… | |
… | |
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 = 1.1; |
69 | our $VERSION = 1.41; |
70 | our @ISA = qw(Exporter); |
70 | our @ISA = qw(Exporter); |
71 | |
71 | |
72 | our @EXPORT = qw(encode_cbor decode_cbor); |
72 | our @EXPORT = qw(encode_cbor decode_cbor); |
73 | |
73 | |
74 | use Exporter; |
74 | use Exporter; |
… | |
… | |
218 | isn't prepared for this will not leak memory. |
218 | isn't prepared for this will not leak memory. |
219 | |
219 | |
220 | If C<$enable> is false (the default), then C<decode> will throw an error |
220 | If C<$enable> is false (the default), then C<decode> will throw an error |
221 | when it encounters a self-referential/cyclic data structure. |
221 | when it encounters a self-referential/cyclic data structure. |
222 | |
222 | |
|
|
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 | |
223 | This option does not affect C<encode> in any way - shared values and |
228 | This option does not affect C<encode> in any way - shared values and |
224 | references will always be decoded properly if present. |
229 | references will always be encoded properly if present. |
225 | |
230 | |
226 | =item $cbor = $cbor->pack_strings ([$enable]) |
231 | =item $cbor = $cbor->pack_strings ([$enable]) |
227 | |
232 | |
228 | =item $enabled = $cbor->get_pack_strings |
233 | =item $enabled = $cbor->get_pack_strings |
229 | |
234 | |
… | |
… | |
242 | the standard CBOR way. |
247 | the standard CBOR way. |
243 | |
248 | |
244 | This option does not affect C<decode> in any way - string references will |
249 | This option does not affect C<decode> in any way - string references will |
245 | always be decoded properly if present. |
250 | always be decoded properly if present. |
246 | |
251 | |
|
|
252 | =item $cbor = $cbor->validate_utf8 ([$enable]) |
|
|
253 | |
|
|
254 | =item $enabled = $cbor->get_validate_utf8 |
|
|
255 | |
|
|
256 | If C<$enable> is true (or missing), then C<decode> will validate that |
|
|
257 | elements (text strings) containing UTF-8 data in fact contain valid UTF-8 |
|
|
258 | data (instead of blindly accepting it). This validation obviously takes |
|
|
259 | extra time during decoding. |
|
|
260 | |
|
|
261 | The concept of "valid UTF-8" used is perl's concept, which is a superset |
|
|
262 | of the official UTF-8. |
|
|
263 | |
|
|
264 | If C<$enable> is false (the default), then C<decode> will blindly accept |
|
|
265 | UTF-8 data, marking them as valid UTF-8 in the resulting data structure |
|
|
266 | regardless of whether thats true or not. |
|
|
267 | |
|
|
268 | Perl isn't too happy about corrupted UTF-8 in strings, but should |
|
|
269 | generally not crash or do similarly evil things. Extensions might be not |
|
|
270 | so forgiving, so it's recommended to turn on this setting if you receive |
|
|
271 | untrusted CBOR. |
|
|
272 | |
|
|
273 | This option does not affect C<encode> in any way - strings that are |
|
|
274 | supposedly valid UTF-8 will simply be dumped into the resulting CBOR |
|
|
275 | string without checking whether that is, in fact, true or not. |
|
|
276 | |
247 | =item $cbor = $cbor->filter ([$cb->($tag, $value)]) |
277 | =item $cbor = $cbor->filter ([$cb->($tag, $value)]) |
248 | |
278 | |
249 | =item $cb_or_undef = $cbor->get_filter |
279 | =item $cb_or_undef = $cbor->get_filter |
250 | |
280 | |
251 | Sets or replaces the tagged value decoding filter (when C<$cb> is |
281 | Sets or replaces the tagged value decoding filter (when C<$cb> is |
… | |
… | |
308 | CBOR::XS->new->decode_prefix ("......") |
338 | CBOR::XS->new->decode_prefix ("......") |
309 | => ("...", 3) |
339 | => ("...", 3) |
310 | |
340 | |
311 | =back |
341 | =back |
312 | |
342 | |
|
|
343 | =head2 INCREMENTAL PARSING |
|
|
344 | |
|
|
345 | In some cases, there is the need for incremental parsing of JSON |
|
|
346 | texts. While this module always has to keep both CBOR text and resulting |
|
|
347 | Perl data structure in memory at one time, it does allow you to parse a |
|
|
348 | CBOR stream incrementally, using a similar to using "decode_prefix" to see |
|
|
349 | if a full CBOR object is available, but is much more efficient. |
|
|
350 | |
|
|
351 | It basically works by parsing as much of a CBOR string as possible - if |
|
|
352 | the CBOR data is not complete yet, the pasrer will remember where it was, |
|
|
353 | to be able to restart when more data has been accumulated. Once enough |
|
|
354 | data is available to either decode a complete CBOR value or raise an |
|
|
355 | error, a real decode will be attempted. |
|
|
356 | |
|
|
357 | A typical use case would be a network protocol that consists of sending |
|
|
358 | and receiving CBOR-encoded messages. The solution that works with CBOR and |
|
|
359 | about anything else is by prepending a length to every CBOR value, so the |
|
|
360 | receiver knows how many octets to read. More compact (and slightly slower) |
|
|
361 | would be to just send CBOR values back-to-back, as C<CBOR::XS> knows where |
|
|
362 | a CBOR value ends, and doesn't need an explicit length. |
|
|
363 | |
|
|
364 | The following methods help with this: |
|
|
365 | |
|
|
366 | =over 4 |
|
|
367 | |
|
|
368 | =item @decoded = $cbor->incr_parse ($buffer) |
|
|
369 | |
|
|
370 | This method attempts to decode exactly one CBOR value from the beginning |
|
|
371 | of the given C<$buffer>. The value is removed from the C<$buffer> on |
|
|
372 | success. When C<$buffer> doesn't contain a complete value yet, it returns |
|
|
373 | nothing. Finally, when the C<$buffer> doesn't start with something |
|
|
374 | that could ever be a valid CBOR value, it raises an exception, just as |
|
|
375 | C<decode> would. In the latter case the decoder state is undefined and |
|
|
376 | must be reset before being able to parse further. |
|
|
377 | |
|
|
378 | This method modifies the C<$buffer> in place. When no CBOR value can be |
|
|
379 | decoded, the decoder stores the current string offset. On the next call, |
|
|
380 | continues decoding at the place where it stopped before. For this to make |
|
|
381 | sense, the C<$buffer> must begin with the same octets as on previous |
|
|
382 | unsuccessful calls. |
|
|
383 | |
|
|
384 | You can call this method in scalar context, in which case it either |
|
|
385 | returns a decoded value or C<undef>. This makes it impossible to |
|
|
386 | distinguish between CBOR null values (which decode to C<undef>) and an |
|
|
387 | unsuccessful decode, which is often acceptable. |
|
|
388 | |
|
|
389 | =item @decoded = $cbor->incr_parse_multiple ($buffer) |
|
|
390 | |
|
|
391 | Same as C<incr_parse>, but attempts to decode as many CBOR values as |
|
|
392 | possible in one go, instead of at most one. Calls to C<incr_parse> and |
|
|
393 | C<incr_parse_multiple> can be interleaved. |
|
|
394 | |
|
|
395 | =item $cbor->incr_reset |
|
|
396 | |
|
|
397 | Resets the incremental decoder. This throws away any saved state, so that |
|
|
398 | subsequent calls to C<incr_parse> or C<incr_parse_multiple> start to parse |
|
|
399 | a new CBOR value from the beginning of the C<$buffer> again. |
|
|
400 | |
|
|
401 | This method can be caled at any time, but it I<must> be called if you want |
|
|
402 | to change your C<$buffer> or there was a decoding error and you want to |
|
|
403 | reuse the C<$cbor> object for future incremental parsings. |
|
|
404 | |
|
|
405 | =back |
|
|
406 | |
313 | |
407 | |
314 | =head1 MAPPING |
408 | =head1 MAPPING |
315 | |
409 | |
316 | This section describes how CBOR::XS maps Perl values to CBOR values and |
410 | This section describes how CBOR::XS maps Perl values to CBOR values and |
317 | vice versa. These mappings are designed to "do the right thing" in most |
411 | vice versa. These mappings are designed to "do the right thing" in most |
… | |
… | |
387 | |
481 | |
388 | =item hash references |
482 | =item hash references |
389 | |
483 | |
390 | Perl hash references become CBOR maps. As there is no inherent ordering in |
484 | Perl hash references become CBOR maps. As there is no inherent ordering in |
391 | hash keys (or CBOR maps), they will usually be encoded in a pseudo-random |
485 | hash keys (or CBOR maps), they will usually be encoded in a pseudo-random |
392 | order. This order can be different each time a hahs is encoded. |
486 | order. This order can be different each time a hash is encoded. |
393 | |
487 | |
394 | Currently, tied hashes will use the indefinite-length format, while normal |
488 | Currently, tied hashes will use the indefinite-length format, while normal |
395 | hashes will use the fixed-length format. |
489 | hashes will use the fixed-length format. |
396 | |
490 | |
397 | =item array references |
491 | =item array references |
… | |
… | |
720 | |
814 | |
721 | These tags are automatically created (and decoded) for serialisable |
815 | These tags are automatically created (and decoded) for serialisable |
722 | objects using the C<FREEZE/THAW> methods (the L<Types::Serialier> object |
816 | objects using the C<FREEZE/THAW> methods (the L<Types::Serialier> object |
723 | serialisation protocol). See L<OBJECT SERIALISATION> for details. |
817 | serialisation protocol). See L<OBJECT SERIALISATION> for details. |
724 | |
818 | |
725 | =item 28, 29 (shareable, sharedref, L <http://cbor.schmorp.de/value-sharing>) |
819 | =item 28, 29 (shareable, sharedref, L<http://cbor.schmorp.de/value-sharing>) |
726 | |
820 | |
727 | These tags are automatically decoded when encountered (and they do not |
821 | These tags are automatically decoded when encountered (and they do not |
728 | result in a cyclic data structure, see C<allow_cycles>), resulting in |
822 | result in a cyclic data structure, see C<allow_cycles>), resulting in |
729 | shared values in the decoded object. They are only encoded, however, when |
823 | shared values in the decoded object. They are only encoded, however, when |
730 | C<allow_sharing> is enabled. |
824 | C<allow_sharing> is enabled. |
… | |
… | |
740 | will be shared, others will not. While non-reference shared values can be |
834 | will be shared, others will not. While non-reference shared values can be |
741 | generated in Perl with some effort, they were considered too unimportant |
835 | generated in Perl with some effort, they were considered too unimportant |
742 | to be supported in the encoder. The decoder, however, will decode these |
836 | to be supported in the encoder. The decoder, however, will decode these |
743 | values as shared values. |
837 | values as shared values. |
744 | |
838 | |
745 | =item 256, 25 (stringref-namespace, stringref, L <http://cbor.schmorp.de/stringref>) |
839 | =item 256, 25 (stringref-namespace, stringref, L<http://cbor.schmorp.de/stringref>) |
746 | |
840 | |
747 | These tags are automatically decoded when encountered. They are only |
841 | These tags are automatically decoded when encountered. They are only |
748 | encoded, however, when C<pack_strings> is enabled. |
842 | encoded, however, when C<pack_strings> is enabled. |
749 | |
843 | |
750 | =item 22098 (indirection, L<http://cbor.schmorp.de/indirection>) |
844 | =item 22098 (indirection, L<http://cbor.schmorp.de/indirection>) |
… | |
… | |
773 | perl core distribution (e.g. L<URI>), it is (currently) up to the user to |
867 | perl core distribution (e.g. L<URI>), it is (currently) up to the user to |
774 | provide these modules. The decoding usually fails with an exception if the |
868 | provide these modules. The decoding usually fails with an exception if the |
775 | required module cannot be loaded. |
869 | required module cannot be loaded. |
776 | |
870 | |
777 | =over 4 |
871 | =over 4 |
|
|
872 | |
|
|
873 | =item 0, 1 (date/time string, seconds since the epoch) |
|
|
874 | |
|
|
875 | These tags are decoded into L<Time::Piece> objects. The corresponding |
|
|
876 | C<Time::Piece::TO_CBOR> method always encodes into tag 1 values currently. |
|
|
877 | |
|
|
878 | The L<Time::Piece> API is generally surprisingly bad, and fractional |
|
|
879 | seconds are only accidentally kept intact, so watch out. On the plus side, |
|
|
880 | the module comes with perl since 5.10, which has to count for something. |
778 | |
881 | |
779 | =item 2, 3 (positive/negative bignum) |
882 | =item 2, 3 (positive/negative bignum) |
780 | |
883 | |
781 | These tags are decoded into L<Math::BigInt> objects. The corresponding |
884 | These tags are decoded into L<Math::BigInt> objects. The corresponding |
782 | C<Math::BigInt::TO_CBOR> method encodes "small" bigints into normal CBOR |
885 | C<Math::BigInt::TO_CBOR> method encodes "small" bigints into normal CBOR |
… | |
… | |
919 | |
1022 | |
920 | |
1023 | |
921 | =head1 LIMITATIONS ON PERLS WITHOUT 64-BIT INTEGER SUPPORT |
1024 | =head1 LIMITATIONS ON PERLS WITHOUT 64-BIT INTEGER SUPPORT |
922 | |
1025 | |
923 | On perls that were built without 64 bit integer support (these are rare |
1026 | On perls that were built without 64 bit integer support (these are rare |
924 | nowadays, even on 32 bit architectures), support for any kind of 64 bit |
1027 | nowadays, even on 32 bit architectures, as all major Perl distributions |
|
|
1028 | are built with 64 bit integer support), support for any kind of 64 bit |
925 | integer in CBOR is very limited - most likely, these 64 bit values will |
1029 | integer in CBOR is very limited - most likely, these 64 bit values will |
926 | be truncated, corrupted, or otherwise not decoded correctly. This also |
1030 | be truncated, corrupted, or otherwise not decoded correctly. This also |
927 | includes string, array and map sizes that are stored as 64 bit integers. |
1031 | includes string, array and map sizes that are stored as 64 bit integers. |
928 | |
1032 | |
929 | |
1033 | |
… | |
… | |
947 | service. I put the contact address into my modules for a reason. |
1051 | service. I put the contact address into my modules for a reason. |
948 | |
1052 | |
949 | =cut |
1053 | =cut |
950 | |
1054 | |
951 | our %FILTER = ( |
1055 | our %FILTER = ( |
952 | # 0 # rfc4287 datetime, utf-8 |
1056 | 0 => sub { # rfc4287 datetime, utf-8 |
953 | # 1 # unix timestamp, any |
1057 | require Time::Piece; |
|
|
1058 | # Time::Piece::Strptime uses the "incredibly flexible date parsing routine" |
|
|
1059 | # from FreeBSD, which can't parse ISO 8601, RFC3339, RFC4287 or much of anything |
|
|
1060 | # else either. Whats incredibe over standard strptime totally escapes me. |
|
|
1061 | # doesn't do fractional times, either. sigh. |
|
|
1062 | # In fact, it's all a lie, it uses whatever strptime it wants, and of course, |
|
|
1063 | # they are all incompatible. The openbsd one simply ignores %z (but according to the |
|
|
1064 | # docs, it would be much more incredibly flexible indeed. If it worked, that is.). |
|
|
1065 | scalar eval { |
|
|
1066 | my $s = $_[1]; |
|
|
1067 | |
|
|
1068 | $s =~ s/Z$/+00:00/; |
|
|
1069 | $s =~ s/(\.[0-9]+)?([+-][0-9][0-9]):([0-9][0-9])$// |
|
|
1070 | or die; |
|
|
1071 | |
|
|
1072 | my $b = $1 - ($2 * 60 + $3) * 60; # fractional part + offset. hopefully |
|
|
1073 | my $d = Time::Piece->strptime ($s, "%Y-%m-%dT%H:%M:%S"); |
|
|
1074 | |
|
|
1075 | Time::Piece::gmtime ($d->epoch + $b) |
|
|
1076 | } || die "corrupted CBOR date/time string ($_[0])"; |
|
|
1077 | }, |
|
|
1078 | |
|
|
1079 | 1 => sub { # seconds since the epoch, possibly fractional |
|
|
1080 | require Time::Piece; |
|
|
1081 | scalar Time::Piece::gmtime (pop) |
|
|
1082 | }, |
954 | |
1083 | |
955 | 2 => sub { # pos bigint |
1084 | 2 => sub { # pos bigint |
956 | require Math::BigInt; |
1085 | require Math::BigInt; |
957 | Math::BigInt->new ("0x" . unpack "H*", pop) |
1086 | Math::BigInt->new ("0x" . unpack "H*", pop) |
958 | }, |
1087 | }, |
… | |
… | |
994 | } |
1123 | } |
995 | |
1124 | |
996 | sub URI::TO_CBOR { |
1125 | sub URI::TO_CBOR { |
997 | my $uri = $_[0]->as_string; |
1126 | my $uri = $_[0]->as_string; |
998 | utf8::upgrade $uri; |
1127 | utf8::upgrade $uri; |
999 | CBOR::XS::tag 32, $uri |
1128 | tag 32, $uri |
1000 | } |
1129 | } |
1001 | |
1130 | |
1002 | sub Math::BigInt::TO_CBOR { |
1131 | sub Math::BigInt::TO_CBOR { |
1003 | if ($_[0] >= -2147483648 && $_[0] <= 2147483647) { |
1132 | if ($_[0] >= -2147483648 && $_[0] <= 2147483647) { |
1004 | $_[0]->numify |
1133 | $_[0]->numify |
1005 | } else { |
1134 | } else { |
1006 | my $hex = substr $_[0]->as_hex, 2; |
1135 | my $hex = substr $_[0]->as_hex, 2; |
1007 | $hex = "0$hex" if 1 & length $hex; # sigh |
1136 | $hex = "0$hex" if 1 & length $hex; # sigh |
1008 | CBOR::XS::tag $_[0] >= 0 ? 2 : 3, pack "H*", $hex |
1137 | tag $_[0] >= 0 ? 2 : 3, pack "H*", $hex |
1009 | } |
1138 | } |
1010 | } |
1139 | } |
1011 | |
1140 | |
1012 | sub Math::BigFloat::TO_CBOR { |
1141 | sub Math::BigFloat::TO_CBOR { |
1013 | my ($m, $e) = $_[0]->parts; |
1142 | my ($m, $e) = $_[0]->parts; |
1014 | CBOR::XS::tag 4, [$e->numify, $m] |
1143 | tag 4, [$e->numify, $m] |
|
|
1144 | } |
|
|
1145 | |
|
|
1146 | sub Time::Piece::TO_CBOR { |
|
|
1147 | tag 1, 0 + $_[0]->epoch |
1015 | } |
1148 | } |
1016 | |
1149 | |
1017 | XSLoader::load "CBOR::XS", $VERSION; |
1150 | XSLoader::load "CBOR::XS", $VERSION; |
1018 | |
1151 | |
1019 | =head1 SEE ALSO |
1152 | =head1 SEE ALSO |