… | |
… | |
48 | Regarding compactness, C<CBOR::XS>-encoded data structures are usually |
48 | Regarding compactness, C<CBOR::XS>-encoded data structures are usually |
49 | about 20% smaller than the same data encoded as (compact) JSON or |
49 | about 20% smaller than the same data encoded as (compact) JSON or |
50 | L<Storable>. |
50 | L<Storable>. |
51 | |
51 | |
52 | In addition to the core CBOR data format, this module implements a |
52 | In addition to the core CBOR data format, this module implements a |
53 | number of extensions, to support cyclic and shared data structures (see |
53 | number of extensions, to support cyclic and shared data structures |
54 | C<allow_sharing>), string deduplication (see C<pack_strings>) and scalar |
54 | (see C<allow_sharing> and C<allow_cycles>), string deduplication (see |
55 | references (always enabled). |
55 | C<pack_strings>) and scalar references (always enabled). |
56 | |
56 | |
57 | The primary goal of this module is to be I<correct> and the secondary goal |
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. |
58 | is to be I<fast>. To reach the latter goal it was written in C. |
59 | |
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 |
… | |
… | |
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 = '1.0'; |
69 | our $VERSION = 1.11; |
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; |
… | |
… | |
180 | reference to the earlier value. |
180 | reference to the earlier value. |
181 | |
181 | |
182 | This means that such values will only be encoded once, and will not result |
182 | This means that such values will only be encoded once, and will not result |
183 | in a deep cloning of the value on decode, in decoders supporting the value |
183 | in a deep cloning of the value on decode, in decoders supporting the value |
184 | sharing extension. This also makes it possible to encode cyclic data |
184 | sharing extension. This also makes it possible to encode cyclic data |
185 | structures. |
185 | structures (which need C<allow_cycles> to ne enabled to be decoded by this |
|
|
186 | module). |
186 | |
187 | |
187 | It is recommended to leave it off unless you know your |
188 | It is recommended to leave it off unless you know your |
188 | communication partner supports the value sharing extensions to CBOR |
189 | communication partner supports the value sharing extensions to CBOR |
189 | (L<http://cbor.schmorp.de/value-sharing>), as without decoder support, the |
190 | (L<http://cbor.schmorp.de/value-sharing>), as without decoder support, the |
190 | resulting data structure might be unusable. |
191 | resulting data structure might be unusable. |
191 | |
192 | |
192 | Detecting shared values incurs a runtime overhead when values are encoded |
193 | Detecting shared values incurs a runtime overhead when values are encoded |
193 | that have a reference counter large than one, and might unnecessarily |
194 | that have a reference counter large than one, and might unnecessarily |
194 | increase the encoded size, as potentially shared values are encode as |
195 | increase the encoded size, as potentially shared values are encode as |
195 | sharable whether or not they are actually shared. |
196 | shareable whether or not they are actually shared. |
196 | |
197 | |
197 | At the moment, only targets of references can be shared (e.g. scalars, |
198 | At the moment, only targets of references can be shared (e.g. scalars, |
198 | arrays or hashes pointed to by a reference). Weirder constructs, such as |
199 | arrays or hashes pointed to by a reference). Weirder constructs, such as |
199 | an array with multiple "copies" of the I<same> string, which are hard but |
200 | an array with multiple "copies" of the I<same> string, which are hard but |
200 | not impossible to create in Perl, are not supported (this is the same as |
201 | not impossible to create in Perl, are not supported (this is the same as |
… | |
… | |
205 | structures cannot be encoded in this mode. |
206 | structures cannot be encoded in this mode. |
206 | |
207 | |
207 | This option does not affect C<decode> in any way - shared values and |
208 | This option does not affect C<decode> in any way - shared values and |
208 | references will always be decoded properly if present. |
209 | references will always be decoded properly if present. |
209 | |
210 | |
|
|
211 | =item $cbor = $cbor->allow_cycles ([$enable]) |
|
|
212 | |
|
|
213 | =item $enabled = $cbor->get_allow_cycles |
|
|
214 | |
|
|
215 | If C<$enable> is true (or missing), then C<decode> will happily decode |
|
|
216 | self-referential (cyclic) data structures. By default these will not be |
|
|
217 | decoded, as they need manual cleanup to avoid memory leaks, so code that |
|
|
218 | isn't prepared for this will not leak memory. |
|
|
219 | |
|
|
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. |
|
|
222 | |
|
|
223 | This option does not affect C<encode> in any way - shared values and |
|
|
224 | references will always be decoded properly if present. |
|
|
225 | |
210 | =item $cbor = $cbor->pack_strings ([$enable]) |
226 | =item $cbor = $cbor->pack_strings ([$enable]) |
211 | |
227 | |
212 | =item $enabled = $cbor->get_pack_strings |
228 | =item $enabled = $cbor->get_pack_strings |
213 | |
229 | |
214 | If C<$enable> is true (or missing), then C<encode> will try not to encode |
230 | If C<$enable> is true (or missing), then C<encode> will try not to encode |
… | |
… | |
225 | If C<$enable> is false (the default), then C<encode> will encode strings |
241 | If C<$enable> is false (the default), then C<encode> will encode strings |
226 | the standard CBOR way. |
242 | the standard CBOR way. |
227 | |
243 | |
228 | This option does not affect C<decode> in any way - string references will |
244 | This option does not affect C<decode> in any way - string references will |
229 | always be decoded properly if present. |
245 | always be decoded properly if present. |
|
|
246 | |
|
|
247 | =item $cbor = $cbor->validate_utf8 ([$enable]) |
|
|
248 | |
|
|
249 | =item $enabled = $cbor->get_validate_utf8 |
|
|
250 | |
|
|
251 | If C<$enable> is true (or missing), then C<decode> will validate that |
|
|
252 | elements (text strings) containing UTF-8 data in fact contain valid UTF-8 |
|
|
253 | data (instead of blindly accepting it). This validation obviously takes |
|
|
254 | extra time during decoding. |
|
|
255 | |
|
|
256 | The concept of "valid UTF-8" used is perl's concept, which is a superset |
|
|
257 | of the official UTF-8. |
|
|
258 | |
|
|
259 | If C<$enable> is false (the default), then C<decode> will blindly accept |
|
|
260 | UTF-8 data, marking them as valid UTF-8 in the resulting data structure |
|
|
261 | regardless of whether thats true or not. |
|
|
262 | |
|
|
263 | Perl isn't too happy about corrupted UTF-8 in strings, but should |
|
|
264 | generally not crash or do similarly evil things. Extensions might be not |
|
|
265 | so forgiving, so it's recommended to turn on this setting if you receive |
|
|
266 | untrusted CBOR. |
|
|
267 | |
|
|
268 | This option does not affect C<encode> in any way - strings that are |
|
|
269 | supposedly valid UTF-8 will simply be dumped into the resulting CBOR |
|
|
270 | string without checking whether that is, in fact, true or not. |
230 | |
271 | |
231 | =item $cbor = $cbor->filter ([$cb->($tag, $value)]) |
272 | =item $cbor = $cbor->filter ([$cb->($tag, $value)]) |
232 | |
273 | |
233 | =item $cb_or_undef = $cbor->get_filter |
274 | =item $cb_or_undef = $cbor->get_filter |
234 | |
275 | |
… | |
… | |
704 | |
745 | |
705 | These tags are automatically created (and decoded) for serialisable |
746 | These tags are automatically created (and decoded) for serialisable |
706 | objects using the C<FREEZE/THAW> methods (the L<Types::Serialier> object |
747 | objects using the C<FREEZE/THAW> methods (the L<Types::Serialier> object |
707 | serialisation protocol). See L<OBJECT SERIALISATION> for details. |
748 | serialisation protocol). See L<OBJECT SERIALISATION> for details. |
708 | |
749 | |
709 | =item 28, 29 (sharable, sharedref, L <http://cbor.schmorp.de/value-sharing>) |
750 | =item 28, 29 (shareable, sharedref, L <http://cbor.schmorp.de/value-sharing>) |
710 | |
751 | |
711 | These tags are automatically decoded when encountered, resulting in |
752 | These tags are automatically decoded when encountered (and they do not |
|
|
753 | result in a cyclic data structure, see C<allow_cycles>), resulting in |
712 | shared values in the decoded object. They are only encoded, however, when |
754 | shared values in the decoded object. They are only encoded, however, when |
713 | C<allow_sharable> is enabled. |
755 | C<allow_sharing> is enabled. |
|
|
756 | |
|
|
757 | Not all shared values can be successfully decoded: values that reference |
|
|
758 | themselves will I<currently> decode as C<undef> (this is not the same |
|
|
759 | as a reference pointing to itself, which will be represented as a value |
|
|
760 | that contains an indirect reference to itself - these will be decoded |
|
|
761 | properly). |
|
|
762 | |
|
|
763 | Note that considerably more shared value data structures can be decoded |
|
|
764 | than will be encoded - currently, only values pointed to by references |
|
|
765 | will be shared, others will not. While non-reference shared values can be |
|
|
766 | generated in Perl with some effort, they were considered too unimportant |
|
|
767 | to be supported in the encoder. The decoder, however, will decode these |
|
|
768 | values as shared values. |
714 | |
769 | |
715 | =item 256, 25 (stringref-namespace, stringref, L <http://cbor.schmorp.de/stringref>) |
770 | =item 256, 25 (stringref-namespace, stringref, L <http://cbor.schmorp.de/stringref>) |
716 | |
771 | |
717 | These tags are automatically decoded when encountered. They are only |
772 | These tags are automatically decoded when encountered. They are only |
718 | encoded, however, when C<pack_strings> is enabled. |
773 | encoded, however, when C<pack_strings> is enabled. |
… | |
… | |
743 | perl core distribution (e.g. L<URI>), it is (currently) up to the user to |
798 | perl core distribution (e.g. L<URI>), it is (currently) up to the user to |
744 | provide these modules. The decoding usually fails with an exception if the |
799 | provide these modules. The decoding usually fails with an exception if the |
745 | required module cannot be loaded. |
800 | required module cannot be loaded. |
746 | |
801 | |
747 | =over 4 |
802 | =over 4 |
|
|
803 | |
|
|
804 | =item 0, 1 (date/time string, seconds since the epoch) |
|
|
805 | |
|
|
806 | These tags are decoded into L<Time::Piece> objects. The corresponding |
|
|
807 | C<Time::Piece::TO_CBOR> method always encodes into tag 1 values currently. |
|
|
808 | |
|
|
809 | The L<Time::Piece> API is generally surprisingly bad, and fractional |
|
|
810 | seconds are only accidentally kept intact, so watch out. On the plus side, |
|
|
811 | the module comes with perl since 5.10, which has to count for something. |
748 | |
812 | |
749 | =item 2, 3 (positive/negative bignum) |
813 | =item 2, 3 (positive/negative bignum) |
750 | |
814 | |
751 | These tags are decoded into L<Math::BigInt> objects. The corresponding |
815 | These tags are decoded into L<Math::BigInt> objects. The corresponding |
752 | C<Math::BigInt::TO_CBOR> method encodes "small" bigints into normal CBOR |
816 | C<Math::BigInt::TO_CBOR> method encodes "small" bigints into normal CBOR |
… | |
… | |
917 | service. I put the contact address into my modules for a reason. |
981 | service. I put the contact address into my modules for a reason. |
918 | |
982 | |
919 | =cut |
983 | =cut |
920 | |
984 | |
921 | our %FILTER = ( |
985 | our %FILTER = ( |
922 | # 0 # rfc4287 datetime, utf-8 |
986 | 0 => sub { # rfc4287 datetime, utf-8 |
923 | # 1 # unix timestamp, any |
987 | require Time::Piece; |
|
|
988 | # Time::Piece::Strptime uses the "incredibly flexible date parsing routine" |
|
|
989 | # from FreeBSD, which can't parse ISO 8601, RFC3339, RFC4287 or much of anything |
|
|
990 | # else either. Whats incredibe over standard strptime totally escapes me. |
|
|
991 | # doesn't do fractional times, either. sigh. |
|
|
992 | scalar eval { |
|
|
993 | my $s = $_[1]; |
|
|
994 | |
|
|
995 | $s =~ s/Z$/+00:00/; |
|
|
996 | $s =~ s/(\.[0-9]+)?([+-][0-9][0-9]):([0-9][0-9])/$2$3/ |
|
|
997 | or die; |
|
|
998 | |
|
|
999 | my $f = $1; # fractional part. hopefully |
|
|
1000 | |
|
|
1001 | my $d = Time::Piece->strptime ($s, "%Y-%m-%dT%H:%M:%S%z"); |
|
|
1002 | |
|
|
1003 | Time::Piece::gmtime ($d->epoch + $f) |
|
|
1004 | } || die "corrupted CBOR date/time string ($_[0])"; |
|
|
1005 | }, |
|
|
1006 | |
|
|
1007 | 1 => sub { # seconds since the epoch, possibly fractional |
|
|
1008 | require Time::Piece; |
|
|
1009 | scalar Time::Piece::gmtime (pop) |
|
|
1010 | }, |
924 | |
1011 | |
925 | 2 => sub { # pos bigint |
1012 | 2 => sub { # pos bigint |
926 | require Math::BigInt; |
1013 | require Math::BigInt; |
927 | Math::BigInt->new ("0x" . unpack "H*", pop) |
1014 | Math::BigInt->new ("0x" . unpack "H*", pop) |
928 | }, |
1015 | }, |
… | |
… | |
964 | } |
1051 | } |
965 | |
1052 | |
966 | sub URI::TO_CBOR { |
1053 | sub URI::TO_CBOR { |
967 | my $uri = $_[0]->as_string; |
1054 | my $uri = $_[0]->as_string; |
968 | utf8::upgrade $uri; |
1055 | utf8::upgrade $uri; |
969 | CBOR::XS::tag 32, $uri |
1056 | tag 32, $uri |
970 | } |
1057 | } |
971 | |
1058 | |
972 | sub Math::BigInt::TO_CBOR { |
1059 | sub Math::BigInt::TO_CBOR { |
973 | if ($_[0] >= -2147483648 && $_[0] <= 2147483647) { |
1060 | if ($_[0] >= -2147483648 && $_[0] <= 2147483647) { |
974 | $_[0]->numify |
1061 | $_[0]->numify |
975 | } else { |
1062 | } else { |
976 | my $hex = substr $_[0]->as_hex, 2; |
1063 | my $hex = substr $_[0]->as_hex, 2; |
977 | $hex = "0$hex" if 1 & length $hex; # sigh |
1064 | $hex = "0$hex" if 1 & length $hex; # sigh |
978 | CBOR::XS::tag $_[0] >= 0 ? 2 : 3, pack "H*", $hex |
1065 | tag $_[0] >= 0 ? 2 : 3, pack "H*", $hex |
979 | } |
1066 | } |
980 | } |
1067 | } |
981 | |
1068 | |
982 | sub Math::BigFloat::TO_CBOR { |
1069 | sub Math::BigFloat::TO_CBOR { |
983 | my ($m, $e) = $_[0]->parts; |
1070 | my ($m, $e) = $_[0]->parts; |
984 | CBOR::XS::tag 4, [$e->numify, $m] |
1071 | tag 4, [$e->numify, $m] |
|
|
1072 | } |
|
|
1073 | |
|
|
1074 | sub Time::Piece::TO_CBOR { |
|
|
1075 | tag 1, $_[0]->epoch |
985 | } |
1076 | } |
986 | |
1077 | |
987 | XSLoader::load "CBOR::XS", $VERSION; |
1078 | XSLoader::load "CBOR::XS", $VERSION; |
988 | |
1079 | |
989 | =head1 SEE ALSO |
1080 | =head1 SEE ALSO |