1 |
NAME |
2 |
CBOR::XS - Concise Binary Object Representation (CBOR, RFC7049) |
3 |
|
4 |
SYNOPSIS |
5 |
use CBOR::XS; |
6 |
|
7 |
$binary_cbor_data = encode_cbor $perl_value; |
8 |
$perl_value = decode_cbor $binary_cbor_data; |
9 |
|
10 |
# OO-interface |
11 |
|
12 |
$coder = CBOR::XS->new; |
13 |
$binary_cbor_data = $coder->encode ($perl_value); |
14 |
$perl_value = $coder->decode ($binary_cbor_data); |
15 |
|
16 |
# prefix decoding |
17 |
|
18 |
my $many_cbor_strings = ...; |
19 |
while (length $many_cbor_strings) { |
20 |
my ($data, $length) = $cbor->decode_prefix ($many_cbor_strings); |
21 |
# data was decoded |
22 |
substr $many_cbor_strings, 0, $length, ""; # remove decoded cbor string |
23 |
} |
24 |
|
25 |
DESCRIPTION |
26 |
WARNING! This module is very new, and not very well tested (that's up to |
27 |
you to do). Furthermore, details of the implementation might change |
28 |
freely before version 1.0. And lastly, the object serialisation protocol |
29 |
depends on a pending IANA assignment, and until that assignment is |
30 |
official, this implementation is not interoperable with other |
31 |
implementations (even future versions of this module) until the |
32 |
assignment is done. |
33 |
|
34 |
You are still invited to try out CBOR, and this module. |
35 |
|
36 |
This module converts Perl data structures to the Concise Binary Object |
37 |
Representation (CBOR) and vice versa. CBOR is a fast binary |
38 |
serialisation format that aims to use a superset of the JSON data model, |
39 |
i.e. when you can represent something in JSON, you should be able to |
40 |
represent it in CBOR. |
41 |
|
42 |
In short, CBOR is a faster and very compact binary alternative to JSON, |
43 |
with the added ability of supporting serialisation of Perl objects. |
44 |
(JSON often compresses better than CBOR though, so if you plan to |
45 |
compress the data later you might want to compare both formats first). |
46 |
|
47 |
The primary goal of this module is to be *correct* and the secondary |
48 |
goal is to be *fast*. To reach the latter goal it was written in C. |
49 |
|
50 |
See MAPPING, below, on how CBOR::XS maps perl values to CBOR values and |
51 |
vice versa. |
52 |
|
53 |
FUNCTIONAL INTERFACE |
54 |
The following convenience methods are provided by this module. They are |
55 |
exported by default: |
56 |
|
57 |
$cbor_data = encode_cbor $perl_scalar |
58 |
Converts the given Perl data structure to CBOR representation. |
59 |
Croaks on error. |
60 |
|
61 |
$perl_scalar = decode_cbor $cbor_data |
62 |
The opposite of "encode_cbor": expects a valid CBOR string to parse, |
63 |
returning the resulting perl scalar. Croaks on error. |
64 |
|
65 |
OBJECT-ORIENTED INTERFACE |
66 |
The object oriented interface lets you configure your own encoding or |
67 |
decoding style, within the limits of supported formats. |
68 |
|
69 |
$cbor = new CBOR::XS |
70 |
Creates a new CBOR::XS object that can be used to de/encode CBOR |
71 |
strings. All boolean flags described below are by default |
72 |
*disabled*. |
73 |
|
74 |
The mutators for flags all return the CBOR object again and thus |
75 |
calls can be chained: |
76 |
|
77 |
#TODO my $cbor = CBOR::XS->new->encode ({a => [1,2]}); |
78 |
|
79 |
$cbor = $cbor->max_depth ([$maximum_nesting_depth]) |
80 |
$max_depth = $cbor->get_max_depth |
81 |
Sets the maximum nesting level (default 512) accepted while encoding |
82 |
or decoding. If a higher nesting level is detected in CBOR data or a |
83 |
Perl data structure, then the encoder and decoder will stop and |
84 |
croak at that point. |
85 |
|
86 |
Nesting level is defined by number of hash- or arrayrefs that the |
87 |
encoder needs to traverse to reach a given point or the number of |
88 |
"{" or "[" characters without their matching closing parenthesis |
89 |
crossed to reach a given character in a string. |
90 |
|
91 |
Setting the maximum depth to one disallows any nesting, so that |
92 |
ensures that the object is only a single hash/object or array. |
93 |
|
94 |
If no argument is given, the highest possible setting will be used, |
95 |
which is rarely useful. |
96 |
|
97 |
Note that nesting is implemented by recursion in C. The default |
98 |
value has been chosen to be as large as typical operating systems |
99 |
allow without crashing. |
100 |
|
101 |
See SECURITY CONSIDERATIONS, below, for more info on why this is |
102 |
useful. |
103 |
|
104 |
$cbor = $cbor->max_size ([$maximum_string_size]) |
105 |
$max_size = $cbor->get_max_size |
106 |
Set the maximum length a CBOR string may have (in bytes) where |
107 |
decoding is being attempted. The default is 0, meaning no limit. |
108 |
When "decode" is called on a string that is longer then this many |
109 |
bytes, it will not attempt to decode the string but throw an |
110 |
exception. This setting has no effect on "encode" (yet). |
111 |
|
112 |
If no argument is given, the limit check will be deactivated (same |
113 |
as when 0 is specified). |
114 |
|
115 |
See SECURITY CONSIDERATIONS, below, for more info on why this is |
116 |
useful. |
117 |
|
118 |
$cbor_data = $cbor->encode ($perl_scalar) |
119 |
Converts the given Perl data structure (a scalar value) to its CBOR |
120 |
representation. |
121 |
|
122 |
$perl_scalar = $cbor->decode ($cbor_data) |
123 |
The opposite of "encode": expects CBOR data and tries to parse it, |
124 |
returning the resulting simple scalar or reference. Croaks on error. |
125 |
|
126 |
($perl_scalar, $octets) = $cbor->decode_prefix ($cbor_data) |
127 |
This works like the "decode" method, but instead of raising an |
128 |
exception when there is trailing garbage after the CBOR string, it |
129 |
will silently stop parsing there and return the number of characters |
130 |
consumed so far. |
131 |
|
132 |
This is useful if your CBOR texts are not delimited by an outer |
133 |
protocol and you need to know where the first CBOR string ends amd |
134 |
the next one starts. |
135 |
|
136 |
CBOR::XS->new->decode_prefix ("......") |
137 |
=> ("...", 3) |
138 |
|
139 |
MAPPING |
140 |
This section describes how CBOR::XS maps Perl values to CBOR values and |
141 |
vice versa. These mappings are designed to "do the right thing" in most |
142 |
circumstances automatically, preserving round-tripping characteristics |
143 |
(what you put in comes out as something equivalent). |
144 |
|
145 |
For the more enlightened: note that in the following descriptions, |
146 |
lowercase *perl* refers to the Perl interpreter, while uppercase *Perl* |
147 |
refers to the abstract Perl language itself. |
148 |
|
149 |
CBOR -> PERL |
150 |
integers |
151 |
CBOR integers become (numeric) perl scalars. On perls without 64 bit |
152 |
support, 64 bit integers will be truncated or otherwise corrupted. |
153 |
|
154 |
byte strings |
155 |
Byte strings will become octet strings in Perl (the byte values |
156 |
0..255 will simply become characters of the same value in Perl). |
157 |
|
158 |
UTF-8 strings |
159 |
UTF-8 strings in CBOR will be decoded, i.e. the UTF-8 octets will be |
160 |
decoded into proper Unicode code points. At the moment, the validity |
161 |
of the UTF-8 octets will not be validated - corrupt input will |
162 |
result in corrupted Perl strings. |
163 |
|
164 |
arrays, maps |
165 |
CBOR arrays and CBOR maps will be converted into references to a |
166 |
Perl array or hash, respectively. The keys of the map will be |
167 |
stringified during this process. |
168 |
|
169 |
null |
170 |
CBOR null becomes "undef" in Perl. |
171 |
|
172 |
true, false, undefined |
173 |
These CBOR values become "Types:Serialiser::true", |
174 |
"Types:Serialiser::false" and "Types::Serialiser::error", |
175 |
respectively. They are overloaded to act almost exactly like the |
176 |
numbers 1 and 0 (for true and false) or to throw an exception on |
177 |
access (for error). See the Types::Serialiser manpage for details. |
178 |
|
179 |
CBOR tag 256 (perl object) |
180 |
The tag value 256 (TODO: pending iana registration) will be used to |
181 |
deserialise a Perl object serialised with "FREEZE". See OBJECT |
182 |
SERIALISATION, below, for details. |
183 |
|
184 |
CBOR tag 55799 (magic header) |
185 |
The tag 55799 is ignored (this tag implements the magic header). |
186 |
|
187 |
other CBOR tags |
188 |
Tagged items consists of a numeric tag and another CBOR value. Tags |
189 |
not handled internally are currently converted into a |
190 |
CBOR::XS::Tagged object, which is simply a blessed array reference |
191 |
consisting of the numeric tag value followed by the (decoded) CBOR |
192 |
value. |
193 |
|
194 |
In the future, support for user-supplied conversions might get |
195 |
added. |
196 |
|
197 |
anything else |
198 |
Anything else (e.g. unsupported simple values) will raise a decoding |
199 |
error. |
200 |
|
201 |
PERL -> CBOR |
202 |
The mapping from Perl to CBOR is slightly more difficult, as Perl is a |
203 |
truly typeless language, so we can only guess which CBOR type is meant |
204 |
by a Perl value. |
205 |
|
206 |
hash references |
207 |
Perl hash references become CBOR maps. As there is no inherent |
208 |
ordering in hash keys (or CBOR maps), they will usually be encoded |
209 |
in a pseudo-random order. |
210 |
|
211 |
Currently, tied hashes will use the indefinite-length format, while |
212 |
normal hashes will use the fixed-length format. |
213 |
|
214 |
array references |
215 |
Perl array references become fixed-length CBOR arrays. |
216 |
|
217 |
other references |
218 |
Other unblessed references are generally not allowed and will cause |
219 |
an exception to be thrown, except for references to the integers 0 |
220 |
and 1, which get turned into false and true in CBOR. |
221 |
|
222 |
CBOR::XS::Tagged objects |
223 |
Objects of this type must be arrays consisting of a single "[tag, |
224 |
value]" pair. The (numerical) tag will be encoded as a CBOR tag, the |
225 |
value will be encoded as appropriate for the value. You cna use |
226 |
"CBOR::XS::tag" to create such objects. |
227 |
|
228 |
Types::Serialiser::true, Types::Serialiser::false, |
229 |
Types::Serialiser::error |
230 |
These special values become CBOR true, CBOR false and CBOR undefined |
231 |
values, respectively. You can also use "\1", "\0" and "\undef" |
232 |
directly if you want. |
233 |
|
234 |
other blessed objects |
235 |
Other blessed objects are serialised via "TO_CBOR" or "FREEZE". See |
236 |
"OBJECT SERIALISATION", below, for details. |
237 |
|
238 |
simple scalars |
239 |
TODO Simple Perl scalars (any scalar that is not a reference) are |
240 |
the most difficult objects to encode: CBOR::XS will encode undefined |
241 |
scalars as CBOR null values, scalars that have last been used in a |
242 |
string context before encoding as CBOR strings, and anything else as |
243 |
number value: |
244 |
|
245 |
# dump as number |
246 |
encode_cbor [2] # yields [2] |
247 |
encode_cbor [-3.0e17] # yields [-3e+17] |
248 |
my $value = 5; encode_cbor [$value] # yields [5] |
249 |
|
250 |
# used as string, so dump as string |
251 |
print $value; |
252 |
encode_cbor [$value] # yields ["5"] |
253 |
|
254 |
# undef becomes null |
255 |
encode_cbor [undef] # yields [null] |
256 |
|
257 |
You can force the type to be a CBOR string by stringifying it: |
258 |
|
259 |
my $x = 3.1; # some variable containing a number |
260 |
"$x"; # stringified |
261 |
$x .= ""; # another, more awkward way to stringify |
262 |
print $x; # perl does it for you, too, quite often |
263 |
|
264 |
You can force the type to be a CBOR number by numifying it: |
265 |
|
266 |
my $x = "3"; # some variable containing a string |
267 |
$x += 0; # numify it, ensuring it will be dumped as a number |
268 |
$x *= 1; # same thing, the choice is yours. |
269 |
|
270 |
You can not currently force the type in other, less obscure, ways. |
271 |
Tell me if you need this capability (but don't forget to explain why |
272 |
it's needed :). |
273 |
|
274 |
Perl values that seem to be integers generally use the shortest |
275 |
possible representation. Floating-point values will use either the |
276 |
IEEE single format if possible without loss of precision, otherwise |
277 |
the IEEE double format will be used. Perls that use formats other |
278 |
than IEEE double to represent numerical values are supported, but |
279 |
might suffer loss of precision. |
280 |
|
281 |
OBJECT SERIALISATION |
282 |
This module knows two way to serialise a Perl object: The CBOR-specific |
283 |
way, and the generic way. |
284 |
|
285 |
Whenever the encoder encounters a Perl object that it cnanot serialise |
286 |
directly (most of them), it will first look up the "TO_CBOR" method on |
287 |
it. |
288 |
|
289 |
If it has a "TO_CBOR" method, it will call it with the object as only |
290 |
argument, and expects exactly one return value, which it will then |
291 |
substitute and encode it in the place of the object. |
292 |
|
293 |
Otherwise, it will look up the "FREEZE" method. If it exists, it will |
294 |
call it with the object as first argument, and the constant string |
295 |
"CBOR" as the second argument, to distinguish it from other serialisers. |
296 |
|
297 |
The "FREEZE" method can return any number of values (i.e. zero or more). |
298 |
These will be encoded as CBOR perl object, together with the classname. |
299 |
|
300 |
If an object supports neither "TO_CBOR" nor "FREEZE", encoding will fail |
301 |
with an error. |
302 |
|
303 |
Objects encoded via "TO_CBOR" cannot be automatically decoded, but |
304 |
objects encoded via "FREEZE" can be decoded using the following |
305 |
protocol: |
306 |
|
307 |
When an encoded CBOR perl object is encountered by the decoder, it will |
308 |
look up the "THAW" method, by using the stored classname, and will fail |
309 |
if the method cannot be found. |
310 |
|
311 |
After the lookup it will call the "THAW" method with the stored |
312 |
classname as first argument, the constant string "CBOR" as second |
313 |
argument, and all values returned by "FREEZE" as remaining arguments. |
314 |
|
315 |
EXAMPLES |
316 |
Here is an example "TO_CBOR" method: |
317 |
|
318 |
sub My::Object::TO_CBOR { |
319 |
my ($obj) = @_; |
320 |
|
321 |
["this is a serialised My::Object object", $obj->{id}] |
322 |
} |
323 |
|
324 |
When a "My::Object" is encoded to CBOR, it will instead encode a simple |
325 |
array with two members: a string, and the "object id". Decoding this |
326 |
CBOR string will yield a normal perl array reference in place of the |
327 |
object. |
328 |
|
329 |
A more useful and practical example would be a serialisation method for |
330 |
the URI module. CBOR has a custom tag value for URIs, namely 32: |
331 |
|
332 |
sub URI::TO_CBOR { |
333 |
my ($self) = @_; |
334 |
my $uri = "$self"; # stringify uri |
335 |
utf8::upgrade $uri; # make sure it will be encoded as UTF-8 string |
336 |
CBOR::XS::tagged 32, "$_[0]" |
337 |
} |
338 |
|
339 |
This will encode URIs as a UTF-8 string with tag 32, which indicates an |
340 |
URI. |
341 |
|
342 |
Decoding such an URI will not (currently) give you an URI object, but |
343 |
instead a CBOR::XS::Tagged object with tag number 32 and the string - |
344 |
exactly what was returned by "TO_CBOR". |
345 |
|
346 |
To serialise an object so it can automatically be deserialised, you need |
347 |
to use "FREEZE" and "THAW". To take the URI module as example, this |
348 |
would be a possible implementation: |
349 |
|
350 |
sub URI::FREEZE { |
351 |
my ($self, $serialiser) = @_; |
352 |
"$self" # encode url string |
353 |
} |
354 |
|
355 |
sub URI::THAW { |
356 |
my ($class, $serialiser, $uri) = @_; |
357 |
|
358 |
$class->new ($uri) |
359 |
} |
360 |
|
361 |
Unlike "TO_CBOR", multiple values can be returned by "FREEZE". For |
362 |
example, a "FREEZE" method that returns "type", "id" and "variant" |
363 |
values would cause an invocation of "THAW" with 5 arguments: |
364 |
|
365 |
sub My::Object::FREEZE { |
366 |
my ($self, $serialiser) = @_; |
367 |
|
368 |
($self->{type}, $self->{id}, $self->{variant}) |
369 |
} |
370 |
|
371 |
sub My::Object::THAW { |
372 |
my ($class, $serialiser, $type, $id, $variant) = @_; |
373 |
|
374 |
$class-<new (type => $type, id => $id, variant => $variant) |
375 |
} |
376 |
|
377 |
MAGIC HEADER |
378 |
There is no way to distinguish CBOR from other formats programmatically. |
379 |
To make it easier to distinguish CBOR from other formats, the CBOR |
380 |
specification has a special "magic string" that can be prepended to any |
381 |
CBOR string without changing it's meaning. |
382 |
|
383 |
This string is available as $CBOR::XS::MAGIC. This module does not |
384 |
prepend this string tot he CBOR data it generates, but it will ignroe it |
385 |
if present, so users can prepend this string as a "file type" indicator |
386 |
as required. |
387 |
|
388 |
THE CBOR::XS::Tagged CLASS |
389 |
CBOR has the concept of tagged values - any CBOR value can be tagged |
390 |
with a numeric 64 bit number, which are centrally administered. |
391 |
|
392 |
"CBOR::XS" handles a few tags internally when en- or decoding. You can |
393 |
also create tags yourself by encoding "CBOR::XS::Tagged" objects, and |
394 |
the decoder will create "CBOR::XS::Tagged" objects itself when it hits |
395 |
an unknown tag. |
396 |
|
397 |
These objects are simply blessed array references - the first member of |
398 |
the array being the numerical tag, the second being the value. |
399 |
|
400 |
You can interact with "CBOR::XS::Tagged" objects in the following ways: |
401 |
|
402 |
$tagged = CBOR::XS::tag $tag, $value |
403 |
This function(!) creates a new "CBOR::XS::Tagged" object using the |
404 |
given $tag (0..2**64-1) to tag the given $value (which can be any |
405 |
Perl value that can be encoded in CBOR, including serialisable Perl |
406 |
objects and "CBOR::XS::Tagged" objects). |
407 |
|
408 |
$tagged->[0] |
409 |
$tagged->[0] = $new_tag |
410 |
$tag = $tagged->tag |
411 |
$new_tag = $tagged->tag ($new_tag) |
412 |
Access/mutate the tag. |
413 |
|
414 |
$tagged->[1] |
415 |
$tagged->[1] = $new_value |
416 |
$value = $tagged->value |
417 |
$new_value = $tagged->value ($new_value) |
418 |
Access/mutate the tagged value. |
419 |
|
420 |
EXAMPLES |
421 |
Here are some examples of "CBOR::XS::Tagged" uses to tag objects. |
422 |
|
423 |
You can look up CBOR tag value and emanings in the IANA registry at |
424 |
<http://www.iana.org/assignments/cbor-tags/cbor-tags.xhtml>. |
425 |
|
426 |
Prepend a magic header ($CBOR::XS::MAGIC): |
427 |
|
428 |
my $cbor = encode_cbor CBOR::XS::tag 55799, $value; |
429 |
# same as: |
430 |
my $cbor = $CBOR::XS::MAGIC . encode_cbor $value; |
431 |
|
432 |
Serialise some URIs and a regex in an array: |
433 |
|
434 |
my $cbor = encode_cbor [ |
435 |
(CBOR::XS::tag 32, "http://www.nethype.de/"), |
436 |
(CBOR::XS::tag 32, "http://software.schmorp.de/"), |
437 |
(CBOR::XS::tag 35, "^[Pp][Ee][Rr][lL]\$"), |
438 |
]; |
439 |
|
440 |
Wrap CBOR data in CBOR: |
441 |
|
442 |
my $cbor_cbor = encode_cbor |
443 |
CBOR::XS::tag 24, |
444 |
encode_cbor [1, 2, 3]; |
445 |
|
446 |
CBOR and JSON |
447 |
CBOR is supposed to implement a superset of the JSON data model, and is, |
448 |
with some coercion, able to represent all JSON texts (something that |
449 |
other "binary JSON" formats such as BSON generally do not support). |
450 |
|
451 |
CBOR implements some extra hints and support for JSON interoperability, |
452 |
and the spec offers further guidance for conversion between CBOR and |
453 |
JSON. None of this is currently implemented in CBOR, and the guidelines |
454 |
in the spec do not result in correct round-tripping of data. If JSON |
455 |
interoperability is improved in the future, then the goal will be to |
456 |
ensure that decoded JSON data will round-trip encoding and decoding to |
457 |
CBOR intact. |
458 |
|
459 |
SECURITY CONSIDERATIONS |
460 |
When you are using CBOR in a protocol, talking to untrusted potentially |
461 |
hostile creatures requires relatively few measures. |
462 |
|
463 |
First of all, your CBOR decoder should be secure, that is, should not |
464 |
have any buffer overflows. Obviously, this module should ensure that and |
465 |
I am trying hard on making that true, but you never know. |
466 |
|
467 |
Second, you need to avoid resource-starving attacks. That means you |
468 |
should limit the size of CBOR data you accept, or make sure then when |
469 |
your resources run out, that's just fine (e.g. by using a separate |
470 |
process that can crash safely). The size of a CBOR string in octets is |
471 |
usually a good indication of the size of the resources required to |
472 |
decode it into a Perl structure. While CBOR::XS can check the size of |
473 |
the CBOR text, it might be too late when you already have it in memory, |
474 |
so you might want to check the size before you accept the string. |
475 |
|
476 |
Third, CBOR::XS recurses using the C stack when decoding objects and |
477 |
arrays. The C stack is a limited resource: for instance, on my amd64 |
478 |
machine with 8MB of stack size I can decode around 180k nested arrays |
479 |
but only 14k nested CBOR objects (due to perl itself recursing deeply on |
480 |
croak to free the temporary). If that is exceeded, the program crashes. |
481 |
To be conservative, the default nesting limit is set to 512. If your |
482 |
process has a smaller stack, you should adjust this setting accordingly |
483 |
with the "max_depth" method. |
484 |
|
485 |
Something else could bomb you, too, that I forgot to think of. In that |
486 |
case, you get to keep the pieces. I am always open for hints, though... |
487 |
|
488 |
Also keep in mind that CBOR::XS might leak contents of your Perl data |
489 |
structures in its error messages, so when you serialise sensitive |
490 |
information you might want to make sure that exceptions thrown by |
491 |
CBOR::XS will not end up in front of untrusted eyes. |
492 |
|
493 |
CBOR IMPLEMENTATION NOTES |
494 |
This section contains some random implementation notes. They do not |
495 |
describe guaranteed behaviour, but merely behaviour as-is implemented |
496 |
right now. |
497 |
|
498 |
64 bit integers are only properly decoded when Perl was built with 64 |
499 |
bit support. |
500 |
|
501 |
Strings and arrays are encoded with a definite length. Hashes as well, |
502 |
unless they are tied (or otherwise magical). |
503 |
|
504 |
Only the double data type is supported for NV data types - when Perl |
505 |
uses long double to represent floating point values, they might not be |
506 |
encoded properly. Half precision types are accepted, but not encoded. |
507 |
|
508 |
Strict mode and canonical mode are not implemented. |
509 |
|
510 |
THREADS |
511 |
This module is *not* guaranteed to be thread safe and there are no plans |
512 |
to change this until Perl gets thread support (as opposed to the |
513 |
horribly slow so-called "threads" which are simply slow and bloated |
514 |
process simulations - use fork, it's *much* faster, cheaper, better). |
515 |
|
516 |
(It might actually work, but you have been warned). |
517 |
|
518 |
BUGS |
519 |
While the goal of this module is to be correct, that unfortunately does |
520 |
not mean it's bug-free, only that I think its design is bug-free. If you |
521 |
keep reporting bugs they will be fixed swiftly, though. |
522 |
|
523 |
Please refrain from using rt.cpan.org or any other bug reporting |
524 |
service. I put the contact address into my modules for a reason. |
525 |
|
526 |
SEE ALSO |
527 |
The JSON and JSON::XS modules that do similar, but human-readable, |
528 |
serialisation. |
529 |
|
530 |
The Types::Serialiser module provides the data model for true, false and |
531 |
error values. |
532 |
|
533 |
AUTHOR |
534 |
Marc Lehmann <schmorp@schmorp.de> |
535 |
http://home.schmorp.de/ |
536 |
|