aboutsummaryrefslogblamecommitdiffstats
path: root/lib/diameter/src/compiler/diameter_spec_util.erl
blob: b60886b678f8bcf164612ca588446334ac970972 (plain) (tree)
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068








































                                                                         
                                                  


                                                                     
                                                  






































































































































































































                                                                           





                                                          
 












                                                                              
 
                       
                                      

                                                                  
 
                            
                                        
                                     

                             
                                                             
                
                                                      



                                      
                                                           

                                                                        
                             













































































































































































































































                                                                              


                                                                    



























































































































































































































































































































































































































































































































































                                                                               
%%
%% %CopyrightBegin%
%%
%% Copyright Ericsson AB 2010-2011. All Rights Reserved.
%%
%% The contents of this file are subject to the Erlang Public License,
%% Version 1.1, (the "License"); you may not use this file except in
%% compliance with the License. You should have received a copy of the
%% Erlang Public License along with this software. If not, it can be
%% retrieved online at http://www.erlang.org/.
%%
%% Software distributed under the License is distributed on an "AS IS"
%% basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
%% the License for the specific language governing rights and limitations
%% under the License.
%%
%% %CopyrightEnd%
%%

%%
%% This module turns a .dia (aka spec) file into the orddict that
%% diameter_codegen.erl in turn morphs into .erl and .hrl files for
%% encode and decode of Diameter messages and AVPs.
%%

-module(diameter_spec_util).

-export([parse/2]).

-define(ERROR(T), erlang:error({T, ?MODULE, ?LINE})).
-define(ATOM, list_to_atom).

%% parse/1
%%
%% Output: orddict()

parse(Path, Options) ->
    put({?MODULE, verbose}, lists:member(verbose, Options)),
    {ok, B} = file:read_file(Path),
    Chunks = chunk(B),
    Spec = make_spec(Chunks),
    true = groups_defined(Spec),  %% sanity checks
    true = customs_defined(Spec), %%
    Full = import_enums(import_groups(import_avps(insert_codes(Spec),
                                                  Options))),
    true = enums_defined(Full),   %% sanity checks
    true = v_flags_set(Spec),
    Full.

%% Optional reports when running verbosely.
report(What, Data) ->
    report(get({?MODULE, verbose}), What, Data).

report(true, Tag, Data) ->
    io:format("##~n## ~p ~p~n", [Tag, Data]);
report(false, _, _) ->
    ok.

%% chunk/1

chunk(B) ->
    chunkify(normalize(binary_to_list(B))).

%% normalize/1
%%
%% Replace CR NL by NL, multiple NL by one, tab by space, and strip
%% comments and leading/trailing space from each line. Precludes
%% semicolons being used for any other purpose than comments.

normalize(Str) ->
    nh(Str, []).

nh([], Acc) ->
    lists:reverse(Acc);

%% Trim leading whitespace.
nh(Str, Acc) ->
    nb(trim(Str), Acc).

%% tab -> space
nb([$\t|Rest], Acc) ->
    nb(Rest, [$\s|Acc]);

%% CR NL -> NL
nb([$\r,$\n|Rest], Acc) ->
    nt(Rest, Acc);

%% Gobble multiple newlines before starting over again.
nb([$\n|Rest], Acc) ->
    nt(Rest, Acc);

%% Comment.
nb([$;|Rest], Acc) ->
    nb(lists:dropwhile(fun(C) -> C /= $\n end, Rest), Acc);

%% Just an ordinary character. Boring ...
nb([C|Rest], Acc) ->
    nb(Rest, [C|Acc]);

nb([] = Str, Acc) ->
    nt(Str, Acc).

%% Discard a subsequent newline.
nt(T, [$\n|_] = Acc) ->
    nh(T, trim(Acc));

%% Trim whitespace from the end of the line before continuing.
nt(T, Acc) ->
    nh(T, [$\n|trim(Acc)]).

trim(S) ->
    lists:dropwhile(fun(C) -> lists:member(C, "\s\t") end, S).

%% chunkify/1
%%
%% Split the spec file into pieces delimited by lines starting with
%% @Tag. Returns a list of {Tag, Args, Chunk} where Chunk is the
%% string extending to the next delimiter. Note that leading
%% whitespace has already been stripped.

chunkify(Str) ->
    %% Drop characters to the start of the first chunk.
    {_, Rest} = split_chunk([$\n|Str]),
    chunkify(Rest, []).

chunkify([], Acc) ->
    lists:reverse(Acc);

chunkify(Rest, Acc) ->
    {H,T} = split_chunk(Rest),
    chunkify(T, [split_tag(H) | Acc]).

split_chunk(Str) ->
    split_chunk(Str, []).

split_chunk([] = Rest, Acc) ->
    {lists:reverse(Acc), Rest};
split_chunk([$@|Rest], [$\n|_] = Acc) ->
    {lists:reverse(Acc), Rest};
split_chunk([C|Rest], Acc) ->
    split_chunk(Rest, [C|Acc]).

%% Expect a tag and its arguments on a single line.
split_tag(Str) ->
    {L, Rest} = get_until($\n, Str),
    [{tag, Tag} | Toks] = diameter_spec_scan:parse(L),
    {Tag, Toks, trim(Rest)}.

get_until(EndT, L) ->
    {H, [EndT | T]} = lists:splitwith(fun(C) -> C =/= EndT end, L),
    {H,T}.

%% ------------------------------------------------------------------------
%% make_spec/1
%%
%% Turn chunks into spec.

make_spec(Chunks) ->
    lists:foldl(fun(T,A) -> report(chunk, T), chunk(T,A) end,
                orddict:new(),
                Chunks).

chunk({T, [X], []}, Dict)
  when T == name;
       T == prefix ->
    store(T, atomize(X), Dict);

chunk({id = T, [{number, I}], []}, Dict) ->
    store(T, I, Dict);

chunk({vendor = T, [{number, I}, N], []}, Dict) ->
    store(T, {I, atomize(N)}, Dict);

%% inherits -> [{Mod, [AvpName, ...]}, ...]
chunk({inherits = T, [_,_|_] = Args, []}, Acc) ->
    Mods = [atomize(A) || A <- Args],
    append_list(T, [{M,[]} || M <- Mods], Acc);
chunk({inherits = T, [Mod], Body}, Acc) ->
    append(T, {atomize(Mod), parse_avp_names(Body)}, Acc);

%% avp_types -> [{AvpName, Code, Type, Flags, Encr}, ...]
chunk({avp_types = T, [], Body}, Acc) ->
    store(T, parse_avp_types(Body), Acc);

%% custom_types -> [{Mod, [AvpName, ...]}, ...]
chunk({custom_types = T, [Mod], Body}, Dict) ->
    [_|_] = Avps = parse_avp_names(Body),
    append(T, {atomize(Mod), Avps}, Dict);

%% messages -> [{MsgName, Code, Type, Appl, Avps}, ...]
chunk({messages = T, [], Body}, Acc) ->
    store(T, parse_messages(Body), Acc);

%% grouped -> [{AvpName, Code, Vendor, Avps}, ...]
chunk({grouped = T, [], Body}, Acc) ->
    store(T, parse_groups(Body), Acc);

%% avp_vendor_id -> [{Id, [AvpName, ...]}, ...]
chunk({avp_vendor_id = T, [{number, I}], Body}, Dict) ->
    [_|_] = Names = parse_avp_names(Body),
    append(T, {I, Names}, Dict);

%% enums -> [{AvpName, [{Value, Name}, ...]}, ...]
chunk({enum, [N], Str}, Dict) ->
    append(enums, {atomize(N), parse_enums(Str)}, Dict);

%% result_codes -> [{ResultName, [{Value, Name}, ...]}, ...]
chunk({result_code, [N], Str}, Dict) ->
    append(result_codes, {atomize(N), parse_enums(Str)}, Dict);

%% commands -> [{Name, Abbrev}, ...]
chunk({commands = T, [], Body}, Dict) ->
    store(T, parse_commands(Body), Dict);

chunk(T, _) ->
    ?ERROR({unknown_tag, T}).

store(Key, Value, Dict) ->
    error == orddict:find(Key, Dict) orelse ?ERROR({duplicate, Key}),
    orddict:store(Key, Value, Dict).
append(Key, Value, Dict) ->
    orddict:append(Key, Value, Dict).
append_list(Key, Values, Dict) ->
    orddict:append_list(Key, Values, Dict).

atomize({tag, T}) ->
    T;
atomize({name, T}) ->
    ?ATOM(T).

get_value(Keys, Spec)
  when is_list(Keys) ->
    [get_value(K, Spec) || K <- Keys];
get_value(Key, Spec) ->
    proplists:get_value(Key, Spec, []).

%% ------------------------------------------------------------------------
%% enums_defined/1
%% groups_defined/1
%% customs_defined/1
%%
%% Ensure that every local enum/grouped/custom is defined as an avp
%% with an appropriate type.

enums_defined(Spec) ->
    Avps = get_value(avp_types, Spec),
    Import = get_value(import_enums, Spec),
    lists:all(fun({N,_}) ->
                      true = enum_defined(N, Avps, Import)
              end,
              get_value(enums, Spec)).

enum_defined(Name, Avps, Import) ->
    case lists:keyfind(Name, 1, Avps) of
        {Name, _, 'Enumerated', _, _} ->
            true;
        {Name, _, T, _, _} ->
            ?ERROR({avp_has_wrong_type, Name, 'Enumerated', T});
        false ->
            lists:any(fun({_,Is}) -> lists:keymember(Name, 1, Is) end, Import)
                orelse ?ERROR({avp_not_defined, Name, 'Enumerated'})
    end.
%% Note that an AVP is imported only if referenced by a message or
%% grouped AVP, so the final branch will fail if an enum definition is
%% extended without this being the case.

groups_defined(Spec) ->
    Avps = get_value(avp_types, Spec),
    lists:all(fun({N,_,_,_}) -> true = group_defined(N, Avps) end,
              get_value(grouped, Spec)).

group_defined(Name, Avps) ->
    case lists:keyfind(Name, 1, Avps) of
        {Name, _, 'Grouped', _, _} ->
            true;
        {Name, _, T, _, _} ->
            ?ERROR({avp_has_wrong_type, Name, 'Grouped', T});
        false ->
            ?ERROR({avp_not_defined, Name, 'Grouped'})
    end.

customs_defined(Spec) ->
    Avps = get_value(avp_types, Spec),
    lists:all(fun(A) -> true = custom_defined(A, Avps) end,
              lists:flatmap(fun last/1, get_value(custom_types, Spec))).

custom_defined(Name, Avps) ->
    case lists:keyfind(Name, 1, Avps) of
        {Name, _, T, _, _} when T == 'Grouped';
                                T == 'Enumerated' ->
            ?ERROR({avp_has_invalid_custom_type, Name, T});
        {Name, _, _, _, _} ->
            true;
        false ->
            ?ERROR({avp_not_defined, Name})
    end.

last({_,Xs}) -> Xs.

%% ------------------------------------------------------------------------
%% v_flags_set/1

v_flags_set(Spec) ->
    Avps = get_value(avp_types, Spec)
        ++ lists:flatmap(fun last/1, get_value(import_avps, Spec)),
    Vs = lists:flatmap(fun last/1, get_value(avp_vendor_id, Spec)),

    lists:all(fun(N) -> vset(N, Avps) end, Vs).

vset(Name, Avps) ->
    A = lists:keyfind(Name, 1, Avps),
    false == A andalso ?ERROR({avp_not_defined, Name}),
    {Name, _Code, _Type, Flags, _Encr} = A,
    lists:member('V', Flags) orelse ?ERROR({v_flag_not_set, A}).

%% ------------------------------------------------------------------------
%% insert_codes/1

insert_codes(Spec) ->
    [Msgs, Cmds] = get_value([messages, commands], Spec),

    %% Code -> [{Name, Flags}, ...]
    Dict = lists:foldl(fun({N,C,Fs,_,_}, D) -> dict:append(C,{N,Fs},D) end,
                       dict:new(),
                       Msgs),

    %% list() of {Code, {ReqName, ReqAbbr}, {AnsName, AnsAbbr}}
    %% If the name and abbreviation are the same then the 2-tuples
    %% are replaced by the common atom()-valued name.
    Codes = dict:fold(fun(C,Ns,A) -> [make_code(C, Ns, Cmds) | A] end,
                      [],
                      dict:erase(-1, Dict)),  %% answer-message

    orddict:store(command_codes, Codes, Spec).

make_code(Code, [_,_] = Ns, Cmds) ->
    {Req, Ans} = make_names(Ns, lists:map(fun({_,Fs}) ->
                                                  lists:member('REQ', Fs)
                                          end,
                                          Ns)),
    {Code, abbrev(Req, Cmds), abbrev(Ans, Cmds)};

make_code(Code, Cs, _) ->
    ?ERROR({missing_request_or_answer, Code, Cs}).

%% 3.3.  Diameter Command Naming Conventions
%%
%%    Diameter command names typically includes one or more English words
%%    followed by the verb Request or Answer.  Each English word is
%%    delimited by a hyphen.  A three-letter acronym for both the request
%%    and answer is also normally provided.

make_names([{Rname,_},{Aname,_}], [true, false]) ->
    {Rname, Aname};
make_names([{Aname,_},{Rname,_}], [false, true]) ->
    {Rname, Aname};
make_names([_,_] = Names, _) ->
    ?ERROR({inconsistent_command_flags, Names}).

abbrev(Name, Cmds) ->
    case abbr(Name, get_value(Name, Cmds)) of
        Name ->
            Name;
        Abbr ->
            {Name, Abbr}
    end.

%% No explicit abbreviation: construct.
abbr(Name, []) ->
    ?ATOM(abbr(string:tokens(atom_to_list(Name), "-")));

%% Abbreviation was specified.
abbr(_Name, Abbr) ->
    Abbr.

%% No hyphens: already abbreviated.
abbr([Abbr]) ->
    Abbr;

%% XX-Request/Answer ==> XXR/XXA
abbr([[_,_] = P, T])
  when T == "Request";
       T == "Answer" ->
    P ++ [hd(T)];

%% XXX-...-YYY-Request/Answer ==> X...YR/X...YA
abbr([_,_|_] = L) ->
    lists:map(fun erlang:hd/1, L).

%% ------------------------------------------------------------------------
%% import_avps/2

import_avps(Spec, Options) ->
    Msgs = get_value(messages, Spec),
    Groups = get_value(grouped, Spec),

    %% Messages and groups require AVP's referenced by them.
    NeededAvps
        = ordsets:from_list(lists:flatmap(fun({_,_,_,_,As}) ->
                                                  [avp_name(A) || A <- As]
                                          end,
                                          Msgs)
                            ++ lists:flatmap(fun({_,_,_,As}) ->
                                                     [avp_name(A) || A <- As]
                                             end,
                                             Groups)),
    MissingAvps = missing_avps(NeededAvps, Spec),

    report(needed, NeededAvps),
    report(missing, MissingAvps),

    Import = inherit(get_value(inherits, Spec), Options),

    report(imported, Import),

    ImportedAvps = lists:map(fun({N,_,_,_,_}) -> N end,
                             lists:flatmap(fun last/1, Import)),

    Unknown = MissingAvps -- ImportedAvps,

    [] == Unknown orelse ?ERROR({undefined_avps, Unknown}),

    orddict:store(import_avps, Import, orddict:erase(inherits, Spec)).

%% missing_avps/2
%%
%% Given a list of AVP names and parsed spec, return the list of
%% AVP's that aren't defined in this spec.

missing_avps(NeededNames, Spec) ->
    Avps = get_value(avp_types, Spec),
    Groups = lists:map(fun({N,_,_,As}) ->
                               {N, [avp_name(A) || A <- As]}
                       end,
                       get_value(grouped, Spec)),
    Names = ordsets:from_list(['AVP' | lists:map(fun({N,_,_,_,_}) -> N end,
                                                 Avps)]),
    missing_avps(NeededNames, [], {Names, Groups}).

avp_name({'<',A,'>'}) -> A;
avp_name({A}) -> A;
avp_name([A]) -> A;
avp_name({_, A}) -> avp_name(A).

missing_avps(NeededNames, MissingNames, {Names, _} = T) ->
    missing(ordsets:filter(fun(N) -> lists:member(N, NeededNames) end, Names),
            ordsets:union(NeededNames, MissingNames),
            T).

%% Nothing found locally.
missing([], MissingNames, _) ->
    MissingNames;

%% Or not. Keep looking for for the AVP's needed by the found AVP's of
%% type Grouped.
missing(FoundNames, MissingNames, {_, Groups} = T) ->
    NeededNames = lists:flatmap(fun({N,As}) ->
                                  choose(lists:member(N, FoundNames), As, [])
                          end,
                          Groups),
    missing_avps(ordsets:from_list(NeededNames),
                 ordsets:subtract(MissingNames, FoundNames),
                 T).

%% inherit/2

inherit(Inherits, Options) ->
    Dirs = [D || {include, D} <- Options] ++ ["."],
    lists:foldl(fun(T,A) -> find_avps(T, A, Dirs) end, [], Inherits).

find_avps({Mod, AvpNames}, Acc, Path) ->
    report(inherit_from, Mod),
    Avps = avps_from_beam(find_beam(Mod, Path), Mod),  %% could be empty
    [{Mod, lists:sort(find_avps(AvpNames, Avps))} | Acc].

find_avps([], Avps) ->
    Avps;
find_avps(Names, Avps) ->
    lists:filter(fun({N,_,_,_,_}) -> lists:member(N, Names) end, Avps).

%% find_beam/2

find_beam(Mod, Dirs)
  when is_atom(Mod) ->
    find_beam(atom_to_list(Mod), Dirs);
find_beam(Mod, Dirs) ->
    Beam = Mod ++ code:objfile_extension(),
    case try_path(Dirs, Beam) of
        {value, Path} ->
            Path;
        false ->
            ?ERROR({beam_not_on_path, Beam, Dirs})
    end.

try_path([D|Ds], Fname) ->
    Path = filename:join(D, Fname),
    case file:read_file_info(Path) of
        {ok, _} ->
            {value, Path};
        _ ->
            try_path(Ds, Fname)
    end;
try_path([], _) ->
    false.

%% avps_from_beam/2

avps_from_beam(Path, Mod) ->
    report(beam, Path),
    ok = load_module(code:is_loaded(Mod), Mod, Path),
    orddict:fetch(avp_types, Mod:dict()).

load_module(false, Mod, Path) ->
    R = filename:rootname(Path, code:objfile_extension()),
    {module, Mod} = code:load_abs(R),
    ok;
load_module({file, _}, _, _) ->
    ok.

choose(true, X, _)  -> X;
choose(false, _, X) -> X.

%% ------------------------------------------------------------------------
%% import_groups/1
%% import_enums/1
%%
%% For each inherited module, store the content of imported AVP's of
%% type grouped/enumerated in a new key.

import_groups(Spec) ->
    orddict:store(import_groups, import(grouped, Spec), Spec).

import_enums(Spec) ->
    orddict:store(import_enums, import(enums, Spec), Spec).

import(Key, Spec) ->
    lists:flatmap(fun(T) -> import_key(Key, T) end,
                  get_value(import_avps, Spec)).

import_key(Key, {Mod, Avps}) ->
    Imports = lists:flatmap(fun(T) ->
                                    choose(lists:keymember(element(1,T),
                                                           1,
                                                           Avps),
                                           [T],
                                           [])
                            end,
                            get_value(Key, Mod:dict())),
    if Imports == [] ->
            [];
       true ->
            [{Mod, Imports}]
    end.

%% ------------------------------------------------------------------------
%% parse_enums/1
%%
%% Enums are specified either as the integer value followed by the
%% name or vice-versa. In the former case the name of the enum is
%% taken to be the string up to the end of line, which may contain
%% whitespace. In the latter case the integer may be parenthesized,
%% specified in hex and followed by an inline comment. This is
%% historical and will likely be changed to require a precise input
%% format.
%%
%% Output: list() of {integer(), atom()}

parse_enums(Str) ->
    lists:flatmap(fun(L) -> parse_enum(trim(L)) end, string:tokens(Str, "\n")).

parse_enum([]) ->
    [];

parse_enum(Str) ->
    REs = [{"^(0[xX][0-9A-Fa-f]+|[0-9]+)\s+(.*?)\s*$", 1, 2},
           {"^(.+?)\s+(0[xX][0-9A-Fa-f]+|[0-9]+)(\s+.*)?$", 2, 1},
           {"^(.+?)\s+\\((0[xX][0-9A-Fa-f]+|[0-9]+)\\)(\s+.*)?$", 2, 1}],
    parse_enum(Str, REs).

parse_enum(Str, REs) ->
    try lists:foreach(fun(R) -> enum(Str, R) end, REs) of
        ok ->
            ?ERROR({bad_enum, Str})
    catch
        throw: {enum, T} ->
            [T]
    end.

enum(Str, {Re, I, N}) ->
    case re:run(Str, Re, [{capture, all_but_first, list}]) of
        {match, Vs} ->
            T = list_to_tuple(Vs),
            throw({enum, {to_int(element(I,T)), ?ATOM(element(N,T))}});
        nomatch ->
            ok
    end.

to_int([$0,X|Hex])
  when X == $x;
       X == $X ->
    {ok, [I], _} = io_lib:fread("~#", "16#" ++ Hex),
    I;
to_int(I) ->
    list_to_integer(I).

%% ------------------------------------------------------------------------
%% parse_messages/1
%%
%% Parse according to the ABNF for message specifications in 3.2 of
%% RFC 3588 (shown below). We require all message and AVP names to
%% start with a digit or uppercase character, except for the base
%% answer-message, which is treated as a special case. Allowing names
%% that start with a digit is more than the RFC specifies but the name
%% doesn't affect what's sent over the wire. (Certains 3GPP standards
%% use names starting with a digit. eg 3GPP-Charging-Id in TS32.299.)

%%
%% Sadly, not even the RFC follows this grammar. In particular, except
%% in the example in 3.2, it wraps each command-name in angle brackets
%% ('<' '>') which makes parsing a sequence of specifications require
%% lookahead: after 'optional' avps have been parsed, it's not clear
%% whether a '<' is a 'fixed' or whether it's the start of a
%% subsequent message until we see whether or not '::=' follows the
%% closing '>'. Require the grammar as specified.
%%
%% Output: list of {Name, Code, Flags, ApplId, Avps}
%%
%%         Name   = atom()
%%         Code   = integer()
%%         Flags  = integer()
%%         ApplId = [] | [integer()]
%%         Avps   = see parse_avps/1

parse_messages(Str) ->
    p_cmd(trim(Str), []).

%%   command-def      = command-name "::=" diameter-message
%%
%%   command-name     = diameter-name
%%
%%   diameter-name    = ALPHA *(ALPHA / DIGIT / "-")
%%
%%   diameter-message = header  [ *fixed] [ *required] [ *optional]
%%                      [ *fixed]
%%
%%   header           = "<" Diameter-Header:" command-id
%%                      [r-bit] [p-bit] [e-bit] [application-id]">"
%%
%% The header spec (and example that follows it) is slightly mangled
%% and, given the examples in the RFC should as follows:
%%
%%   header           = "<" "Diameter Header:" command-id
%%                      [r-bit] [p-bit] [e-bit] [application-id]">"
%%
%% This is what's required/parsed below, modulo whitespace. This is
%% also what's specified in the current draft standard at
%% http://ftp.ietf.org/drafts/wg/dime.
%%
%% Note that the grammar specifies the order fixed, required,
%% optional. In practise there seems to be little difference between
%% the latter two since qualifiers can be used to change the
%% semantics. For example 1*[XXX] and *1{YYY} specify 1 or more of the
%% optional avp XXX and 0 or 1 of the required avp YYY, making the
%% iotional avp required and the required avp optional. The current
%% draft addresses this somewhat by requiring that min for a qualifier
%% on an optional avp must be 0 if present. It doesn't say anything
%% about required avps however, so specifying a min of 0 would still
%% be possible. The draft also does away with the trailing *fixed.
%%
%% What will be parsed here will treat required and optional
%% interchangeably. That is. only require that required/optional
%% follow and preceed fixed, not that optional avps must follow
%% required ones. We already have several specs for which this parsing
%% is necessary and there seems to be no harm in accepting it.

p_cmd("", Acc) ->
    lists:reverse(Acc);

p_cmd(Str, Acc) ->
    {Next, Rest} = split_def(Str),
    report(command, Next),
    p_cmd(Rest, [p_cmd(Next) | Acc]).

p_cmd("answer-message" ++ Str) ->
    p_header([{name, 'answer-message'} | diameter_spec_scan:parse(Str)]);

p_cmd(Str) ->
    p_header(diameter_spec_scan:parse(Str)).

%% p_header/1

p_header(['<', {name, _} = N, '>' | Toks]) ->
    p_header([N | Toks]);

p_header([{name, 'answer-message' = N}, '::=',
          '<', {name, "Diameter"}, {name, "Header"}, ':', {tag, code},
          ',', {name, "ERR"}, '[', {name, "PXY"}, ']', '>'
          | Toks]) ->
    {N, -1, ['ERR', 'PXY'], [], parse_avps(Toks)};

p_header([{name, Name}, '::=',
          '<', {name, "Diameter"}, {name, "Header"}, ':', {number, Code}
          | Toks]) ->
    {Flags, Rest} = p_flags(Toks),
    {ApplId, [C|_] = R} = p_appl(Rest),
    '>' == C orelse ?ERROR({invalid_flag, {Name, Code, Flags, ApplId}, R}),
    {?ATOM(Name), Code, Flags, ApplId, parse_avps(tl(R))};

p_header(Toks) ->
    ?ERROR({invalid_header, Toks}).

%%   application-id   = 1*DIGIT
%%
%%   command-id       = 1*DIGIT
%%                      ; The Command Code assigned to the command
%%
%%   r-bit            = ", REQ"
%%                      ; If present, the 'R' bit in the Command
%%                      ; Flags is set, indicating that the message
%%                      ; is a request, as opposed to an answer.
%%
%%   p-bit            = ", PXY"
%%                      ; If present, the 'P' bit in the Command
%%                      ; Flags is set, indicating that the message
%%                      ; is proxiable.
%%
%%   e-bit            = ", ERR"
%%                      ; If present, the 'E' bit in the Command
%%                      ; Flags is set, indicating that the answer
%%                      ; message contains a Result-Code AVP in
%%                      ; the "protocol error" class.

p_flags(Toks) ->
    lists:foldl(fun p_flags/2, {[], Toks}, ["REQ", "PXY", "ERR"]).

p_flags(N, {Acc, [',', {name, N} | Toks]}) ->
    {[?ATOM(N) | Acc], Toks};

p_flags(_, T) ->
    T.

%% The RFC doesn't specify ',' before application-id but this seems a
%% bit inconsistent. Accept a comma if it exists.
p_appl([',', {number, I} | Toks]) ->
    {[I], Toks};
p_appl([{number, I} | Toks]) ->
    {[I], Toks};
p_appl(Toks) ->
    {[], Toks}.

%% parse_avps/1
%%
%% Output: list() of Avp | {Qual, Avp}
%%
%%         Qual = '*' | {Min, '*'} | {'*', Max} | {Min, Max}
%%         Avp  = {'<', Name, '>'} | {Name} | [Name]
%%
%%         Min, Max = integer() >= 0

parse_avps(Toks) ->
    p_avps(Toks, ['<', '|', '<'], []).
%% The list corresponds to the delimiters expected at the front, middle
%% and back of the avp specification, '|' representing '{' and '['.

%%   fixed            = [qual] "<" avp-spec ">"
%%                      ; Defines the fixed position of an AVP
%%
%%   required         = [qual] "{" avp-spec "}"
%%                      ; The AVP MUST be present and can appear
%%                      ; anywhere in the message.
%%
%%   optional         = [qual] "[" avp-name "]"
%%                      ; The avp-name in the 'optional' rule cannot
%%                      ; evaluate to any AVP Name which is included
%%                      ; in a fixed or required rule.  The AVP can
%%                      ; appear anywhere in the message.
%%
%%   qual             = [min] "*" [max]
%%                      ; See ABNF conventions, RFC 2234 Section 6.6.
%%                      ; The absence of any qualifiers depends on whether
%%                      ; it precedes a fixed, required, or optional
%%                      ; rule.  If a fixed or required rule has no
%%                      ; qualifier, then exactly one such AVP MUST
%%                      ; be present.  If an optional rule has no
%%                      ; qualifier, then 0 or 1 such AVP may be
%%                      ; present.
%%                      ;
%%                      ; NOTE:  "[" and "]" have a different meaning
%%                      ; than in ABNF (see the optional rule, above).
%%                      ; These braces cannot be used to express
%%                      ; optional fixed rules (such as an optional
%%                      ; ICV at the end).  To do this, the convention
%%                      ; is '0*1fixed'.
%%
%%   min              = 1*DIGIT
%%                      ; The minimum number of times the element may
%%                      ; be present.  The default value is zero.
%%
%%   max              = 1*DIGIT
%%                      ; The maximum number of times the element may
%%                      ; be present.  The default value is infinity.  A
%%                      ; value of zero implies the AVP MUST NOT be
%%                      ; present.
%%
%%   avp-spec         = diameter-name
%%                      ; The avp-spec has to be an AVP Name, defined
%%                      ; in the base or extended Diameter
%%                      ; specifications.
%%
%%   avp-name         = avp-spec / "AVP"
%%                      ; The string "AVP" stands for *any* arbitrary
%%                      ; AVP Name, which does not conflict with the
%%                      ; required or fixed position AVPs defined in
%%                      ; the command code definition.
%%

p_avps([], _, Acc) ->
    lists:reverse(Acc);

p_avps(Toks, Delim, Acc) ->
    {Qual, Rest} = p_qual(Toks),
    {Avp, R, D} = p_avp(Rest, Delim),
    T = if Qual == false ->
                Avp;
           true ->
                {Qual, Avp}
        end,
    p_avps(R, D, [T | Acc]).

p_qual([{number, Min}, '*', {number, Max} | Toks]) ->
    {{Min, Max}, Toks};
p_qual([{number, Min}, '*' = Max | Toks]) ->
    {{Min, Max}, Toks};
p_qual(['*' = Min, {number, Max} | Toks]) ->
    {{Min, Max}, Toks};
p_qual(['*' = Q | Toks]) ->
    {Q, Toks};
p_qual(Toks) ->
    {false, Toks}.

p_avp([B, {name, Name}, E | Toks], [_|_] = Delim) ->
    {avp(B, ?ATOM(Name), E),
     Toks,
     delim(choose(B == '<', B, '|'), Delim)};
p_avp(Toks, Delim) ->
    ?ERROR({invalid_avp, Toks, Delim}).

avp('<' = B, Name, '>' = E) ->
    {B, Name, E};
avp('{', Name, '}') ->
    {Name};
avp('[', Name, ']') ->
    [Name];
avp(B, Name, E) ->
    ?ERROR({invalid_avp, B, Name, E}).

delim(B, D) ->
    if B == hd(D) -> D; true -> tl(D) end.

%% split_def/1
%%
%% Strip one command definition off head of a string.

split_def(Str) ->
    sdh(Str, []).

%% Look for the "::=" starting off the definition.
sdh("", _) ->
    ?ERROR({missing, '::='});
sdh("::=" ++ Rest, Acc) ->
    sdb(Rest, [$=,$:,$:|Acc]);
sdh([C|Rest], Acc) ->
    sdh(Rest, [C|Acc]).

%% Look for the "::=" starting off the following definition.
sdb("::=" ++ _ = Rest, Acc) ->
    sdt(trim(Acc), Rest);
sdb("" = Rest, Acc) ->
    sd(Acc, Rest);
sdb([C|Rest], Acc) ->
    sdb(Rest, [C|Acc]).

%% Put name characters of the subsequent specification back into Rest.
sdt([C|Acc], Rest)
  when C /= $\n, C /= $\s ->
    sdt(Acc, [C|Rest]);

sdt(Acc, Rest) ->
    sd(Acc, Rest).

sd(Acc, Rest) ->
    {trim(lists:reverse(Acc)), Rest}.
%% Note that Rest is already trimmed of leading space.

%% ------------------------------------------------------------------------
%% parse_groups/1
%%
%% Parse according to the ABNF for message specifications in 4.4 of
%% RFC 3588 (shown below). Again, allow names starting with a digit
%% and also require "AVP Header" without "-" since this is what
%% the RFC uses in all examples.
%%
%% Output: list of {Name, Code, Vendor, Avps}
%%
%%         Name   = atom()
%%         Code   = integer()
%%         Vendor = [] | [integer()]
%%         Avps   = see parse_avps/1

parse_groups(Str) ->
    p_group(trim(Str), []).

%%      grouped-avp-def  = name "::=" avp
%%
%%      name-fmt         = ALPHA *(ALPHA / DIGIT / "-")
%%
%%      name             = name-fmt
%%                         ; The name has to be the name of an AVP,
%%                         ; defined in the base or extended Diameter
%%                         ; specifications.
%%
%%      avp              = header  [ *fixed] [ *required] [ *optional]
%%                         [ *fixed]
%%
%%      header           = "<" "AVP-Header:" avpcode [vendor] ">"
%%
%%      avpcode          = 1*DIGIT
%%                         ; The AVP Code assigned to the Grouped AVP
%%
%%      vendor           = 1*DIGIT
%%                         ; The Vendor-ID assigned to the Grouped AVP.
%%                         ; If absent, the default value of zero is
%%                         ; used.

p_group("", Acc) ->
    lists:reverse(Acc);

p_group(Str, Acc) ->
    {Next, Rest} = split_def(Str),
    report(group, Next),
    p_group(Rest, [p_group(diameter_spec_scan:parse(Next)) | Acc]).

p_group([{name, Name}, '::=', '<', {name, "AVP"}, {name, "Header"},
         ':', {number, Code}
         | Toks]) ->
    {Id, [C|_] = R} = p_vendor(Toks),
    C == '>' orelse ?ERROR({invalid_group_header, R}),
    {?ATOM(Name), Code, Id, parse_avps(tl(R))};

p_group(Toks) ->
    ?ERROR({invalid_group, Toks}).

p_vendor([{number, I} | Toks]) ->
    {[I], Toks};
p_vendor(Toks) ->
    {[], Toks}.

%% ------------------------------------------------------------------------
%% parse_avp_names/1

parse_avp_names(Str) ->
    [p_name(N) || N <- diameter_spec_scan:parse(Str)].

p_name({name, N}) ->
    ?ATOM(N);
p_name(T) ->
    ?ERROR({invalid_avp_name, T}).

%% ------------------------------------------------------------------------
%% parse_avp_types/1
%%
%% Output: list() of {Name, Code, Type, Flags, Encr}

parse_avp_types(Str) ->
    p_avp_types(Str, []).

p_avp_types(Str, Acc) ->
    p_type(diameter_spec_scan:split(Str, 3), Acc).

p_type({[],[]}, Acc) ->
    lists:reverse(Acc);

p_type({[{name, Name}, {number, Code}, {name, Type}], Str}, Acc) ->
    {Flags, Encr, Rest} = try
                              p_avp_flags(trim(Str), [])
                          catch
                              throw: {?MODULE, Reason} ->
                                  ?ERROR({invalid_avp_type, Reason})
                          end,
    p_avp_types(Rest, [{?ATOM(Name), Code, ?ATOM(type(Type)), Flags, Encr}
                       | Acc]);

p_type(T, _) ->
    ?ERROR({invalid_avp_type, T}).

p_avp_flags([C|Str], Acc)
  when C == $M;
       C == $P;
       C == $V ->
    p_avp_flags(Str, [?ATOM([C]) | Acc]);
%% Could support lowercase here if there's a use for distinguishing
%% between Must and Should in the future in deciding whether or not
%% to set a flag.

p_avp_flags([$-|Str], Acc) ->
    %% Require encr on same line as flags if specified.
    {H,T} = lists:splitwith(fun(C) -> C /= $\n end, Str),

    {[{name, [$X|X]} | Toks], Rest} = diameter_spec_scan:split([$X|H], 2),

    "" == X orelse throw({?MODULE, {invalid_avp_flag, Str}}),

    Encr = case Toks of
               [] ->
                   "-";
               [{_, E}] ->
                   (E == "Y" orelse E == "N")
                       orelse throw({?MODULE, {invalid_encr, E}}),
                   E
           end,

    Flags = ordsets:from_list(lists:reverse(Acc)),

    {Flags, ?ATOM(Encr), Rest ++ T};

p_avp_flags(Str, Acc) ->
    p_avp_flags([$-|Str], Acc).

type("DiamIdent")   -> "DiameterIdentity";  %% RFC 3588
type("DiamURI")     -> "DiameterURI";       %% RFC 3588
type("IPFltrRule")  -> "IPFilterRule";      %% RFC 4005
type("QoSFltrRule") -> "QoSFilterRule";     %% RFC 4005
type(N)
  when N == "OctetString";
       N == "Integer32";
       N == "Integer64";
       N == "Unsigned32";
       N == "Unsigned64";
       N == "Float32";
       N == "Float64";
       N == "Grouped";
       N == "Enumerated";
       N == "Address";
       N == "Time";
       N == "UTF8String";
       N == "DiameterIdentity";
       N == "DiameterURI";
       N == "IPFilterRule";
       N == "QoSFilterRule" ->
    N;
type(N) ->
    ?ERROR({invalid_avp_type, N}).

%% ------------------------------------------------------------------------
%% parse_commands/1

parse_commands(Str) ->
    p_abbr(diameter_spec_scan:parse(Str), []).

 p_abbr([{name, Name}, {name, Abbrev} | Toks], Acc)
  when length(Abbrev) < length(Name) ->
    p_abbr(Toks, [{?ATOM(Name), ?ATOM(Abbrev)} | Acc]);

p_abbr([], Acc) ->
    lists:reverse(Acc);

p_abbr(T, _) ->
    ?ERROR({invalid_command, T}).