da2797702f
Added test for UTF-7 encoding availability. Added test for input buffer underflow in charsetFilteredOutputStream. Refactored charset conversion tests and removed useless tests.
149 lines
4.8 KiB
C++
149 lines
4.8 KiB
C++
//
|
|
// VMime library (http://www.vmime.org)
|
|
// Copyright (C) 2002-2013 Vincent Richard <vincent@vmime.org>
|
|
//
|
|
// 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 3 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, write to the Free Software Foundation, Inc.,
|
|
// 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
|
|
//
|
|
// Linking this library statically or dynamically with other modules is making
|
|
// a combined work based on this library. Thus, the terms and conditions of
|
|
// the GNU General Public License cover the whole combination.
|
|
//
|
|
|
|
#include "tests/testUtils.hpp"
|
|
|
|
#include "charsetTestSuites.hpp"
|
|
|
|
|
|
VMIME_TEST_SUITE_BEGIN(charsetTest)
|
|
|
|
VMIME_TEST_LIST_BEGIN
|
|
// Test valid input
|
|
VMIME_TEST(testConvertStringValid)
|
|
VMIME_TEST(testConvertStreamValid)
|
|
VMIME_TEST(testEncodingHebrew1255)
|
|
|
|
// IDNA
|
|
VMIME_TEST(testEncodeIDNA)
|
|
VMIME_TEST(testDecodeIDNA)
|
|
|
|
VMIME_TEST(testUTF7Support)
|
|
VMIME_TEST_LIST_END
|
|
|
|
|
|
void testConvertStringValid()
|
|
{
|
|
for (unsigned int i = 0 ; i < charsetTestSuitesCount ; ++i)
|
|
{
|
|
const charsetTestSuiteStruct& entry = charsetTestSuites[i];
|
|
|
|
std::ostringstream testName;
|
|
testName << i << ": " << entry.fromCharset << " -> " << entry.toCharset;
|
|
|
|
const unsigned int inLength = (entry.fromLength == 0 ? strlen(entry.fromBytes) : entry.fromLength);
|
|
vmime::string in(entry.fromBytes, entry.fromBytes + inLength);
|
|
|
|
const unsigned int outLength = (entry.toLength == 0 ? strlen(entry.toBytes) : entry.toLength);
|
|
vmime::string expectedOut(entry.toBytes, entry.toBytes + outLength);
|
|
|
|
vmime::string actualOut;
|
|
|
|
vmime::charset::convert
|
|
(in, actualOut, entry.fromCharset, entry.toCharset);
|
|
|
|
VASSERT_EQ(testName.str(), toHex(expectedOut), toHex(actualOut));
|
|
}
|
|
}
|
|
|
|
void testConvertStreamValid()
|
|
{
|
|
for (unsigned int i = 0 ; i < charsetTestSuitesCount ; ++i)
|
|
{
|
|
const charsetTestSuiteStruct& entry = charsetTestSuites[i];
|
|
|
|
std::ostringstream testName;
|
|
testName << i << ": " << entry.fromCharset << " -> " << entry.toCharset;
|
|
|
|
const unsigned int inLength = (entry.fromLength == 0 ? strlen(entry.fromBytes) : entry.fromLength);
|
|
vmime::string in(entry.fromBytes, entry.fromBytes + inLength);
|
|
|
|
const unsigned int outLength = (entry.toLength == 0 ? strlen(entry.toBytes) : entry.toLength);
|
|
vmime::string expectedOut(entry.toBytes, entry.toBytes + outLength);
|
|
|
|
vmime::string actualOut;
|
|
vmime::utility::outputStreamStringAdapter os(actualOut);
|
|
|
|
vmime::utility::inputStreamStringAdapter is(in);
|
|
|
|
vmime::charset::convert
|
|
(is, os, entry.fromCharset, entry.toCharset);
|
|
|
|
os.flush();
|
|
|
|
VASSERT_EQ(testName.str(), toHex(expectedOut), toHex(actualOut));
|
|
}
|
|
}
|
|
|
|
void testEncodingHebrew1255()
|
|
{
|
|
// hewbrew string in windows-1255 charset
|
|
const char data[] = "\xe9\xf9\xf7\xf8\xe9\xf9\xf8\xf7\xe9\xe9\xf9";
|
|
vmime::word w = vmime::word(data, "windows-1255");
|
|
vmime::string encoded = w.generate();
|
|
// less than 60% ascii, base64 received
|
|
VASSERT_EQ("1", "=?windows-1255?B?6fn3+On5+Pfp6fk=?=", encoded);
|
|
}
|
|
|
|
static const vmime::string convertHelper
|
|
(const vmime::string& in, const vmime::charset& csrc, const vmime::charset& cdest)
|
|
{
|
|
vmime::string out;
|
|
vmime::charset::convert(in, out, csrc, cdest);
|
|
|
|
return out;
|
|
}
|
|
|
|
void testEncodeIDNA()
|
|
{
|
|
VASSERT_EQ("1", "xn--espaol-zwa", convertHelper("español", "utf-8", "idna"));
|
|
|
|
// Tests from ICANN
|
|
VASSERT_EQ("2.1", "xn--hxajbheg2az3al", convertHelper("παράδειγμα", "utf-8", "idna"));
|
|
VASSERT_EQ("2.2", "xn--jxalpdlp", convertHelper("δοκιμή", "utf-8", "idna"));
|
|
|
|
VASSERT_EQ("3.1", "xn--mgbh0fb", convertHelper("مثال", "utf-8", "idna"));
|
|
VASSERT_EQ("3.2", "xn--kgbechtv", convertHelper("إختبار", "utf-8", "idna"));
|
|
}
|
|
|
|
void testDecodeIDNA()
|
|
{
|
|
VASSERT_EQ("1", "español", convertHelper("xn--espaol-zwa", "idna", "utf-8"));
|
|
|
|
// Tests from ICANN
|
|
VASSERT_EQ("2.1", "παράδειγμα", convertHelper("xn--hxajbheg2az3al", "idna", "utf-8"));
|
|
VASSERT_EQ("2.2", "δοκιμή", convertHelper("xn--jxalpdlp", "idna", "utf-8"));
|
|
|
|
VASSERT_EQ("3.1", "مثال", convertHelper("xn--mgbh0fb", "idna", "utf-8"));
|
|
VASSERT_EQ("3.2", "إختبار", convertHelper("xn--kgbechtv", "idna", "utf-8"));
|
|
}
|
|
|
|
void testUTF7Support()
|
|
{
|
|
// Ensure UTF-7 is supported, because it is used for IMAP
|
|
VASSERT_EQ("1", "VMime +ACY UTF-7 encoding", convertHelper("VMime & UTF-7 encoding", "utf-8", "utf-7"));
|
|
}
|
|
|
|
VMIME_TEST_SUITE_END
|
|
|