|
|
1 | =head1 NAME |
|
|
2 | |
| 1 | =encoding utf-8 |
3 | =encoding utf-8 |
| 2 | |
|
|
| 3 | =head1 NAME |
|
|
| 4 | |
4 | |
| 5 | JSON::XS - JSON serialising/deserialising, done correctly and fast |
5 | JSON::XS - JSON serialising/deserialising, done correctly and fast |
| 6 | |
6 | |
| 7 | JSON::XS - 正しくて高速な JSON シリアライザ/デシリアライザ |
7 | JSON::XS - 正しくて高速な JSON シリアライザ/デシリアライザ |
| 8 | (http://fleur.hio.jp/perldoc/mix/lib/JSON/XS.html) |
8 | (http://fleur.hio.jp/perldoc/mix/lib/JSON/XS.html) |
| … | |
… | |
| 681 | => ([], 3) |
681 | => ([], 3) |
| 682 | |
682 | |
| 683 | =back |
683 | =back |
| 684 | |
684 | |
| 685 | |
685 | |
|
|
686 | =head1 INCREMENTAL PARSING |
|
|
687 | |
|
|
688 | [This section is still EXPERIMENTAL] |
|
|
689 | |
|
|
690 | In some cases, there is the need for incremental parsing of JSON |
|
|
691 | texts. While this module always has to keep both JSON text and resulting |
|
|
692 | Perl data structure in memory at one time, it does allow you to parse a |
|
|
693 | JSON stream incrementally. It does so by accumulating text until it has |
|
|
694 | a full JSON object, which it then can decode. This process is similar to |
|
|
695 | using C<decode_prefix> to see if a full JSON object is available, but is |
|
|
696 | much more efficient (JSON::XS will only attempt to parse the JSON text |
|
|
697 | once it is sure it has enough text to get a decisive result, using a very |
|
|
698 | simple but truly incremental parser). |
|
|
699 | |
|
|
700 | The following two methods deal with this. |
|
|
701 | |
|
|
702 | =over 4 |
|
|
703 | |
|
|
704 | =item [void, scalar or list context] = $json->incr_parse ([$string]) |
|
|
705 | |
|
|
706 | This is the central parsing function. It can both append new text and |
|
|
707 | extract objects from the stream accumulated so far (both of these |
|
|
708 | functions are optional). |
|
|
709 | |
|
|
710 | If C<$string> is given, then this string is appended to the already |
|
|
711 | existing JSON fragment stored in the C<$json> object. |
|
|
712 | |
|
|
713 | After that, if the function is called in void context, it will simply |
|
|
714 | return without doing anything further. This can be used to add more text |
|
|
715 | in as many chunks as you want. |
|
|
716 | |
|
|
717 | If the method is called in scalar context, then it will try to extract |
|
|
718 | exactly I<one> JSON object. If that is successful, it will return this |
|
|
719 | object, otherwise it will return C<undef>. This is the most common way of |
|
|
720 | using the method. |
|
|
721 | |
|
|
722 | And finally, in list context, it will try to extract as many objects |
|
|
723 | from the stream as it can find and return them, or the empty list |
|
|
724 | otherwise. For this to work, there must be no separators between the JSON |
|
|
725 | objects or arrays, instead they must be concatenated back-to-back. |
|
|
726 | |
|
|
727 | =item $lvalue_string = $json->incr_text |
|
|
728 | |
|
|
729 | This method returns the currently stored JSON fragment as an lvalue, that |
|
|
730 | is, you can manipulate it. This I<only> works when a preceding call to |
|
|
731 | C<incr_parse> in I<scalar context> successfully returned an object. Under |
|
|
732 | all other circumstances you must not call this function (I mean it. |
|
|
733 | although in simple tests it might actually work, it I<will> fail under |
|
|
734 | real world conditions). As a special exception, you can also call this |
|
|
735 | method before having parsed anything. |
|
|
736 | |
|
|
737 | This function is useful in two cases: a) finding the trailing text after a |
|
|
738 | JSON object or b) parsing multiple JSON objects separated by non-JSON text |
|
|
739 | (such as commas). |
|
|
740 | |
|
|
741 | =back |
|
|
742 | |
|
|
743 | =head2 LIMITATIONS |
|
|
744 | |
|
|
745 | All options that affect decoding are supported, except |
|
|
746 | C<allow_nonref>. The reason for this is that it cannot be made to |
|
|
747 | work sensibly: JSON objects and arrays are self-delimited, i.e. you can concatenate |
|
|
748 | them back to back and still decode them perfectly. This does not hold true |
|
|
749 | for JSON numbers, however. |
|
|
750 | |
|
|
751 | For example, is the string C<1> a single JSON number, or is it simply the |
|
|
752 | start of C<12>? Or is C<12> a single JSON number, or the concatenation |
|
|
753 | of C<1> and C<2>? In neither case you can tell, and this is why JSON::XS |
|
|
754 | takes the conservative route and disallows this case. |
|
|
755 | |
|
|
756 | =head2 EXAMPLES |
|
|
757 | |
|
|
758 | Some examples will make all this clearer. First, a simple example that |
|
|
759 | works similarly to C<decode_prefix>: We want to decode the JSON object at |
|
|
760 | the start of a string and identify the portion after the JSON object: |
|
|
761 | |
|
|
762 | my $text = "[1,2,3] hello"; |
|
|
763 | |
|
|
764 | my $json = new JSON::XS; |
|
|
765 | |
|
|
766 | my $obj = $json->incr_parse ($text) |
|
|
767 | or die "expected JSON object or array at beginning of string"; |
|
|
768 | |
|
|
769 | my $tail = $json->incr_text; |
|
|
770 | # $tail now contains " hello" |
|
|
771 | |
|
|
772 | Easy, isn't it? |
|
|
773 | |
|
|
774 | Now for a more complicated example: Imagine a hypothetical protocol where |
|
|
775 | you read some requests from a TCP stream, and each request is a JSON |
|
|
776 | array, without any separation between them (in fact, it is often useful to |
|
|
777 | use newlines as "separators", as these get interpreted as whitespace at |
|
|
778 | the start of the JSON text, which makes it possible to test said protocol |
|
|
779 | with C<telnet>...). |
|
|
780 | |
|
|
781 | Here is how you'd do it (it is trivial to write this in an event-based |
|
|
782 | manner): |
|
|
783 | |
|
|
784 | my $json = new JSON::XS; |
|
|
785 | |
|
|
786 | # read some data from the socket |
|
|
787 | while (sysread $socket, my $buf, 4096) { |
|
|
788 | |
|
|
789 | # split and decode as many requests as possible |
|
|
790 | for my $request ($json->incr_parse ($buf)) { |
|
|
791 | # act on the $request |
|
|
792 | } |
|
|
793 | } |
|
|
794 | |
|
|
795 | Another complicated example: Assume you have a string with JSON objects |
|
|
796 | or arrays, all separated by (optional) comma characters (e.g. C<[1],[2], |
|
|
797 | [3]>). To parse them, we have to skip the commas between the JSON texts, |
|
|
798 | and here is where the lvalue-ness of C<incr_text> comes in useful: |
|
|
799 | |
|
|
800 | my $text = "[1],[2], [3]"; |
|
|
801 | my $json = new JSON::XS; |
|
|
802 | |
|
|
803 | # void context, so no parsing done |
|
|
804 | $json->incr_parse ($text); |
|
|
805 | |
|
|
806 | # now extract as many objects as possible. note the |
|
|
807 | # use of scalar context so incr_text can be called. |
|
|
808 | while (my $obj = $json->incr_parse) { |
|
|
809 | # do something with $obj |
|
|
810 | |
|
|
811 | # now skip the optional comma |
|
|
812 | $json->incr_text =~ s/^ \s* , //x; |
|
|
813 | } |
|
|
814 | |
|
|
815 | Now lets go for a very complex example: Assume that you have a gigantic |
|
|
816 | JSON array-of-objects, many gigabytes in size, and you want to parse it, |
|
|
817 | but you cannot load it into memory fully (this has actually happened in |
|
|
818 | the real world :). |
|
|
819 | |
|
|
820 | Well, you lost, you have to implement your own JSON parser. But JSON::XS |
|
|
821 | can still help you: You implement a (very simple) array parser and let |
|
|
822 | JSON decode the array elements, which are all full JSON objects on their |
|
|
823 | own (this wouldn't work if the array elements could be JSON numbers, for |
|
|
824 | example): |
|
|
825 | |
|
|
826 | my $json = new JSON::XS; |
|
|
827 | |
|
|
828 | # open the monster |
|
|
829 | open my $fh, "<bigfile.json" |
|
|
830 | or die "bigfile: $!"; |
|
|
831 | |
|
|
832 | # first parse the initial "[" |
|
|
833 | for (;;) { |
|
|
834 | sysread $fh, my $buf, 65536 |
|
|
835 | or die "read error: $!"; |
|
|
836 | $json->incr_parse ($buf); # void context, so no parsing |
|
|
837 | |
|
|
838 | # Exit the loop once we found and removed(!) the initial "[". |
|
|
839 | # In essence, we are (ab-)using the $json object as a simple scalar |
|
|
840 | # we append data to. |
|
|
841 | last if $json->incr_text =~ s/^ \s* \[ //x; |
|
|
842 | } |
|
|
843 | |
|
|
844 | # now we have the skipped the initial "[", so continue |
|
|
845 | # parsing all the elements. |
|
|
846 | for (;;) { |
|
|
847 | # in this loop we read data until we got a single JSON object |
|
|
848 | for (;;) { |
|
|
849 | if (my $obj = $json->incr_parse) { |
|
|
850 | # do something with $obj |
|
|
851 | last; |
|
|
852 | } |
|
|
853 | |
|
|
854 | # add more data |
|
|
855 | sysread $fh, my $buf, 65536 |
|
|
856 | or die "read error: $!"; |
|
|
857 | $json->incr_parse ($buf); # void context, so no parsing |
|
|
858 | } |
|
|
859 | |
|
|
860 | # in this loop we read data until we either found and parsed the |
|
|
861 | # separating "," between elements, or the final "]" |
|
|
862 | for (;;) { |
|
|
863 | # first skip whitespace |
|
|
864 | $json->incr_text =~ s/^\s*//; |
|
|
865 | |
|
|
866 | # if we find "]", we are done |
|
|
867 | if ($json->incr_text =~ s/^\]//) { |
|
|
868 | print "finished.\n"; |
|
|
869 | exit; |
|
|
870 | } |
|
|
871 | |
|
|
872 | # if we find ",", we can continue with the next element |
|
|
873 | if ($json->incr_text =~ s/^,//) { |
|
|
874 | last; |
|
|
875 | } |
|
|
876 | |
|
|
877 | # if we find anything else, we have a parse error! |
|
|
878 | if (length $json->incr_text) { |
|
|
879 | die "parse error near ", $json->incr_text; |
|
|
880 | } |
|
|
881 | |
|
|
882 | # else add more data |
|
|
883 | sysread $fh, my $buf, 65536 |
|
|
884 | or die "read error: $!"; |
|
|
885 | $json->incr_parse ($buf); # void context, so no parsing |
|
|
886 | } |
|
|
887 | |
|
|
888 | This is a complex example, but most of the complexity comes from the fact |
|
|
889 | that we are trying to be correct (bear with me if I am wrong, I never ran |
|
|
890 | the above example :). |
|
|
891 | |
|
|
892 | |
|
|
893 | |
| 686 | =head1 MAPPING |
894 | =head1 MAPPING |
| 687 | |
895 | |
| 688 | This section describes how JSON::XS maps Perl values to JSON values and |
896 | This section describes how JSON::XS maps Perl values to JSON values and |
| 689 | vice versa. These mappings are designed to "do the right thing" in most |
897 | vice versa. These mappings are designed to "do the right thing" in most |
| 690 | circumstances automatically, preserving round-tripping characteristics |
898 | circumstances automatically, preserving round-tripping characteristics |
| … | |
… | |
| 825 | my $x = "3"; # some variable containing a string |
1033 | my $x = "3"; # some variable containing a string |
| 826 | $x += 0; # numify it, ensuring it will be dumped as a number |
1034 | $x += 0; # numify it, ensuring it will be dumped as a number |
| 827 | $x *= 1; # same thing, the choice is yours. |
1035 | $x *= 1; # same thing, the choice is yours. |
| 828 | |
1036 | |
| 829 | You can not currently force the type in other, less obscure, ways. Tell me |
1037 | You can not currently force the type in other, less obscure, ways. Tell me |
| 830 | if you need this capability (but don't forget to explain why its needed |
1038 | if you need this capability (but don't forget to explain why it's needed |
| 831 | :). |
1039 | :). |
| 832 | |
1040 | |
| 833 | =back |
1041 | =back |
| 834 | |
1042 | |
| 835 | |
1043 | |
| … | |
… | |
| 837 | |
1045 | |
| 838 | The interested reader might have seen a number of flags that signify |
1046 | The interested reader might have seen a number of flags that signify |
| 839 | encodings or codesets - C<utf8>, C<latin1> and C<ascii>. There seems to be |
1047 | encodings or codesets - C<utf8>, C<latin1> and C<ascii>. There seems to be |
| 840 | some confusion on what these do, so here is a short comparison: |
1048 | some confusion on what these do, so here is a short comparison: |
| 841 | |
1049 | |
| 842 | C<utf8> controls wether the JSON text created by C<encode> (and expected |
1050 | C<utf8> controls whether the JSON text created by C<encode> (and expected |
| 843 | by C<decode>) is UTF-8 encoded or not, while C<latin1> and C<ascii> only |
1051 | by C<decode>) is UTF-8 encoded or not, while C<latin1> and C<ascii> only |
| 844 | control wether C<encode> escapes character values outside their respective |
1052 | control whether C<encode> escapes character values outside their respective |
| 845 | codeset range. Neither of these flags conflict with each other, although |
1053 | codeset range. Neither of these flags conflict with each other, although |
| 846 | some combinations make less sense than others. |
1054 | some combinations make less sense than others. |
| 847 | |
1055 | |
| 848 | Care has been taken to make all flags symmetrical with respect to |
1056 | Care has been taken to make all flags symmetrical with respect to |
| 849 | C<encode> and C<decode>, that is, texts encoded with any combination of |
1057 | C<encode> and C<decode>, that is, texts encoded with any combination of |
| … | |
… | |
| 925 | as UTF-8, ISO-8859-1, ASCII, KOI8-R or most about any character set and |
1133 | as UTF-8, ISO-8859-1, ASCII, KOI8-R or most about any character set and |
| 926 | 8-bit-encoding, and still get the same data structure back. This is useful |
1134 | 8-bit-encoding, and still get the same data structure back. This is useful |
| 927 | when your channel for JSON transfer is not 8-bit clean or the encoding |
1135 | when your channel for JSON transfer is not 8-bit clean or the encoding |
| 928 | might be mangled in between (e.g. in mail), and works because ASCII is a |
1136 | might be mangled in between (e.g. in mail), and works because ASCII is a |
| 929 | proper subset of most 8-bit and multibyte encodings in use in the world. |
1137 | proper subset of most 8-bit and multibyte encodings in use in the world. |
| 930 | |
|
|
| 931 | =back |
|
|
| 932 | |
|
|
| 933 | |
|
|
| 934 | =head1 COMPARISON |
|
|
| 935 | |
|
|
| 936 | As already mentioned, this module was created because none of the existing |
|
|
| 937 | JSON modules could be made to work correctly. First I will describe the |
|
|
| 938 | problems (or pleasures) I encountered with various existing JSON modules, |
|
|
| 939 | followed by some benchmark values. JSON::XS was designed not to suffer |
|
|
| 940 | from any of these problems or limitations. |
|
|
| 941 | |
|
|
| 942 | =over 4 |
|
|
| 943 | |
|
|
| 944 | =item JSON 2.xx |
|
|
| 945 | |
|
|
| 946 | A marvellous piece of engineering, this module either uses JSON::XS |
|
|
| 947 | directly when available (so will be 100% compatible with it, including |
|
|
| 948 | speed), or it uses JSON::PP, which is basically JSON::XS translated to |
|
|
| 949 | Pure Perl, which should be 100% compatible with JSON::XS, just a bit |
|
|
| 950 | slower. |
|
|
| 951 | |
|
|
| 952 | You cannot really lose by using this module, especially as it tries very |
|
|
| 953 | hard to work even with ancient Perl versions, while JSON::XS does not. |
|
|
| 954 | |
|
|
| 955 | =item JSON 1.07 |
|
|
| 956 | |
|
|
| 957 | Slow (but very portable, as it is written in pure Perl). |
|
|
| 958 | |
|
|
| 959 | Undocumented/buggy Unicode handling (how JSON handles Unicode values is |
|
|
| 960 | undocumented. One can get far by feeding it Unicode strings and doing |
|
|
| 961 | en-/decoding oneself, but Unicode escapes are not working properly). |
|
|
| 962 | |
|
|
| 963 | No round-tripping (strings get clobbered if they look like numbers, e.g. |
|
|
| 964 | the string C<2.0> will encode to C<2.0> instead of C<"2.0">, and that will |
|
|
| 965 | decode into the number 2. |
|
|
| 966 | |
|
|
| 967 | =item JSON::PC 0.01 |
|
|
| 968 | |
|
|
| 969 | Very fast. |
|
|
| 970 | |
|
|
| 971 | Undocumented/buggy Unicode handling. |
|
|
| 972 | |
|
|
| 973 | No round-tripping. |
|
|
| 974 | |
|
|
| 975 | Has problems handling many Perl values (e.g. regex results and other magic |
|
|
| 976 | values will make it croak). |
|
|
| 977 | |
|
|
| 978 | Does not even generate valid JSON (C<{1,2}> gets converted to C<{1:2}> |
|
|
| 979 | which is not a valid JSON text. |
|
|
| 980 | |
|
|
| 981 | Unmaintained (maintainer unresponsive for many months, bugs are not |
|
|
| 982 | getting fixed). |
|
|
| 983 | |
|
|
| 984 | =item JSON::Syck 0.21 |
|
|
| 985 | |
|
|
| 986 | Very buggy (often crashes). |
|
|
| 987 | |
|
|
| 988 | Very inflexible (no human-readable format supported, format pretty much |
|
|
| 989 | undocumented. I need at least a format for easy reading by humans and a |
|
|
| 990 | single-line compact format for use in a protocol, and preferably a way to |
|
|
| 991 | generate ASCII-only JSON texts). |
|
|
| 992 | |
|
|
| 993 | Completely broken (and confusingly documented) Unicode handling (Unicode |
|
|
| 994 | escapes are not working properly, you need to set ImplicitUnicode to |
|
|
| 995 | I<different> values on en- and decoding to get symmetric behaviour). |
|
|
| 996 | |
|
|
| 997 | No round-tripping (simple cases work, but this depends on whether the scalar |
|
|
| 998 | value was used in a numeric context or not). |
|
|
| 999 | |
|
|
| 1000 | Dumping hashes may skip hash values depending on iterator state. |
|
|
| 1001 | |
|
|
| 1002 | Unmaintained (maintainer unresponsive for many months, bugs are not |
|
|
| 1003 | getting fixed). |
|
|
| 1004 | |
|
|
| 1005 | Does not check input for validity (i.e. will accept non-JSON input and |
|
|
| 1006 | return "something" instead of raising an exception. This is a security |
|
|
| 1007 | issue: imagine two banks transferring money between each other using |
|
|
| 1008 | JSON. One bank might parse a given non-JSON request and deduct money, |
|
|
| 1009 | while the other might reject the transaction with a syntax error. While a |
|
|
| 1010 | good protocol will at least recover, that is extra unnecessary work and |
|
|
| 1011 | the transaction will still not succeed). |
|
|
| 1012 | |
|
|
| 1013 | =item JSON::DWIW 0.04 |
|
|
| 1014 | |
|
|
| 1015 | Very fast. Very natural. Very nice. |
|
|
| 1016 | |
|
|
| 1017 | Undocumented Unicode handling (but the best of the pack. Unicode escapes |
|
|
| 1018 | still don't get parsed properly). |
|
|
| 1019 | |
|
|
| 1020 | Very inflexible. |
|
|
| 1021 | |
|
|
| 1022 | No round-tripping. |
|
|
| 1023 | |
|
|
| 1024 | Does not generate valid JSON texts (key strings are often unquoted, empty keys |
|
|
| 1025 | result in nothing being output) |
|
|
| 1026 | |
|
|
| 1027 | Does not check input for validity. |
|
|
| 1028 | |
1138 | |
| 1029 | =back |
1139 | =back |
| 1030 | |
1140 | |
| 1031 | |
1141 | |
| 1032 | =head2 JSON and YAML |
1142 | =head2 JSON and YAML |
| … | |
… | |
| 1193 | =head1 THREADS |
1303 | =head1 THREADS |
| 1194 | |
1304 | |
| 1195 | This module is I<not> guaranteed to be thread safe and there are no |
1305 | This module is I<not> guaranteed to be thread safe and there are no |
| 1196 | plans to change this until Perl gets thread support (as opposed to the |
1306 | plans to change this until Perl gets thread support (as opposed to the |
| 1197 | horribly slow so-called "threads" which are simply slow and bloated |
1307 | horribly slow so-called "threads" which are simply slow and bloated |
| 1198 | process simulations - use fork, its I<much> faster, cheaper, better). |
1308 | process simulations - use fork, it's I<much> faster, cheaper, better). |
| 1199 | |
1309 | |
| 1200 | (It might actually work, but you have been warned). |
1310 | (It might actually work, but you have been warned). |
| 1201 | |
1311 | |
| 1202 | |
1312 | |
| 1203 | =head1 BUGS |
1313 | =head1 BUGS |
| 1204 | |
1314 | |
| 1205 | While the goal of this module is to be correct, that unfortunately does |
1315 | While the goal of this module is to be correct, that unfortunately does |
| 1206 | not mean its bug-free, only that I think its design is bug-free. It is |
1316 | not mean it's bug-free, only that I think its design is bug-free. It is |
| 1207 | still relatively early in its development. If you keep reporting bugs they |
1317 | still relatively early in its development. If you keep reporting bugs they |
| 1208 | will be fixed swiftly, though. |
1318 | will be fixed swiftly, though. |
| 1209 | |
1319 | |
| 1210 | Please refrain from using rt.cpan.org or any other bug reporting |
1320 | Please refrain from using rt.cpan.org or any other bug reporting |
| 1211 | service. I put the contact address into my modules for a reason. |
1321 | service. I put the contact address into my modules for a reason. |
| … | |
… | |
| 1233 | "--" => sub { $_[0] = ${$_[0]} - 1 }, |
1343 | "--" => sub { $_[0] = ${$_[0]} - 1 }, |
| 1234 | fallback => 1; |
1344 | fallback => 1; |
| 1235 | |
1345 | |
| 1236 | 1; |
1346 | 1; |
| 1237 | |
1347 | |
|
|
1348 | =head1 SEE ALSO |
|
|
1349 | |
|
|
1350 | The F<json_xs> command line utility for quick experiments. |
|
|
1351 | |
| 1238 | =head1 AUTHOR |
1352 | =head1 AUTHOR |
| 1239 | |
1353 | |
| 1240 | Marc Lehmann <schmorp@schmorp.de> |
1354 | Marc Lehmann <schmorp@schmorp.de> |
| 1241 | http://home.schmorp.de/ |
1355 | http://home.schmorp.de/ |
| 1242 | |
1356 | |
