The lejson from mellegar

lejson's Introduction

Le'Json - Lightweight Erlang Json library

Requires R17 since it produces maps on decode. Not yet handling unicode code points like "\uD8B3" (arabic) correctly. Other than that it seem to keep with the standard (rfc4627).

Latest Tag: 0.5.0

Exports

-spec lejson:decode(iolist()) -> list() | map() | {error, not_json}.
-spec lejson:encode(list() | map()) -> binary().

Example usage

1> l(lejson).
ok
2> M = lejson:decode(<<"{\"boolean\": [true, false],
                         \"neg_num\": -12,
                         \"floats\": [-22.3, -22.3e-12, 22.3E-12,
                                      22.3E+4, 22.3E+4, 22.3E4],
                         \"null\": null,
                         \"pos_int\": 6789,
                         \"string_value\": \"value\",
                         \"utf_value\": \"\\uC3B8 and \\uc2a9\",
                         \"arabic\": \"\\uD8B3\\ud8b5\\ud8b8\",
                         \"more unicode\": \" \\uD834 \\uDD1E \",
                         \"array\": [{\"object_inside_array\": 1}],
                         \"nested_array\": [[[79]]],
                         \"another_array\": [1,2,3,[1,[2],3],12]}">>).
#{<<"another_array">> => [1,2,3,[1,[2],3],12],
  <<"arabic">> => <<216,179,216,181,216,184>>,
  <<"array">> => [#{<<"object_inside_array">> => 1}],
  <<"boolean">> => [true,false],
  <<"floats">> => [-22.3,-2.23e-11,2.23e-11,2.23e5,2.23e5,2.23e5],
  <<"more unicode">> => <<32,216,52,32,221,30,32>>,
  <<"neg_num">> => -12,
  <<"nested_array">> => [["O"]],
  <<"null">> => null,
  <<"pos_int">> => 6789,
  <<"string_value">> => <<"value">>,
  <<"utf_value">> => <<"ø and ©"/utf8>>}
3> lejson:encode(M)
<<"{\"another_array\": [1, 2, 3, [1, [2], 3], 12],\"arabic\": \""...>>