diff options
Diffstat (limited to 'lib/stdlib/src')
-rw-r--r-- | lib/stdlib/src/beam_lib.erl | 3 | ||||
-rw-r--r-- | lib/stdlib/src/binary.erl | 28 | ||||
-rw-r--r-- | lib/stdlib/src/c.erl | 4 | ||||
-rw-r--r-- | lib/stdlib/src/epp.erl | 18 | ||||
-rw-r--r-- | lib/stdlib/src/erl_lint.erl | 8 | ||||
-rw-r--r-- | lib/stdlib/src/ets.erl | 11 | ||||
-rw-r--r-- | lib/stdlib/src/gen.erl | 80 | ||||
-rw-r--r-- | lib/stdlib/src/gen_server.erl | 11 | ||||
-rw-r--r-- | lib/stdlib/src/gen_statem.erl | 13 | ||||
-rw-r--r-- | lib/stdlib/src/io_lib.erl | 58 | ||||
-rw-r--r-- | lib/stdlib/src/io_lib_format.erl | 43 | ||||
-rw-r--r-- | lib/stdlib/src/lists.erl | 19 | ||||
-rw-r--r-- | lib/stdlib/src/string.erl | 8 | ||||
-rw-r--r-- | lib/stdlib/src/uri_string.erl | 65 |
14 files changed, 199 insertions, 170 deletions
diff --git a/lib/stdlib/src/beam_lib.erl b/lib/stdlib/src/beam_lib.erl index 06c15fceda..24349c74e8 100644 --- a/lib/stdlib/src/beam_lib.erl +++ b/lib/stdlib/src/beam_lib.erl @@ -148,7 +148,8 @@ chunks(File, Chunks, Options) -> try read_chunk_data(File, Chunks, Options) catch Error -> Error end. --spec all_chunks(beam()) -> {'ok', 'beam_lib', [{chunkid(), dataB()}]}. +-spec all_chunks(beam()) -> + {'ok', 'beam_lib', [{chunkid(), dataB()}]} | {'error', 'beam_lib', info_rsn()}. all_chunks(File) -> read_all_chunks(File). diff --git a/lib/stdlib/src/binary.erl b/lib/stdlib/src/binary.erl index 6a64133b45..7d0e42489e 100644 --- a/lib/stdlib/src/binary.erl +++ b/lib/stdlib/src/binary.erl @@ -47,23 +47,39 @@ at(_, _) -> -spec bin_to_list(Subject) -> [byte()] when Subject :: binary(). -bin_to_list(_) -> - erlang:nif_error(undef). +bin_to_list(Subject) -> + binary_to_list(Subject). -spec bin_to_list(Subject, PosLen) -> [byte()] when Subject :: binary(), PosLen :: part(). -bin_to_list(_, _) -> - erlang:nif_error(undef). +bin_to_list(Subject, {Pos, Len}) -> + bin_to_list(Subject, Pos, Len); +bin_to_list(_Subject, _BadArg) -> + erlang:error(badarg). -spec bin_to_list(Subject, Pos, Len) -> [byte()] when Subject :: binary(), Pos :: non_neg_integer(), Len :: integer(). -bin_to_list(_, _, _) -> - erlang:nif_error(undef). +bin_to_list(Subject, Pos, Len) when not is_binary(Subject); + not is_integer(Pos); + not is_integer(Len) -> + %% binary_to_list/3 allows bitstrings as long as the slice fits, and we + %% want to badarg when Pos/Len aren't integers instead of raising badarith + %% when adjusting args for binary_to_list/3. + erlang:error(badarg); +bin_to_list(Subject, Pos, 0) when Pos >= 0, Pos =< byte_size(Subject) -> + %% binary_to_list/3 doesn't handle this case. + []; +bin_to_list(_Subject, _Pos, 0) -> + erlang:error(badarg); +bin_to_list(Subject, Pos, Len) when Len < 0 -> + bin_to_list(Subject, Pos + Len, -Len); +bin_to_list(Subject, Pos, Len) when Len > 0 -> + binary_to_list(Subject, Pos + 1, Pos + Len). -spec compile_pattern(Pattern) -> cp() when Pattern :: binary() | [binary()]. diff --git a/lib/stdlib/src/c.erl b/lib/stdlib/src/c.erl index 3597e61c26..13f78841aa 100644 --- a/lib/stdlib/src/c.erl +++ b/lib/stdlib/src/c.erl @@ -564,7 +564,7 @@ display_info(Pid) -> Other end, Reds = fetch(reductions, Info), - LM = length(fetch(messages, Info)), + LM = fetch(message_queue_len, Info), HS = fetch(heap_size, Info), SS = fetch(stack_size, Info), iformat(w(Pid), mfa_string(Call), @@ -886,7 +886,7 @@ portinfo(Id) -> procline(Name, Info, Pid) -> Call = initial_call(Info), Reds = fetch(reductions, Info), - LM = length(fetch(messages, Info)), + LM = fetch(message_queue_len, Info), procformat(io_lib:format("~tw",[Name]), io_lib:format("~w",[Pid]), io_lib:format("~ts",[mfa_string(Call)]), diff --git a/lib/stdlib/src/epp.erl b/lib/stdlib/src/epp.erl index 00e6a10d8a..77cc88eb08 100644 --- a/lib/stdlib/src/epp.erl +++ b/lib/stdlib/src/epp.erl @@ -1,7 +1,7 @@ %% %% %CopyrightBegin% %% -%% Copyright Ericsson AB 1996-2017. All Rights Reserved. +%% Copyright Ericsson AB 1996-2018. All Rights Reserved. %% %% Licensed under the Apache License, Version 2.0 (the "License"); %% you may not use this file except in compliance with the License. @@ -1197,21 +1197,21 @@ skip_else(_Else, From, St, Sis) -> %% macro_expansion(Tokens, Anno) %% Extract the macro parameters and the expansion from a macro definition. -macro_pars([{')',_Lp}, {',',Ld}|Ex], Args) -> - {ok, {lists:reverse(Args), macro_expansion(Ex, Ld)}}; -macro_pars([{var,_,Name}, {')',_Lp}, {',',Ld}|Ex], Args) -> +macro_pars([{')',_Lp}, {',',_Ld}=Comma|Ex], Args) -> + {ok, {lists:reverse(Args), macro_expansion(Ex, Comma)}}; +macro_pars([{var,_,Name}, {')',_Lp}, {',',_Ld}=Comma|Ex], Args) -> false = lists:member(Name, Args), %Prolog is nice - {ok, {lists:reverse([Name|Args]), macro_expansion(Ex, Ld)}}; + {ok, {lists:reverse([Name|Args]), macro_expansion(Ex, Comma)}}; macro_pars([{var,_L,Name}, {',',_}|Ts], Args) -> false = lists:member(Name, Args), macro_pars(Ts, [Name|Args]). -macro_expansion([{')',_Lp},{dot,_Ld}], _Anno0) -> []; -macro_expansion([{dot,_}=Dot], _Anno0) -> +macro_expansion([{')',_Lp},{dot,_Ld}], _T0) -> []; +macro_expansion([{dot,_}=Dot], _T0) -> throw({error,loc(Dot),missing_parenthesis}); -macro_expansion([T|Ts], _Anno0) -> +macro_expansion([T|Ts], _T0) -> [T|macro_expansion(Ts, T)]; -macro_expansion([], Anno0) -> throw({error,loc(Anno0),premature_end}). +macro_expansion([], T0) -> throw({error,loc(T0),premature_end}). %% expand_macros(Tokens, St) %% expand_macro(Tokens, MacroToken, RestTokens) diff --git a/lib/stdlib/src/erl_lint.erl b/lib/stdlib/src/erl_lint.erl index 1930c462e8..9a62d21d34 100644 --- a/lib/stdlib/src/erl_lint.erl +++ b/lib/stdlib/src/erl_lint.erl @@ -3971,6 +3971,8 @@ extract_sequence(3, [$.,_|Fmt], Need) -> extract_sequence(4, Fmt, Need); extract_sequence(3, Fmt, Need) -> extract_sequence(4, Fmt, Need); +extract_sequence(4, [$t, $l | Fmt], Need) -> + extract_sequence(4, [$l, $t | Fmt], Need); extract_sequence(4, [$t, $c | Fmt], Need) -> extract_sequence(5, [$c|Fmt], Need); extract_sequence(4, [$t, $s | Fmt], Need) -> @@ -3987,8 +3989,14 @@ extract_sequence(4, [$t, C | _Fmt], _Need) -> {error,"invalid control ~t" ++ [C]}; extract_sequence(4, [$l, $p | Fmt], Need) -> extract_sequence(5, [$p|Fmt], Need); +extract_sequence(4, [$l, $t, $p | Fmt], Need) -> + extract_sequence(5, [$p|Fmt], Need); extract_sequence(4, [$l, $P | Fmt], Need) -> extract_sequence(5, [$P|Fmt], Need); +extract_sequence(4, [$l, $t, $P | Fmt], Need) -> + extract_sequence(5, [$P|Fmt], Need); +extract_sequence(4, [$l, $t, C | _Fmt], _Need) -> + {error,"invalid control ~lt" ++ [C]}; extract_sequence(4, [$l, C | _Fmt], _Need) -> {error,"invalid control ~l" ++ [C]}; extract_sequence(4, Fmt, Need) -> diff --git a/lib/stdlib/src/ets.erl b/lib/stdlib/src/ets.erl index 039ab45868..6a559f0be5 100644 --- a/lib/stdlib/src/ets.erl +++ b/lib/stdlib/src/ets.erl @@ -73,7 +73,8 @@ select_count/2, select_delete/2, select_replace/2, select_reverse/1, select_reverse/2, select_reverse/3, setopts/2, slot/2, take/2, - update_counter/3, update_counter/4, update_element/3]). + update_counter/3, update_counter/4, update_element/3, + whereis/1]). %% internal exports -export([internal_request_all/0]). @@ -145,6 +146,7 @@ give_away(_, _, _) -> InfoList :: [InfoTuple], InfoTuple :: {compressed, boolean()} | {heir, pid() | none} + | {id, tid()} | {keypos, pos_integer()} | {memory, non_neg_integer()} | {name, atom()} @@ -162,7 +164,7 @@ info(_) -> -spec info(Tab, Item) -> Value | undefined when Tab :: tab(), - Item :: compressed | fixed | heir | keypos | memory + Item :: compressed | fixed | heir | id | keypos | memory | name | named_table | node | owner | protection | safe_fixed | safe_fixed_monotonic_time | size | stats | type | write_concurrency | read_concurrency, @@ -512,6 +514,11 @@ update_counter(_, _, _, _) -> update_element(_, _, _) -> erlang:nif_error(undef). +-spec whereis(TableName) -> tid() | undefined when + TableName :: atom(). +whereis(_) -> + erlang:nif_error(undef). + %%% End of BIFs -opaque comp_match_spec() :: reference(). diff --git a/lib/stdlib/src/gen.erl b/lib/stdlib/src/gen.erl index 0e6f49d99f..2e6223d2bb 100644 --- a/lib/stdlib/src/gen.erl +++ b/lib/stdlib/src/gen.erl @@ -157,52 +157,27 @@ call(Process, Label, Request, Timeout) Fun = fun(Pid) -> do_call(Pid, Label, Request, Timeout) end, do_for_proc(Process, Fun). -do_call(Process, Label, Request, Timeout) -> - try erlang:monitor(process, Process) of - Mref -> - %% If the monitor/2 call failed to set up a connection to a - %% remote node, we don't want the '!' operator to attempt - %% to set up the connection again. (If the monitor/2 call - %% failed due to an expired timeout, '!' too would probably - %% have to wait for the timeout to expire.) Therefore, - %% use erlang:send/3 with the 'noconnect' option so that it - %% will fail immediately if there is no connection to the - %% remote node. - - catch erlang:send(Process, {Label, {self(), Mref}, Request}, - [noconnect]), - receive - {Mref, Reply} -> - erlang:demonitor(Mref, [flush]), - {ok, Reply}; - {'DOWN', Mref, _, _, noconnection} -> - Node = get_node(Process), - exit({nodedown, Node}); - {'DOWN', Mref, _, _, Reason} -> - exit(Reason) - after Timeout -> - erlang:demonitor(Mref, [flush]), - exit(timeout) - end - catch - error:_ -> - %% Node (C/Java?) is not supporting the monitor. - %% The other possible case -- this node is not distributed - %% -- should have been handled earlier. - %% Do the best possible with monitor_node/2. - %% This code may hang indefinitely if the Process - %% does not exist. It is only used for featureweak remote nodes. - Node = get_node(Process), - monitor_node(Node, true), - receive - {nodedown, Node} -> - monitor_node(Node, false), - exit({nodedown, Node}) - after 0 -> - Tag = make_ref(), - Process ! {Label, {self(), Tag}, Request}, - wait_resp(Node, Tag, Timeout) - end +do_call(Process, Label, Request, Timeout) when is_atom(Process) =:= false -> + Mref = erlang:monitor(process, Process), + + %% OTP-21: + %% Auto-connect is asynchronous. But we still use 'noconnect' to make sure + %% we send on the monitored connection, and not trigger a new auto-connect. + %% + erlang:send(Process, {Label, {self(), Mref}, Request}, [noconnect]), + + receive + {Mref, Reply} -> + erlang:demonitor(Mref, [flush]), + {ok, Reply}; + {'DOWN', Mref, _, _, noconnection} -> + Node = get_node(Process), + exit({nodedown, Node}); + {'DOWN', Mref, _, _, Reason} -> + exit(Reason) + after Timeout -> + erlang:demonitor(Mref, [flush]), + exit(timeout) end. get_node(Process) -> @@ -217,19 +192,6 @@ get_node(Process) -> node(Process) end. -wait_resp(Node, Tag, Timeout) -> - receive - {Tag, Reply} -> - monitor_node(Node, false), - {ok, Reply}; - {nodedown, Node} -> - monitor_node(Node, false), - exit({nodedown, Node}) - after Timeout -> - monitor_node(Node, false), - exit(timeout) - end. - %% %% Send a reply to the client. %% diff --git a/lib/stdlib/src/gen_server.erl b/lib/stdlib/src/gen_server.erl index 77a46419f6..f29314d0a2 100644 --- a/lib/stdlib/src/gen_server.erl +++ b/lib/stdlib/src/gen_server.erl @@ -437,12 +437,11 @@ decode_msg(Msg, Parent, Name, State, Mod, Time, HibernateAfterTimeout, Debug, Hi %%% Send/receive functions %%% --------------------------------------------------- do_send(Dest, Msg) -> - case catch erlang:send(Dest, Msg, [noconnect]) of - noconnect -> - spawn(erlang, send, [Dest,Msg]); - Other -> - Other - end. + try erlang:send(Dest, Msg) + catch + error:_ -> ok + end, + ok. do_multi_call(Nodes, Name, Req, infinity) -> Tag = make_ref(), diff --git a/lib/stdlib/src/gen_statem.erl b/lib/stdlib/src/gen_statem.erl index f95b2ea9cd..9dc360a289 100644 --- a/lib/stdlib/src/gen_statem.erl +++ b/lib/stdlib/src/gen_statem.erl @@ -641,16 +641,11 @@ replies([]) -> %% Might actually not send the message in case of caught exception send(Proc, Msg) -> - try erlang:send(Proc, Msg, [noconnect]) of - noconnect -> - _ = spawn(erlang, send, [Proc,Msg]), - ok; - ok -> - ok + try erlang:send(Proc, Msg) catch - _:_ -> - ok - end. + error:_ -> ok + end, + ok. %% Here the init_it/6 and enter_loop/5,6,7 functions converge enter(Module, Opts, State, Data, Server, Actions, Parent) -> diff --git a/lib/stdlib/src/io_lib.erl b/lib/stdlib/src/io_lib.erl index cacd9f2524..e37c13093b 100644 --- a/lib/stdlib/src/io_lib.erl +++ b/lib/stdlib/src/io_lib.erl @@ -931,7 +931,7 @@ limit_term(Term, Depth) -> limit(_, 0) -> '...'; limit([H|T]=L, D) -> if - D =:= 1 -> '...'; + D =:= 1 -> ['...']; true -> case printable_list(L) of true -> L; @@ -944,7 +944,7 @@ limit(Term, D) when is_map(Term) -> limit({}=T, _D) -> T; limit(T, D) when is_tuple(T) -> if - D =:= 1 -> '...'; + D =:= 1 -> {'...'}; true -> list_to_tuple([limit(element(1, T), D-1)| limit_tail(tl(tuple_to_list(T)), D-1)]) @@ -961,32 +961,29 @@ limit_tail(Other, D) -> %% Cannot limit maps properly since there is no guarantee that %% maps:from_list() creates a map with the same internal ordering of -%% the selected associations as in Map. +%% the selected associations as in Map. Instead of subtracting one +%% from the depth as the map associations are traversed (as is done +%% for tuples and lists), the same depth is applied to each and every +%% (returned) association. limit_map(Map, D) -> - limit_map(maps:iterator(Map), D, []). + %% Keep one extra association to make sure the final ',...' is included. + limit_map_body(maps:iterator(Map), D + 1, D, []). -limit_map(_I, 0, Acc) -> +limit_map_body(_I, 0, _D0, Acc) -> maps:from_list(Acc); -limit_map(I, D, Acc) -> +limit_map_body(I, D, D0, Acc) -> case maps:next(I) of {K, V, NextI} -> - limit_map(NextI, D-1, [{K,V} | Acc]); + limit_map_body(NextI, D-1, D0, [limit_map_assoc(K, V, D0) | Acc]); none -> maps:from_list(Acc) end. -%% maps:from_list(limit_map_body(erts_internal:maps_to_list(Map, D), D)). +limit_map_assoc(K, V, D) -> + %% Keep keys as are to avoid creating duplicated keys. + {K, limit(V, D - 1)}. -%% limit_map_body(_, 0) -> [{'...', '...'}]; -%% limit_map_body([], _) -> []; -%% limit_map_body([{K,V}], D) -> [limit_map_assoc(K, V, D)]; -%% limit_map_body([{K,V}|KVs], D) -> -%% [limit_map_assoc(K, V, D) | limit_map_body(KVs, D-1)]. - -%% limit_map_assoc(K, V, D) -> -%% {limit(K, D-1), limit(V, D-1)}. - -limit_bitstring(B, _D) -> B. %% Keeps all printable binaries. +limit_bitstring(B, _D) -> B. % Keeps all printable binaries. test_limit(_, 0) -> throw(limit); test_limit([H|T]=L, D) when is_integer(D) -> @@ -1022,18 +1019,21 @@ test_limit_tuple(T, I, Sz, D) -> test_limit(element(I, T), D-1), test_limit_tuple(T, I+1, Sz, D-1). -test_limit_map(_Map, _D) -> ok. -%% test_limit_map_body(erts_internal:maps_to_list(Map, D), D). +test_limit_map(Map, D) -> + test_limit_map_body(maps:iterator(Map), D). -%% test_limit_map_body(_, 0) -> throw(limit); -%% test_limit_map_body([], _) -> ok; -%% test_limit_map_body([{K,V}], D) -> test_limit_map_assoc(K, V, D); -%% test_limit_map_body([{K,V}|KVs], D) -> -%% test_limit_map_assoc(K, V, D), -%% test_limit_map_body(KVs, D-1). +test_limit_map_body(_I, 0) -> throw(limit); % cannot happen +test_limit_map_body(I, D) -> + case maps:next(I) of + {K, V, NextI} -> + test_limit_map_assoc(K, V, D), + test_limit_map_body(NextI, D-1); + none -> + ok + end. -%% test_limit_map_assoc(K, V, D) -> -%% test_limit(K, D-1), -%% test_limit(V, D-1). +test_limit_map_assoc(K, V, D) -> + test_limit(K, D - 1), + test_limit(V, D - 1). test_limit_bitstring(_, _) -> ok. diff --git a/lib/stdlib/src/io_lib_format.erl b/lib/stdlib/src/io_lib_format.erl index e345810ca0..64edbf1824 100644 --- a/lib/stdlib/src/io_lib_format.erl +++ b/lib/stdlib/src/io_lib_format.erl @@ -95,7 +95,7 @@ print([]) -> []. print(C, F, Ad, P, Pad, Encoding, Strings) -> - [$~] ++ print_field_width(F, Ad) ++ print_precision(P) ++ + [$~] ++ print_field_width(F, Ad) ++ print_precision(P, Pad) ++ print_pad_char(Pad) ++ print_encoding(Encoding) ++ print_strings(Strings) ++ [C]. @@ -103,8 +103,9 @@ print_field_width(none, _Ad) -> ""; print_field_width(F, left) -> integer_to_list(-F); print_field_width(F, right) -> integer_to_list(F). -print_precision(none) -> ""; -print_precision(P) -> [$. | integer_to_list(P)]. +print_precision(none, $\s) -> ""; +print_precision(none, _Pad) -> "."; % pad must be second dot +print_precision(P, _Pad) -> [$. | integer_to_list(P)]. print_pad_char($\s) -> ""; % default, no need to make explicit print_pad_char(Pad) -> [$., Pad]. @@ -126,25 +127,23 @@ collect_cseq(Fmt0, Args0) -> {F,Ad,Fmt1,Args1} = field_width(Fmt0, Args0), {P,Fmt2,Args2} = precision(Fmt1, Args1), {Pad,Fmt3,Args3} = pad_char(Fmt2, Args2), - {Encoding,Fmt4,Args4} = encoding(Fmt3, Args3), - {Strings,Fmt5,Args5} = strings(Fmt4, Args4), - {C,As,Fmt6,Args6} = collect_cc(Fmt5, Args5), - FormatSpec = #{control_char => C, args => As, width => F, adjust => Ad, - precision => P, pad_char => Pad, encoding => Encoding, - strings => Strings}, - {FormatSpec,Fmt6,Args6}. - -encoding([$t|Fmt],Args) -> - true = hd(Fmt) =/= $l, - {unicode,Fmt,Args}; -encoding(Fmt,Args) -> - {latin1,Fmt,Args}. - -strings([$l|Fmt],Args) -> - true = hd(Fmt) =/= $t, - {false,Fmt,Args}; -strings(Fmt,Args) -> - {true,Fmt,Args}. + Spec0 = #{width => F, + adjust => Ad, + precision => P, + pad_char => Pad, + encoding => latin1, + strings => true}, + {Spec1,Fmt4} = modifiers(Fmt3, Spec0), + {C,As,Fmt5,Args4} = collect_cc(Fmt4, Args3), + Spec2 = Spec1#{control_char => C, args => As}, + {Spec2,Fmt5,Args4}. + +modifiers([$t|Fmt], Spec) -> + modifiers(Fmt, Spec#{encoding => unicode}); +modifiers([$l|Fmt], Spec) -> + modifiers(Fmt, Spec#{strings => false}); +modifiers(Fmt, Spec) -> + {Spec, Fmt}. field_width([$-|Fmt0], Args0) -> {F,Fmt,Args} = field_value(Fmt0, Args0), diff --git a/lib/stdlib/src/lists.erl b/lib/stdlib/src/lists.erl index af9d63ddd6..06c90c0280 100644 --- a/lib/stdlib/src/lists.erl +++ b/lib/stdlib/src/lists.erl @@ -1,7 +1,7 @@ %% %% %CopyrightBegin% %% -%% Copyright Ericsson AB 1996-2016. All Rights Reserved. +%% Copyright Ericsson AB 1996-2018. All Rights Reserved. %% %% Licensed under the Apache License, Version 2.0 (the "License"); %% you may not use this file except in compliance with the License. @@ -38,8 +38,8 @@ -export([all/2,any/2,map/2,flatmap/2,foldl/3,foldr/3,filter/2, partition/2,zf/2,filtermap/2, - mapfoldl/3,mapfoldr/3,foreach/2,takewhile/2,dropwhile/2,splitwith/2, - split/2, + mapfoldl/3,mapfoldr/3,foreach/2,takewhile/2,dropwhile/2, + search/2, splitwith/2,split/2, join/2]). %%% BIFs @@ -1399,6 +1399,19 @@ dropwhile(Pred, [Hd|Tail]=Rest) -> end; dropwhile(Pred, []) when is_function(Pred, 1) -> []. +-spec search(Pred, List) -> {value, Value} | false when + Pred :: fun((T) -> boolean()), + List :: [T], + Value :: T. + +search(Pred, [Hd|Tail]) -> + case Pred(Hd) of + true -> {value, Hd}; + false -> search(Pred, Tail) + end; +search(Pred, []) when is_function(Pred, 1) -> + false. + -spec splitwith(Pred, List) -> {List1, List2} when Pred :: fun((T) -> boolean()), List :: [T], diff --git a/lib/stdlib/src/string.erl b/lib/stdlib/src/string.erl index e01bb7d85e..4e89819e41 100644 --- a/lib/stdlib/src/string.erl +++ b/lib/stdlib/src/string.erl @@ -420,10 +420,12 @@ to_number(_, Number, Rest, _, Tail) -> %% Return the remaining string with prefix removed or else nomatch -spec prefix(String::unicode:chardata(), Prefix::unicode:chardata()) -> 'nomatch' | unicode:chardata(). -prefix(Str, []) -> Str; prefix(Str, Prefix0) -> - Prefix = unicode:characters_to_list(Prefix0), - case prefix_1(Str, Prefix) of + Result = case unicode:characters_to_list(Prefix0) of + [] -> Str; + Prefix -> prefix_1(Str, Prefix) + end, + case Result of [] when is_binary(Str) -> <<>>; Res -> Res end. diff --git a/lib/stdlib/src/uri_string.erl b/lib/stdlib/src/uri_string.erl index a84679c595..28d36ea229 100644 --- a/lib/stdlib/src/uri_string.erl +++ b/lib/stdlib/src/uri_string.erl @@ -227,7 +227,7 @@ %% External API %%------------------------------------------------------------------------- -export([compose_query/1, compose_query/2, - dissect_query/1, normalize/1, parse/1, + dissect_query/1, normalize/1, normalize/2, parse/1, recompose/1, transcode/2]). -export_type([error/0, uri_map/0, uri_string/0]). @@ -292,18 +292,36 @@ %%------------------------------------------------------------------------- %% Normalize URIs %%------------------------------------------------------------------------- --spec normalize(URIString) -> NormalizedURI when - URIString :: uri_string(), - NormalizedURI :: uri_string(). -normalize(URIString) -> - %% Percent-encoding normalization and case normalization for - %% percent-encoded triplets are achieved by running parse and - %% recompose on the input URI string. - recompose( - normalize_path_segment( - normalize_scheme_based( - normalize_case( - parse(URIString))))). +-spec normalize(URI) -> NormalizedURI when + URI :: uri_string() | uri_map(), + NormalizedURI :: uri_string() + | error(). +normalize(URIMap) -> + normalize(URIMap, []). + + +-spec normalize(URI, Options) -> NormalizedURI when + URI :: uri_string() | uri_map(), + Options :: [return_map], + NormalizedURI :: uri_string() | uri_map(). +normalize(URIMap, []) when is_map(URIMap) -> + recompose(normalize_map(URIMap)); +normalize(URIMap, [return_map]) when is_map(URIMap) -> + normalize_map(URIMap); +normalize(URIString, []) -> + case parse(URIString) of + Value when is_map(Value) -> + recompose(normalize_map(Value)); + Error -> + Error + end; +normalize(URIString, [return_map]) -> + case parse(URIString) of + Value when is_map(Value) -> + normalize_map(Value); + Error -> + Error + end. %%------------------------------------------------------------------------- @@ -385,7 +403,8 @@ transcode(URIString, Options) when is_list(URIString) -> %%------------------------------------------------------------------------- %% Functions for working with the query part of a URI as a list %% of key/value pairs. -%% HTML5 - 4.10.22.6 URL-encoded form data +%% HTML 5.2 - 4.10.21.6 URL-encoded form data - WHATWG URL (10 Jan 2018) - UTF-8 +%% HTML 5.0 - 4.10.22.6 URL-encoded form data - non UTF-8 %%------------------------------------------------------------------------- %%------------------------------------------------------------------------- @@ -393,7 +412,7 @@ transcode(URIString, Options) when is_list(URIString) -> %% (application/x-www-form-urlencoded encoding algorithm) %%------------------------------------------------------------------------- -spec compose_query(QueryList) -> QueryString when - QueryList :: [{uri_string(), uri_string()}], + QueryList :: [{unicode:chardata(), unicode:chardata()}], QueryString :: uri_string() | error(). compose_query(List) -> @@ -401,7 +420,7 @@ compose_query(List) -> -spec compose_query(QueryList, Options) -> QueryString when - QueryList :: [{uri_string(), uri_string()}], + QueryList :: [{unicode:chardata(), unicode:chardata()}], Options :: [{encoding, atom()}], QueryString :: uri_string() | error(). @@ -432,7 +451,7 @@ compose_query([], _Options, IsList, Acc) -> %%------------------------------------------------------------------------- -spec dissect_query(QueryString) -> QueryList when QueryString :: uri_string(), - QueryList :: [{uri_string(), uri_string()}] + QueryList :: [{unicode:chardata(), unicode:chardata()}] | error(). dissect_query(<<>>) -> []; @@ -1755,7 +1774,8 @@ get_separator(_L) -> <<"&">>. -%% HTML5 - 4.10.22.6 URL-encoded form data - encoding +%% HTML 5.2 - 4.10.21.6 URL-encoded form data - WHATWG URL (10 Jan 2018) - UTF-8 +%% HTML 5.0 - 4.10.22.6 URL-encoded form data - encoding (non UTF-8) form_urlencode(Cs, [{encoding, latin1}]) when is_list(Cs) -> B = convert_to_binary(Cs, utf8, utf8), html5_byte_encode(base10_encode(B)); @@ -1850,7 +1870,8 @@ dissect_query_value(<<>>, IsList, Acc, Key, Value) -> lists:reverse([{K,V}|Acc]). -%% Form-urldecode input based on RFC 1866 [8.2.1] +%% HTML 5.2 - 4.10.21.6 URL-encoded form data - WHATWG URL (10 Jan 2018) - UTF-8 +%% HTML 5.0 - 4.10.22.6 URL-encoded form data - decoding (non UTF-8) form_urldecode(true, B) -> Result = base10_decode(form_urldecode(B, <<>>)), convert_to_list(Result, utf8); @@ -1903,6 +1924,12 @@ base10_decode_unicode(<<H,_/binary>>, _, _) -> %% Helper functions for normalize %%------------------------------------------------------------------------- +normalize_map(URIMap) -> + normalize_path_segment( + normalize_scheme_based( + normalize_case(URIMap))). + + %% 6.2.2.1. Case Normalization normalize_case(#{scheme := Scheme, host := Host} = Map) -> Map#{scheme => to_lower(Scheme), |