-module(neotoma_parse).
-export([parse/1,file/1]).
-define(p_anything,true).
-define(p_charclass,true).
-define(p_choose,true).
-define(p_label,true).
-define(p_not,true).
-define(p_one_or_more,true).
-define(p_optional,true).
-define(p_scan,true).
-define(p_seq,true).
-define(p_string,true).
-define(p_zero_or_more,true).


% insert escapes into a string
-spec escape_string(string()) -> string().
escape_string(String) -> escape_string(String, []).

-spec escape_string(string(), string()) -> string().
escape_string([], Output) ->
  lists:reverse(Output);
escape_string([H|T], Output) ->
  escape_string(T,
    case H of
        $/  -> [$/,$\\|Output];
        $\" -> [$\",$\\|Output];     % " comment inserted to help some editors with highlighting the generated parser
        $\' -> [$\',$\\|Output];     % ' comment inserted to help some editors with highlighting the generated parser
        $\b -> [$b,$\\|Output];
        $\d -> [$d,$\\|Output];
        $\e -> [$e,$\\|Output];
        $\f -> [$f,$\\|Output];
        $\n -> [$n,$\\|Output];
        $\r -> [$r,$\\|Output];
        $\s -> [$s,$\\|Output];
        $\t -> [$t,$\\|Output];
        $\v -> [$v,$\\|Output];
        _   -> [H|Output]
    end).

-spec add_lhs(binary(), index()) -> true.
add_lhs(Symbol, Index) ->
  case ets:lookup(memo_table_name(), lhs) of
    [] ->
      ets:insert(memo_table_name(), {lhs, [{Symbol,Index}]});
    [{lhs, L}] when is_list(L) ->
      ets:insert(memo_table_name(), {lhs, [{Symbol,Index}|L]})
  end.

-spec add_nt(binary(), index()) -> true | ok.
add_nt(Symbol, Index) ->
  case ets:lookup(memo_table_name(), nts) of
    [] ->
      ets:insert(memo_table_name(), {nts, [{Symbol,Index}]});
    [{nts, L}] when is_list(L) ->
      case proplists:is_defined(Symbol, L) of
        true ->
          ok;
        _ ->
          ets:insert(memo_table_name(), {nts, [{Symbol,Index}|L]})
      end
  end.

-spec verify_rules() -> ok | no_return().
verify_rules() ->
  [{lhs, LHS}] = ets:lookup(memo_table_name(), lhs),
  [{nts, NTs}] = ets:lookup(memo_table_name(), nts),
  [Root|NonRoots] = lists:reverse(LHS),
  lists:foreach(fun({Sym,Idx}) ->
                    case proplists:is_defined(Sym, NTs) of
                      true ->
                        ok;
                      _ ->
                        io:format("neotoma warning: rule '~s' is unused. ~p~n", [Sym,Idx])
                    end
                end, NonRoots),
  lists:foreach(fun({S,I}) ->
                    case proplists:is_defined(S, LHS) of
                      true ->
                        ok;
                      _ ->
                        io:format("neotoma error: nonterminal '~s' has no reduction. (found at ~p) No parser will be generated!~n", [S,I]),
                        exit({neotoma, {no_reduction, list_to_atom(binary_to_list(S))}})
                    end
                end, NTs),
    Root.

-spec used_combinator(atom()) -> true.
used_combinator(C) ->
    case ets:lookup(memo_table_name(), combinators) of
        [] ->
            ets:insert(memo_table_name(), {combinators, ordsets:from_list([C])});
        [{combinators, Cs}] ->
            ets:insert(memo_table_name(), {combinators, ordsets:add_element(C, Cs)})
    end.

-spec used_transform_variables(binary()) -> [ 'Node' | 'Idx' ].
used_transform_variables(Transform) ->
  Code = unicode:characters_to_list(Transform),
  {ok, Tokens, _} = erl_scan:string(Code),
  used_transform_variables(Tokens, []).

used_transform_variables([{var, _, Name}|Tokens], Acc) ->
  used_transform_variables(Tokens, case Name of
                                    'Node' -> [Name | Acc];
                                    'Idx'  -> [Name | Acc];
                                    _      -> Acc
                                  end);
used_transform_variables([_|Tokens], Acc) ->
  used_transform_variables(Tokens, Acc);
used_transform_variables([], Acc) ->
  lists:usort(Acc).

-spec file(file:name()) -> any().
file(Filename) -> case file:read_file(Filename) of {ok,Bin} -> parse(Bin); Err -> Err end.

-spec parse(binary() | list()) -> any().
parse(List) when is_list(List) -> parse(unicode:characters_to_binary(List));
parse(Input) when is_binary(Input) ->
  _ = setup_memo(),
  Result = case 'rules'(Input,{{line,1},{column,1}}) of
             {AST, <<>>, _Index} -> AST;
             Any -> Any
           end,
  release_memo(), Result.

-spec 'rules'(input(), index()) -> parse_result().
'rules'(Input, Index) ->
  p(Input, Index, 'rules', fun(I,D) -> (p_seq([p_optional(fun 'space'/2), fun 'declaration_sequence'/2, p_optional(fun 'space'/2), p_optional(fun 'code_block'/2), p_optional(fun 'space'/2)]))(I,D) end, fun(Node, _Idx) ->
  RootRule = verify_rules(),
  Rules = unicode:characters_to_binary(lists:map(fun(R) -> [R, "\n\n"] end, lists:nth(2, Node))),
  Code = case lists:nth(4, Node) of
             {code, Block} -> Block;
             _ -> []
         end,
  [{rules, Rules},
   {code, Code},
   {root, RootRule},
   {transform, ets:lookup(memo_table_name(),gen_transform)},
   {combinators, ets:lookup_element(memo_table_name(), combinators, 2)}]

 end).

-spec 'declaration_sequence'(input(), index()) -> parse_result().
'declaration_sequence'(Input, Index) ->
  p(Input, Index, 'declaration_sequence', fun(I,D) -> (p_seq([p_label('head', fun 'declaration'/2), p_label('tail', p_zero_or_more(p_seq([fun 'space'/2, fun 'declaration'/2])))]))(I,D) end, fun(Node, _Idx) ->
  FirstRule = proplists:get_value(head, Node),
  OtherRules =  [I || [_,I] <- proplists:get_value(tail, Node, [])],
  [FirstRule|OtherRules]
 end).

-spec 'declaration'(input(), index()) -> parse_result().
'declaration'(Input, Index) ->
  p(Input, Index, 'declaration', fun(I,D) -> (p_seq([fun 'nonterminal'/2, p_zero_or_more(fun 'space'/2), p_string(<<"<-">>), p_zero_or_more(fun 'space'/2), fun 'parsing_expression'/2, p_optional(fun 'space'/2), p_optional(fun 'code_block'/2), p_optional(fun 'space'/2), p_string(<<";">>)]))(I,D) end, fun(Node, _Idx) ->
  [{nonterminal,Symbol}|Tail] = Node,
  add_lhs(Symbol, Index),
  Transform = case lists:nth(6,Tail) of
                  {code, CodeBlock} -> CodeBlock;
                  _ ->
                      ets:insert_new(memo_table_name(),{gen_transform, true}),
                      ["transform('",Symbol,"', Node, Idx)"]
                  end,
  TransformArgs = case used_transform_variables(Transform) of
    []              -> "_Node, _Idx";
    ['Idx']         -> "_Node, Idx";
    ['Node']        -> "Node, _Idx";
    ['Idx', 'Node'] -> "Node, Idx"
  end,
  ["-spec '", Symbol, "'(input(), index()) -> parse_result().\n",
   "'",Symbol,"'","(Input, Index) ->\n  ",
        "p(Input, Index, '",Symbol,"', fun(I,D) -> (",
        lists:nth(4, Tail),
        ")(I,D) end, fun(", TransformArgs, ") ->",Transform," end)."]
 end).

-spec 'parsing_expression'(input(), index()) -> parse_result().
'parsing_expression'(Input, Index) ->
  p(Input, Index, 'parsing_expression', fun(I,D) -> (p_choose([fun 'choice'/2, fun 'sequence'/2, fun 'primary'/2]))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'choice'(input(), index()) -> parse_result().
'choice'(Input, Index) ->
  p(Input, Index, 'choice', fun(I,D) -> (p_seq([p_label('head', fun 'alternative'/2), p_label('tail', p_one_or_more(p_seq([fun 'space'/2, p_string(<<"\/">>), fun 'space'/2, fun 'alternative'/2])))]))(I,D) end, fun(Node, _Idx) ->
  Tail = [lists:last(S) || S <- proplists:get_value(tail, Node)],
  Head = proplists:get_value(head, Node),
  Statements = [[", ", TS] ||  TS <- Tail],
  used_combinator(p_choose),
  ["p_choose([", Head, Statements, "])"]
 end).

-spec 'alternative'(input(), index()) -> parse_result().
'alternative'(Input, Index) ->
  p(Input, Index, 'alternative', fun(I,D) -> (p_choose([fun 'sequence'/2, fun 'labeled_primary'/2]))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'primary'(input(), index()) -> parse_result().
'primary'(Input, Index) ->
  p(Input, Index, 'primary', fun(I,D) -> (p_choose([p_seq([fun 'prefix'/2, fun 'atomic'/2]), p_seq([fun 'atomic'/2, fun 'suffix'/2]), fun 'atomic'/2]))(I,D) end, fun(Node, _Idx) ->
case Node of
  [Atomic, one_or_more] ->
        used_combinator(p_one_or_more),
        used_combinator(p_scan),
        ["p_one_or_more(", Atomic, ")"];
  [Atomic, zero_or_more] ->
        used_combinator(p_zero_or_more),
        used_combinator(p_scan),
        ["p_zero_or_more(", Atomic, ")"];
  [Atomic, optional] ->
        used_combinator(p_optional),
        ["p_optional(", Atomic, ")"];
  [assert, Atomic] ->
        used_combinator(p_assert),
        ["p_assert(", Atomic, ")"];
  [not_, Atomic] ->
        used_combinator(p_not),
        ["p_not(", Atomic, ")"];
  _ -> Node
end
 end).

-spec 'sequence'(input(), index()) -> parse_result().
'sequence'(Input, Index) ->
  p(Input, Index, 'sequence', fun(I,D) -> (p_seq([p_label('head', fun 'labeled_primary'/2), p_label('tail', p_one_or_more(p_seq([fun 'space'/2, fun 'labeled_primary'/2])))]))(I,D) end, fun(Node, _Idx) ->
  Tail = [lists:nth(2, S) || S <- proplists:get_value(tail, Node)],
  Head = proplists:get_value(head, Node),
  Statements = [[", ", TS] || TS <- Tail],
  used_combinator(p_seq),
  ["p_seq([", Head, Statements, "])"]
 end).

-spec 'labeled_primary'(input(), index()) -> parse_result().
'labeled_primary'(Input, Index) ->
  p(Input, Index, 'labeled_primary', fun(I,D) -> (p_seq([p_optional(fun 'label'/2), fun 'primary'/2]))(I,D) end, fun(Node, _Idx) ->
  case hd(Node) of
    [] -> lists:nth(2, Node);
    Label ->
          used_combinator(p_label),
          ["p_label('",  Label, "', ", lists:nth(2, Node), ")"]
  end
 end).

-spec 'label'(input(), index()) -> parse_result().
'label'(Input, Index) ->
  p(Input, Index, 'label', fun(I,D) -> (p_seq([fun 'alpha_char'/2, p_zero_or_more(fun 'alphanumeric_char'/2), p_string(<<":">>)]))(I,D) end, fun(Node, _Idx) ->
  lists:sublist(Node, length(Node)-1)
 end).

-spec 'suffix'(input(), index()) -> parse_result().
'suffix'(Input, Index) ->
  p(Input, Index, 'suffix', fun(I,D) -> (p_choose([fun 'repetition_suffix'/2, fun 'optional_suffix'/2]))(I,D) end, fun(Node, _Idx) ->
  case Node of
    <<"*">> -> zero_or_more;
    <<"+">> -> one_or_more;
    <<"?">> -> optional
  end
 end).

-spec 'optional_suffix'(input(), index()) -> parse_result().
'optional_suffix'(Input, Index) ->
  p(Input, Index, 'optional_suffix', fun(I,D) -> (p_string(<<"?">>))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'repetition_suffix'(input(), index()) -> parse_result().
'repetition_suffix'(Input, Index) ->
  p(Input, Index, 'repetition_suffix', fun(I,D) -> (p_choose([p_string(<<"+">>), p_string(<<"*">>)]))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'prefix'(input(), index()) -> parse_result().
'prefix'(Input, Index) ->
  p(Input, Index, 'prefix', fun(I,D) -> (p_choose([p_string(<<"&">>), p_string(<<"!">>)]))(I,D) end, fun(Node, _Idx) ->
  case Node of
    <<"&">> -> assert;
    <<"!">> -> not_
  end
 end).

-spec 'atomic'(input(), index()) -> parse_result().
'atomic'(Input, Index) ->
  p(Input, Index, 'atomic', fun(I,D) -> (p_choose([fun 'terminal'/2, fun 'nonterminal'/2, fun 'parenthesized_expression'/2]))(I,D) end, fun(Node, _Idx) ->
case Node of
  {nonterminal, Symbol} ->
                [<<"fun '">>, Symbol, <<"'/2">>];
  _ -> Node
end
 end).

-spec 'parenthesized_expression'(input(), index()) -> parse_result().
'parenthesized_expression'(Input, Index) ->
  p(Input, Index, 'parenthesized_expression', fun(I,D) -> (p_seq([p_string(<<"(">>), p_optional(fun 'space'/2), fun 'parsing_expression'/2, p_optional(fun 'space'/2), p_string(<<")">>)]))(I,D) end, fun(Node, _Idx) ->lists:nth(3, Node) end).

-spec 'nonterminal'(input(), index()) -> parse_result().
'nonterminal'(Input, Index) ->
  p(Input, Index, 'nonterminal', fun(I,D) -> (p_seq([fun 'alpha_char'/2, p_zero_or_more(fun 'alphanumeric_char'/2)]))(I,D) end, fun(Node, Idx) ->
  Symbol = unicode:characters_to_binary(Node),
  add_nt(Symbol, Idx),
  {nonterminal, Symbol}
 end).

-spec 'terminal'(input(), index()) -> parse_result().
'terminal'(Input, Index) ->
  p(Input, Index, 'terminal', fun(I,D) -> (p_choose([fun 'regexp_string'/2, fun 'quoted_string'/2, fun 'character_class'/2, fun 'anything_symbol'/2]))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'regexp_string'(input(), index()) -> parse_result().
'regexp_string'(Input, Index) ->
  p(Input, Index, 'regexp_string', fun(I,D) -> (p_seq([p_string(<<"#">>), p_label('string', p_one_or_more(p_seq([p_not(p_string(<<"#">>)), p_choose([p_string(<<"\\#">>), p_anything()])]))), p_string(<<"#">>)]))(I,D) end, fun(Node, _Idx) ->
  used_combinator(p_regexp),
  ["p_regexp(<<\"",
	% Escape \ and " as they are used in erlang string. Other sumbol stay as is.
	%  \ -> \\
	%  " -> \"
   re:replace(proplists:get_value(string, Node), "\"|\\\\", "\\\\&", [{return, binary}, global]),
   "\">>)"]
 end).

-spec 'quoted_string'(input(), index()) -> parse_result().
'quoted_string'(Input, Index) ->
  p(Input, Index, 'quoted_string', fun(I,D) -> (p_choose([fun 'single_quoted_string'/2, fun 'double_quoted_string'/2]))(I,D) end, fun(Node, _Idx) ->
  used_combinator(p_string),
  lists:flatten(["p_string(<<\"",
   escape_string(unicode:characters_to_list(proplists:get_value(string, Node))),
   "\">>)"])
 end).

-spec 'double_quoted_string'(input(), index()) -> parse_result().
'double_quoted_string'(Input, Index) ->
  p(Input, Index, 'double_quoted_string', fun(I,D) -> (p_seq([p_string(<<"\"">>), p_label('string', p_zero_or_more(p_seq([p_not(p_string(<<"\"">>)), p_choose([p_string(<<"\\\\">>), p_string(<<"\\\"">>), p_anything()])]))), p_string(<<"\"">>)]))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'single_quoted_string'(input(), index()) -> parse_result().
'single_quoted_string'(Input, Index) ->
  p(Input, Index, 'single_quoted_string', fun(I,D) -> (p_seq([p_string(<<"\'">>), p_label('string', p_zero_or_more(p_seq([p_not(p_string(<<"\'">>)), p_choose([p_string(<<"\\\\">>), p_string(<<"\\\'">>), p_anything()])]))), p_string(<<"\'">>)]))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'character_class'(input(), index()) -> parse_result().
'character_class'(Input, Index) ->
  p(Input, Index, 'character_class', fun(I,D) -> (p_seq([p_string(<<"[">>), p_label('characters', p_one_or_more(p_seq([p_not(p_string(<<"]">>)), p_choose([p_seq([p_string(<<"\\\\">>), p_anything()]), p_seq([p_not(p_string(<<"\\\\">>)), p_anything()])])]))), p_string(<<"]">>)]))(I,D) end, fun(Node, _Idx) ->
  used_combinator(p_charclass),
  ["p_charclass(<<\"[",
   escape_string(unicode:characters_to_list(proplists:get_value(characters, Node))),
   "]\">>)"]
 end).

-spec 'anything_symbol'(input(), index()) -> parse_result().
'anything_symbol'(Input, Index) ->
  p(Input, Index, 'anything_symbol', fun(I,D) -> (p_string(<<".">>))(I,D) end, fun(_Node, _Idx) -> used_combinator(p_anything), <<"p_anything()">>  end).

-spec 'alpha_char'(input(), index()) -> parse_result().
'alpha_char'(Input, Index) ->
  p(Input, Index, 'alpha_char', fun(I,D) -> (p_charclass(<<"[A-Za-z_]">>))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'alphanumeric_char'(input(), index()) -> parse_result().
'alphanumeric_char'(Input, Index) ->
  p(Input, Index, 'alphanumeric_char', fun(I,D) -> (p_choose([fun 'alpha_char'/2, p_charclass(<<"[0-9]">>)]))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'space'(input(), index()) -> parse_result().
'space'(Input, Index) ->
  p(Input, Index, 'space', fun(I,D) -> (p_one_or_more(p_choose([fun 'white'/2, fun 'comment_to_eol'/2])))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'comment_to_eol'(input(), index()) -> parse_result().
'comment_to_eol'(Input, Index) ->
  p(Input, Index, 'comment_to_eol', fun(I,D) -> (p_seq([p_not(p_string(<<"%{">>)), p_string(<<"%">>), p_zero_or_more(p_seq([p_not(p_string(<<"\n">>)), p_anything()]))]))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'white'(input(), index()) -> parse_result().
'white'(Input, Index) ->
  p(Input, Index, 'white', fun(I,D) -> (p_charclass(<<"[\s\t\n\r]">>))(I,D) end, fun(Node, _Idx) ->Node end).

-spec 'code_block'(input(), index()) -> parse_result().
'code_block'(Input, Index) ->
  p(Input, Index, 'code_block', fun(I,D) -> (p_choose([p_seq([p_string(<<"%{">>), p_label('code', p_one_or_more(p_choose([p_string(<<"\\%">>), p_string(<<"$%">>), p_seq([p_not(p_string(<<"%}">>)), p_anything()])]))), p_string(<<"%}">>)]), p_seq([p_string(<<"`">>), p_label('code', p_one_or_more(p_choose([p_string(<<"\\`">>), p_string(<<"$`">>), p_seq([p_not(p_string(<<"`">>)), p_anything()])]))), p_string(<<"`">>)]), p_string(<<"~">>)]))(I,D) end, fun(Node, _Idx) ->
   case Node of
       <<"~">> -> {code, <<"Node">>};
       _   -> {code, proplists:get_value('code', Node)}
   end
 end).


-file("peg_includes.hrl", 1).
-type index() :: {{line, pos_integer()}, {column, pos_integer()}}.
-type input() :: binary().
-type parse_failure() :: {fail, term()}.
-type parse_success() :: {term(), input(), index()}.
-type parse_result() :: parse_failure() | parse_success().
-type parse_fun() :: fun((input(), index()) -> parse_result()).
-type xform_fun() :: fun((input(), index()) -> term()).

-spec p(input(), index(), atom(), parse_fun(), xform_fun()) -> parse_result().
p(Inp, StartIndex, Name, ParseFun, TransformFun) ->
  case get_memo(StartIndex, Name) of      % See if the current reduction is memoized
    {ok, Memo} -> %Memo;                     % If it is, return the stored result
      Memo;
    _ ->                                        % If not, attempt to parse
      Result = case ParseFun(Inp, StartIndex) of
        {fail,_} = Failure ->                       % If it fails, memoize the failure
          Failure;
        {Match, InpRem, NewIndex} ->               % If it passes, transform and memoize the result.
          Transformed = TransformFun(Match, StartIndex),
          {Transformed, InpRem, NewIndex}
      end,
      memoize(StartIndex, Name, Result),
      Result
  end.

-spec setup_memo() -> ets:tid().
setup_memo() ->
  put({parse_memo_table, ?MODULE}, ets:new(?MODULE, [set])).

-spec release_memo() -> true.
release_memo() ->
  ets:delete(memo_table_name()).

-spec memoize(index(), atom(), parse_result()) -> true.
memoize(Index, Name, Result) ->
  Memo = case ets:lookup(memo_table_name(), Index) of
              [] -> [];
              [{Index, Plist}] -> Plist
         end,
  ets:insert(memo_table_name(), {Index, [{Name, Result}|Memo]}).

-spec get_memo(index(), atom()) -> {ok, term()} | {error, not_found}.
get_memo(Index, Name) ->
  case ets:lookup(memo_table_name(), Index) of
    [] -> {error, not_found};
    [{Index, Plist}] ->
      case proplists:lookup(Name, Plist) of
        {Name, Result}  -> {ok, Result};
        _  -> {error, not_found}
      end
    end.

-spec memo_table_name() -> ets:tid().
memo_table_name() ->
    get({parse_memo_table, ?MODULE}).

-ifdef(p_eof).
-spec p_eof() -> parse_fun().
p_eof() ->
  fun(<<>>, Index) -> {eof, [], Index};
     (_, Index) -> {fail, {expected, eof, Index}} end.
-endif.

-ifdef(p_optional).
-spec p_optional(parse_fun()) -> parse_fun().
p_optional(P) ->
  fun(Input, Index) ->
      case P(Input, Index) of
        {fail,_} -> {[], Input, Index};
        {_, _, _} = Success -> Success
      end
  end.
-endif.

-ifdef(p_not).
-spec p_not(parse_fun()) -> parse_fun().
p_not(P) ->
  fun(Input, Index)->
      case P(Input,Index) of
        {fail,_} ->
          {[], Input, Index};
        {Result, _, _} -> {fail, {expected, {no_match, Result},Index}}
      end
  end.
-endif.

-ifdef(p_assert).
-spec p_assert(parse_fun()) -> parse_fun().
p_assert(P) ->
  fun(Input,Index) ->
      case P(Input,Index) of
        {fail,_} = Failure-> Failure;
        _ -> {[], Input, Index}
      end
  end.
-endif.

-ifdef(p_seq).
-spec p_seq([parse_fun()]) -> parse_fun().
p_seq(P) ->
  fun(Input, Index) ->
      p_all(P, Input, Index, [])
  end.

-spec p_all([parse_fun()], input(), index(), [term()]) -> parse_result().
p_all([], Inp, Index, Accum ) -> {lists:reverse( Accum ), Inp, Index};
p_all([P|Parsers], Inp, Index, Accum) ->
  case P(Inp, Index) of
    {fail, _} = Failure -> Failure;
    {Result, InpRem, NewIndex} -> p_all(Parsers, InpRem, NewIndex, [Result|Accum])
  end.
-endif.

-ifdef(p_choose).
-spec p_choose([parse_fun()]) -> parse_fun().
p_choose(Parsers) ->
  fun(Input, Index) ->
      p_attempt(Parsers, Input, Index, none)
  end.

-spec p_attempt([parse_fun()], input(), index(), none | parse_failure()) -> parse_result().
p_attempt([], _Input, _Index, Failure) -> Failure;
p_attempt([P|Parsers], Input, Index, FirstFailure)->
  case P(Input, Index) of
    {fail, _} = Failure ->
      case FirstFailure of
        none -> p_attempt(Parsers, Input, Index, Failure);
        _ -> p_attempt(Parsers, Input, Index, FirstFailure)
      end;
    Result -> Result
  end.
-endif.

-ifdef(p_zero_or_more).
-spec p_zero_or_more(parse_fun()) -> parse_fun().
p_zero_or_more(P) ->
  fun(Input, Index) ->
      p_scan(P, Input, Index, [])
  end.
-endif.

-ifdef(p_one_or_more).
-spec p_one_or_more(parse_fun()) -> parse_fun().
p_one_or_more(P) ->
  fun(Input, Index)->
      Result = p_scan(P, Input, Index, []),
      case Result of
        {[_|_], _, _} ->
          Result;
        _ ->
          {fail, {expected, Failure, _}} = P(Input,Index),
          {fail, {expected, {at_least_one, Failure}, Index}}
      end
  end.
-endif.

-ifdef(p_label).
-spec p_label(atom(), parse_fun()) -> parse_fun().
p_label(Tag, P) ->
  fun(Input, Index) ->
      case P(Input, Index) of
        {fail,_} = Failure ->
           Failure;
        {Result, InpRem, NewIndex} ->
          {{Tag, Result}, InpRem, NewIndex}
      end
  end.
-endif.

-ifdef(p_scan).
-spec p_scan(parse_fun(), input(), index(), [term()]) -> {[term()], input(), index()}.
p_scan(_, <<>>, Index, Accum) -> {lists:reverse(Accum), <<>>, Index};
p_scan(P, Inp, Index, Accum) ->
  case P(Inp, Index) of
    {fail,_} -> {lists:reverse(Accum), Inp, Index};
    {Result, InpRem, NewIndex} -> p_scan(P, InpRem, NewIndex, [Result | Accum])
  end.
-endif.

-ifdef(p_string).
-spec p_string(binary()) -> parse_fun().
p_string(S) ->
    Length = erlang:byte_size(S),
    fun(Input, Index) ->
      try
          <<S:Length/binary, Rest/binary>> = Input,
          {S, Rest, p_advance_index(S, Index)}
      catch
          error:{badmatch,_} -> {fail, {expected, {string, S}, Index}}
      end
    end.
-endif.

-ifdef(p_anything).
-spec p_anything() -> parse_fun().
p_anything() ->
  fun(<<>>, Index) -> {fail, {expected, any_character, Index}};
     (Input, Index) when is_binary(Input) ->
          <<C/utf8, Rest/binary>> = Input,
          {<<C/utf8>>, Rest, p_advance_index(<<C/utf8>>, Index)}
  end.
-endif.

-ifdef(p_charclass).
-spec p_charclass(string() | binary()) -> parse_fun().
p_charclass(Class) ->
    {ok, RE} = re:compile(Class, [unicode, dotall]),
    fun(Inp, Index) ->
            case re:run(Inp, RE, [anchored]) of
                {match, [{0, Length}|_]} ->
                    {Head, Tail} = erlang:split_binary(Inp, Length),
                    {Head, Tail, p_advance_index(Head, Index)};
                _ -> {fail, {expected, {character_class, binary_to_list(Class)}, Index}}
            end
    end.
-endif.

-ifdef(p_regexp).
-spec p_regexp(binary()) -> parse_fun().
p_regexp(Regexp) ->
    {ok, RE} = re:compile(Regexp, [unicode, dotall, anchored]),
    fun(Inp, Index) ->
        case re:run(Inp, RE) of
            {match, [{0, Length}|_]} ->
                {Head, Tail} = erlang:split_binary(Inp, Length),
                {Head, Tail, p_advance_index(Head, Index)};
            _ -> {fail, {expected, {regexp, binary_to_list(Regexp)}, Index}}
        end
    end.
-endif.

-ifdef(line).
-spec line(index() | term()) -> pos_integer() | undefined.
line({{line,L},_}) -> L;
line(_) -> undefined.
-endif.

-ifdef(column).
-spec column(index() | term()) -> pos_integer() | undefined.
column({_,{column,C}}) -> C;
column(_) -> undefined.
-endif.

-spec p_advance_index(input() | unicode:charlist() | pos_integer(), index()) -> index().
p_advance_index(MatchedInput, Index) when is_list(MatchedInput) orelse is_binary(MatchedInput)-> % strings
  lists:foldl(fun p_advance_index/2, Index, unicode:characters_to_list(MatchedInput));
p_advance_index(MatchedInput, Index) when is_integer(MatchedInput) -> % single characters
  {{line, Line}, {column, Col}} = Index,
  case MatchedInput of
    $\n -> {{line, Line+1}, {column, 1}};
    _ -> {{line, Line}, {column, Col+1}}
  end.