From 516231de7317bbaf0f83a27047924bb690a217e5 Mon Sep 17 00:00:00 2001 From: Jonathan Wakely Date: Thu, 16 Mar 2017 15:27:57 +0000 Subject: [PATCH] Fix encoding() and max_length() values for codecvt facets * src/c++11/codecvt.cc (codecvt) (codecvt, __codecvt_utf8_base) (__codecvt_utf8_base, __codecvt_utf8_base) (__codecvt_utf16_base, __codecvt_utf16_base) (__codecvt_utf16_base, __codecvt_utf8_utf16_base) (__codecvt_utf8_utf16_base) (__codecvt_utf8_utf16_base): Fix do_encoding() and do_max_length() return values. * testsuite/22_locale/codecvt/codecvt_utf16/members.cc: New test. * testsuite/22_locale/codecvt/codecvt_utf8/members.cc: New test. * testsuite/22_locale/codecvt/codecvt_utf8_utf16/members.cc: New test. From-SVN: r246201 --- libstdc++-v3/ChangeLog | 12 ++ libstdc++-v3/src/c++11/codecvt.cc | 122 +++++++++++++----- .../testsuite/22_locale/codecvt/char16_t.cc | 2 +- .../codecvt/codecvt_utf16/members.cc | 81 ++++++++++++ .../22_locale/codecvt/codecvt_utf8/members.cc | 81 ++++++++++++ .../codecvt/codecvt_utf8_utf16/members.cc | 76 +++++++++++ 6 files changed, 341 insertions(+), 33 deletions(-) create mode 100644 libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf16/members.cc create mode 100644 libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf8/members.cc create mode 100644 libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf8_utf16/members.cc diff --git a/libstdc++-v3/ChangeLog b/libstdc++-v3/ChangeLog index 83f74efa390..9facce84994 100644 --- a/libstdc++-v3/ChangeLog +++ b/libstdc++-v3/ChangeLog @@ -1,5 +1,17 @@ 2017-03-16 Jonathan Wakely + * src/c++11/codecvt.cc (codecvt) + (codecvt, __codecvt_utf8_base) + (__codecvt_utf8_base, __codecvt_utf8_base) + (__codecvt_utf16_base, __codecvt_utf16_base) + (__codecvt_utf16_base, __codecvt_utf8_utf16_base) + (__codecvt_utf8_utf16_base) + (__codecvt_utf8_utf16_base): Fix do_encoding() and + do_max_length() return values. + * testsuite/22_locale/codecvt/codecvt_utf16/members.cc: New test. + * testsuite/22_locale/codecvt/codecvt_utf8/members.cc: New test. + * testsuite/22_locale/codecvt/codecvt_utf8_utf16/members.cc: New test. + PR libstdc++/79980 * include/bits/locale_conv.h (__do_str_codecvt): Set __count on error path. diff --git a/libstdc++-v3/src/c++11/codecvt.cc b/libstdc++-v3/src/c++11/codecvt.cc index a50804cc164..9c917257567 100644 --- a/libstdc++-v3/src/c++11/codecvt.cc +++ b/libstdc++-v3/src/c++11/codecvt.cc @@ -72,8 +72,8 @@ namespace // Multibyte sequences can have "header" consisting of Byte Order Mark const unsigned char utf8_bom[3] = { 0xEF, 0xBB, 0xBF }; - const unsigned char utf16_bom[4] = { 0xFE, 0xFF }; - const unsigned char utf16le_bom[4] = { 0xFF, 0xFE }; + const unsigned char utf16_bom[2] = { 0xFE, 0xFF }; + const unsigned char utf16le_bom[2] = { 0xFF, 0xFE }; template inline bool @@ -695,7 +695,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int codecvt::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-8 is not a fixed-width encoding bool codecvt::do_always_noconv() const throw() @@ -713,9 +713,9 @@ do_length(state_type&, const extern_type* __from, int codecvt::do_max_length() const throw() { - // Any valid UTF-8 sequence of 3 bytes fits in a single 16-bit code unit, - // whereas 4 byte sequences require two 16-bit code units. - return 3; + // A single character (one or two UTF-16 code units) requires + // up to four UTF-8 code units. + return 4; } // Define members of codecvt specialization. @@ -766,7 +766,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int codecvt::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-8 is not a fixed-width encoding bool codecvt::do_always_noconv() const throw() @@ -783,7 +783,11 @@ do_length(state_type&, const extern_type* __from, int codecvt::do_max_length() const throw() -{ return 4; } +{ + // A single character (one UTF-32 code unit) requires + // up to 4 UTF-8 code units. + return 4; +} // Define members of codecvt_utf8 base class implementation. // Converts from UTF-8 to UCS-2. @@ -835,7 +839,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int __codecvt_utf8_base::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-8 is not a fixed-width encoding bool __codecvt_utf8_base::do_always_noconv() const throw() @@ -852,7 +856,14 @@ do_length(state_type&, const extern_type* __from, int __codecvt_utf8_base::do_max_length() const throw() -{ return 3; } +{ + // A single UCS-2 character requires up to three UTF-8 code units. + // (UCS-2 cannot represent characters that use four UTF-8 code units). + int max = 3; + if (_M_mode & consume_header) + max += sizeof(utf8_bom); + return max; +} // Define members of codecvt_utf8 base class implementation. // Converts from UTF-8 to UTF-32 (aka UCS-4). @@ -900,7 +911,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int __codecvt_utf8_base::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-8 is not a fixed-width encoding bool __codecvt_utf8_base::do_always_noconv() const throw() @@ -917,7 +928,13 @@ do_length(state_type&, const extern_type* __from, int __codecvt_utf8_base::do_max_length() const throw() -{ return 4; } +{ + // A single UCS-4 character requires up to four UTF-8 code units. + int max = 4; + if (_M_mode & consume_header) + max += sizeof(utf8_bom); + return max; +} #ifdef _GLIBCXX_USE_WCHAR_T // Define members of codecvt_utf8 base class implementation. @@ -992,7 +1009,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int __codecvt_utf8_base::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-8 is not a fixed-width encoding bool __codecvt_utf8_base::do_always_noconv() const throw() @@ -1015,7 +1032,16 @@ do_length(state_type&, const extern_type* __from, int __codecvt_utf8_base::do_max_length() const throw() -{ return 4; } +{ +#if __SIZEOF_WCHAR_T__ == 2 + int max = 3; // See __codecvt_utf8_base::do_max_length() +#else + int max = 4; // See __codecvt_utf8_base::do_max_length() +#endif + if (_M_mode & consume_header) + max += sizeof(utf8_bom); + return max; +} #endif // Define members of codecvt_utf16 base class implementation. @@ -1070,7 +1096,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int __codecvt_utf16_base::do_encoding() const throw() -{ return 1; } +{ return 0; } // UTF-16 is not a fixed-width encoding bool __codecvt_utf16_base::do_always_noconv() const throw() @@ -1089,7 +1115,14 @@ do_length(state_type&, const extern_type* __from, int __codecvt_utf16_base::do_max_length() const throw() -{ return 3; } +{ + // A single UCS-2 character requires one UTF-16 code unit (so two chars). + // (UCS-2 cannot represent characters that use multiple UTF-16 code units). + int max = 2; + if (_M_mode & consume_header) + max += sizeof(utf16_bom); + return max; +} // Define members of codecvt_utf16 base class implementation. // Converts from UTF-16 to UTF-32 (aka UCS-4). @@ -1143,7 +1176,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int __codecvt_utf16_base::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-16 is not a fixed-width encoding bool __codecvt_utf16_base::do_always_noconv() const throw() @@ -1162,7 +1195,14 @@ do_length(state_type&, const extern_type* __from, int __codecvt_utf16_base::do_max_length() const throw() -{ return 4; } +{ + // A single UCS-4 character requires one or two UTF-16 code units + // (so up to four chars). + int max = 4; + if (_M_mode & consume_header) + max += sizeof(utf16_bom); + return max; +} #ifdef _GLIBCXX_USE_WCHAR_T // Define members of codecvt_utf16 base class implementation. @@ -1237,7 +1277,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int __codecvt_utf16_base::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-16 is not a fixed-width encoding bool __codecvt_utf16_base::do_always_noconv() const throw() @@ -1261,7 +1301,16 @@ do_length(state_type&, const extern_type* __from, int __codecvt_utf16_base::do_max_length() const throw() -{ return 4; } +{ +#if __SIZEOF_WCHAR_T__ == 2 + int max = 2; // See __codecvt_utf16_base::do_max_length() +#else + int max = 4; // See __codecvt_utf16_base::do_max_length() +#endif + if (_M_mode & consume_header) + max += sizeof(utf16_bom); + return max; +} #endif // Define members of codecvt_utf8_utf16 base class implementation. @@ -1314,7 +1363,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int __codecvt_utf8_utf16_base::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-8 is not a fixed-width encoding bool __codecvt_utf8_utf16_base::do_always_noconv() const throw() @@ -1332,9 +1381,12 @@ do_length(state_type&, const extern_type* __from, int __codecvt_utf8_utf16_base::do_max_length() const throw() { - // Any valid UTF-8 sequence of 3 bytes fits in a single 16-bit code unit, - // whereas 4 byte sequences require two 16-bit code units. - return 3; + // A single character can be 1 or 2 UTF-16 code units, + // requiring up to 4 UTF-8 code units. + int max = 4; + if (_M_mode & consume_header) + max += sizeof(utf8_bom); + return max; } // Define members of codecvt_utf8_utf16 base class implementation. @@ -1387,7 +1439,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int __codecvt_utf8_utf16_base::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-8 is not a fixed-width encoding bool __codecvt_utf8_utf16_base::do_always_noconv() const throw() @@ -1405,9 +1457,12 @@ do_length(state_type&, const extern_type* __from, int __codecvt_utf8_utf16_base::do_max_length() const throw() { - // Any valid UTF-8 sequence of 3 bytes fits in a single 16-bit code unit, - // whereas 4 byte sequences require two 16-bit code units. - return 3; + // A single character can be 1 or 2 UTF-16 code units, + // requiring up to 4 UTF-8 code units. + int max = 4; + if (_M_mode & consume_header) + max += sizeof(utf8_bom); + return max; } #ifdef _GLIBCXX_USE_WCHAR_T @@ -1461,7 +1516,7 @@ do_in(state_type&, const extern_type* __from, const extern_type* __from_end, int __codecvt_utf8_utf16_base::do_encoding() const throw() -{ return 0; } +{ return 0; } // UTF-8 is not a fixed-width encoding bool __codecvt_utf8_utf16_base::do_always_noconv() const throw() @@ -1479,9 +1534,12 @@ do_length(state_type&, const extern_type* __from, int __codecvt_utf8_utf16_base::do_max_length() const throw() { - // Any valid UTF-8 sequence of 3 bytes fits in a single 16-bit code unit, - // whereas 4 byte sequences require two 16-bit code units. - return 3; + // A single character can be 1 or 2 UTF-16 code units, + // requiring up to 4 UTF-8 code units. + int max = 4; + if (_M_mode & consume_header) + max += sizeof(utf8_bom); + return max; } #endif diff --git a/libstdc++-v3/testsuite/22_locale/codecvt/char16_t.cc b/libstdc++-v3/testsuite/22_locale/codecvt/char16_t.cc index b40fc655a3a..3288e77c514 100644 --- a/libstdc++-v3/testsuite/22_locale/codecvt/char16_t.cc +++ b/libstdc++-v3/testsuite/22_locale/codecvt/char16_t.cc @@ -34,7 +34,7 @@ test01() const codecvt_c16* const cvt = &use_facet(loc_c); VERIFY(!cvt->always_noconv()); - VERIFY(cvt->max_length() == 3); + VERIFY(cvt->max_length() == 4); VERIFY(cvt->encoding() == 0); const char u8dat[] = u8"H\U000000E4ll\U000000F6 \U0001F63F \U000056FD " diff --git a/libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf16/members.cc b/libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf16/members.cc new file mode 100644 index 00000000000..993c86082c1 --- /dev/null +++ b/libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf16/members.cc @@ -0,0 +1,81 @@ +// Copyright (C) 2017 Free Software Foundation, Inc. +// +// This file is part of the GNU ISO C++ Library. This library is free +// software; you can redistribute it and/or modify it under the +// terms of the GNU General Public License as published by the +// Free Software Foundation; either version 3, or (at your option) +// any later version. + +// This library is distributed in the hope that it will be useful, +// but WITHOUT ANY WARRANTY; without even the implied warranty of +// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +// GNU General Public License for more details. + +// You should have received a copy of the GNU General Public License along +// with this library; see the file COPYING3. If not see +// . + +// { dg-do run { target c++11 } } + +#include +#include + +const int bomlen = 2; // UTF-16 BOM is 16 bits + +void +test01() +{ + const int maxlen = 2; + + std::codecvt_utf16 c; + VERIFY( c.always_noconv() == false ); + VERIFY( c.encoding() == 0 ); + VERIFY( c.max_length() == maxlen ); + + std::codecvt_utf16 c_bom; + VERIFY( c_bom.always_noconv() == false ); + VERIFY( c_bom.encoding() == 0 ); + VERIFY( c_bom.max_length() == (maxlen + bomlen) ); +} + +void +test02() +{ + const int maxlen = 4; + + std::codecvt_utf16 c; + VERIFY( c.always_noconv() == false ); + VERIFY( c.encoding() == 0 ); + VERIFY( c.max_length() == maxlen ); + + std::codecvt_utf16 c_bom; + VERIFY( c_bom.always_noconv() == false ); + VERIFY( c_bom.encoding() == 0 ); + VERIFY( c_bom.max_length() == (maxlen + bomlen) ); +} + +void +test03() +{ +#ifdef _GLIBCXX_USE_WCHAR_T + const int maxlen = sizeof(wchar_t) == 4 ? 4 : 2; + + std::codecvt_utf16 c; + VERIFY( c.always_noconv() == false ); + VERIFY( c.encoding() == 0 ); + VERIFY( c.max_length() == maxlen ); + + std::codecvt_utf16 c_bom; + VERIFY( c_bom.always_noconv() == false ); + VERIFY( c_bom.encoding() == 0 ); + VERIFY( c_bom.max_length() == (maxlen + bomlen) ); +#endif +} + +int +main() +{ + test01(); + test02(); + test03(); +} diff --git a/libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf8/members.cc b/libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf8/members.cc new file mode 100644 index 00000000000..baeb049861a --- /dev/null +++ b/libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf8/members.cc @@ -0,0 +1,81 @@ +// Copyright (C) 2017 Free Software Foundation, Inc. +// +// This file is part of the GNU ISO C++ Library. This library is free +// software; you can redistribute it and/or modify it under the +// terms of the GNU General Public License as published by the +// Free Software Foundation; either version 3, or (at your option) +// any later version. + +// This library is distributed in the hope that it will be useful, +// but WITHOUT ANY WARRANTY; without even the implied warranty of +// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +// GNU General Public License for more details. + +// You should have received a copy of the GNU General Public License along +// with this library; see the file COPYING3. If not see +// . + +// { dg-do run { target c++11 } } + +#include +#include + +const int bomlen = 3; // UTF-8 BOM is 24 bits + +void +test01() +{ + const int maxlen = 3; + + std::codecvt_utf8 c; + VERIFY( c.always_noconv() == false ); + VERIFY( c.encoding() == 0 ); + VERIFY( c.max_length() == maxlen ); + + std::codecvt_utf8 c_bom; + VERIFY( c_bom.always_noconv() == false ); + VERIFY( c_bom.encoding() == 0 ); + VERIFY( c_bom.max_length() == (maxlen + bomlen) ); +} + +void +test02() +{ + const int maxlen = 4; + + std::codecvt_utf8 c; + VERIFY( c.always_noconv() == false ); + VERIFY( c.encoding() == 0 ); + VERIFY( c.max_length() == maxlen ); + + std::codecvt_utf8 c_bom; + VERIFY( c_bom.always_noconv() == false ); + VERIFY( c_bom.encoding() == 0 ); + VERIFY( c_bom.max_length() == (maxlen + bomlen) ); +} + +void +test03() +{ +#ifdef _GLIBCXX_USE_WCHAR_T + const int maxlen = sizeof(wchar_t) == 4 ? 4 : 3; + + std::codecvt_utf8 c; + VERIFY( c.always_noconv() == false ); + VERIFY( c.encoding() == 0 ); + VERIFY( c.max_length() == maxlen ); + + std::codecvt_utf8 c_bom; + VERIFY( c_bom.always_noconv() == false ); + VERIFY( c_bom.encoding() == 0 ); + VERIFY( c_bom.max_length() == (maxlen + bomlen) ); +#endif +} + +int +main() +{ + test01(); + test02(); + test03(); +} diff --git a/libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf8_utf16/members.cc b/libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf8_utf16/members.cc new file mode 100644 index 00000000000..8fcdfff2cc1 --- /dev/null +++ b/libstdc++-v3/testsuite/22_locale/codecvt/codecvt_utf8_utf16/members.cc @@ -0,0 +1,76 @@ +// Copyright (C) 2017 Free Software Foundation, Inc. +// +// This file is part of the GNU ISO C++ Library. This library is free +// software; you can redistribute it and/or modify it under the +// terms of the GNU General Public License as published by the +// Free Software Foundation; either version 3, or (at your option) +// any later version. + +// This library is distributed in the hope that it will be useful, +// but WITHOUT ANY WARRANTY; without even the implied warranty of +// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +// GNU General Public License for more details. + +// You should have received a copy of the GNU General Public License along +// with this library; see the file COPYING3. If not see +// . + +// { dg-do run { target c++11 } } + +#include +#include + +const int bomlen = 3; // UTF-8 BOM is 24 bits +const int maxlen = 4; + +void +test01() +{ + std::codecvt_utf8_utf16 c; + VERIFY( c.always_noconv() == false ); + VERIFY( c.encoding() == 0 ); + VERIFY( c.max_length() == maxlen ); + + std::codecvt_utf8_utf16 c_bom; + VERIFY( c_bom.always_noconv() == false ); + VERIFY( c_bom.encoding() == 0 ); + VERIFY( c_bom.max_length() == (maxlen + bomlen) ); +} + +void +test02() +{ + std::codecvt_utf8_utf16 c; + VERIFY( c.always_noconv() == false ); + VERIFY( c.encoding() == 0 ); + VERIFY( c.max_length() == maxlen ); + + std::codecvt_utf8_utf16 c_bom; + VERIFY( c_bom.always_noconv() == false ); + VERIFY( c_bom.encoding() == 0 ); + VERIFY( c_bom.max_length() == (maxlen + bomlen) ); +} + +void +test03() +{ +#ifdef _GLIBCXX_USE_WCHAR_T + std::codecvt_utf8_utf16 c; + VERIFY( c.always_noconv() == false ); + VERIFY( c.encoding() == 0 ); + VERIFY( c.max_length() == maxlen ); + + std::codecvt_utf8_utf16 c_bom; + VERIFY( c_bom.always_noconv() == false ); + VERIFY( c_bom.encoding() == 0 ); + VERIFY( c_bom.max_length() == (maxlen + bomlen) ); +#endif +} + +int +main() +{ + test01(); + test02(); + test03(); +} -- 2.30.2