aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorBjörn Gustavsson <[email protected]>2014-03-19 11:51:46 +0100
committerBjörn Gustavsson <[email protected]>2014-03-19 11:51:46 +0100
commit17944ea74120ac831b85d0c3cfe2be419285c6d7 (patch)
tree933fb986a8833cb7e0c42eefd00f3f414bf0c225
parent539614ad82cfd0ced75e9f43fa76b013d4598d01 (diff)
parentbfe7e45cda6591dc31405ed1be961f079cc541c9 (diff)
downloadotp-17944ea74120ac831b85d0c3cfe2be419285c6d7.tar.gz
otp-17944ea74120ac831b85d0c3cfe2be419285c6d7.tar.bz2
otp-17944ea74120ac831b85d0c3cfe2be419285c6d7.zip
Merge branch 'bjorn/compiler/utf8-warning/OTP-11791'
* bjorn/compiler/utf8-warning/OTP-11791: Don't fail compilation for modules that contain invalid UTF-8 epp: Make it possible to specify a default encoding
-rw-r--r--lib/compiler/src/compile.erl59
-rw-r--r--lib/compiler/test/error_SUITE.erl28
-rw-r--r--lib/compiler/test/warnings_SUITE.erl38
-rw-r--r--lib/stdlib/doc/src/epp.xml56
-rw-r--r--lib/stdlib/src/epp.erl147
-rw-r--r--lib/stdlib/test/epp_SUITE.erl78
6 files changed, 341 insertions, 65 deletions
diff --git a/lib/compiler/src/compile.erl b/lib/compiler/src/compile.erl
index 9030dd998b..c7d91070f6 100644
--- a/lib/compiler/src/compile.erl
+++ b/lib/compiler/src/compile.erl
@@ -234,7 +234,9 @@ format_error({crash,Pass,Reason}) ->
format_error({bad_return,Pass,Reason}) ->
io_lib:format("internal error in ~p;\nbad return value: ~ts", [Pass,format_error_reason(Reason)]);
format_error({module_name,Mod,Filename}) ->
- io_lib:format("Module name '~s' does not match file name '~ts'", [Mod,Filename]).
+ io_lib:format("Module name '~s' does not match file name '~ts'", [Mod,Filename]);
+format_error(reparsing_invalid_unicode) ->
+ "Non-UTF-8 character(s) detected, but no encoding declared. Encode the file in UTF-8 or add \"%% coding: latin-1\" at the beginning of the file. Retrying with latin-1 encoding.".
format_error_reason({Reason, Stack}) when is_list(Stack) ->
StackFun = fun
@@ -792,20 +794,59 @@ no_native_compilation(BeamFile, #compile{options=Opts0}) ->
_ -> false
end.
-parse_module(St) ->
- Opts = St#compile.options,
- Cwd = ".",
- IncludePath = [Cwd, St#compile.dir|inc_paths(Opts)],
- R = epp:parse_file(St#compile.ifile, IncludePath, pre_defs(Opts)),
+parse_module(St0) ->
+ case do_parse_module(utf8, St0) of
+ {ok,_}=Ret ->
+ Ret;
+ {error,_}=Ret ->
+ Ret;
+ {invalid_unicode,File,Line} ->
+ case do_parse_module(latin1, St0) of
+ {ok,St} ->
+ Es = [{File,[{Line,?MODULE,reparsing_invalid_unicode}]}],
+ {ok,St#compile{warnings=Es++St#compile.warnings}};
+ {error,St} ->
+ Es = [{File,[{Line,?MODULE,reparsing_invalid_unicode}]}],
+ {error,St#compile{errors=Es++St#compile.errors}}
+ end
+ end.
+
+do_parse_module(DefEncoding, #compile{ifile=File,options=Opts,dir=Dir}=St) ->
+ R = epp:parse_file(File,
+ [{includes,[".",Dir|inc_paths(Opts)]},
+ {macros,pre_defs(Opts)},
+ {default_encoding,DefEncoding},
+ extra]),
case R of
- {ok,Forms} ->
- Encoding = epp:read_encoding(St#compile.ifile),
- {ok,St#compile{code=Forms,encoding=Encoding}};
+ {ok,Forms,Extra} ->
+ Encoding = proplists:get_value(encoding, Extra),
+ case find_invalid_unicode(Forms, File) of
+ none ->
+ {ok,St#compile{code=Forms,encoding=Encoding}};
+ {invalid_unicode,_,_}=Ret ->
+ case Encoding of
+ none ->
+ Ret;
+ _ ->
+ {ok,St#compile{code=Forms,encoding=Encoding}}
+ end
+ end;
{error,E} ->
Es = [{St#compile.ifile,[{none,?MODULE,{epp,E}}]}],
{error,St#compile{errors=St#compile.errors ++ Es}}
end.
+find_invalid_unicode([H|T], File0) ->
+ case H of
+ {attribute,_,file,{File,_}} ->
+ find_invalid_unicode(T, File);
+ {error,{Line,file_io_server,invalid_unicode}} ->
+ {invalid_unicode,File0,Line};
+ _Other ->
+ find_invalid_unicode(T, File0)
+ end;
+find_invalid_unicode([], _) -> none.
+
parse_core(St) ->
case file:read_file(St#compile.ifile) of
{ok,Bin} ->
diff --git a/lib/compiler/test/error_SUITE.erl b/lib/compiler/test/error_SUITE.erl
index 5cdf429a5f..bd877bb528 100644
--- a/lib/compiler/test/error_SUITE.erl
+++ b/lib/compiler/test/error_SUITE.erl
@@ -23,7 +23,7 @@
-export([all/0, suite/0,groups/0,init_per_suite/1, end_per_suite/1,
init_per_group/2,end_per_group/2,
head_mismatch_line/1,warnings_as_errors/1, bif_clashes/1,
- transforms/1,forbidden_maps/1]).
+ transforms/1,forbidden_maps/1,bad_utf8/1]).
%% Used by transforms/1 test case.
-export([parse_transform/2]).
@@ -36,7 +36,8 @@ all() ->
groups() ->
[{p,test_lib:parallel(),
- [head_mismatch_line,warnings_as_errors,bif_clashes,transforms,forbidden_maps]}].
+ [head_mismatch_line,warnings_as_errors,bif_clashes,
+ transforms,forbidden_maps,bad_utf8]}].
init_per_suite(Config) ->
Config.
@@ -254,6 +255,23 @@ forbidden_maps(Config) when is_list(Config) ->
[] = run2(Config, Ts1),
ok.
+bad_utf8(Config) ->
+ Ts = [{bad_utf8,
+ %% If coding is specified explicitly as utf-8, there should be
+ %% a compilation error; we must not fallback to parsing the
+ %% file in latin-1 mode.
+ <<"%% coding: utf-8
+ %% Bj",246,"rn
+ t() -> \"",246,"\".
+ ">>,
+ [],
+ {error,[{2,epp,cannot_parse},
+ {2,file_io_server,invalid_unicode}],
+ []}
+ }],
+ [] = run2(Config, Ts),
+ ok.
+
run(Config, Tests) ->
?line File = test_filename(Config),
@@ -318,6 +336,7 @@ run_test(Test0, File, Warnings, WriteBeam) ->
?line compile:file(File, [binary,report|Warnings]),
%% Test result of compilation.
+ io:format("~p\n", [Opts]),
?line Res = case compile:file(File, Opts) of
{ok,Mod,_,[{_File,Ws}]} ->
%io:format("compile:file(~s,~p) ->~n~p~n",
@@ -335,6 +354,11 @@ run_test(Test0, File, Warnings, WriteBeam) ->
%io:format("compile:file(~s,~p) ->~n~p~n",
% [File,Opts,_ZZ]),
{error,Es,Ws};
+ {error,[{XFile,Es1},{XFile,Es2}],Ws} = _ZZ
+ when is_list(XFile) ->
+ %io:format("compile:file(~s,~p) ->~n~p~n",
+ % [File,Opts,_ZZ]),
+ {error,Es1++Es2,Ws};
{error,Es,[{_File,Ws}]} = _ZZ->
%io:format("compile:file(~s,~p) ->~n~p~n",
% [File,Opts,_ZZ]),
diff --git a/lib/compiler/test/warnings_SUITE.erl b/lib/compiler/test/warnings_SUITE.erl
index de56a59e12..c3b02819f9 100644
--- a/lib/compiler/test/warnings_SUITE.erl
+++ b/lib/compiler/test/warnings_SUITE.erl
@@ -37,8 +37,9 @@
-export([pattern/1,pattern2/1,pattern3/1,pattern4/1,
guard/1,bad_arith/1,bool_cases/1,bad_apply/1,
- files/1,effect/1,bin_opt_info/1,bin_construction/1, comprehensions/1,
- maps/1,redundant_boolean_clauses/1]).
+ files/1,effect/1,bin_opt_info/1,bin_construction/1,
+ comprehensions/1,maps/1,redundant_boolean_clauses/1,
+ latin1_fallback/1]).
% Default timetrap timeout (set in init_per_testcase).
-define(default_timeout, ?t:minutes(2)).
@@ -63,7 +64,7 @@ groups() ->
[pattern,pattern2,pattern3,pattern4,guard,
bad_arith,bool_cases,bad_apply,files,effect,
bin_opt_info,bin_construction,comprehensions,maps,
- redundant_boolean_clauses]}].
+ redundant_boolean_clauses,latin1_fallback]}].
init_per_suite(Config) ->
Config.
@@ -591,6 +592,37 @@ redundant_boolean_clauses(Config) when is_list(Config) ->
run(Config, Ts),
ok.
+latin1_fallback(Conf) when is_list(Conf) ->
+ DataDir = ?privdir,
+ IncFile = filename:join(DataDir, "include_me.hrl"),
+ file:write_file(IncFile, <<"%% ",246," in include file\n">>),
+ Ts1 = [{latin1_fallback1,
+ %% Test that the compiler fall backs to latin-1 with
+ %% a warning if a file has no encoding and does not
+ %% contain correct UTF-8 sequences.
+ <<"%% Bj",246,"rn
+ t(_) -> \"",246,"\";
+ t(x) -> ok.
+ ">>,
+ [],
+ {warnings,[{1,compile,reparsing_invalid_unicode},
+ {3,sys_core_fold,{nomatch_shadow,2}}]}}],
+ [] = run(Conf, Ts1),
+
+ Ts2 = [{latin1_fallback2,
+ %% Test that the compiler fall backs to latin-1 with
+ %% a warning if a file has no encoding and does not
+ %% contain correct UTF-8 sequences.
+ <<"
+
+ -include(\"include_me.hrl\").
+ ">>,
+ [],
+ {warnings,[{1,compile,reparsing_invalid_unicode}]}
+ }],
+ [] = run(Conf, Ts2),
+ ok.
+
%%%
%%% End of test cases.
%%%
diff --git a/lib/stdlib/doc/src/epp.xml b/lib/stdlib/doc/src/epp.xml
index cf33530395..50baad04a9 100644
--- a/lib/stdlib/doc/src/epp.xml
+++ b/lib/stdlib/doc/src/epp.xml
@@ -64,11 +64,29 @@
</datatypes>
<funcs>
<func>
+ <name name="open" arity="1"/>
+ <fsummary>Open a file for preprocessing</fsummary>
+ <desc>
+ <p>Opens a file for preprocessing.</p>
+ <p>If <c>extra</c> is given in
+ <c><anno>Options</anno></c>, the return value will be
+ <c>{ok, <anno>Epp</anno>, <anno>Extra</anno>}</c> instead
+ of <c>{ok, <anno>Epp</anno>}</c>.</p>
+ </desc>
+ </func>
+ <func>
<name name="open" arity="2"/>
+ <fsummary>Open a file for preprocessing</fsummary>
+ <desc>
+ <p>Equivalent to <c>epp:open([{name, FileName}, {includes, IncludePath}])</c>.</p>
+ </desc>
+ </func>
+ <func>
<name name="open" arity="3"/>
<fsummary>Open a file for preprocessing</fsummary>
<desc>
- <p>Opens a file for preprocessing.</p>
+ <p>Equivalent to <c>epp:open([{name, FileName}, {includes, IncludePath},
+ {macros, PredefMacros}])</c>.</p>
</desc>
</func>
<func>
@@ -89,12 +107,24 @@
</desc>
</func>
<func>
- <name name="parse_file" arity="3"/>
+ <name name="parse_file" arity="2"/>
<fsummary>Preprocess and parse an Erlang source file</fsummary>
<desc>
<p>Preprocesses and parses an Erlang source file.
- Note that the tuple <c>{eof, <anno>Line</anno>}</c> returned at end-of-file is
- included as a "form".</p>
+ Note that the tuple <c>{eof, <anno>Line</anno>}</c> returned
+ at end-of-file is included as a "form".</p>
+ <p>If <c>extra</c> is given in
+ <c><anno>Options</anno></c>, the return value will be
+ <c>{ok, [<anno>Form</anno>], <anno>Extra</anno>}</c> instead
+ of <c>{ok, [<anno>Form</anno>]}</c>.</p>
+ </desc>
+ </func>
+ <func>
+ <name name="parse_file" arity="3"/>
+ <fsummary>Preprocess and parse an Erlang source file</fsummary>
+ <desc>
+ <p>Equivalent to <c>epp:parse_file(FileName, [{includes, IncludePath},
+ {macros, PredefMacros}])</c>.</p>
</desc>
</func>
<func>
@@ -111,7 +141,7 @@
<p>Returns a string representation of an encoding. The string
is recognized by <c>read_encoding/1,2</c>,
<c>read_encoding_from_binary/1,2</c>, and
- <c>set_encoding/1</c> as a valid encoding.</p>
+ <c>set_encoding/1,2</c> as a valid encoding.</p>
</desc>
</func>
<func>
@@ -157,6 +187,22 @@
</desc>
</func>
<func>
+ <name name="set_encoding" arity="2"/>
+ <fsummary>Read and set the encoding of an IO device</fsummary>
+ <desc>
+ <p>Reads the <seealso marker="#encoding">encoding</seealso> from
+ an IO device and sets the encoding of the device
+ accordingly. The position of the IO device referenced by
+ <c><anno>File</anno></c> is not affected. If no valid
+ encoding can be read from the IO device the encoding of the
+ IO device is set to the
+ <seealso marker="#encoding">encoding</seealso> given by
+ <c><anno>Default</anno></c>.</p>
+ <p>Returns the read encoding, or <c>none</c> if no valid
+ encoding was found.</p>
+ </desc>
+ </func>
+ <func>
<name name="format_error" arity="1"/>
<fsummary>Format an error descriptor</fsummary>
<desc>
diff --git a/lib/stdlib/src/epp.erl b/lib/stdlib/src/epp.erl
index 68e079b7e5..d212a55b47 100644
--- a/lib/stdlib/src/epp.erl
+++ b/lib/stdlib/src/epp.erl
@@ -20,12 +20,12 @@
%% An Erlang code preprocessor.
--export([open/2,open/3,open/5,close/1,format_error/1]).
+-export([open/1, open/2,open/3,open/5,close/1,format_error/1]).
-export([scan_erl_form/1,parse_erl_form/1,macro_defs/1]).
--export([parse_file/1, parse_file/3]).
+-export([parse_file/1, parse_file/2, parse_file/3]).
-export([default_encoding/0, encoding_to_string/1,
read_encoding_from_binary/1, read_encoding_from_binary/2,
- set_encoding/1, read_encoding/1, read_encoding/2]).
+ set_encoding/1, set_encoding/2, read_encoding/1, read_encoding/2]).
-export([interpret_file_attribute/1]).
-export([normalize_typed_record_fields/1,restore_typed_record_fields/1]).
@@ -37,9 +37,11 @@
-type epp_handle() :: pid().
-type source_encoding() :: latin1 | utf8.
+-define(DEFAULT_ENCODING, utf8).
+
%% Epp state record.
-record(epp, {file, %Current file
- location, %Current location
+ location=1, %Current location
delta, %Offset from Location (-file)
name="", %Current file name
name2="", %-"-, modified by -file
@@ -48,6 +50,7 @@
path=[], %Include-path
macs = dict:new() :: dict:dict(),%Macros (don't care locations)
uses = dict:new() :: dict:dict(),%Macro use structure
+ default_encoding = ?DEFAULT_ENCODING :: source_encoding(),
pre_opened = false :: boolean()
}).
@@ -58,6 +61,7 @@
%%% distinction in the internal representation would simplify the code
%%% a little.
+%% open(Options)
%% open(FileName, IncludePath)
%% open(FileName, IncludePath, PreDefMacros)
%% open(FileName, IoDevice, StartLocation, IncludePath, PreDefMacros)
@@ -65,6 +69,7 @@
%% scan_erl_form(Epp)
%% parse_erl_form(Epp)
%% parse_file(Epp)
+%% parse_file(FileName, Options)
%% parse_file(FileName, IncludePath, PreDefMacros)
%% macro_defs(Epp)
@@ -87,14 +92,43 @@ open(Name, Path) ->
ErrorDescriptor :: term().
open(Name, Path, Pdm) ->
- Self = self(),
- Epp = spawn(fun() -> server(Self, Name, Path, Pdm) end),
- epp_request(Epp).
+ internal_open([{name, Name}, {includes, Path}, {macros, Pdm}], #epp{}).
open(Name, File, StartLocation, Path, Pdm) ->
- Self = self(),
- Epp = spawn(fun() -> server(Self, Name, File, StartLocation,Path,Pdm) end),
- epp_request(Epp).
+ internal_open([{name, Name}, {includes, Path}, {macros, Pdm}],
+ #epp{file=File, pre_opened=true, location=StartLocation}).
+
+-spec open(Options) ->
+ {'ok', Epp} | {'ok', Epp, Extra} | {'error', ErrorDescriptor} when
+ Options :: [{'default_encoding', DefEncoding :: source_encoding()} |
+ {'includes', IncludePath :: [DirectoryName :: file:name()]} |
+ {'macros', PredefMacros :: macros()} |
+ {'name',FileName :: file:name()} |
+ 'extra'],
+ Epp :: epp_handle(),
+ Extra :: [{'encoding', source_encoding() | 'none'}],
+ ErrorDescriptor :: term().
+
+open(Options) ->
+ internal_open(Options, #epp{}).
+
+internal_open(Options, St) ->
+ case proplists:get_value(name, Options) of
+ undefined ->
+ erlang:error(badarg);
+ Name ->
+ Self = self(),
+ Epp = spawn(fun() -> server(Self, Name, Options, St) end),
+ case epp_request(Epp) of
+ {ok, Pid, Encoding} ->
+ case proplists:get_bool(extra, Options) of
+ true -> {ok, Pid, [{encoding, Encoding}]};
+ false -> {ok, Pid}
+ end;
+ Other ->
+ Other
+ end
+ end.
-spec close(Epp) -> 'ok' when
Epp :: epp_handle().
@@ -170,9 +204,6 @@ format_error({'NYI',What}) ->
io_lib:format("not yet implemented '~s'", [What]);
format_error(E) -> file:format_error(E).
-%% parse_file(FileName, IncludePath, [PreDefMacro]) ->
-%% {ok,[Form]} | {error,OpenError}
-
-spec parse_file(FileName, IncludePath, PredefMacros) ->
{'ok', [Form]} | {error, OpenError} when
FileName :: file:name(),
@@ -184,17 +215,40 @@ format_error(E) -> file:format_error(E).
OpenError :: file:posix() | badarg | system_limit.
parse_file(Ifile, Path, Predefs) ->
- case open(Ifile, Path, Predefs) of
+ parse_file(Ifile, [{includes, Path}, {macros, Predefs}]).
+
+-spec parse_file(FileName, Options) ->
+ {'ok', [Form]} | {'ok', [Form], Extra} | {error, OpenError} when
+ FileName :: file:name(),
+ Options :: [{'includes', IncludePath :: [DirectoryName :: file:name()]} |
+ {'macros', PredefMacros :: macros()} |
+ {'default_encoding', DefEncoding :: source_encoding()} |
+ 'extra'],
+ Form :: erl_parse:abstract_form() | {'error', ErrorInfo} | {'eof',Line},
+ Line :: erl_scan:line(),
+ ErrorInfo :: erl_scan:error_info() | erl_parse:error_info(),
+ Extra :: [{'encoding', source_encoding() | 'none'}],
+ OpenError :: file:posix() | badarg | system_limit.
+
+parse_file(Ifile, Options) ->
+ case internal_open([{name, Ifile} | Options], #epp{}) of
{ok,Epp} ->
Forms = parse_file(Epp),
close(Epp),
{ok,Forms};
+ {ok,Epp,Extra} ->
+ Forms = parse_file(Epp),
+ close(Epp),
+ {ok,Forms,Extra};
{error,E} ->
{error,E}
end.
-%% parse_file(Epp) ->
-%% [Form]
+-spec parse_file(Epp) -> [Form] when
+ Epp :: epp_handle(),
+ Form :: erl_parse:abstract_form() | {'error', ErrorInfo} | {'eof',Line},
+ Line :: erl_scan:line(),
+ ErrorInfo :: erl_scan:error_info() | erl_parse:error_info().
parse_file(Epp) ->
case parse_erl_form(Epp) of
@@ -219,8 +273,6 @@ parse_file(Epp) ->
[{eof,Location}]
end.
--define(DEFAULT_ENCODING, utf8).
-
-spec default_encoding() -> source_encoding().
default_encoding() ->
@@ -258,9 +310,16 @@ read_encoding(Name, Options) ->
File :: io:device(). % pid(); raw files don't work
set_encoding(File) ->
+ set_encoding(File, ?DEFAULT_ENCODING).
+
+-spec set_encoding(File, Default) -> source_encoding() | none when
+ Default :: source_encoding(),
+ File :: io:device(). % pid(); raw files don't work
+
+set_encoding(File, Default) ->
Encoding = read_encoding_from_file(File, true),
Enc = case Encoding of
- none -> default_encoding();
+ none -> Default;
Encoding -> Encoding
end,
ok = io:setopts(File, [{encoding, Enc}]),
@@ -446,35 +505,37 @@ restore_typed_record_fields([{attribute,La,type,{{record,Record},Fields,[]}}|
restore_typed_record_fields([Form|Forms]) ->
[Form|restore_typed_record_fields(Forms)].
-%% server(StarterPid, FileName, Path, PreDefMacros)
-
-server(Pid, Name, Path, Pdm) ->
+server(Pid, Name, Options, #epp{pre_opened=PreOpened}=St) ->
process_flag(trap_exit, true),
- case file:open(Name, [read]) of
- {ok,File} ->
- Location = 1,
- init_server(Pid, Name, File, Location, Path, Pdm, false);
- {error,E} ->
- epp_reply(Pid, {error,E})
+ case PreOpened of
+ false ->
+ case file:open(Name, [read]) of
+ {ok,File} ->
+ init_server(Pid, Name, Options, St#epp{file = File});
+ {error,E} ->
+ epp_reply(Pid, {error,E})
+ end;
+ true ->
+ init_server(Pid, Name, Options, St)
end.
-%% server(StarterPid, FileName, IoDevice, Location, Path, PreDefMacros)
-server(Pid, Name, File, AtLocation, Path, Pdm) ->
- process_flag(trap_exit, true),
- init_server(Pid, Name, File, AtLocation, Path, Pdm, true).
-
-init_server(Pid, Name, File, AtLocation, Path, Pdm, Pre) ->
+init_server(Pid, Name, Options, St0) ->
+ Pdm = proplists:get_value(macros, Options, []),
Ms0 = predef_macros(Name),
case user_predef(Pdm, Ms0) of
{ok,Ms1} ->
- _ = set_encoding(File),
- epp_reply(Pid, {ok,self()}),
+ #epp{file = File, location = AtLocation} = St0,
+ DefEncoding = proplists:get_value(default_encoding, Options,
+ ?DEFAULT_ENCODING),
+ Encoding = set_encoding(File, DefEncoding),
+ epp_reply(Pid, {ok,self(),Encoding}),
%% ensure directory of current source file is
%% first in path
- Path1 = [filename:dirname(Name) | Path],
- St = #epp{file=File, location=AtLocation, delta=0,
- name=Name, name2=Name, path=Path1, macs=Ms1,
- pre_opened = Pre},
+ Path = [filename:dirname(Name) |
+ proplists:get_value(includes, Options, [])],
+ St = St0#epp{delta=0, name=Name, name2=Name,
+ path=Path, macs=Ms1,
+ default_encoding=DefEncoding},
From = wait_request(St),
enter_file_reply(From, Name, AtLocation, AtLocation),
wait_req_scan(St);
@@ -600,9 +661,11 @@ enter_file2(NewF, Pname, From, St0, AtLocation) ->
%% the path) must be dropped, otherwise the path used within the current
%% file will depend on the order of file inclusions in the parent files
Path = [filename:dirname(Pname) | tl(St0#epp.path)],
- _ = set_encoding(NewF),
+ DefEncoding = St0#epp.default_encoding,
+ _ = set_encoding(NewF, DefEncoding),
#epp{file=NewF,location=Loc,name=Pname,name2=Pname,delta=0,
- sstk=[St0|St0#epp.sstk],path=Path,macs=Ms}.
+ sstk=[St0|St0#epp.sstk],path=Path,macs=Ms,
+ default_encoding=DefEncoding}.
enter_file_reply(From, Name, Location, AtLocation) ->
Attr = loc_attr(AtLocation),
diff --git a/lib/stdlib/test/epp_SUITE.erl b/lib/stdlib/test/epp_SUITE.erl
index 0b4726c07a..b17e8bd186 100644
--- a/lib/stdlib/test/epp_SUITE.erl
+++ b/lib/stdlib/test/epp_SUITE.erl
@@ -26,7 +26,7 @@
pmod/1, not_circular/1, skip_header/1, otp_6277/1, otp_7702/1,
otp_8130/1, overload_mac/1, otp_8388/1, otp_8470/1, otp_8503/1,
otp_8562/1, otp_8665/1, otp_8911/1, otp_10302/1, otp_10820/1,
- otp_11728/1]).
+ otp_11728/1, encoding/1]).
-export([epp_parse_erl_form/2]).
@@ -68,7 +68,8 @@ all() ->
{group, variable}, otp_4870, otp_4871, otp_5362, pmod,
not_circular, skip_header, otp_6277, otp_7702, otp_8130,
overload_mac, otp_8388, otp_8470, otp_8503, otp_8562,
- otp_8665, otp_8911, otp_10302, otp_10820, otp_11728].
+ otp_8665, otp_8911, otp_10302, otp_10820, otp_11728,
+ encoding].
groups() ->
[{upcase_mac, [], [upcase_mac_1, upcase_mac_2]},
@@ -123,10 +124,22 @@ include_local(Config) when is_list(Config) ->
%%% regular epp:parse_file, the test case will time out, and then epp
%%% server will go on growing until we dump core.
epp_parse_file(File, Inc, Predef) ->
- {ok, Epp} = epp:open(File, Inc, Predef),
+ List = do_epp_parse_file(fun() ->
+ epp:open(File, Inc, Predef)
+ end),
+ List = do_epp_parse_file(fun() ->
+ Opts = [{name, File},
+ {includes, Inc},
+ {macros, Predef}],
+ epp:open(Opts)
+ end),
+ {ok, List}.
+
+do_epp_parse_file(Open) ->
+ {ok, Epp} = Open(),
List = collect_epp_forms(Epp),
epp:close(Epp),
- {ok, List}.
+ List.
collect_epp_forms(Epp) ->
Result = epp_parse_erl_form(Epp),
@@ -1413,6 +1426,63 @@ otp_11728(Config) when is_list(Config) ->
_ = file:delete(ErlFile),
ok.
+%% Check the new API for setting the default encoding.
+encoding(Config) when is_list(Config) ->
+ Dir = ?config(priv_dir, Config),
+ ErlFile = filename:join(Dir, "encoding.erl"),
+
+ %% Try a latin-1 file with no encoding given.
+ C1 = <<"-module(encoding).
+ %% ",246,"
+ ">>,
+ ok = file:write_file(ErlFile, C1),
+ {ok,[{attribute,1,file,_},
+ {attribute,1,module,encoding},
+ {error,_},
+ {error,{2,epp,cannot_parse}},
+ {eof,2}]} = epp:parse_file(ErlFile, []),
+ {ok,[{attribute,1,file,_},
+ {attribute,1,module,encoding},
+ {eof,3}]} =
+ epp:parse_file(ErlFile, [{default_encoding,latin1}]),
+ {ok,[{attribute,1,file,_},
+ {attribute,1,module,encoding},
+ {eof,3}],[{encoding,none}]} =
+ epp:parse_file(ErlFile, [{default_encoding,latin1},extra]),
+
+ %% Try a latin-1 file with encoding given in a comment.
+ C2 = <<"-module(encoding).
+ %% encoding: latin-1
+ %% ",246,"
+ ">>,
+ ok = file:write_file(ErlFile, C2),
+ {ok,[{attribute,1,file,_},
+ {attribute,1,module,encoding},
+ {eof,4}]} =
+ epp:parse_file(ErlFile, []),
+ {ok,[{attribute,1,file,_},
+ {attribute,1,module,encoding},
+ {eof,4}]} =
+ epp:parse_file(ErlFile, [{default_encoding,latin1}]),
+ {ok,[{attribute,1,file,_},
+ {attribute,1,module,encoding},
+ {eof,4}]} =
+ epp:parse_file(ErlFile, [{default_encoding,utf8}]),
+ {ok,[{attribute,1,file,_},
+ {attribute,1,module,encoding},
+ {eof,4}],[{encoding,latin1}]} =
+ epp:parse_file(ErlFile, [extra]),
+ {ok,[{attribute,1,file,_},
+ {attribute,1,module,encoding},
+ {eof,4}],[{encoding,latin1}]} =
+ epp:parse_file(ErlFile, [{default_encoding,latin1},extra]),
+ {ok,[{attribute,1,file,_},
+ {attribute,1,module,encoding},
+ {eof,4}],[{encoding,latin1}]} =
+ epp:parse_file(ErlFile, [{default_encoding,utf8},extra]),
+ ok.
+
+
check(Config, Tests) ->
eval_tests(Config, fun check_test/2, Tests).