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