… | |
… | |
38 | with the added ability of supporting serialisation of Perl objects. (JSON |
38 | with the added ability of supporting serialisation of Perl objects. (JSON |
39 | 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 |
40 | data later and speed is less important you might want to compare both |
40 | data later and speed is less important you might want to compare both |
41 | formats first). |
41 | formats first). |
42 | |
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. |
|
|
45 | |
43 | 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, |
44 | 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 |
45 | 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 |
46 | data, the worse L<Storable> performs in comparison. |
49 | data, the worse L<Storable> performs in comparison. |
47 | |
50 | |
… | |
… | |
52 | In addition to the core CBOR data format, this module implements a |
55 | In addition to the core CBOR data format, this module implements a |
53 | number of extensions, to support cyclic and shared data structures |
56 | number of extensions, to support cyclic and shared data structures |
54 | (see C<allow_sharing> and C<allow_cycles>), string deduplication (see |
57 | (see C<allow_sharing> and C<allow_cycles>), string deduplication (see |
55 | C<pack_strings>) and scalar references (always enabled). |
58 | C<pack_strings>) and scalar references (always enabled). |
56 | |
59 | |
57 | The primary goal of this module is to be I<correct> and the secondary goal |
|
|
58 | is to be I<fast>. To reach the latter goal it was written in C. |
|
|
59 | |
|
|
60 | 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 |
61 | vice versa. |
61 | vice versa. |
62 | |
62 | |
63 | =cut |
63 | =cut |
64 | |
64 | |
65 | package CBOR::XS; |
65 | package CBOR::XS; |
66 | |
66 | |
67 | use common::sense; |
67 | use common::sense; |
68 | |
68 | |
69 | our $VERSION = 1.51; |
69 | our $VERSION = 1.87; |
70 | our @ISA = qw(Exporter); |
70 | our @ISA = qw(Exporter); |
71 | |
71 | |
72 | our @EXPORT = qw(encode_cbor decode_cbor); |
72 | our @EXPORT = qw(encode_cbor decode_cbor); |
73 | |
73 | |
74 | use Exporter; |
74 | use Exporter; |
… | |
… | |
112 | |
112 | |
113 | 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 |
114 | be chained: |
114 | be chained: |
115 | |
115 | |
116 | 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 | } |
117 | |
144 | |
118 | =item $cbor = $cbor->max_depth ([$maximum_nesting_depth]) |
145 | =item $cbor = $cbor->max_depth ([$maximum_nesting_depth]) |
119 | |
146 | |
120 | =item $max_depth = $cbor->get_max_depth |
147 | =item $max_depth = $cbor->get_max_depth |
121 | |
148 | |
… | |
… | |
137 | |
164 | |
138 | 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 |
139 | 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 |
140 | crashing. |
167 | crashing. |
141 | |
168 | |
142 | 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. |
143 | |
170 | |
144 | =item $cbor = $cbor->max_size ([$maximum_string_size]) |
171 | =item $cbor = $cbor->max_size ([$maximum_string_size]) |
145 | |
172 | |
146 | =item $max_size = $cbor->get_max_size |
173 | =item $max_size = $cbor->get_max_size |
147 | |
174 | |
… | |
… | |
152 | effect on C<encode> (yet). |
179 | effect on C<encode> (yet). |
153 | |
180 | |
154 | 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 |
155 | C<0> is specified). |
182 | C<0> is specified). |
156 | |
183 | |
157 | 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. |
158 | |
185 | |
159 | =item $cbor = $cbor->allow_unknown ([$enable]) |
186 | =item $cbor = $cbor->allow_unknown ([$enable]) |
160 | |
187 | |
161 | =item $enabled = $cbor->get_allow_unknown |
188 | =item $enabled = $cbor->get_allow_unknown |
162 | |
189 | |
… | |
… | |
189 | communication partner supports the value sharing extensions to CBOR |
216 | communication partner supports the value sharing extensions to CBOR |
190 | (L<http://cbor.schmorp.de/value-sharing>), as without decoder support, the |
217 | (L<http://cbor.schmorp.de/value-sharing>), as without decoder support, the |
191 | resulting data structure might be unusable. |
218 | resulting data structure might be unusable. |
192 | |
219 | |
193 | Detecting shared values incurs a runtime overhead when values are encoded |
220 | Detecting shared values incurs a runtime overhead when values are encoded |
194 | that have a reference counter large than one, and might unnecessarily |
221 | that have a reference counter larger than one, and might unnecessarily |
195 | increase the encoded size, as potentially shared values are encode as |
222 | increase the encoded size, as potentially shared values are encoded as |
196 | shareable whether or not they are actually shared. |
223 | shareable whether or not they are actually shared. |
197 | |
224 | |
198 | At the moment, only targets of references can be shared (e.g. scalars, |
225 | At the moment, only targets of references can be shared (e.g. scalars, |
199 | arrays or hashes pointed to by a reference). Weirder constructs, such as |
226 | arrays or hashes pointed to by a reference). Weirder constructs, such as |
200 | an array with multiple "copies" of the I<same> string, which are hard but |
227 | an array with multiple "copies" of the I<same> string, which are hard but |
… | |
… | |
218 | isn't prepared for this will not leak memory. |
245 | isn't prepared for this will not leak memory. |
219 | |
246 | |
220 | If C<$enable> is false (the default), then C<decode> will throw an error |
247 | If C<$enable> is false (the default), then C<decode> will throw an error |
221 | when it encounters a self-referential/cyclic data structure. |
248 | when it encounters a self-referential/cyclic data structure. |
222 | |
249 | |
223 | FUTURE DIRECTION: the motivation behind this option is to avoid I<real> |
|
|
224 | cycles - future versions of this module might chose to decode cyclic data |
|
|
225 | structures using weak references when this option is off, instead of |
|
|
226 | throwing an error. |
|
|
227 | |
|
|
228 | This option does not affect C<encode> in any way - shared values and |
250 | This option does not affect C<encode> in any way - shared values and |
229 | references will always be encoded properly if present. |
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. |
230 | |
288 | |
231 | =item $cbor = $cbor->pack_strings ([$enable]) |
289 | =item $cbor = $cbor->pack_strings ([$enable]) |
232 | |
290 | |
233 | =item $enabled = $cbor->get_pack_strings |
291 | =item $enabled = $cbor->get_pack_strings |
234 | |
292 | |
… | |
… | |
286 | strings as CBOR byte strings. |
344 | strings as CBOR byte strings. |
287 | |
345 | |
288 | This option does not affect C<decode> in any way. |
346 | This option does not affect C<decode> in any way. |
289 | |
347 | |
290 | This option has similar advantages and disadvantages as C<text_keys>. In |
348 | This option has similar advantages and disadvantages as C<text_keys>. In |
291 | addition, this option effectively removes the ability to encode byte |
349 | addition, this option effectively removes the ability to automatically |
292 | strings, which might break some C<FREEZE> and C<TO_CBOR> methods that rely |
350 | encode byte strings, which might break some C<FREEZE> and C<TO_CBOR> |
293 | on this, such as bignum encoding, so this option is mainly useful for very |
351 | methods that rely on this. |
294 | simple data. |
352 | |
|
|
353 | A workaround is to use explicit type casts, which are unaffected by this option. |
295 | |
354 | |
296 | =item $cbor = $cbor->validate_utf8 ([$enable]) |
355 | =item $cbor = $cbor->validate_utf8 ([$enable]) |
297 | |
356 | |
298 | =item $enabled = $cbor->get_validate_utf8 |
357 | =item $enabled = $cbor->get_validate_utf8 |
299 | |
358 | |
… | |
… | |
337 | replace the tagged value in the decoded data structure, or no values, |
396 | replace the tagged value in the decoded data structure, or no values, |
338 | which will result in default handling, which currently means the decoder |
397 | which will result in default handling, which currently means the decoder |
339 | creates a C<CBOR::XS::Tagged> object to hold the tag and the value. |
398 | creates a C<CBOR::XS::Tagged> object to hold the tag and the value. |
340 | |
399 | |
341 | When the filter is cleared (the default state), the default filter |
400 | When the filter is cleared (the default state), the default filter |
342 | function, C<CBOR::XS::default_filter>, is used. This function simply looks |
401 | function, C<CBOR::XS::default_filter>, is used. This function simply |
343 | up the tag in the C<%CBOR::XS::FILTER> hash. If an entry exists it must be |
402 | looks up the tag in the C<%CBOR::XS::FILTER> hash. If an entry exists |
344 | a code reference that is called with tag and value, and is responsible for |
403 | it must be a code reference that is called with tag and value, and is |
345 | decoding the value. If no entry exists, it returns no values. |
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. |
346 | |
418 | |
347 | Example: decode all tags not handled internally into C<CBOR::XS::Tagged> |
419 | Example: decode all tags not handled internally into C<CBOR::XS::Tagged> |
348 | objects, with no other special handling (useful when working with |
420 | objects, with no other special handling (useful when working with |
349 | potentially "unsafe" CBOR data). |
421 | potentially "unsafe" CBOR data). |
350 | |
422 | |
… | |
… | |
357 | my ($tag, $value); |
429 | my ($tag, $value); |
358 | |
430 | |
359 | "tag 1347375694 value $value" |
431 | "tag 1347375694 value $value" |
360 | }; |
432 | }; |
361 | |
433 | |
|
|
434 | Example: provide your own filter function that looks up tags in your own |
|
|
435 | hash: |
|
|
436 | |
|
|
437 | my %my_filter = ( |
|
|
438 | 998347484 => sub { |
|
|
439 | my ($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 regarding security). |
|
|
452 | |
|
|
453 | CBOR::XS->new->filter (\&CBOR::XS::safe_filter)->decode ($cbor_data); |
|
|
454 | |
362 | =item $cbor_data = $cbor->encode ($perl_scalar) |
455 | =item $cbor_data = $cbor->encode ($perl_scalar) |
363 | |
456 | |
364 | 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 |
365 | representation. |
458 | representation. |
366 | |
459 | |
… | |
… | |
375 | 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 |
376 | 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. |
377 | |
470 | |
378 | 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 |
379 | 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 |
380 | 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. |
381 | |
475 | |
382 | CBOR::XS->new->decode_prefix ("......") |
476 | CBOR::XS->new->decode_prefix ("......") |
383 | => ("...", 3) |
477 | => ("...", 3) |
384 | |
478 | |
385 | =back |
479 | =back |
… | |
… | |
391 | Perl data structure in memory at one time, it does allow you to parse a |
485 | Perl data structure in memory at one time, it does allow you to parse a |
392 | CBOR stream incrementally, using a similar to using "decode_prefix" to see |
486 | CBOR stream incrementally, using a similar to using "decode_prefix" to see |
393 | if a full CBOR object is available, but is much more efficient. |
487 | if a full CBOR object is available, but is much more efficient. |
394 | |
488 | |
395 | It basically works by parsing as much of a CBOR string as possible - if |
489 | It basically works by parsing as much of a CBOR string as possible - if |
396 | the CBOR data is not complete yet, the pasrer will remember where it was, |
490 | the CBOR data is not complete yet, the parser will remember where it was, |
397 | to be able to restart when more data has been accumulated. Once enough |
491 | to be able to restart when more data has been accumulated. Once enough |
398 | data is available to either decode a complete CBOR value or raise an |
492 | data is available to either decode a complete CBOR value or raise an |
399 | error, a real decode will be attempted. |
493 | error, a real decode will be attempted. |
400 | |
494 | |
401 | A typical use case would be a network protocol that consists of sending |
495 | A typical use case would be a network protocol that consists of sending |
… | |
… | |
553 | create such objects. |
647 | create such objects. |
554 | |
648 | |
555 | =item Types::Serialiser::true, Types::Serialiser::false, Types::Serialiser::error |
649 | =item Types::Serialiser::true, Types::Serialiser::false, Types::Serialiser::error |
556 | |
650 | |
557 | These special values become CBOR true, CBOR false and CBOR undefined |
651 | These special values become CBOR true, CBOR false and CBOR undefined |
558 | values, respectively. You can also use C<\1>, C<\0> and C<\undef> directly |
652 | values, respectively. |
559 | if you want. |
|
|
560 | |
653 | |
561 | =item other blessed objects |
654 | =item other blessed objects |
562 | |
655 | |
563 | 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 |
564 | L<TAG HANDLING AND EXTENSIONS> for specific classes handled by this |
657 | L<TAG HANDLING AND EXTENSIONS> for specific classes handled by this |
… | |
… | |
589 | "$x"; # stringified |
682 | "$x"; # stringified |
590 | $x .= ""; # another, more awkward way to stringify |
683 | $x .= ""; # another, more awkward way to stringify |
591 | print $x; # perl does it for you, too, quite often |
684 | print $x; # perl does it for you, too, quite often |
592 | |
685 | |
593 | You can force whether a string is encoded as byte or text string by using |
686 | You can force whether a string is encoded as byte or text string by using |
594 | C<utf8::upgrade> and C<utf8::downgrade> (if C<text_strings> is disabled): |
687 | C<utf8::upgrade> and C<utf8::downgrade> (if C<text_strings> is disabled). |
595 | |
688 | |
596 | utf8::upgrade $x; # encode $x as text string |
689 | utf8::upgrade $x; # encode $x as text string |
597 | utf8::downgrade $x; # encode $x as byte string |
690 | utf8::downgrade $x; # encode $x as byte string |
|
|
691 | |
|
|
692 | More options are available, see L<TYPE CASTS>, below, and the C<text_keys> |
|
|
693 | and C<text_strings> options. |
598 | |
694 | |
599 | Perl doesn't define what operations up- and downgrade strings, so if the |
695 | Perl doesn't define what operations up- and downgrade strings, so if the |
600 | difference between byte and text is important, you should up- or downgrade |
696 | difference between byte and text is important, you should up- or downgrade |
601 | your string as late as possible before encoding. You can also force the |
697 | your string as late as possible before encoding. You can also force the |
602 | use of CBOR text strings by using C<text_keys> or C<text_strings>. |
698 | use of CBOR text strings by using C<text_keys> or C<text_strings>. |
… | |
… | |
617 | format will be used. Perls that use formats other than IEEE double to |
713 | format will be used. Perls that use formats other than IEEE double to |
618 | represent numerical values are supported, but might suffer loss of |
714 | represent numerical values are supported, but might suffer loss of |
619 | precision. |
715 | precision. |
620 | |
716 | |
621 | =back |
717 | =back |
|
|
718 | |
|
|
719 | =head2 TYPE CASTS |
|
|
720 | |
|
|
721 | B<EXPERIMENTAL>: As an experimental extension, C<CBOR::XS> allows you to |
|
|
722 | force specific CBOR types to be used when encoding. That allows you to |
|
|
723 | encode types not normally accessible (e.g. half floats) as well as force |
|
|
724 | string types even when C<text_strings> is in effect. |
|
|
725 | |
|
|
726 | Type forcing is done by calling a special "cast" function which keeps a |
|
|
727 | copy of the value and returns a new value that can be handed over to any |
|
|
728 | CBOR encoder function. |
|
|
729 | |
|
|
730 | The following casts are currently available (all of which are unary |
|
|
731 | operators, that is, have a prototype of C<$>): |
|
|
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 | } |
622 | |
819 | |
623 | =head2 OBJECT SERIALISATION |
820 | =head2 OBJECT SERIALISATION |
624 | |
821 | |
625 | This module implements both a CBOR-specific and the generic |
822 | This module implements both a CBOR-specific and the generic |
626 | L<Types::Serialier> object serialisation protocol. The following |
823 | L<Types::Serialier> object serialisation protocol. The following |
… | |
… | |
978 | CBOR intact. |
1175 | CBOR intact. |
979 | |
1176 | |
980 | |
1177 | |
981 | =head1 SECURITY CONSIDERATIONS |
1178 | =head1 SECURITY CONSIDERATIONS |
982 | |
1179 | |
983 | 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 |
984 | 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): |
985 | |
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 | |
986 | 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 |
987 | 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 |
988 | trying hard on making that true, but you never know. |
1199 | on making that true, but you never know. |
989 | |
1200 | |
|
|
1201 | =item CBOR::XS can invoke almost arbitrary callbacks during decoding |
|
|
1202 | |
990 | Second, CBOR::XS supports object serialisation - decoding CBOR can cause |
1203 | CBOR::XS supports object serialisation - decoding CBOR can cause calls |
991 | calls to I<any> C<THAW> method in I<any> package that exists in your |
1204 | to I<any> C<THAW> method in I<any> package that exists in your process |
992 | process (that is, CBOR::XS will not try to load modules, but any existing |
1205 | (that is, CBOR::XS will not try to load modules, but any existing C<THAW> |
993 | C<THAW> method or function can be called, so they all have to be secure). |
1206 | method or function can be called, so they all have to be secure). |
994 | |
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 | |
995 | Third, you need to avoid resource-starving attacks. That means you should |
1244 | You need to avoid resource-starving attacks. That means you should limit |
996 | 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 |
997 | 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 |
998 | 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 |
999 | 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 |
1000 | 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 |
1001 | 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 |
1002 | 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. |
1003 | |
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 | |
1004 | Fourth, 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 |
1005 | 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 |
1006 | 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 |
1007 | 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 |
1008 | 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, |
1009 | 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 |
1010 | 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> |
1011 | 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... |
1012 | |
1288 | |
1013 | 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 |
1014 | 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... |
1015 | |
1291 | |
1016 | Also keep in mind that CBOR::XS might leak contents of your Perl data |
1292 | =back |
1017 | structures in its error messages, so when you serialise sensitive |
|
|
1018 | information you might want to make sure that exceptions thrown by CBOR::XS |
|
|
1019 | will not end up in front of untrusted eyes. |
|
|
1020 | |
1293 | |
1021 | |
1294 | |
1022 | =head1 BIGNUM SECURITY CONSIDERATIONS |
1295 | =head1 BIGNUM SECURITY CONSIDERATIONS |
1023 | |
1296 | |
1024 | CBOR::XS provides a C<TO_CBOR> method for both L<Math::BigInt> and |
1297 | CBOR::XS provides a C<TO_CBOR> method for both L<Math::BigInt> and |
… | |
… | |
1073 | =head1 LIMITATIONS ON PERLS WITHOUT 64-BIT INTEGER SUPPORT |
1346 | =head1 LIMITATIONS ON PERLS WITHOUT 64-BIT INTEGER SUPPORT |
1074 | |
1347 | |
1075 | On perls that were built without 64 bit integer support (these are rare |
1348 | On perls that were built without 64 bit integer support (these are rare |
1076 | nowadays, even on 32 bit architectures, as all major Perl distributions |
1349 | nowadays, even on 32 bit architectures, as all major Perl distributions |
1077 | are built with 64 bit integer support), support for any kind of 64 bit |
1350 | are built with 64 bit integer support), support for any kind of 64 bit |
1078 | integer in CBOR is very limited - most likely, these 64 bit values will |
1351 | value in CBOR is very limited - most likely, these 64 bit values will |
1079 | be truncated, corrupted, or otherwise not decoded correctly. This also |
1352 | be truncated, corrupted, or otherwise not decoded correctly. This also |
1080 | includes string, array and map sizes that are stored as 64 bit integers. |
1353 | includes string, float, array and map sizes that are stored as 64 bit |
|
|
1354 | integers. |
1081 | |
1355 | |
1082 | |
1356 | |
1083 | =head1 THREADS |
1357 | =head1 THREADS |
1084 | |
1358 | |
1085 | 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 |
… | |
… | |
1098 | |
1372 | |
1099 | 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 |
1100 | 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. |
1101 | |
1375 | |
1102 | =cut |
1376 | =cut |
|
|
1377 | |
|
|
1378 | # clumsy and slow hv_store-in-hash helper function |
|
|
1379 | sub _hv_store { |
|
|
1380 | $_[0]{$_[1]} = $_[2]; |
|
|
1381 | } |
1103 | |
1382 | |
1104 | our %FILTER = ( |
1383 | our %FILTER = ( |
1105 | 0 => sub { # rfc4287 datetime, utf-8 |
1384 | 0 => sub { # rfc4287 datetime, utf-8 |
1106 | require Time::Piece; |
1385 | require Time::Piece; |
1107 | # Time::Piece::Strptime uses the "incredibly flexible date parsing routine" |
1386 | # Time::Piece::Strptime uses the "incredibly flexible date parsing routine" |
… | |
… | |
1180 | # 34 # base64 rfc46484, utf-8 |
1459 | # 34 # base64 rfc46484, utf-8 |
1181 | # 35 # regex pcre/ecma262, utf-8 |
1460 | # 35 # regex pcre/ecma262, utf-8 |
1182 | # 36 # mime message rfc2045, utf-8 |
1461 | # 36 # mime message rfc2045, utf-8 |
1183 | ); |
1462 | ); |
1184 | |
1463 | |
1185 | sub CBOR::XS::default_filter { |
1464 | sub default_filter { |
1186 | &{ $FILTER{$_[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 } |
1187 | } |
1472 | } |
1188 | |
1473 | |
1189 | sub URI::TO_CBOR { |
1474 | sub URI::TO_CBOR { |
1190 | my $uri = $_[0]->as_string; |
1475 | my $uri = $_[0]->as_string; |
1191 | utf8::upgrade $uri; |
1476 | utf8::upgrade $uri; |