From 300c5466a7c9cfe3ed22bba2a88ba21058406402 Mon Sep 17 00:00:00 2001
From: Hans Bolinder <hasse@erlang.org>
Date: Thu, 4 Oct 2012 15:58:26 +0200
Subject: [stdlib, kernel] Introduce Unicode support for Erlang source files

Expect modifications, additions and corrections.

There is a kludge in file_io_server and
erl_scan:continuation_location() that's not so pleasing.
---
 lib/stdlib/test/epp_SUITE.erl | 82 +++++++++++++++++++++++++++++++++++++++----
 1 file changed, 76 insertions(+), 6 deletions(-)

(limited to 'lib/stdlib/test/epp_SUITE.erl')

diff --git a/lib/stdlib/test/epp_SUITE.erl b/lib/stdlib/test/epp_SUITE.erl
index 77c615d6d9..606bbbcbb2 100644
--- a/lib/stdlib/test/epp_SUITE.erl
+++ b/lib/stdlib/test/epp_SUITE.erl
@@ -25,7 +25,7 @@
 	 variable_1/1, otp_4870/1, otp_4871/1, otp_5362/1,
          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_8562/1, otp_8665/1, otp_8911/1, otp_10302/1]).
 
 -export([epp_parse_erl_form/2]).
 
@@ -67,7 +67,7 @@ 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_8665, otp_8911, otp_10302].
 
 groups() -> 
     [{upcase_mac, [], [upcase_mac_1, upcase_mac_2]},
@@ -582,12 +582,13 @@ otp_8130(suite) ->
 otp_8130(Config) when is_list(Config) ->
     true = os:putenv("epp_inc1", "stdlib"),
     Ts = [{otp_8130_1,
-           %% The scanner handles UNICODE in a special way. Hopefully
-           %% temporarily.
            <<"-define(M(A), ??A). "
              "t() ->  "
-             "   \"{ 34 , [ $1 , 2730 ] , \\\"34\\\" , X . a , 2730 }\" = "
-             "        ?M({34,\"1\\x{aaa}\",\"34\",X.a,$\\x{aaa}}), ok. ">>,
+             "   L = \"{ 34 , \\\"1\\\\x{AAA}\\\" , \\\"34\\\" , X . a , $\\\\x{AAA} }\", "
+             "   R = ?M({34,\"1\\x{aaa}\",\"34\",X.a,$\\x{aaa}}),"
+             "   Lt = erl_scan:string(L, 1, [unicode]),"
+             "   Rt = erl_scan:string(R, 1, [unicode]),"
+             "   Lt = Rt, ok. ">>,
           ok},
 
           {otp_8130_2,
@@ -1284,6 +1285,75 @@ otp_8665(Config) when is_list(Config) ->
     ?line [] = compile(Config, Cs),
     ok.
 
+otp_10302(doc) ->
+    "OTP-10302. Unicode characters scanner/parser.";
+otp_10302(suite) ->
+    [];
+otp_10302(Config) when is_list(Config) ->
+    %% Two messages (one too many). Keeps otp_4871 happy.
+    Cs = [{otp_8562,
+           <<"%% coding: utf-8\n \n \x{E4}">>,
+           {errors,[{3,epp,cannot_parse},
+                    {3,file_io_server,invalid_unicode}],[]}}
+         ],
+    [] = compile(Config, Cs),
+    Dir = ?config(priv_dir, Config),
+    File = filename:join(Dir, "otp_10302.erl"),
+    utf8 = encoding("coding: utf-8", File),
+    utf8 = encoding("coding: UTF-8", File),
+    latin1 = encoding("coding: Latin-1", File),
+    latin1 = encoding("coding: latin-1", File),
+    none = encoding_com("coding: utf-8", File),
+    none = encoding_com("\n\n%% coding: utf-8", File),
+    none = encoding_nocom("\n\n coding: utf-8", File),
+    utf8 = encoding_com("\n%% coding: utf-8", File),
+    utf8 = encoding_nocom("\n coding: utf-8", File),
+    none = encoding("coding: \nutf-8", File),
+    latin1 = encoding("Encoding :  latin-1", File),
+    utf8 = encoding("ccoding: UTF-8", File),
+    utf8 = encoding("coding= utf-8", File),
+    utf8 = encoding_com(" %% coding= utf-8", File),
+    utf8 = encoding("coding  =  utf-8", File),
+    none = encoding("coding: utf-16 coding: utf-8", File), %first is bad
+    none = encoding("Coding: utf-8", File),    %capital c
+    utf8 = encoding("-*- coding: utf-8 -*-", File),
+    utf8 = encoding("-*-coding= utf-8-*-", File),
+    utf8 = encoding("codingcoding= utf-8", File),
+    ok = prefix("coding: utf-8", File, utf8),
+
+    "coding: latin-1" = epp:encoding_to_string(latin1),
+    "coding: utf-8" = epp:encoding_to_string(utf8),
+    true = lists:member(epp:default_encoding(), [latin1, utf8]),
+
+    ok.
+
+prefix(S, File, Enc) ->
+    prefix(0, S, File, Enc).
+
+prefix(100, _S, _File, _) ->
+    ok;
+prefix(N, S, File, Enc) ->
+    Enc = encoding(lists:duplicate(N, $\s) ++ S, File),
+    prefix(N+1, S, File, Enc).
+
+encoding(Enc, File) ->
+    E = encoding_com("%% " ++ Enc, File),
+    none = encoding_com(Enc, File),
+    E = encoding_nocom(Enc, File).
+
+encoding_com(Enc, File) ->
+    ok = file:write_file(File, Enc),
+    {ok, Fd} = file:open(File, [read]),
+    E = epp:set_encoding(Fd),
+    ok = file:close(Fd),
+    E = epp:read_encoding(File).
+
+encoding_nocom(Enc, File) ->
+    ok = file:write_file(File, Enc),
+    {ok, Fd} = file:open(File, [read]),
+    ok = file:close(Fd),
+    epp:read_encoding(File, [{in_comment_only, false}]).
+
 check(Config, Tests) ->
     eval_tests(Config, fun check_test/2, Tests).
 
-- 
cgit v1.2.3