aboutsummaryrefslogtreecommitdiffstats
path: root/lib/kernel
diff options
context:
space:
mode:
Diffstat (limited to 'lib/kernel')
-rw-r--r--lib/kernel/doc/src/file.xml28
-rw-r--r--lib/kernel/src/dist_util.erl3
-rw-r--r--lib/kernel/src/file.erl7
-rw-r--r--lib/kernel/src/file_io_server.erl18
-rw-r--r--lib/kernel/src/group.erl8
-rw-r--r--lib/kernel/src/user.erl190
6 files changed, 142 insertions, 112 deletions
diff --git a/lib/kernel/doc/src/file.xml b/lib/kernel/doc/src/file.xml
index b2a259080d..536b98b5f5 100644
--- a/lib/kernel/doc/src/file.xml
+++ b/lib/kernel/doc/src/file.xml
@@ -110,7 +110,7 @@
<desc>
<p>As returned by
<seealso marker="#open/2">file:open/2</seealso>,
- a process handling IO protocols.</p>
+ a process handling I/O-protocols.</p>
</desc>
</datatype>
<datatype>
@@ -261,6 +261,9 @@
{person, "pelle", 30}.</code>
<pre>1> <input>file:consult("f.txt").</input>
{ok,[{person,"kalle",25},{person,"pelle",30}]}</pre>
+ <p>The encoding of of <c><anno>Filename</anno></c> can be set
+ by a comment as described in <seealso
+ marker="stdlib:epp#encoding">epp(3)</seealso>.</p>
</desc>
</func>
<func>
@@ -399,6 +402,9 @@
of the error.</p>
</item>
</taglist>
+ <p>The encoding of of <c><anno>Filename</anno></c> can be set
+ by a comment as described in <seealso
+ marker="stdlib:epp#encoding">epp(3)</seealso>.</p>
</desc>
</func>
<func>
@@ -610,7 +616,7 @@
<name name="open" arity="2"/>
<fsummary>Open a file</fsummary>
<desc>
- <p>Opens the file <c><anno>Filename</anno></c> in the mode determined
+ <p>Opens the file <c><anno>File</anno></c> in the mode determined
by <c><anno>Modes</anno></c>, which may contain one or more of the
following items:</p>
<taglist>
@@ -767,6 +773,10 @@
<p>The Encoding can be changed for a file "on the fly" by using the <seealso marker="stdlib:io#setopts/2">io:setopts/2</seealso> function, why a file can be analyzed in latin1 encoding for i.e. a BOM, positioned beyond the BOM and then be set for the right encoding before further reading.See the <seealso marker="stdlib:unicode">unicode(3)</seealso> module for functions identifying BOM's.</p>
<p>This option is not allowed on <c>raw</c> files.</p>
</item>
+ <tag><c>ram</c></tag>
+ <item>
+ <p><c>File</c> must be <c>iodata()</c>. Returns an <c>fd()</c> which lets the <c>file</c> module operate on the data in-memory as if it is a file.</p>
+ </item>
</taglist>
<p>Returns:</p>
<taglist>
@@ -861,6 +871,9 @@
the error.</p>
</item>
</taglist>
+ <p>The encoding of of <c><anno>Filename</anno></c> can be set
+ by a comment as described in <seealso
+ marker="stdlib:epp#encoding">epp(3)</seealso>.</p>
</desc>
</func>
<func>
@@ -902,6 +915,9 @@
of the error.</p>
</item>
</taglist>
+ <p>The encoding of of <c><anno>Filename</anno></c> can be set
+ by a comment as described in <seealso
+ marker="stdlib:epp#encoding">epp(3)</seealso>.</p>
</desc>
</func>
<func>
@@ -971,7 +987,10 @@
of the error.</p>
</item>
</taglist>
- </desc>
+ <p>The encoding of of <c><anno>Filename</anno></c> can be set
+ by a comment as described in <seealso
+ marker="stdlib:epp#encoding">epp(3)</seealso>.</p>
+ </desc>
</func>
<func>
<name name="path_script" arity="3"/>
@@ -1502,6 +1521,9 @@
of the error.</p>
</item>
</taglist>
+ <p>The encoding of of <c><anno>Filename</anno></c> can be set
+ by a comment as described in <seealso
+ marker="stdlib:epp#encoding">epp(3)</seealso>.</p>
</desc>
</func>
<func>
diff --git a/lib/kernel/src/dist_util.erl b/lib/kernel/src/dist_util.erl
index f0d54a2f3e..e3511988a6 100644
--- a/lib/kernel/src/dist_util.erl
+++ b/lib/kernel/src/dist_util.erl
@@ -757,7 +757,8 @@ setup_timer(Pid, Timeout) ->
end.
reset_timer(Timer) ->
- Timer ! {self(), reset}.
+ Timer ! {self(), reset},
+ ok.
cancel_timer(Timer) ->
unlink(Timer),
diff --git a/lib/kernel/src/file.erl b/lib/kernel/src/file.erl
index 22af38c598..de3eaad5a1 100644
--- a/lib/kernel/src/file.erl
+++ b/lib/kernel/src/file.erl
@@ -397,9 +397,10 @@ raw_write_file_info(Name, #file_info{} = Info) ->
%% Contemporary mode specification - list of options
--spec open(Filename, Modes) -> {ok, IoDevice} | {error, Reason} when
+-spec open(File, Modes) -> {ok, IoDevice} | {error, Reason} when
+ File :: Filename | iodata(),
Filename :: name(),
- Modes :: [mode()],
+ Modes :: [mode() | ram],
IoDevice :: io_device(),
Reason :: posix() | badarg | system_limit.
@@ -1314,6 +1315,7 @@ sendfile_send(Sock, Data, Old) ->
%%% Helpers
consult_stream(Fd) ->
+ _ = epp:set_encoding(Fd),
consult_stream(Fd, 1, []).
consult_stream(Fd, Line, Acc) ->
@@ -1327,6 +1329,7 @@ consult_stream(Fd, Line, Acc) ->
end.
eval_stream(Fd, Handling, Bs) ->
+ _ = epp:set_encoding(Fd),
eval_stream(Fd, Handling, 1, undefined, [], Bs).
eval_stream(Fd, H, Line, Last, E, Bs) ->
diff --git a/lib/kernel/src/file_io_server.erl b/lib/kernel/src/file_io_server.erl
index 0bff56cf46..acaffe1e41 100644
--- a/lib/kernel/src/file_io_server.erl
+++ b/lib/kernel/src/file_io_server.erl
@@ -40,6 +40,8 @@ format_error({_Line, ?MODULE, Reason}) ->
io_lib:format("~w", [Reason]);
format_error({_Line, Mod, Reason}) ->
Mod:format_error(Reason);
+format_error(invalid_unicode) ->
+ io_lib:format("cannot translate from UTF-8", []);
format_error(ErrorId) ->
erl_posix_msg:message(ErrorId).
@@ -549,7 +551,7 @@ get_chars_notempty(Mod, Func, XtraArg, S, OutEnc,
<<>> ->
get_chars_apply(Mod, Func, XtraArg, S, OutEnc, State, eof);
_ ->
- {stop,invalid_unicode,{error,invalid_unicode},State}
+ {stop,invalid_unicode,invalid_unicode_error(Mod, Func, XtraArg, S),State}
end;
{error,Reason}=Error ->
{stop,Reason,Error,State}
@@ -616,12 +618,22 @@ get_chars_apply(Mod, Func, XtraArg, S0, OutEnc,
end
catch
exit:ExReason ->
- {stop,ExReason,{error,err_func(Mod, Func, XtraArg)},State};
+ {stop,ExReason,invalid_unicode_error(Mod, Func, XtraArg, S0),State};
error:ErrReason ->
{stop,ErrReason,{error,err_func(Mod, Func, XtraArg)},State}
end.
-
+%% A hack that tries to inform the caller about the position where the
+%% error occured.
+invalid_unicode_error(Mod, Func, XtraArg, S) ->
+ try
+ {erl_scan,tokens,_Args} = XtraArg,
+ Location = erl_scan:continuation_location(S),
+ {error,{Location, ?MODULE, invalid_unicode},Location}
+ catch
+ _:_ ->
+ {error,err_func(Mod, Func, XtraArg)}
+ end.
%% Convert error code to make it look as before
err_func(io_lib, get_until, {_,F,_}) ->
diff --git a/lib/kernel/src/group.erl b/lib/kernel/src/group.erl
index f92c6f7208..4d2e31a429 100644
--- a/lib/kernel/src/group.erl
+++ b/lib/kernel/src/group.erl
@@ -424,7 +424,7 @@ get_password_chars(Drv,Buf) ->
end.
get_chars(Prompt, M, F, Xa, Drv, Buf, Encoding) ->
- Pbs = prompt_bytes(Prompt),
+ Pbs = prompt_bytes(Prompt, Encoding),
get_chars_loop(Pbs, M, F, Xa, Drv, Buf, start, Encoding).
get_chars_loop(Pbs, M, F, Xa, Drv, Buf0, State, Encoding) ->
@@ -688,9 +688,9 @@ edit_password([Char|Cs],Chars) ->
edit_password(Cs,[Char|Chars]).
%% prompt_bytes(Prompt)
-%% Return a flat list of bytes for the Prompt.
-prompt_bytes(Prompt) ->
- lists:flatten(io_lib:format_prompt(Prompt)).
+%% Return a flat list of characters for the Prompt.
+prompt_bytes(Prompt, Encoding) ->
+ lists:flatten(io_lib:format_prompt(Prompt, Encoding)).
cast(L, binary,latin1) when is_list(L) ->
list_to_binary(L);
diff --git a/lib/kernel/src/user.erl b/lib/kernel/src/user.erl
index 88f32df20b..d6449d9e5e 100644
--- a/lib/kernel/src/user.erl
+++ b/lib/kernel/src/user.erl
@@ -81,7 +81,7 @@ server(PortName,PortSettings) ->
run(P) ->
put(read_mode,list),
- put(unicode,false),
+ put(encoding,latin1),
case init:get_argument(noshell) of
%% non-empty list -> noshell
{ok, [_|_]} ->
@@ -191,39 +191,27 @@ do_io_request(Req, From, ReplyAs, Port, Q0) ->
%% New in R13B
%% Encoding option (unicode/latin1)
io_request({put_chars,unicode,Chars}, Port, Q) -> % Binary new in R9C
- put_chars(wrap_characters_to_binary(Chars,unicode,
- case get(unicode) of
- true -> unicode;
- _ -> latin1
- end), Port, Q);
+ put_chars(wrap_characters_to_binary(Chars,unicode, get(encoding)), Port, Q);
io_request({put_chars,unicode,Mod,Func,Args}, Port, Q) ->
Result = case catch apply(Mod,Func,Args) of
Data when is_list(Data); is_binary(Data) ->
- wrap_characters_to_binary(Data,unicode,
- case get(unicode) of
- true -> unicode;
- _ -> latin1
- end);
+ wrap_characters_to_binary(Data,unicode,get(encoding));
Undef ->
Undef
end,
put_chars(Result, Port, Q);
io_request({put_chars,latin1,Chars}, Port, Q) -> % Binary new in R9C
- Data = case get(unicode) of
- true ->
+ Data = case get(encoding) of
+ unicode ->
unicode:characters_to_binary(Chars,latin1,unicode);
- false ->
+ latin1 ->
erlang:iolist_to_binary(Chars)
end,
put_chars(Data, Port, Q);
io_request({put_chars,latin1,Mod,Func,Args}, Port, Q) ->
Result = case catch apply(Mod,Func,Args) of
Data when is_list(Data); is_binary(Data) ->
- unicode:characters_to_binary(Data,latin1,
- case get(unicode) of
- true -> unicode;
- _ -> latin1
- end);
+ unicode:characters_to_binary(Data,latin1,get(encoding));
Undef ->
Undef
end,
@@ -351,9 +339,9 @@ check_valid_opts(_) ->
do_setopts(Opts, _Port, Q) ->
case proplists:get_value(encoding,Opts) of
Valid when Valid =:= unicode; Valid =:= utf8 ->
- put(unicode,true);
+ put(encoding,unicode);
latin1 ->
- put(unicode,false);
+ put(encoding,latin1);
undefined ->
ok
end,
@@ -370,12 +358,7 @@ do_setopts(Opts, _Port, Q) ->
getopts(_Port,Q) ->
Bin = {binary, get(read_mode) =:= binary},
- Uni = {encoding, case get(unicode) of
- true ->
- unicode;
- _ ->
- latin1
- end},
+ Uni = {encoding, get(encoding)},
{ok,[Bin,Uni],Q}.
@@ -575,31 +558,32 @@ binrev(L, T) ->
%% end
%% end
%% end.
-%% get_chars(Prompt, Module, Function, XtraArg, Port, Queue)
+%% get_chars(Prompt, Module, Function, XtraArg, Port, Queue, Encoding)
%% Gets characters from the input port until the applied function
%% returns {stop,Result,RestBuf}. Does not block output until input
-%% has been received.
+%% has been received. Encoding is the encoding of the data sent to
+%% the client and to Function.
%% Returns:
%% {Status,Result,NewQueue}
%% {exit,Reason}
%% Entry function.
-get_chars(Prompt, M, F, Xa, Port, Q, Fmt) ->
+get_chars(Prompt, M, F, Xa, Port, Q, Enc) ->
prompt(Port, Prompt),
case {get(eof),queue:is_empty(Q)} of
{true,true} ->
{ok,eof,Q};
_ ->
- get_chars(Prompt, M, F, Xa, Port, Q, start, Fmt)
+ get_chars(Prompt, M, F, Xa, Port, Q, start, Enc)
end.
%% First loop. Wait for port data. Respond to output requests.
-get_chars(Prompt, M, F, Xa, Port, Q, State, Fmt) ->
+get_chars(Prompt, M, F, Xa, Port, Q, State, Enc) ->
case queue:is_empty(Q) of
true ->
receive
{Port,{data,Bytes}} ->
- get_chars_bytes(State, M, F, Xa, Port, Q, Bytes, Fmt);
+ get_chars_bytes(State, M, F, Xa, Port, Q, Bytes, Enc);
{Port, eof} ->
put(eof, true),
{ok, eof, []};
@@ -610,41 +594,41 @@ get_chars(Prompt, M, F, Xa, Port, Q, State, Fmt) ->
do_io_request(Req, From, ReplyAs, Port,
queue:new()), %Keep Q over this call
%% No prompt.
- get_chars(Prompt, M, F, Xa, Port, Q, State, Fmt);
+ get_chars(Prompt, M, F, Xa, Port, Q, State, Enc);
{io_request,From,ReplyAs,Request} when is_pid(From) ->
get_chars_req(Prompt, M, F, Xa, Port, Q, State,
- Request, From, ReplyAs, Fmt);
+ Request, From, ReplyAs, Enc);
{'EXIT',From,What} when node(From) =:= node() ->
{exit,What}
end;
false ->
- get_chars_apply(State, M, F, Xa, Port, Q, Fmt)
+ get_chars_apply(State, M, F, Xa, Port, Q, Enc)
end.
get_chars_req(Prompt, M, F, XtraArg, Port, Q, State,
- Req, From, ReplyAs, Fmt) ->
+ Req, From, ReplyAs, Enc) ->
do_io_request(Req, From, ReplyAs, Port, queue:new()), %Keep Q over this call
prompt(Port, Prompt),
- get_chars(Prompt, M, F, XtraArg, Port, Q, State, Fmt).
+ get_chars(Prompt, M, F, XtraArg, Port, Q, State, Enc).
%% Second loop. Pass data to client as long as it wants more.
%% A ^G in data interrupts loop if 'noshell' is not undefined.
-get_chars_bytes(State, M, F, Xa, Port, Q, Bytes, Fmt) ->
+get_chars_bytes(State, M, F, Xa, Port, Q, Bytes, Enc) ->
case get(shell) of
noshell ->
- get_chars_apply(State, M, F, Xa, Port, queue:snoc(Q, Bytes),Fmt);
+ get_chars_apply(State, M, F, Xa, Port, queue:snoc(Q, Bytes),Enc);
_ ->
case contains_ctrl_g_or_ctrl_c(Bytes) of
false ->
get_chars_apply(State, M, F, Xa, Port,
- queue:snoc(Q, Bytes),Fmt);
+ queue:snoc(Q, Bytes),Enc);
_ ->
throw(new_shell)
end
end.
-get_chars_apply(State0, M, F, Xa, Port, Q, Fmt) ->
- case catch M:F(State0, cast(queue:head(Q),Fmt), Fmt, Xa) of
+get_chars_apply(State0, M, F, Xa, Port, Q, Enc) ->
+ case catch M:F(State0, cast(queue:head(Q),Enc), Enc, Xa) of
{stop,Result,<<>>} ->
{ok,Result,queue:tail(Q)};
{stop,Result,[]} ->
@@ -653,32 +637,32 @@ get_chars_apply(State0, M, F, Xa, Port, Q, Fmt) ->
{ok,Result,queue:tail(Q)};
{stop,Result,Buf} ->
{ok,Result,queue:cons(Buf, queue:tail(Q))};
- {'EXIT',_} ->
+ {'EXIT',_Why} ->
{error,{error,err_func(M, F, Xa)},queue:new()};
State1 ->
- get_chars_more(State1, M, F, Xa, Port, queue:tail(Q), Fmt)
+ get_chars_more(State1, M, F, Xa, Port, queue:tail(Q), Enc)
end.
-get_chars_more(State, M, F, Xa, Port, Q, Fmt) ->
+get_chars_more(State, M, F, Xa, Port, Q, Enc) ->
case queue:is_empty(Q) of
true ->
case get(eof) of
undefined ->
receive
{Port,{data,Bytes}} ->
- get_chars_bytes(State, M, F, Xa, Port, Q, Bytes, Fmt);
+ get_chars_bytes(State, M, F, Xa, Port, Q, Bytes, Enc);
{Port,eof} ->
put(eof, true),
get_chars_apply(State, M, F, Xa, Port,
- queue:snoc(Q, eof), Fmt);
+ queue:snoc(Q, eof), Enc);
{'EXIT',From,What} when node(From) =:= node() ->
{exit,What}
end;
_ ->
- get_chars_apply(State, M, F, Xa, Port, queue:snoc(Q, eof), Fmt)
+ get_chars_apply(State, M, F, Xa, Port, queue:snoc(Q, eof), Enc)
end;
false ->
- get_chars_apply(State, M, F, Xa, Port, Q, Fmt)
+ get_chars_apply(State, M, F, Xa, Port, Q, Enc)
end.
@@ -689,11 +673,10 @@ get_chars_more(State, M, F, Xa, Port, Q, Fmt) ->
prompt(_Port, '') -> ok;
prompt(Port, Prompt) ->
- put_port(wrap_characters_to_binary(io_lib:format_prompt(Prompt),unicode,
- case get(unicode) of
- true -> unicode;
- _ -> latin1
- end), Port).
+ Encoding = get(encoding),
+ put_port(wrap_characters_to_binary(io_lib:format_prompt(Prompt, Encoding),
+ unicode, Encoding),
+ Port).
%% Convert error code to make it look as before
err_func(io_lib, get_until, {_,F,_}) ->
@@ -710,56 +693,65 @@ contains_ctrl_g_or_ctrl_c(BinOrList)->
end.
%% Convert a buffer between list and binary
-cast(Data, _Format) when is_atom(Data) ->
+cast(Data, _Encoding) when is_atom(Data) ->
Data;
-cast(Data, Format) ->
- cast(Data, get(read_mode), Format, get(unicode)).
+cast(Data, Encoding) ->
+ IoEncoding = get(encoding),
+ cast(Data, get(read_mode), IoEncoding, Encoding).
-cast(B, binary, latin1, false) when is_binary(B) ->
+cast(B, binary, latin1, latin1) when is_binary(B) ->
B;
-cast(B, binary, latin1, true) when is_binary(B) ->
- unicode:characters_to_binary(B, unicode, latin1);
-cast(L, binary, latin1, false) ->
- erlang:iolist_to_binary(L);
-cast(L, binary, latin1, true) ->
- case unicode:characters_to_binary(
- erlang:iolist_to_binary(L),unicode,latin1) of % may fail
- {error,_,_} -> exit({no_translation, unicode, latin1});
- Else -> Else
+cast(L, binary, latin1, latin1) ->
+ case catch erlang:iolist_to_binary(L) of
+ Bin when is_binary(Bin) -> Bin;
+ _ -> exit({no_translation, latin1, latin1})
+ end;
+cast(Data, binary, unicode, latin1) when is_binary(Data); is_list(Data) ->
+ case catch unicode:characters_to_binary(Data, unicode, latin1) of
+ Bin when is_binary(Bin) -> Bin;
+ _ -> exit({no_translation, unicode, latin1})
+ end;
+cast(Data, binary, latin1, unicode) when is_binary(Data); is_list(Data) ->
+ case catch unicode:characters_to_binary(Data, latin1, unicode) of
+ Bin when is_binary(Bin) -> Bin;
+ _ -> exit({no_translation, latin1, unicode})
end;
-cast(B, binary, unicode, true) when is_binary(B) ->
+cast(B, binary, unicode, unicode) when is_binary(B) ->
B;
-cast(B, binary, unicode, false) when is_binary(B) ->
- unicode:characters_to_binary(B,latin1,unicode);
-cast(L, binary, unicode, true) ->
- % possibly a list containing UTF-8 encoded characters
- unicode:characters_to_binary(erlang:iolist_to_binary(L));
-cast(L, binary, unicode, false) ->
- unicode:characters_to_binary(L, latin1, unicode);
-cast(L, list, latin1, UniTerm) ->
- case UniTerm of
- true -> % Convert input characters to protocol format (i.e latin1)
- case unicode:characters_to_list(
- erlang:iolist_to_binary(L),unicode) of % may fail
- {error,_,_} -> exit({no_translation, unicode, latin1});
- Else -> [ case X of
- High when High > 255 ->
- exit({no_translation, unicode, latin1});
- Low ->
- Low
- end || X <- Else ]
- end;
- _ ->
- binary_to_list(erlang:iolist_to_binary(L))
+cast(L, binary, unicode, unicode) ->
+ case catch unicode:characters_to_binary(L, unicode) of
+ Bin when is_binary(Bin) -> Bin;
+ _ -> exit({no_translation, unicode, unicode})
end;
-cast(L, list, unicode, UniTerm) ->
- unicode:characters_to_list(erlang:iolist_to_binary(L),
- case UniTerm of
- true -> unicode;
- _ -> latin1
- end);
-cast(Other, _, _,_) ->
- Other.
+cast(B, list, latin1, latin1) when is_binary(B) ->
+ binary_to_list(B);
+cast(L, list, latin1, latin1) ->
+ case catch erlang:iolist_to_binary(L) of
+ Bin when is_binary(Bin) -> binary_to_list(Bin);
+ _ -> exit({no_translation, latin1, latin1})
+ end;
+cast(Data, list, unicode, latin1) when is_binary(Data); is_list(Data) ->
+ case catch unicode:characters_to_list(Data, unicode) of
+ Chars when is_list(Chars) ->
+ [ case X of
+ High when High > 255 ->
+ exit({no_translation, unicode, latin1});
+ Low ->
+ Low
+ end || X <- Chars ];
+ _ ->
+ exit({no_translation, unicode, latin1})
+ end;
+cast(Data, list, latin1, unicode) when is_binary(Data); is_list(Data) ->
+ case catch unicode:characters_to_list(Data, latin1) of
+ Chars when is_list(Chars) -> Chars;
+ _ -> exit({no_translation, latin1, unicode})
+ end;
+cast(Data, list, unicode, unicode) when is_binary(Data); is_list(Data) ->
+ case catch unicode:characters_to_list(Data, unicode) of
+ Chars when is_list(Chars) -> Chars;
+ _ -> exit({no_translation, unicode, unicode})
+ end.
wrap_characters_to_binary(Chars,unicode,latin1) ->
case unicode:characters_to_binary(Chars,unicode,latin1) of
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-28 16:32:22 +0000