 CBOR::XS will use the Math::BigInt, Math::BigFloat and Math::BigRat
1018 libraries to represent encode/decode bignums. These can be very
1019 slow (as in, centuries of CPU time) and can even crash your program
1020 (and are generally not very trustworthy). See the next section for
1021 details.
1022 Data breaches: leaking information in error messages
1024 CBOR::XS might leak contents of your Perl data structures in its
1025 error messages, so when you serialise sensitive information you
1026 might want to make sure that exceptions thrown by CBOR::XS will not
1027 end up in front of untrusted eyes.
1028 Something else...
1030 Something else could bomb you, too, that I forgot to think of. In
1031 that case, you get to keep the pieces. I am always open for hints,
1032 though...
1033
1035 CBOR::XS provides a "TO_CBOR" method for both Math::BigInt and
1036 Math::BigFloat that tries to encode the number in the simplest possible
1037 way, that is, either a CBOR integer, a CBOR bigint/decimal fraction
1038 (tag 4) or an arbitrary-exponent decimal fraction (tag 264). Rational
1039 numbers (Math::BigRat, tag 30) can also contain bignums as members.
1040 CBOR::XS will also understand base-2 bigfloat or arbitrary-exponent
1042 bigfloats (tags 5 and 265), but it will never generate these on its
1043 own.
1044 Using the built-in Math::BigInt::Calc support, encoding and decoding
1046 decimal fractions is generally fast. Decoding bigints can be slow for
1047 very big numbers (tens of thousands of digits, something that could
1048 potentially be caught by limiting the size of CBOR texts), and decoding
1049 bigfloats or arbitrary-exponent bigfloats can be extremely slow
1050 (minutes, decades) for large exponents (roughly 40 bit and longer).
1051 Additionally, Math::BigInt can take advantage of other bignum
1053 libraries, such as Math::GMP, which cannot handle big floats with large
1054 exponents, and might simply abort or crash your program, due to their
1055 code quality.
1056 This can be a concern if you want to parse untrusted CBOR. If it is,
1058 you might want to disable decoding of tag 2 (bigint) and 3 (negative
1059 bigint) types. You should also disable types 5 and 265, as these can be
1060 slow even without bigints.
1061 Disabling bigints will also partially or fully disable types that rely
1063 on them, e.g. rational numbers that use bignums.
1064
1066 This section contains some random implementation notes. They do not
1067 describe guaranteed behaviour, but merely behaviour as-is implemented
1068 right now.
1069 64 bit integers are only properly decoded when Perl was built with 64
1071 bit support.
1072 Strings and arrays are encoded with a definite length. Hashes as well,
1074 unless they are tied (or otherwise magical).
1075 Only the double data type is supported for NV data types - when Perl
1077 uses long double to represent floating point values, they might not be
1078 encoded properly. Half precision types are accepted, but not encoded.
1079 Strict mode and canonical mode are not implemented.
1081
1083 On perls that were built without 64 bit integer support (these are rare
1084 nowadays, even on 32 bit architectures, as all major Perl distributions
1085 are built with 64 bit integer support), support for any kind of 64 bit
1086 integer in CBOR is very limited - most likely, these 64 bit values will
1087 be truncated, corrupted, or otherwise not decoded correctly. This also
1088 includes string, array and map sizes that are stored as 64 bit
1089 integers.
1090
1092 This module is not guaranteed to be thread safe and there are no plans
1093 to change this until Perl gets thread support (as opposed to the
1094 horribly slow so-called "threads" which are simply slow and bloated
1095 process simulations - use fork, it's much faster, cheaper, better).
1096 (It might actually work, but you have been warned).
1098
1100 While the goal of this module is to be correct, that unfortunately does
1101 not mean it's bug-free, only that I think its design is bug-free. If
1102 you keep reporting bugs they will be fixed swiftly, though.
1103 Please refrain from using rt.cpan.org or any other bug reporting
1105 service. I put the contact address into my modules for a reason.
1106
1108 The JSON and JSON::XS modules that do similar, but human-readable,
1109 serialisation.
1110 The Types::Serialiser module provides the data model for true, false
1112 and error values.
1113
1115 Marc Lehmann <schmorp@schmorp.de>
1116 http://home.schmorp.de/
1117perl v5.30.0 2019-07-26 XS(3)