revert change from strings/keys being presented as binaries to being presented as lists of characters. ~30% speedup

src/jsx.erl

@@ -282,10 +282,10 @@ multi_test_result() ->
         [{literal, false}],
         [{literal, null}],
         [start_object, end_object],
-        [start_object, {key, <<"key">>}, {string, <<"value">>}, end_object],
+        [start_object, {key, "key"}, {string, "value"}, end_object],
         [start_array, end_array],
         [start_array, {integer, 1}, {integer, 2}, {integer, 3}, end_array],
-        [{string, <<"hope this works">>}]
+        [{string, "hope this works"}]
     ].
 
     

src/jsx_common.hrl

@@ -43,15 +43,15 @@
         | utf32
         | {utf32, little}
     }.
-
+    
 
 -type jsx_event() :: start_object
     | end_object
     | start_array
     | end_array
     | end_json
-    | {key, binary()}
-    | {string, binary()}
+    | {key, list()}
+    | {string, list()}
     | {integer, integer()}
     | {float, float()}
     | {literal, true}

src/jsx_decoder.hrl

@@ -357,11 +357,11 @@ key(Bin, Stack, Opts) ->
 string(Bin, Stack, Opts) -> string(Bin, Stack, Opts, []).
 
 string(<<?quote/?utfx, Rest/binary>>, [key|_] = Stack, Opts, Acc) ->
-    {jsx, {key, unicode:characters_to_binary(lists:reverse(Acc))}, fun() ->
+    {jsx, {key, lists:reverse(Acc)}, fun() ->
         colon(Rest, Stack, Opts)
     end};
 string(<<?quote/?utfx, Rest/binary>>, Stack, Opts, Acc) ->
-    {jsx, {string, unicode:characters_to_binary(lists:reverse(Acc))}, fun() -> 
+    {jsx, {string, lists:reverse(Acc)}, fun() -> 
         maybe_done(Rest, Stack, Opts)
     end};
 string(<<?rsolidus/?utfx, Rest/binary>>, Stack, Opts, Acc) ->
@@ -1228,7 +1228,7 @@ check_bad(List) ->
     ).
 
 check_replaced(List) ->
-    lists:dropwhile(fun({_, [{string, <<16#fffd/utf8>>}|_]}) ->
+    lists:dropwhile(fun({_, [{string, [16#fffd]}|_]}) ->
                 true
             ; (_) -> 
                 false 

src/jsx_terms.erl

@@ -163,8 +163,8 @@ collect(_, _, _) -> erlang:error(badarg).
     
 
 %% helper functions for converting jsx events to term format 
-event({string, String}, _Opts) -> String;
-event({key, Key}, _Opts) -> Key;
+event({string, String}, _Opts) -> unicode:characters_to_binary(String);
+event({key, Key}, _Opts) -> unicode:characters_to_binary(Key);
 event({integer, Integer}, _Opts) -> Integer;
 event({float, Float}, _Opts) -> Float;
 event({literal, Literal}, _Opts) -> Literal.

test/cases/array.test

@@ -1,8 +1,8 @@
 {name, "array"}.
 {jsx, [start_array,
- {string,<<"foo">>},
- {string,<<"bar">>},
- {string,<<"baz">>},
+ {string,"foo"},
+ {string,"bar"},
+ {string,"baz"},
  start_array,
  {literal,true},
  end_array,start_array,
@@ -15,14 +15,14 @@
  {literal,null},
  {float,0.7},
  start_object,
- {key,<<"key">>},
- {string,<<"value">>},
+ {key,"key"},
+ {string,"value"},
  end_object,start_array,start_object,end_object,
  {literal,null},
  {literal,null},
  {literal,null},
  start_array,end_array,end_array,
- {string,<<"\n\r\\">>},
+ {string,"\n\r\\"},
  start_array,
  {integer,-1},
  end_array,end_array,end_json]}.

test/cases/encoded_surrogates.test

@@ -1,3 +1,3 @@
 {name, "encoded_surrogates"}.
-{jsx, [start_array,{string,<<66560/utf8>>},end_array,end_json]}.
+{jsx, [start_array,{string,66560]},end_array,end_json]}.
 {json, "encoded_surrogates.json"}.

test/cases/escaped_noncharacter_ext_replaced.test

@@ -1,4 +1,4 @@
 {name, "escaped noncharacter (extended)"}.
-{jsx, [{string, <<16#fffd/utf8>>}, end_json]}.
+{jsx, [{string, [16#fffd]}, end_json]}.
 {json, "escaped_noncharacter_ext.json"}.
 {jsx_flags, [loose_unicode]}.

test/cases/escaped_noncharacter_replaced.test

@@ -1,4 +1,4 @@
 {name, "escaped noncharacter replacement"}.
-{jsx, [{string,<<16#fffd/utf8>>},end_json]}.
+{jsx, [{string,[16#fffd]},end_json]}.
 {json, "escaped_noncharacter_replaced.json"}.
 {jsx_flags, [loose_unicode]}.

test/cases/escaped_nullbyte_replaced.test

@@ -1,4 +1,4 @@
 {name, "escaped nullbyte replaced"}.
-{jsx, [{string,<<16#fffd/utf8>>},end_json]}.
+{jsx, [{string,[16#fffd]},end_json]}.
 {json, "escaped_nullbyte_replaced.json"}.
 {jsx_flags, [loose_unicode]}.

test/cases/exp.test

@@ -4,9 +4,9 @@
  end_array,
  {float,2.0e7},
  start_object,
- {key,<<"key">>},
+ {key,"key"},
  {float,2.0e7},
- {key,<<"another key">>},
+ {key,"another key"},
  {float,2.0e7},
  end_object,
  {float,4.2e70},

test/cases/fraction.test

@@ -4,14 +4,14 @@
  end_array,
  {float,2.0},
  start_object,
- {key,<<"key">>},
+ {key,"key"},
  {float,2.0e7},
- {key,<<"another key">>},
+ {key,"another key"},
  {float,2.0e7},
  end_object,start_object,
- {key,<<"key">>},
+ {key,"key"},
  {float,2.0},
- {key,<<"another key">>},
+ {key,"another key"},
  {float,2.0},
  end_object,
  {float,4.321},

test/cases/integer.test

@@ -4,9 +4,9 @@
  end_array,
  {integer,20},
  start_object,
- {key,<<"key">>},
+ {key,"key"},
  {integer,20},
- {key,<<"another key">>},
+ {key,"another key"},
  {integer,20},
  end_object,
  {integer,42},

test/cases/multibyte_utf.test

@@ -1,3 +1,3 @@
 {name, "multibyte_utf"}.
-{jsx, [start_array,{string,<<32,119070/utf8,32>>},end_array,end_json]}.
+{jsx, [start_array,{string,[32,119070,32]},end_array,end_json]}.
 {json, "multibyte_utf.json"}.

test/cases/naked_string.test

@@ -1,3 +1,3 @@
 {name, "naked_string"}.
-{jsx, [{string,<<"this is a naked string">>},end_json]}.
+{jsx, [{string,"this is a naked string"},end_json]}.
 {json, "naked_string.json"}.

test/cases/negative_zero.test

@@ -4,9 +4,9 @@
  end_array,
  {integer,0},
  start_object,
- {key,<<"key">>},
+ {key,"key"},
  {integer,0},
- {key,<<"another key">>},
+ {key,"another key"},
  {integer,0},
  end_object,
  {integer,0},

test/cases/noncharacter_replaced.test

@@ -1,4 +1,4 @@
 {name, "noncharacter replaced"}.
-{jsx, [{string,<<16#fffd/utf8>>},end_json]}.
+{jsx, [{string,[16#fffd/utf8]},end_json]}.
 {json, "noncharacter_replaced.json"}.
 {jsx_flags, [loose_unicode]}.

test/cases/nullbyte_replaced.test

@@ -1,4 +1,4 @@
 {name, "nullbyte replaced"}.
-{jsx, [{string,<<16#fffd/utf8>>},end_json]}.
+{jsx, [{string,[16#fffd]},end_json]}.
 {json, "nullbyte_replaced.json"}.
 {jsx_flags, [loose_unicode]}.

test/cases/object.test

@@ -1,22 +1,22 @@
 {name, "object"}.
 {jsx, [start_object,
- {key,<<"foo">>},
- {string,<<"bar">>},
- {key,<<"baz">>},
+ {key,"foo"},
+ {string,"bar"},
+ {key,"baz"},
  {literal,true},
- {key,<<"false">>},
+ {key,"false"},
  {literal,null},
- {key,<<"object">>},
+ {key,"object"},
  start_object,
- {key,<<"key">>},
- {string,<<"value">>},
+ {key,"key"},
+ {string,"value"},
  end_object,
- {key,<<"list">>},
+ {key,"list"},
  start_array,
  {literal,null},
  {literal,null},
  {literal,null},
  start_array,end_array,
- {string,<<"\n\r\\">>},
+ {string,"\n\r\\"},
  end_array,end_object,end_json]}.
 {json, "object.json"}.

test/cases/string.test

@@ -1,5 +1,5 @@
 {name, "string"}.
 {jsx, [start_array,
- {string,<<"this is a random string with \n embedded escapes in it">>},
+ {string,"this is a random string with \n embedded escapes in it"},
  end_array,end_json]}.
 {json, "string.json"}.

test/cases/string_escapes.test

@@ -1,12 +1,12 @@
 {name, "string_escapes"}.
 {jsx, [start_array,
- {string,<<"\"">>},
- {string,<<"\\">>},
- {string,<<"/">>},
- {string,<<"\b">>},
- {string,<<"\f">>},
- {string,<<"\n">>},
- {string,<<"\r">>},
- {string,<<"\t">>},
+ {string,"\""},
+ {string,"\\"},
+ {string,"/"},
+ {string,"\b"},
+ {string,"\f"},
+ {string,"\n"},
+ {string,"\r"},
+ {string,"\t"},
  end_array,end_json]}.
 {json, "string_escapes.json"}.

test/cases/unicode_to_codepoint.test

@@ -1,6 +1,6 @@
 {name, "unicode_to_codepoint"}.
 {jsx, [start_array,
- {string,<<"arabic letter alef: ">>},
- {string,<<1575/utf8>>},
+ {string,"arabic letter alef: "},
+ {string,[1575]},
  end_array,end_json]}.
 {json, "unicode_to_codepoint.json"}.

test/cases/unpaired_surrogate_replaced.test

@@ -1,4 +1,4 @@
 {name, "unpaired surrogate replaced"}.
-{jsx, [start_array,{string,<<16#fffd/utf8, "blah">>},end_array,end_json]}.
+{jsx, [start_array,{string,16#fffd/utf8 ++ "blah"},end_array,end_json]}.
 {json, "unpaired_surrogate_replaced.json"}.
 {jsx_flags, [loose_unicode]}.

test/cases/zero.test

@@ -4,9 +4,9 @@
  end_array,
  {integer,0},
  start_object,
- {key,<<"key">>},
+ {key,"key"},
  {integer,0},
- {key,<<"another key">>},
+ {key,"another key"},
  {integer,0},
  end_object,
  {integer,0},