diff options
22 files changed, 594 insertions, 226 deletions
diff --git a/erts/doc/src/notes.xml b/erts/doc/src/notes.xml index 08f6732036..8ae87ac280 100644 --- a/erts/doc/src/notes.xml +++ b/erts/doc/src/notes.xml @@ -32,6 +32,26 @@ <p>This document describes the changes made to the ERTS application.</p> +<section><title>Erts 8.3.4</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + Add option to make SIGTERM trigger the OS default + behaviour instead of doing a gracefull shutdown. To + activate this bahviour the environment variable + ERL_ZZ_SIGTERM_KILL should be set to "true". This option + only works in OTP 19 as OTP 20 will have a different way + to deal with SIGTERM.</p> + <p> + Own Id: OTP-14418 Aux Id: ERIERL-15 </p> + </item> + </list> + </section> + +</section> + <section><title>Erts 8.3.3</title> <section><title>Fixed Bugs and Malfunctions</title> diff --git a/erts/emulator/beam/erl_process.c b/erts/emulator/beam/erl_process.c index d4385e3987..285fa05d63 100644 --- a/erts/emulator/beam/erl_process.c +++ b/erts/emulator/beam/erl_process.c @@ -13932,7 +13932,11 @@ erts_continue_exit_process(Process *p) erts_set_gc_state(p, 1); state = erts_smp_atomic32_read_acqb(&p->state); - if (state & ERTS_PSFLG_ACTIVE_SYS) { + if (state & ERTS_PSFLG_ACTIVE_SYS +#ifdef ERTS_DIRTY_SCHEDULERS + || p->dirty_sys_tasks +#endif + ) { if (cleanup_sys_tasks(p, state, CONTEXT_REDS) >= CONTEXT_REDS/2) goto yield; } diff --git a/erts/emulator/drivers/unix/unix_efile.c b/erts/emulator/drivers/unix/unix_efile.c index 0acc2432a7..f8341f788a 100644 --- a/erts/emulator/drivers/unix/unix_efile.c +++ b/erts/emulator/drivers/unix/unix_efile.c @@ -969,17 +969,21 @@ efile_sendfile(Efile_error* errInfo, int in_fd, int out_fd, fdrec.sfv_len = SENDFILE_CHUNK_SIZE; else fdrec.sfv_len = *nbytes; + retval = sendfilev(out_fd, &fdrec, 1, &len); - /* Sometimes sendfilev can return -1 and still send data. - When that happens we just pretend that no error happend. */ - if (retval != -1 || errno == EAGAIN || errno == EINTR || - len != 0) { + if (retval == -1 && errno == EINVAL) { + /* On some solaris versions (I've seen it on SunOS 5.10), + using a sfv_len larger then a filesize will result in + a -1 && errno == EINVAL return. We translate this so + a successful send of the data.*/ + retval = len; + } + + if (retval != -1 || errno == EAGAIN || errno == EINTR) { *offset += len; *nbytes -= len; written += len; - if (errno != EAGAIN && errno != EINTR && len != 0) - retval = len; } } while (len == SENDFILE_CHUNK_SIZE); #elif defined(__DARWIN__) diff --git a/erts/emulator/test/binary_SUITE.erl b/erts/emulator/test/binary_SUITE.erl index 355be7a36d..4d17276e5c 100644 --- a/erts/emulator/test/binary_SUITE.erl +++ b/erts/emulator/test/binary_SUITE.erl @@ -1358,17 +1358,19 @@ do_trapping(N, Bif, ArgFun) -> io:format("N=~p: Do ~p ~s gc.\n", [N, Bif, case N rem 2 of 0 -> "with"; 1 -> "without" end]), Pid = spawn(?MODULE,trapping_loop,[Bif, ArgFun, 1000, self()]), receive ok -> ok end, - receive after 100 -> ok end, Ref = make_ref(), case N rem 2 of - 0 -> erlang:garbage_collect(Pid, [{async,Ref}]), - receive after 100 -> ok end; + 0 -> + erlang:garbage_collect(Pid, [{async,Ref}]), + receive after 1 -> ok end; 1 -> void end, - exit(Pid,kill), + exit(Pid, kill), case N rem 2 of - 0 -> receive {garbage_collect, Ref, _} -> ok end; - 1 -> void + 0 -> + receive {garbage_collect, Ref, _} -> ok end; + 1 -> + void end, receive after 1 -> ok end, do_trapping(N-1, Bif, ArgFun). diff --git a/lib/eldap/test/eldap_basic_SUITE.erl b/lib/eldap/test/eldap_basic_SUITE.erl index ac3447cfe6..4bfb0dd291 100644 --- a/lib/eldap/test/eldap_basic_SUITE.erl +++ b/lib/eldap/test/eldap_basic_SUITE.erl @@ -119,7 +119,10 @@ init_per_suite(Config) -> {ldaps_server, LDAPS_server} | Config]. end_per_suite(_Config) -> - ssl:stop(). + try ssl:stop() + catch + _:_ -> ok + end. init_per_group(return_values, Config) -> diff --git a/lib/inets/src/http_client/httpc_handler.erl b/lib/inets/src/http_client/httpc_handler.erl index 89c17a8679..4b2bcc7242 100644 --- a/lib/inets/src/http_client/httpc_handler.erl +++ b/lib/inets/src/http_client/httpc_handler.erl @@ -1175,17 +1175,20 @@ handle_empty_queue(Session, ProfileName, TimeOut, State) -> %% If a pipline | keep_alive session has been idle for some time is not %% closed by the server, the client may want to close it. NewState = activate_queue_timeout(TimeOut, State), - update_session(ProfileName, Session, #session.queue_length, 0), - %% Note mfa will be initialized when a new request - %% arrives. - {noreply, - NewState#state{request = undefined, - mfa = undefined, - status_line = undefined, - headers = undefined, - body = undefined - } - }. + case update_session(ProfileName, Session, #session.queue_length, 0) of + {stop, Reason} -> + {stop, {shutdown, Reason}, State}; + _ -> + %% Note mfa will be initialized when a new request + %% arrives. + {noreply, + NewState#state{request = undefined, + mfa = undefined, + status_line = undefined, + headers = undefined, + body = undefined + }} + end. receive_response(Request, Session, Data, State) -> NewState = init_wait_for_response_state(Request, State), @@ -1677,7 +1680,7 @@ update_session(ProfileName, #session{id = SessionId} = Session, Pos, Value) -> Session2 = erlang:setelement(Pos, Session, Value), insert_session(Session2, ProfileName); error:badarg -> - exit(normal); %% Manager has been shutdown + {stop, normal}; T:E -> %% Unexpected this must be an error! Stacktrace = erlang:get_stacktrace(), @@ -1697,14 +1700,14 @@ update_session(ProfileName, #session{id = SessionId} = Session, Pos, Value) -> Session, (catch httpc_manager:lookup_session(SessionId, ProfileName)), T, E]), - exit({failed_updating_session, - [{profile, ProfileName}, - {session_id, SessionId}, - {pos, Pos}, - {value, Value}, - {etype, T}, - {error, E}, - {stacktrace, Stacktrace}]}) + {stop, {failed_updating_session, + [{profile, ProfileName}, + {session_id, SessionId}, + {pos, Pos}, + {value, Value}, + {etype, T}, + {error, E}, + {stacktrace, Stacktrace}]}} end. diff --git a/lib/kernel/test/os_SUITE.erl b/lib/kernel/test/os_SUITE.erl index 3cbb75a633..1233e362f2 100644 --- a/lib/kernel/test/os_SUITE.erl +++ b/lib/kernel/test/os_SUITE.erl @@ -319,31 +319,38 @@ perf_counter_api(_Config) -> true = is_integer(os:perf_counter()), true = os:perf_counter() > 0, - T1 = os:perf_counter(), + Conv = fun(T1, T2) -> + erlang:convert_time_unit(T2 - T1, perf_counter, nanosecond) + end, + + do_perf_counter_test([], Conv, 120000000, 80000000), + do_perf_counter_test([1000], fun(T1, T2) -> T2 - T1 end, 120, 80). + +do_perf_counter_test(CntArgs, Conv, Upper, Lower) -> + %% We run the test multiple times to try to get a somewhat + %% stable value... what does this test? That the + %% calculate_perf_counter_unit in sys_time.c works somewhat ok. + do_perf_counter_test(CntArgs, Conv, Upper, Lower, 10). + +do_perf_counter_test(CntArgs, _Conv, Upper, Lower, 0) -> + ct:fail("perf_counter_test ~p ~p ~p",[CntArgs, Upper, Lower]); +do_perf_counter_test(CntArgs, Conv, Upper, Lower, Iters) -> + + T1 = apply(os, perf_counter, CntArgs), timer:sleep(100), - T2 = os:perf_counter(), - TsDiff = erlang:convert_time_unit(T2 - T1, perf_counter, nanosecond), - ct:pal("T1: ~p~n" + T2 = apply(os, perf_counter, CntArgs), + TsDiff = Conv(T1, T2), + ct:log("T1: ~p~n" "T2: ~p~n" "TsDiff: ~p~n", [T1,T2,TsDiff]), - %% We allow a 15% diff - true = TsDiff < 115000000, - true = TsDiff > 85000000, - - T1Ms = os:perf_counter(1000), - timer:sleep(100), - T2Ms = os:perf_counter(1000), - MsDiff = T2Ms - T1Ms, - ct:pal("T1Ms: ~p~n" - "T2Ms: ~p~n" - "MsDiff: ~p~n", - [T1Ms,T2Ms,MsDiff]), - - %% We allow a 15% diff - true = MsDiff < 115, - true = MsDiff > 85. + if + TsDiff < Upper, TsDiff > Lower -> + ok; + true -> + do_perf_counter_test(CntArgs, Conv, Upper, Lower, Iters-1) + end. %% Util functions diff --git a/lib/kernel/test/sendfile_SUITE.erl b/lib/kernel/test/sendfile_SUITE.erl index 2673c38494..e839959623 100644 --- a/lib/kernel/test/sendfile_SUITE.erl +++ b/lib/kernel/test/sendfile_SUITE.erl @@ -100,13 +100,13 @@ init_per_testcase(TC,Config) when TC == t_sendfile_recvduring; %% Check if sendfile is supported on this platform case catch sendfile_send(Send) of ok -> - Config; + init_per_testcase(t_sendfile, Config); Error -> ct:log("Error: ~p",[Error]), {skip,"Not supported"} end; init_per_testcase(_Tc,Config) -> - Config. + Config ++ [{sendfile_opts,[{use_threads,false}]}]. t_sendfile_small(Config) when is_list(Config) -> diff --git a/lib/mnesia/src/mnesia_controller.erl b/lib/mnesia/src/mnesia_controller.erl index 17b47c059e..93a7e21f4e 100644 --- a/lib/mnesia/src/mnesia_controller.erl +++ b/lib/mnesia/src/mnesia_controller.erl @@ -1888,7 +1888,7 @@ info_format(Tab, Size, Mem, Media) -> StrT = mnesia_lib:pad_name(atom_to_list(Tab), 15, []), StrS = mnesia_lib:pad_name(integer_to_list(Size), 8, []), StrM = mnesia_lib:pad_name(integer_to_list(Mem), 8, []), - io:format("~s: with ~s records occupying ~s ~s~n", + io:format("~ts: with ~s records occupying ~s ~s~n", [StrT, StrS, StrM, Media]). %% Handle early arrived messages diff --git a/lib/observer/src/observer_lib.erl b/lib/observer/src/observer_lib.erl index 7352af936c..c7ee294719 100644 --- a/lib/observer/src/observer_lib.erl +++ b/lib/observer/src/observer_lib.erl @@ -636,12 +636,12 @@ user_term_multiline(Parent, Title, Default) -> parse_string(Str) -> try - Tokens = case erl_scan:string(Str) of + Tokens = case erl_scan:string(Str, 1, [text]) of {ok, Ts, _} -> Ts; {error, {_SLine, SMod, SError}, _} -> throw(io_lib:format("~s", [SMod:format_error(SError)])) end, - case erl_parse:parse_term(Tokens) of + case lib:extended_parse_term(Tokens) of {error, {_PLine, PMod, PError}} -> throw(io_lib:format("~s", [PMod:format_error(PError)])); Res -> Res diff --git a/lib/ssl/src/dtls_connection.erl b/lib/ssl/src/dtls_connection.erl index 2de947d8b4..f338471829 100644 --- a/lib/ssl/src/dtls_connection.erl +++ b/lib/ssl/src/dtls_connection.erl @@ -115,7 +115,7 @@ send_handshake_flight(#state{socket = Socket, {Encoded, ConnectionStates} = encode_handshake_flight(lists:reverse(Flight), Version, 1400, Epoch, ConnectionStates0), send(Transport, Socket, Encoded), - start_flight(State0#state{connection_states = ConnectionStates}); + {State0#state{connection_states = ConnectionStates}, []}; send_handshake_flight(#state{socket = Socket, transport_cb = Transport, @@ -129,7 +129,7 @@ send_handshake_flight(#state{socket = Socket, {EncChangeCipher, ConnectionStates} = encode_change_cipher(ChangeCipher, Version, Epoch, ConnectionStates1), send(Transport, Socket, [HsBefore, EncChangeCipher]), - start_flight(State0#state{connection_states = ConnectionStates}); + {State0#state{connection_states = ConnectionStates}, []}; send_handshake_flight(#state{socket = Socket, transport_cb = Transport, @@ -145,7 +145,7 @@ send_handshake_flight(#state{socket = Socket, {HsAfter, ConnectionStates} = encode_handshake_flight(lists:reverse(Flight1), Version, 1400, Epoch, ConnectionStates2), send(Transport, Socket, [HsBefore, EncChangeCipher, HsAfter]), - start_flight(State0#state{connection_states = ConnectionStates}); + {State0#state{connection_states = ConnectionStates}, []}; send_handshake_flight(#state{socket = Socket, transport_cb = Transport, @@ -159,7 +159,7 @@ send_handshake_flight(#state{socket = Socket, {HsAfter, ConnectionStates} = encode_handshake_flight(lists:reverse(Flight1), Version, 1400, Epoch, ConnectionStates1), send(Transport, Socket, [EncChangeCipher, HsAfter]), - start_flight(State0#state{connection_states = ConnectionStates}). + {State0#state{connection_states = ConnectionStates}, []}. queue_change_cipher(ChangeCipher, #state{flight_buffer = Flight, connection_states = ConnectionStates0} = State) -> @@ -235,12 +235,14 @@ init([Role, Host, Port, Socket, Options, User, CbInfo]) -> end. callback_mode() -> - state_functions. + [state_functions, state_enter]. %%-------------------------------------------------------------------- %% State functions %%-------------------------------------------------------------------- +init(enter, _, State) -> + {keep_state, State}; init({call, From}, {start, Timeout}, #state{host = Host, port = Port, role = client, ssl_options = SslOpts, @@ -282,6 +284,8 @@ init({call, _} = Type, Event, #state{role = server} = State) -> init(Type, Event, State) -> ssl_connection:init(Type, Event, State, ?MODULE). +error(enter, _, State) -> + {keep_state, State}; error({call, From}, {start, _Timeout}, {Error, State}) -> {stop_and_reply, normal, {reply, From, {error, Error}}, State}; error({call, From}, Msg, State) -> @@ -295,6 +299,11 @@ error(_, _, _) -> #state{}) -> gen_statem:state_function_result(). %%-------------------------------------------------------------------- +hello(enter, _, #state{role = server} = State) -> + {keep_state, State}; +hello(enter, _, #state{role = client} = State0) -> + {State, Actions} = handle_flight_timer(State0), + {keep_state, State, Actions}; hello(internal, #client_hello{cookie = <<>>, client_version = Version} = Hello, #state{role = server, transport_cb = Transport, @@ -374,6 +383,9 @@ hello(state_timeout, Event, State) -> hello(Type, Event, State) -> ssl_connection:hello(Type, Event, State, ?MODULE). +abbreviated(enter, _, State0) -> + {State, Actions} = handle_flight_timer(State0), + {keep_state, State, Actions}; abbreviated(info, Event, State) -> handle_info(Event, abbreviated, State); abbreviated(internal = Type, @@ -391,6 +403,9 @@ abbreviated(state_timeout, Event, State) -> abbreviated(Type, Event, State) -> ssl_connection:abbreviated(Type, Event, State, ?MODULE). +certify(enter, _, State0) -> + {State, Actions} = handle_flight_timer(State0), + {keep_state, State, Actions}; certify(info, Event, State) -> handle_info(Event, certify, State); certify(internal = Type, #server_hello_done{} = Event, State) -> @@ -400,6 +415,9 @@ certify(state_timeout, Event, State) -> certify(Type, Event, State) -> ssl_connection:certify(Type, Event, State, ?MODULE). +cipher(enter, _, State0) -> + {State, Actions} = handle_flight_timer(State0), + {keep_state, State, Actions}; cipher(info, Event, State) -> handle_info(Event, cipher, State); cipher(internal = Type, #change_cipher_spec{type = <<1>>} = Event, @@ -417,6 +435,8 @@ cipher(state_timeout, Event, State) -> cipher(Type, Event, State) -> ssl_connection:cipher(Type, Event, State, ?MODULE). +connection(enter, _, State) -> + {keep_state, State}; connection(info, Event, State) -> handle_info(Event, connection, State); connection(internal, #hello_request{}, #state{host = Host, port = Port, @@ -449,6 +469,9 @@ connection(internal, #client_hello{}, #state{role = server, allow_renegotiate = connection(Type, Event, State) -> ssl_connection:connection(Type, Event, State, ?MODULE). +%%TODO does this make sense for DTLS ? +downgrade(enter, _, State) -> + {keep_state, State}; downgrade(Type, Event, State) -> ssl_connection:downgrade(Type, Event, State, ?MODULE). @@ -750,14 +773,16 @@ next_event(connection = StateName, no_record, {#ssl_tls{epoch = Epoch, type = ?HANDSHAKE, version = _Version}, State1} = _Record when Epoch == CurrentEpoch-1 -> - {State, MoreActions} = send_handshake_flight(State1, CurrentEpoch), - {next_state, StateName, State, Actions ++ MoreActions}; + {State2, MoreActions} = send_handshake_flight(State1, CurrentEpoch), + {NextRecord, State} = next_record(State2), + next_event(StateName, NextRecord, State, Actions ++ MoreActions); %% From FLIGHT perspective CHANGE_CIPHER_SPEC is treated as a handshake {#ssl_tls{epoch = Epoch, type = ?CHANGE_CIPHER_SPEC, version = _Version}, State1} = _Record when Epoch == CurrentEpoch-1 -> - {State, MoreActions} = send_handshake_flight(State1, CurrentEpoch), - {next_state, StateName, State, Actions ++ MoreActions}; + {State2, MoreActions} = send_handshake_flight(State1, CurrentEpoch), + {NextRecord, State} = next_record(State2), + next_event(StateName, NextRecord, State, Actions ++ MoreActions); {#ssl_tls{epoch = _Epoch, version = _Version}, State1} -> %% TODO maybe buffer later epoch @@ -774,14 +799,16 @@ next_event(connection = StateName, Record, #ssl_tls{epoch = Epoch, type = ?HANDSHAKE, version = _Version} when Epoch == CurrentEpoch-1 -> - {State, MoreActions} = send_handshake_flight(State0, CurrentEpoch), - {next_state, StateName, State, Actions ++ MoreActions}; + {State1, MoreActions} = send_handshake_flight(State0, CurrentEpoch), + {NextRecord, State} = next_record(State1), + next_event(StateName, NextRecord, State, Actions ++ MoreActions); %% From FLIGHT perspective CHANGE_CIPHER_SPEC is treated as a handshake #ssl_tls{epoch = Epoch, type = ?CHANGE_CIPHER_SPEC, version = _Version} when Epoch == CurrentEpoch-1 -> - {State, MoreActions} = send_handshake_flight(State0, CurrentEpoch), - {next_state, StateName, State, Actions ++ MoreActions}; + {State1, MoreActions} = send_handshake_flight(State0, CurrentEpoch), + {NextRecord, State} = next_record(State1), + next_event(StateName, NextRecord, State, Actions ++ MoreActions); _ -> next_event(StateName, no_record, State0, Actions) end; @@ -841,13 +868,13 @@ next_flight(Flight) -> change_cipher_spec => undefined, handshakes_after_change_cipher_spec => []}. -start_flight(#state{transport_cb = gen_udp, - flight_state = {retransmit, Timeout}} = State) -> +handle_flight_timer(#state{transport_cb = gen_udp, + flight_state = {retransmit, Timeout}} = State) -> start_retransmision_timer(Timeout, State); -start_flight(#state{transport_cb = gen_udp, +handle_flight_timer(#state{transport_cb = gen_udp, flight_state = connection} = State) -> {State, []}; -start_flight(State) -> +handle_flight_timer(State) -> %% No retransmision needed i.e DTLS over SCTP {State#state{flight_state = reliable}, []}. diff --git a/lib/xmerl/doc/src/notes.xml b/lib/xmerl/doc/src/notes.xml index ef4831e6ec..56856d026e 100644 --- a/lib/xmerl/doc/src/notes.xml +++ b/lib/xmerl/doc/src/notes.xml @@ -32,6 +32,27 @@ <p>This document describes the changes made to the Xmerl application.</p> +<section><title>Xmerl 1.3.14</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> A couple of bugs are fixed in the sax parser + (xmerl_sax_parser). </p> <list> <item>The continuation + function was not called correctly when the XML directive + was fragmented. </item> <item>When the event callback + modules (xmerl_sax_old_dom and xmerl_sax_simple) got an + endDocument event at certain conditions the parser + crashed.</item> <item>Replaced internal ets table with + map to avoid table leakage.</item> </list> + <p> + Own Id: OTP-14430</p> + </item> + </list> + </section> + +</section> + <section><title>Xmerl 1.3.13</title> <section><title>Fixed Bugs and Malfunctions</title> diff --git a/lib/xmerl/src/xmerl_sax_old_dom.erl b/lib/xmerl/src/xmerl_sax_old_dom.erl index fefcf03fce..411121370f 100644 --- a/lib/xmerl/src/xmerl_sax_old_dom.erl +++ b/lib/xmerl/src/xmerl_sax_old_dom.erl @@ -127,9 +127,10 @@ build_dom(endDocument, State#xmerl_sax_old_dom_state{dom=[Decl, Current#xmlElement{ content=lists:reverse(C) }]}; - _ -> - %%?dbg("~p\n", [D]), - ?error("we're not at end the document when endDocument event is encountered.") + _ -> + %% endDocument is also sent by the parser when a fault occur to tell + %% the event receiver that no more input will be sent + State end; %% Element diff --git a/lib/xmerl/src/xmerl_sax_parser.erl b/lib/xmerl/src/xmerl_sax_parser.erl index 1aef6c58c4..e383c4c349 100644 --- a/lib/xmerl/src/xmerl_sax_parser.erl +++ b/lib/xmerl/src/xmerl_sax_parser.erl @@ -64,7 +64,7 @@ %% Description: Parse file containing an XML document. %%---------------------------------------------------------------------- file(Name,Options) -> - case file:open(Name, [raw, read,binary]) of + case file:open(Name, [raw, read_ahead, read,binary]) of {error, Reason} -> {error,{Name, file:format_error(Reason)}}; {ok, FD} -> @@ -120,21 +120,22 @@ stream(Xml, Options, InputType) when is_binary(Xml), is_list(Options) -> normal -> parse end, - case detect_charset(Xml, State) of - {error, Reason} -> {fatal_error, - { - State#xmerl_sax_parser_state.current_location, - State#xmerl_sax_parser_state.entity, - 1 - }, - Reason, - [], - State#xmerl_sax_parser_state.event_state}; - {Xml1, State1} -> - parse_binary(Xml1, - State1#xmerl_sax_parser_state{input_type = InputType}, - ParseFunction) - end + try + {Xml1, State1} = detect_charset(Xml, State), + parse_binary(Xml1, + State1#xmerl_sax_parser_state{input_type = InputType}, + ParseFunction) + catch + throw:{fatal_error, {State2, Reason}} -> + {fatal_error, + { + State2#xmerl_sax_parser_state.current_location, + State2#xmerl_sax_parser_state.entity, + 1 + }, + Reason, [], + State2#xmerl_sax_parser_state.event_state} + end end. %%---------------------------------------------------------------------- @@ -156,8 +157,8 @@ parse_binary(Xml, #xmerl_sax_parser_state{encoding={utf16,big}}=State, F) -> xmerl_sax_parser_utf16be:F(Xml, State); parse_binary(Xml, #xmerl_sax_parser_state{encoding=latin1}=State, F) -> xmerl_sax_parser_latin1:F(Xml, State); -parse_binary(_, #xmerl_sax_parser_state{encoding=Enc}, _) -> - {error, lists:flatten(io_lib:format("Charcter set ~p not supported", [Enc]))}. +parse_binary(_, #xmerl_sax_parser_state{encoding=Enc}, State) -> + ?fatal_error(State, lists:flatten(io_lib:format("Charcter set ~p not supported", [Enc]))). %%---------------------------------------------------------------------- %% Function: initial_state/0 @@ -211,8 +212,7 @@ parse_options([{entity, Entity} |Options], State) -> parse_options([skip_external_dtd |Options], State) -> parse_options(Options, State#xmerl_sax_parser_state{skip_external_dtd = true}); parse_options([O |_], _State) -> - {error, - lists:flatten(io_lib:format("Option: ~p not supported", [O]))}. + {error, lists:flatten(io_lib:format("Option: ~p not supported", [O]))}. check_encoding_option(E) when E==utf8; E=={utf16,little}; E=={utf16,big}; @@ -230,16 +230,10 @@ check_encoding_option(E) -> %% Output: {utf8|utf16le|utf16be|iso8859, Xml, State} %% Description: Detects which character set is used in a binary stream. %%---------------------------------------------------------------------- -detect_charset(<<>>, #xmerl_sax_parser_state{continuation_fun = undefined} = _) -> - {error, "Can't detect character encoding due to no indata"}; -detect_charset(<<>>, #xmerl_sax_parser_state{continuation_fun = CFun, - continuation_state = CState} = State) -> - case CFun(CState) of - {<<>>, _} -> - {error, "Can't detect character encoding due to lack of indata"}; - {NewBytes, NewContState} -> - detect_charset(NewBytes, State#xmerl_sax_parser_state{continuation_state = NewContState}) - end; +detect_charset(<<>>, #xmerl_sax_parser_state{continuation_fun = undefined} = State) -> + ?fatal_error(State, "Can't detect character encoding due to lack of indata"); +detect_charset(<<>>, State) -> + cf(<<>>, State, fun detect_charset/2); detect_charset(Bytes, State) -> case unicode:bom_to_encoding(Bytes) of {latin1, 0} -> @@ -249,25 +243,47 @@ detect_charset(Bytes, State) -> {RealBytes, State#xmerl_sax_parser_state{encoding=Enc}} end. +detect_charset_1(<<16#00>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); +detect_charset_1(<<16#00, 16#3C>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); +detect_charset_1(<<16#00, 16#3C, 16#00>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); detect_charset_1(<<16#00, 16#3C, 16#00, 16#3F, _/binary>> = Xml, State) -> {Xml, State#xmerl_sax_parser_state{encoding={utf16, big}}}; +detect_charset_1(<<16#3C>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); +detect_charset_1(<<16#3C, 16#00>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); +detect_charset_1(<<16#3C, 16#00, 16#3F>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); detect_charset_1(<<16#3C, 16#00, 16#3F, 16#00, _/binary>> = Xml, State) -> {Xml, State#xmerl_sax_parser_state{encoding={utf16, little}}}; -detect_charset_1(<<16#3C, 16#3F, 16#78, 16#6D, 16#6C, Xml2/binary>> = Xml, State) -> - case parse_xml_directive(Xml2) of +detect_charset_1(<<16#3C>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); +detect_charset_1(<<16#3C, 16#3F>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); +detect_charset_1(<<16#3C, 16#3F, 16#78>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); +detect_charset_1(<<16#3C, 16#3F, 16#78, 16#6D>> = Xml, State) -> + cf(Xml, State, fun detect_charset_1/2); +detect_charset_1(<<16#3C, 16#3F, 16#78, 16#6D, 16#6C, Xml2/binary>>, State) -> + {Xml3, State1} = read_until_end_of_xml_directive(Xml2, State), + case parse_xml_directive(Xml3) of {error, Reason} -> - {error, Reason}; + ?fatal_error(State, Reason); AttrList -> case lists:keysearch("encoding", 1, AttrList) of {value, {_, E}} -> case convert_encoding(E) of {error, Reason} -> - {error, Reason}; + ?fatal_error(State, Reason); Enc -> - {Xml, State#xmerl_sax_parser_state{encoding=Enc}} + {<<16#3C, 16#3F, 16#78, 16#6D, 16#6C, Xml3/binary>>, + State1#xmerl_sax_parser_state{encoding=Enc}} end; _ -> - {Xml, State} + {<<16#3C, 16#3F, 16#78, 16#6D, 16#6C, Xml3/binary>>, State1} end end; detect_charset_1(Xml, State) -> @@ -377,7 +393,7 @@ parse_value_1(<<C, Rest/binary>>, Stop, Acc) -> parse_value_1(Rest, Stop, [C |Acc]). %%====================================================================== -%%Default functions +%% Default functions %%====================================================================== %%---------------------------------------------------------------------- %% Function: default_event_cb(Event, LineNo, State) -> Result @@ -393,7 +409,7 @@ default_event_cb(_Event, _LineNo, State) -> %%---------------------------------------------------------------------- %% Function: default_continuation_cb(IoDevice) -> Result %% IoDevice = iodevice() -%% Output: Result = {[char()], State} +%% Output: Result = {binary(), IoDevice} %% Description: Default continuation callback reading blocks. %%---------------------------------------------------------------------- default_continuation_cb(IoDevice) -> @@ -403,3 +419,82 @@ default_continuation_cb(IoDevice) -> {ok, FileBin} -> {FileBin, IoDevice} end. + +%%---------------------------------------------------------------------- +%% Function: read_until_end_of_xml_directive(Rest, State) -> Result +%% Rest = binary() +%% Output: Result = {binary(), State} +%% Description: Reads a utf8 or latin1 until it finds '?>' +%%---------------------------------------------------------------------- +read_until_end_of_xml_directive(Rest, State) -> + case binary:match(Rest, <<"?>">>) of + nomatch -> + case cf(Rest, State) of + {<<>>, _} -> + ?fatal_error(State, "Can't detect character encoding due to lack of indata"); + {NewBytes, NewState} -> + read_until_end_of_xml_directive(NewBytes, NewState) + end; + _ -> + {Rest, State} + end. + + +%%---------------------------------------------------------------------- +%% Function : cf(Rest, State) -> Result +%% Parameters: Rest = binary() +%% State = #xmerl_sax_parser_state{} +%% NextCall = fun() +%% Result : {Rest, State} +%% Description: Function that uses provided fun to read another chunk from +%% input stream and calls the fun in NextCall. +%%---------------------------------------------------------------------- +cf(_Rest, #xmerl_sax_parser_state{continuation_fun = undefined} = State) -> + ?fatal_error(State, "Continuation function undefined"); +cf(Rest, #xmerl_sax_parser_state{continuation_fun = CFun, continuation_state = CState} = State) -> + Result = + try + CFun(CState) + catch + throw:ErrorTerm -> + ?fatal_error(State, ErrorTerm); + exit:Reason -> + ?fatal_error(State, {'EXIT', Reason}) + end, + case Result of + {<<>>, _} -> + ?fatal_error(State, "Can't detect character encoding due to lack of indata"); + {NewBytes, NewContState} -> + {<<Rest/binary, NewBytes/binary>>, + State#xmerl_sax_parser_state{continuation_state = NewContState}} + end. + +%%---------------------------------------------------------------------- +%% Function : cf(Rest, State, NextCall) -> Result +%% Parameters: Rest = binary() +%% State = #xmerl_sax_parser_state{} +%% NextCall = fun() +%% Result : {Rest, State} +%% Description: Function that uses provided fun to read another chunk from +%% input stream and calls the fun in NextCall. +%%---------------------------------------------------------------------- +cf(_Rest, #xmerl_sax_parser_state{continuation_fun = undefined} = State, _) -> + ?fatal_error(State, "Continuation function undefined"); +cf(Rest, #xmerl_sax_parser_state{continuation_fun = CFun, continuation_state = CState} = State, + NextCall) -> + Result = + try + CFun(CState) + catch + throw:ErrorTerm -> + ?fatal_error(State, ErrorTerm); + exit:Reason -> + ?fatal_error(State, {'EXIT', Reason}) + end, + case Result of + {<<>>, _} -> + ?fatal_error(State, "Can't detect character encoding due to lack of indata"); + {NewBytes, NewContState} -> + NextCall(<<Rest/binary, NewBytes/binary>>, + State#xmerl_sax_parser_state{continuation_state = NewContState}) + end. diff --git a/lib/xmerl/src/xmerl_sax_parser_base.erlsrc b/lib/xmerl/src/xmerl_sax_parser_base.erlsrc index f3470b2809..1dca9608cb 100644 --- a/lib/xmerl/src/xmerl_sax_parser_base.erlsrc +++ b/lib/xmerl/src/xmerl_sax_parser_base.erlsrc @@ -64,38 +64,42 @@ %% Description: Parsing XML from input stream. %%---------------------------------------------------------------------- parse(Xml, State) -> - RefTable = ets:new(xmerl_sax_entity_refs, [private]), - - State1 = event_callback(startDocument, State), - - case catch parse_document(Xml, State1#xmerl_sax_parser_state{ref_table=RefTable}) of - {ok, Rest, State2} -> - State3 = event_callback(endDocument, State2), - ets:delete(RefTable), - case check_if_rest_ok(State3#xmerl_sax_parser_state.input_type, Rest) of - true -> - {ok, State3#xmerl_sax_parser_state.event_state, Rest}; - false -> - format_error(fatal_error, State3, "Input found after legal document") - end; - {fatal_error, {State2, Reason}} -> - State3 = event_callback(endDocument, State2), - ets:delete(RefTable), - format_error(fatal_error, State3, Reason); - {event_receiver_error, State2, {Tag, Reason}} -> - State3 = event_callback(endDocument, State2), - ets:delete(RefTable), - format_error(Tag, State3, Reason); - {endDocument, Rest, State2} -> - State3 = event_callback(endDocument, State2), - ets:delete(RefTable), - {ok, State3#xmerl_sax_parser_state.event_state, Rest}; - Other -> - _State2 = event_callback(endDocument, State1), - ets:delete(RefTable), - {fatal_error, Other} + RefTable = maps:new(), + + try + State1 = event_callback(startDocument, State), + Result = parse_document(Xml, State1#xmerl_sax_parser_state{ref_table=RefTable}), + handle_end_document(Result) + catch + throw:Exception -> + handle_end_document(Exception); + _:OtherError -> + handle_end_document({other, OtherError, State}) end. + % case catch parse_document(Xml, State1#xmerl_sax_parser_state{ref_table=RefTable}) of + % {ok, Rest, State2} -> + % State3 = event_callback(endDocument, State2), + % case check_if_rest_ok(State3#xmerl_sax_parser_state.input_type, Rest) of + % true -> + % {ok, State3#xmerl_sax_parser_state.event_state, Rest}; + % false -> + % format_error(fatal_error, State3, "Input found after legal document") + % end; + % {fatal_error, {State2, Reason}} -> + % State3 = event_callback(endDocument, State2), + % format_error(fatal_error, State3, Reason); + % {event_receiver_error, State2, {Tag, Reason}} -> + % State3 = event_callback(endDocument, State2), + % format_error(Tag, State3, Reason); + % {endDocument, Rest, State2} -> + % State3 = event_callback(endDocument, State2), + % {ok, State3#xmerl_sax_parser_state.event_state, Rest}; + % Other -> + % _State2 = event_callback(endDocument, State1), + % {fatal_error, Other} + % end. + %%---------------------------------------------------------------------- %% Function: parse_dtd(Xml, State) -> Result %% Input: Xml = string() | binary() @@ -105,38 +109,120 @@ parse(Xml, State) -> %% Description: Parsing XML DTD from input stream. %%---------------------------------------------------------------------- parse_dtd(Xml, State) -> - RefTable = ets:new(xmerl_sax_entity_refs, [private]), - - State1 = event_callback(startDocument, State), - - case catch parse_external_entity_1(Xml, State1#xmerl_sax_parser_state{ref_table=RefTable}) of - {fatal_error, {State2, Reason}} -> - State3 = event_callback(endDocument, State2), - ets:delete(RefTable), - format_error(fatal_error, State3, Reason); - {event_receiver_error, State2, {Tag, Reason}} -> - State3 = event_callback(endDocument, State2), - format_error(Tag, State3, Reason); - {Rest, State2} when is_record(State2, xmerl_sax_parser_state) -> - State3 = event_callback(endDocument, State2), - ets:delete(RefTable), - {ok, State3#xmerl_sax_parser_state.event_state, Rest}; - {endDocument, Rest, State2} when is_record(State2, xmerl_sax_parser_state) -> - State3 = event_callback(endDocument, State2), - ets:delete(RefTable), - {ok, State3#xmerl_sax_parser_state.event_state, Rest}; - Other -> - _State2 = event_callback(endDocument, State1), - ets:delete(RefTable), - {fatal_error, Other} + RefTable = maps:new(), + + try + State1 = event_callback(startDocument, State), + Result = parse_external_entity_1(Xml, State1#xmerl_sax_parser_state{ref_table=RefTable}), + handle_end_document(Result) + catch + throw:Exception -> + handle_end_document(Exception); + _:OtherError -> + handle_end_document({other, OtherError, State}) end. + % case catch parse_external_entity_1(Xml, State1#xmerl_sax_parser_state{ref_table=RefTable}) of + % {fatal_error, {State2, Reason}} -> + % State3 = event_callback(endDocument, State2), + % format_error(fatal_error, State3, Reason); + % {event_receiver_error, State2, {Tag, Reason}} -> + % State3 = event_callback(endDocument, State2), + % format_error(Tag, State3, Reason); + % {Rest, State2} when is_record(State2, xmerl_sax_parser_state) -> + % State3 = event_callback(endDocument, State2), + % {ok, State3#xmerl_sax_parser_state.event_state, Rest}; + % {endDocument, Rest, State2} when is_record(State2, xmerl_sax_parser_state) -> + % State3 = event_callback(endDocument, State2), + % {ok, State3#xmerl_sax_parser_state.event_state, Rest}; + % Other -> + % _State2 = event_callback(endDocument, State1), + % {fatal_error, Other} + % end. + + %%====================================================================== %% Internal functions %%====================================================================== %%---------------------------------------------------------------------- +%% Function: handle_end_document(ParserResult) -> Result +%% Input: ParseResult = term() +%% Output: Result = {ok, Rest, EventState} | +%% EventState = term() +%% Description: Ends the parsing and formats output +%%---------------------------------------------------------------------- +handle_end_document({ok, Rest, State}) -> + %%ok case from parse + try + State1 = event_callback(endDocument, State), + case check_if_rest_ok(State1#xmerl_sax_parser_state.input_type, Rest) of + true -> + {ok, State1#xmerl_sax_parser_state.event_state, Rest}; + false -> + format_error(fatal_error, State1, "Input found after legal document") + end + catch + throw:{event_receiver_error, State2, {Tag, Reason}} -> + format_error(Tag, State2, Reason); + _:Other -> + {fatal_error, Other} + end; +handle_end_document({endDocument, Rest, State}) -> + %% ok case from parse and parse_dtd + try + State1 = event_callback(endDocument, State), + {ok, State1#xmerl_sax_parser_state.event_state, Rest} + catch + throw:{event_receiver_error, State2, {Tag, Reason}} -> + format_error(Tag, State2, Reason); + _:Other -> + {fatal_error, Other} + end; +handle_end_document({fatal_error, {State, Reason}}) -> + try + State1 = event_callback(endDocument, State), + format_error(fatal_error, State1, Reason) + catch + throw:{event_receiver_error, State2, {Tag, Reason}} -> + format_error(Tag, State2, Reason); + _:Other -> + {fatal_error, Other} + end; +handle_end_document({event_receiver_error, State, {Tag, Reason}}) -> + try + State1 = event_callback(endDocument, State), + format_error(Tag, State1, Reason) + catch + throw:{event_receiver_error, State2, {Tag, Reason}} -> + format_error(Tag, State2, Reason); + _:Other -> + {fatal_error, Other} + end; +handle_end_document({Rest, State}) when is_record(State, xmerl_sax_parser_state) -> + %%ok case from parse_dtd + try + State1 = event_callback(endDocument, State), + {ok, State1#xmerl_sax_parser_state.event_state, Rest} + catch + throw:{event_receiver_error, State2, {Tag, Reason}} -> + format_error(Tag, State2, Reason); + _:Other -> + {fatal_error, Other} + end; +handle_end_document({other, Error, State}) -> + try + _State1 = event_callback(endDocument, State), + {fatal_error, Error} + catch + throw:{event_receiver_error, State2, {Tag, Reason}} -> + format_error(Tag, State2, Reason); + _:Other -> + {fatal_error, Other} + end. + +%%---------------------------------------------------------------------- %% Function: parse_document(Rest, State) -> Result %% Input: Rest = string() | binary() %% State = #xmerl_sax_parser_state{} @@ -439,6 +525,7 @@ check_if_rest_ok(stream, _) -> check_if_rest_ok(_, _) -> false. + %%---------------------------------------------------------------------- %% Function: parse_pi_1(Rest, State) -> Result %% Input: Rest = string() | binary() @@ -883,11 +970,11 @@ send_end_prefix_mapping_event([{Prefix, _Uri} |Ns], State) -> parse_eq(?STRING_EMPTY, State) -> cf(?STRING_EMPTY, State, fun parse_eq/2); parse_eq(?STRING_REST("=", Rest), State) -> - {Rest, State}; + {Rest, State}; parse_eq(?STRING_UNBOUND_REST(C, _) = Bytes, State) when ?is_whitespace(C) -> - {_WS, Rest, State1} = - whitespace(Bytes, State, []), - parse_eq(Rest, State1); + {_WS, Rest, State1} = + whitespace(Bytes, State, []), + parse_eq(Rest, State1); parse_eq(Bytes, State) -> unicode_incomplete_check([Bytes, State, fun parse_eq/2], "expecting = or whitespace"). @@ -905,11 +992,11 @@ parse_eq(Bytes, State) -> parse_att_value(?STRING_EMPTY, State) -> cf(?STRING_EMPTY, State, fun parse_att_value/2); parse_att_value(?STRING_UNBOUND_REST(C, Rest), State) when C == $'; C == $" -> - parse_att_value(Rest, State, C, []); + parse_att_value(Rest, State, C, []); parse_att_value(?STRING_UNBOUND_REST(C, _) = Bytes, State) when ?is_whitespace(C) -> - {_WS, Rest, State1} = - whitespace(Bytes, State, []), - parse_att_value(Rest, State1); + {_WS, Rest, State1} = + whitespace(Bytes, State, []), + parse_att_value(Rest, State1); parse_att_value(Bytes, State) -> unicode_incomplete_check([Bytes, State, fun parse_att_value/2], "\', \" or whitespace expected"). @@ -1360,17 +1447,20 @@ parse_pe_reference_1(Bytes, State, Name) -> %%---------------------------------------------------------------------- -%% Function: insert_reference(Reference, State) -> Result -%% Parameters: Reference = string() +%% Function: insert_reference(Name, Ref, State) -> Result +%% Parameters: Name = string() +%% Ref = {Type, Value} +%% Type = atom() +%% Value = term() %% State = #xmerl_sax_parser_state{} %% Result : %%---------------------------------------------------------------------- -insert_reference({Name, Type, Value}, Table) -> - case ets:lookup(Table, Name) of - [{Name, _, _}] -> - ok; +insert_reference(Name, Value, #xmerl_sax_parser_state{ref_table = Map} = State) -> + case maps:find(Name, Map) of + error -> + State#xmerl_sax_parser_state{ref_table = maps:put(Name, Value, Map)}; _ -> - ets:insert(Table, {Name, Type, Value}) + State end. @@ -1391,8 +1481,8 @@ look_up_reference("apos", _, _) -> look_up_reference("quot", _, _) -> {internal_general, "quot", "\""}; look_up_reference(Name, HaveToExist, State) -> - case ets:lookup(State#xmerl_sax_parser_state.ref_table, Name) of - [{Name, Type, Value}] -> + case maps:find(Name, State#xmerl_sax_parser_state.ref_table) of + {ok, {Type, Value}} -> {Type, Name, Value}; _ -> case HaveToExist of @@ -1474,7 +1564,7 @@ parse_system_litteral(?STRING_EMPTY, State, Stop, Acc) -> parse_system_litteral(?STRING_UNBOUND_REST(Stop, Rest), State, Stop, Acc) -> {lists:reverse(Acc), Rest, State}; parse_system_litteral(?STRING_UNBOUND_REST(C, Rest), State, Stop, Acc) -> - parse_system_litteral(Rest, State, Stop, [C |Acc]); + parse_system_litteral(Rest, State, Stop, [C |Acc]); parse_system_litteral(Bytes, State, Stop, Acc) -> unicode_incomplete_check([Bytes, State, Stop, Acc, fun parse_system_litteral/4], undefined). @@ -1646,9 +1736,11 @@ parse_external_entity(State, _PubId, SysId) -> end_tags = []}, - EventState = handle_external_entity(ExtRef, State1), + {EventState, RefTable} = handle_external_entity(ExtRef, State1), - NewState = event_callback({endEntity, SysId}, SaveState#xmerl_sax_parser_state{event_state=EventState}), + NewState = event_callback({endEntity, SysId}, + SaveState#xmerl_sax_parser_state{event_state=EventState, + ref_table=RefTable}), NewState#xmerl_sax_parser_state{file_type=normal}. @@ -1675,7 +1767,8 @@ handle_external_entity({file, FileToOpen}, State) -> entity=filename:basename(FileToOpen), input_type=file}), ok = file:close(FD), - EntityState#xmerl_sax_parser_state.event_state + {EntityState#xmerl_sax_parser_state.event_state, + EntityState#xmerl_sax_parser_state.ref_table} end; handle_external_entity({http, Url}, State) -> @@ -1695,7 +1788,9 @@ handle_external_entity({http, Url}, State) -> input_type=file}), ok = file:close(FD), ok = file:delete(TmpFile), - EntityState#xmerl_sax_parser_state.event_state + {EntityState#xmerl_sax_parser_state.event_state, + EntityState#xmerl_sax_parser_state.ref_table} + end catch throw:{error, Error} -> @@ -1716,7 +1811,7 @@ handle_external_entity({Tag, _Url}, State) -> parse_external_entity_1(?STRING_EMPTY, #xmerl_sax_parser_state{file_type=Type} = State) -> case catch cf(?STRING_EMPTY, State, fun parse_external_entity_1/2) of {Rest, State1} when is_record(State1, xmerl_sax_parser_state) -> - {Rest, State}; + {Rest, State1}; {fatal_error, {State1, "No more bytes"}} when Type == dtd; Type == entity -> {?STRING_EMPTY, State1}; Other -> @@ -2442,24 +2537,24 @@ parse_entity_def(?STRING_EMPTY, State, Name) -> cf(?STRING_EMPTY, State, Name, fun parse_entity_def/3); parse_entity_def(?STRING_UNBOUND_REST(C, Rest), State, Name) when C == $'; C == $" -> {Value, Rest1, State1} = parse_entity_value(Rest, State, C, []), - insert_reference({Name, internal_general, Value}, State1#xmerl_sax_parser_state.ref_table), - State2 = event_callback({internalEntityDecl, Name, Value}, State1), - {_WS, Rest2, State3} = whitespace(Rest1, State2, []), - parse_def_end(Rest2, State3); + State2 = insert_reference(Name, {internal_general, Value}, State1), + State3 = event_callback({internalEntityDecl, Name, Value}, State2), + {_WS, Rest2, State4} = whitespace(Rest1, State3, []), + parse_def_end(Rest2, State4); parse_entity_def(?STRING_UNBOUND_REST(C, _) = Rest, State, Name) when C == $S; C == $P -> {PubId, SysId, Rest1, State1} = parse_external_id(Rest, State, false), {Ndata, Rest2, State2} = parse_ndata(Rest1, State1), case Ndata of undefined -> - insert_reference({Name, external_general, {PubId, SysId}}, - State2#xmerl_sax_parser_state.ref_table), - State3 = event_callback({externalEntityDecl, Name, PubId, SysId}, State2), - {Rest2, State3}; + State3 = insert_reference(Name, {external_general, {PubId, SysId}}, + State2), + State4 = event_callback({externalEntityDecl, Name, PubId, SysId}, State3), + {Rest2, State4}; _ -> - insert_reference({Name, unparsed, {PubId, SysId, Ndata}}, - State2#xmerl_sax_parser_state.ref_table), - State3 = event_callback({unparsedEntityDecl, Name, PubId, SysId, Ndata}, State2), - {Rest2, State3} + State3 = insert_reference(Name, {unparsed, {PubId, SysId, Ndata}}, + State2), + State4 = event_callback({unparsedEntityDecl, Name, PubId, SysId, Ndata}, State3), + {Rest2, State4} end; parse_entity_def(Bytes, State, Name) -> unicode_incomplete_check([Bytes, State, Name, fun parse_entity_def/3], @@ -2636,19 +2731,19 @@ parse_pe_def(?STRING_EMPTY, State, Name) -> parse_pe_def(?STRING_UNBOUND_REST(C, Rest), State, Name) when C == $'; C == $" -> {Value, Rest1, State1} = parse_entity_value(Rest, State, C, []), Name1 = "%" ++ Name, - insert_reference({Name1, internal_parameter, Value}, - State1#xmerl_sax_parser_state.ref_table), - State2 = event_callback({internalEntityDecl, Name1, Value}, State1), - {_WS, Rest2, State3} = whitespace(Rest1, State2, []), - parse_def_end(Rest2, State3); + State2 = insert_reference(Name1, {internal_parameter, Value}, + State1), + State3 = event_callback({internalEntityDecl, Name1, Value}, State2), + {_WS, Rest2, State4} = whitespace(Rest1, State3, []), + parse_def_end(Rest2, State4); parse_pe_def(?STRING_UNBOUND_REST(C, _) = Bytes, State, Name) when C == $S; C == $P -> {PubId, SysId, Rest1, State1} = parse_external_id(Bytes, State, false), Name1 = "%" ++ Name, - insert_reference({Name1, external_parameter, {PubId, SysId}}, - State1#xmerl_sax_parser_state.ref_table), - State2 = event_callback({externalEntityDecl, Name1, PubId, SysId}, State1), - {_WS, Rest2, State3} = whitespace(Rest1, State2, []), - parse_def_end(Rest2, State3); + State2 = insert_reference(Name1, {external_parameter, {PubId, SysId}}, + State1), + State3 = event_callback({externalEntityDecl, Name1, PubId, SysId}, State2), + {_WS, Rest2, State4} = whitespace(Rest1, State3, []), + parse_def_end(Rest2, State4); parse_pe_def(Bytes, State, Name) -> unicode_incomplete_check([Bytes, State, Name, fun parse_pe_def/3], "\", \', SYSTEM or PUBLIC expected"). diff --git a/lib/xmerl/src/xmerl_sax_simple_dom.erl b/lib/xmerl/src/xmerl_sax_simple_dom.erl index 7eb3afd499..d842bd982b 100644 --- a/lib/xmerl/src/xmerl_sax_simple_dom.erl +++ b/lib/xmerl/src/xmerl_sax_simple_dom.erl @@ -129,8 +129,9 @@ build_dom(endDocument, State#xmerl_sax_simple_dom_state{dom=[Decl, {Tag, Attributes, lists:reverse(Content)}]}; _ -> - ?dbg("~p\n", [D]), - ?error("we're not at end the document when endDocument event is encountered.") + %% endDocument is also sent by the parser when a fault occur to tell + %% the event receiver that no more input will be sent + State end; %% Element diff --git a/lib/xmerl/test/Makefile b/lib/xmerl/test/Makefile index e5c89f84b9..3204f081ba 100644 --- a/lib/xmerl/test/Makefile +++ b/lib/xmerl/test/Makefile @@ -126,5 +126,6 @@ release_tests_spec: opt @tar cfh - xmerl_xsd_MS2002-01-16_SUITE_data | (cd "$(RELSYSDIR)"; tar xf -) @tar cfh - xmerl_xsd_NIST2002-01-16_SUITE_data | (cd "$(RELSYSDIR)"; tar xf -) @tar cfh - xmerl_xsd_Sun2002-01-16_SUITE_data | (cd "$(RELSYSDIR)"; tar xf -) + @tar cfh - xmerl_sax_SUITE_data | (cd "$(RELSYSDIR)"; tar xf -) @tar cfh - xmerl_sax_stream_SUITE_data | (cd "$(RELSYSDIR)"; tar xf -) chmod -R u+w "$(RELSYSDIR)" diff --git a/lib/xmerl/test/xmerl_sax_SUITE.erl b/lib/xmerl/test/xmerl_sax_SUITE.erl index eb9cefe0df..68b9bcc4a2 100644 --- a/lib/xmerl/test/xmerl_sax_SUITE.erl +++ b/lib/xmerl/test/xmerl_sax_SUITE.erl @@ -33,7 +33,6 @@ %%====================================================================== %% External functions %%====================================================================== - %%---------------------------------------------------------------------- %% Initializations %%---------------------------------------------------------------------- @@ -42,12 +41,15 @@ all() -> [{group, bugs}]. groups() -> - [{bugs, [], [ticket_8213, ticket_8214, ticket_11551]}]. + [{bugs, [], [ticket_8213, ticket_8214, ticket_11551, + fragmented_xml_directive, + old_dom_event_fun_endDocument_bug, + event_fun_endDocument_error_test, + event_fun_startDocument_error_test]}]. -%%---------------------------------------------------------------------- +%%====================================================================== %% Tests -%%---------------------------------------------------------------------- - +%%====================================================================== %%---------------------------------------------------------------------- %% Test Case %% ID: ticket_8213 @@ -56,7 +58,6 @@ ticket_8213(_Config) -> {ok,ok,[]} = xmerl_sax_parser:stream("<elem/>", [{event_fun, fun (_E,_,_) -> ok end}]), ok. - %%---------------------------------------------------------------------- %% Test Case %% ID: ticket_8214 @@ -99,3 +100,66 @@ ticket_11551(_Config) -> <a>hej</a>">>, {ok, undefined, <<"\n\n<?xml", _/binary>>} = xmerl_sax_parser:stream(Stream3, []), ok. + +%%---------------------------------------------------------------------- +%% Test Case +%% ID: fragmented_xml_directive +%% Test of fragmented xml directive by reading one byte per continuation ca +fragmented_xml_directive(Config) -> + DataDir = proplists:get_value(data_dir, Config), + Name = filename:join(DataDir, "test_data_1.xml"), + {ok, Fd} = file:open(Name, [raw, read,binary]), + Cf = fun cf_fragmented_xml_directive/1, + {ok, undefined, _} = xmerl_sax_parser:stream(<<>>, + [{continuation_fun, Cf}, + {continuation_state, Fd}]), + ok. + +%%---------------------------------------------------------------------- +%% Test Case +%% ID: old_dom_event_fun_endDocument_bug +%% The old_dom backend previous generateded an uncatched exception +%% instead of the correct fatal_error from the parser. +old_dom_event_fun_endDocument_bug(_Config) -> + %% Stream contains bad characters, + {fatal_error, _, _, _, _} = + xmerl_sax_parser:stream([60,63,120,109,108,32,118,101,114,115,105,111,110,61,39,49,46,48,39,32,101,110,99,111,100,105,110,103,61,39,117,116,102,45,56,39,63,62,60, + 99,111,109,109,97,110,100,62,60,104,101,97,100,101,114,62,60,116,114,97,110,115,97,99,116,105,111,110,73,100,62,49,60,47,116,114,97,110, + 115,97,99,116,105,111,110,73,100,62,60,47,104,101,97,100,101,114,62,60,98,111,100,121,62,95,226,130,172,59,60,60,47,98,111,100,121,62,60, + 47,99,111,109,109,97,110,100,62,60,47,120,49,95,49,62], + [{event_fun,fun xmerl_sax_old_dom:event/3}, + {event_state,xmerl_sax_old_dom:initial_state()}]), + ok. + +%%---------------------------------------------------------------------- +%% Test Case +%% ID: event_fun_endDocument_error_test +event_fun_endDocument_error_test(_Config) -> + Stream = <<"<?xml version=\"1.0\" encoding=\"utf-8\"?><a>hej</a>">>, + Ef = fun(endDocument, _ , _) -> throw({event_error, "endDocument error"}); + (_, _, S) -> S + end, + {event_error, _, _, _, _} = xmerl_sax_parser:stream(Stream, [{event_fun, Ef}]), + ok. + +%%---------------------------------------------------------------------- +%% Test Case +%% ID: event_fun_startDocument_error_test +event_fun_startDocument_error_test(_Config) -> + Stream = <<"<?xml version=\"1.0\" encoding=\"utf-8\"?><a>hej</a>">>, + Ef = fun(startDocument, _ , _) -> throw({event_error, "endDocument error"}); + (_, _, S) -> S + end, + {event_error, _, _, _, _} = xmerl_sax_parser:stream(Stream, [{event_fun, Ef}]), + ok. + +%%====================================================================== +%% Internal functions +%%====================================================================== +cf_fragmented_xml_directive(IoDevice) -> + case file:read(IoDevice, 1) of + eof -> + {<<>>, IoDevice}; + {ok, FileBin} -> + {FileBin, IoDevice} + end. diff --git a/lib/xmerl/test/xmerl_sax_SUITE_data/test_data_1.xml b/lib/xmerl/test/xmerl_sax_SUITE_data/test_data_1.xml new file mode 100644 index 0000000000..efbaee6b81 --- /dev/null +++ b/lib/xmerl/test/xmerl_sax_SUITE_data/test_data_1.xml @@ -0,0 +1,4 @@ +<?xml version="1.0" encoding="utf-8"?> +<a> +Hej +</a> diff --git a/lib/xmerl/test/xmerl_sax_stream_SUITE.erl b/lib/xmerl/test/xmerl_sax_stream_SUITE.erl index a306eb66a2..7315f67374 100644 --- a/lib/xmerl/test/xmerl_sax_stream_SUITE.erl +++ b/lib/xmerl/test/xmerl_sax_stream_SUITE.erl @@ -40,7 +40,8 @@ all() -> [ one_document, two_documents, - one_document_and_junk + one_document_and_junk, + end_of_stream ]. %%---------------------------------------------------------------------- @@ -62,6 +63,9 @@ end_per_testcase(_Func, _Config) -> %%---------------------------------------------------------------------- %% Tests %%---------------------------------------------------------------------- + +%%---------------------------------------------------------------------- +%% Send One doc over stream one_document(Config) -> Port = 11111, @@ -107,6 +111,8 @@ one_document(Config) -> end, ok. +%%---------------------------------------------------------------------- +%% Send Two doc over stream two_documents(Config) -> Port = 11111, @@ -152,6 +158,8 @@ two_documents(Config) -> end, ok. +%%---------------------------------------------------------------------- +%% Send one doc and then junk on stream one_document_and_junk(Config) -> Port = 11111, @@ -196,6 +204,13 @@ one_document_and_junk(Config) -> ct:fail("Timeout") end, ok. + +%%---------------------------------------------------------------------- +%% Test of continuation when end of stream +end_of_stream(Config) -> + Stream = <<"<?xml version=\"1.0\" encoding=\"utf-8\"?><a>hej</a>">>, + {ok, undefined, <<>>} = xmerl_sax_parser:stream(Stream, []), + ok. %%---------------------------------------------------------------------- %% Utility functions diff --git a/lib/xmerl/vsn.mk b/lib/xmerl/vsn.mk index 1515a4e37d..4e741d59a8 100644 --- a/lib/xmerl/vsn.mk +++ b/lib/xmerl/vsn.mk @@ -1 +1 @@ -XMERL_VSN = 1.3.13 +XMERL_VSN = 1.3.14 diff --git a/otp_versions.table b/otp_versions.table index 62eaa62fed..e86ace6e02 100644 --- a/otp_versions.table +++ b/otp_versions.table @@ -1,3 +1,4 @@ +OTP-19.3.5 : erts-8.3.4 xmerl-1.3.14 # asn1-4.0.4 common_test-1.14 compiler-7.0.4 cosEvent-2.2.1 cosEventDomain-1.2.1 cosFileTransfer-1.2.1 cosNotification-1.2.2 cosProperty-1.2.1 cosTime-1.2.2 cosTransactions-1.3.2 crypto-3.7.4 debugger-4.2.1 dialyzer-3.1.1 diameter-1.12.2 edoc-0.8.1 eldap-1.2.2 erl_docgen-0.6.1 erl_interface-3.9.3 et-1.6 eunit-2.3.2 gs-1.6.2 hipe-3.15.4 ic-4.4.2 inets-6.3.9 jinterface-1.7.1 kernel-5.2 megaco-3.18.1 mnesia-4.14.3 observer-2.3.1 odbc-2.12 orber-3.8.2 os_mon-2.4.2 otp_mibs-1.1.1 parsetools-2.1.4 percept-0.9 public_key-1.4 reltool-0.7.3 runtime_tools-1.11.1 sasl-3.0.3 snmp-5.2.5 ssh-4.4.2 ssl-8.1.3 stdlib-3.3 syntax_tools-2.1.1 tools-2.9.1 typer-0.9.12 wx-1.8 : OTP-19.3.4 : inets-6.3.9 ssl-8.1.3 # asn1-4.0.4 common_test-1.14 compiler-7.0.4 cosEvent-2.2.1 cosEventDomain-1.2.1 cosFileTransfer-1.2.1 cosNotification-1.2.2 cosProperty-1.2.1 cosTime-1.2.2 cosTransactions-1.3.2 crypto-3.7.4 debugger-4.2.1 dialyzer-3.1.1 diameter-1.12.2 edoc-0.8.1 eldap-1.2.2 erl_docgen-0.6.1 erl_interface-3.9.3 erts-8.3.3 et-1.6 eunit-2.3.2 gs-1.6.2 hipe-3.15.4 ic-4.4.2 jinterface-1.7.1 kernel-5.2 megaco-3.18.1 mnesia-4.14.3 observer-2.3.1 odbc-2.12 orber-3.8.2 os_mon-2.4.2 otp_mibs-1.1.1 parsetools-2.1.4 percept-0.9 public_key-1.4 reltool-0.7.3 runtime_tools-1.11.1 sasl-3.0.3 snmp-5.2.5 ssh-4.4.2 stdlib-3.3 syntax_tools-2.1.1 tools-2.9.1 typer-0.9.12 wx-1.8 xmerl-1.3.13 : OTP-19.3.3 : dialyzer-3.1.1 erts-8.3.3 inets-6.3.8 # asn1-4.0.4 common_test-1.14 compiler-7.0.4 cosEvent-2.2.1 cosEventDomain-1.2.1 cosFileTransfer-1.2.1 cosNotification-1.2.2 cosProperty-1.2.1 cosTime-1.2.2 cosTransactions-1.3.2 crypto-3.7.4 debugger-4.2.1 diameter-1.12.2 edoc-0.8.1 eldap-1.2.2 erl_docgen-0.6.1 erl_interface-3.9.3 et-1.6 eunit-2.3.2 gs-1.6.2 hipe-3.15.4 ic-4.4.2 jinterface-1.7.1 kernel-5.2 megaco-3.18.1 mnesia-4.14.3 observer-2.3.1 odbc-2.12 orber-3.8.2 os_mon-2.4.2 otp_mibs-1.1.1 parsetools-2.1.4 percept-0.9 public_key-1.4 reltool-0.7.3 runtime_tools-1.11.1 sasl-3.0.3 snmp-5.2.5 ssh-4.4.2 ssl-8.1.2 stdlib-3.3 syntax_tools-2.1.1 tools-2.9.1 typer-0.9.12 wx-1.8 xmerl-1.3.13 : OTP-19.3.2 : erts-8.3.2 # asn1-4.0.4 common_test-1.14 compiler-7.0.4 cosEvent-2.2.1 cosEventDomain-1.2.1 cosFileTransfer-1.2.1 cosNotification-1.2.2 cosProperty-1.2.1 cosTime-1.2.2 cosTransactions-1.3.2 crypto-3.7.4 debugger-4.2.1 dialyzer-3.1 diameter-1.12.2 edoc-0.8.1 eldap-1.2.2 erl_docgen-0.6.1 erl_interface-3.9.3 et-1.6 eunit-2.3.2 gs-1.6.2 hipe-3.15.4 ic-4.4.2 inets-6.3.7 jinterface-1.7.1 kernel-5.2 megaco-3.18.1 mnesia-4.14.3 observer-2.3.1 odbc-2.12 orber-3.8.2 os_mon-2.4.2 otp_mibs-1.1.1 parsetools-2.1.4 percept-0.9 public_key-1.4 reltool-0.7.3 runtime_tools-1.11.1 sasl-3.0.3 snmp-5.2.5 ssh-4.4.2 ssl-8.1.2 stdlib-3.3 syntax_tools-2.1.1 tools-2.9.1 typer-0.9.12 wx-1.8 xmerl-1.3.13 : |