kicad/qa/common/test_utf8.cpp

143 lines
3.6 KiB
C++

/*
* This program source code file is part of KiCad, a free EDA CAD application.
*
* Copyright (C) 2018 KiCad Developers, see CHANGELOG.TXT for contributors.
*
* This program 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 2
* of the License, or (at your option) any later version.
*
* This program 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 program; if not, you may find one here:
* http://www.gnu.org/licenses/old-licenses/gpl-2.0.html
* or you may search the http://www.gnu.org website for the version 2 license,
* or you may write to the Free Software Foundation, Inc.,
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
*/
#include <boost/test/unit_test.hpp>
#include <boost/test/test_case_template.hpp>
#include <utf8.h>
#include <algorithm>
struct Utf8Fixture
{
};
/**
* Declares a struct as the Boost test fixture.
*/
BOOST_FIXTURE_TEST_SUITE( Utf8, Utf8Fixture )
/**
* Check direct and copy construction from std::string
*/
BOOST_AUTO_TEST_CASE( Utf8AndStdString )
{
std::string str { "input" };
UTF8 utf8_inited { "input" };
UTF8 utf8_copied_from_stdstr = str;
BOOST_CHECK_EQUAL( utf8_inited, utf8_copied_from_stdstr );
UTF8 utf8_copied_from_utf8 = utf8_inited;
BOOST_CHECK_EQUAL( utf8_inited, utf8_copied_from_utf8 );
}
/**
* Check direct and copy construction from wxString
*/
BOOST_AUTO_TEST_CASE( Utf8AndWx )
{
UTF8 utf8_inited { "input" };
wxString wx_inited { "input" };
// Check that we can copy convert WxString and compare
wxString wx_copied_from_utf8 = utf8_inited;
BOOST_CHECK_EQUAL( wx_inited, wx_copied_from_utf8 );
// Check we can copy-construct from a WxString
UTF8 utf8_copied_from_wxstring = wx_inited;
BOOST_CHECK_EQUAL( utf8_inited, utf8_copied_from_wxstring );
}
/**
* UTF8::uni_iter null tests
*/
BOOST_AUTO_TEST_CASE( UniIterNull )
{
UTF8::uni_iter it;
const UTF8::uni_iter null;
// Check nulls are equivalent
BOOST_CHECK( it == null );
// check null string start == end
UTF8 uNull { "" };
BOOST_CHECK( uNull.ubegin() == uNull.uend() );
}
/**
* UTF8::uni_iter increment tests
*/
BOOST_AUTO_TEST_CASE( UniIterIncrement )
{
UTF8 u0 { "inp\ua740t" };
UTF8::uni_iter it;
const UTF8::uni_iter begin = u0.ubegin();
const UTF8::uni_iter end = u0.uend();
// Check copy-construction and equality operator
it = begin;
BOOST_CHECK( it == begin );
BOOST_CHECK( it >= begin );
// Check de-referencing
BOOST_CHECK_EQUAL( *it, 'i' );
// postfix increment - normal char and inequality operators
it++;
BOOST_CHECK( it != begin );
BOOST_CHECK( it > begin );
BOOST_CHECK( !( begin >= it ) );
BOOST_CHECK( it < end );
BOOST_CHECK( it <= end );
BOOST_CHECK_EQUAL( *it, 'n' );
// prefix increment - normal char
++it;
BOOST_CHECK_EQUAL( *it, 'p' );
// increment to a unicode char
++it;
BOOST_CHECK_EQUAL( *it, 0xA740 );
// and again to the next char - normal again
++it;
BOOST_CHECK_EQUAL( *it, 't' );
// and now we should be at the end
++it;
BOOST_CHECK( it == end );
BOOST_CHECK( it <= end );
BOOST_CHECK( !( it < end ) );
}
BOOST_AUTO_TEST_SUITE_END()