[80] | 1 | // |
---|
| 2 | // TextConverter.cpp |
---|
| 3 | // |
---|
| 4 | // $Id: //poco/1.3/Foundation/src/TextConverter.cpp#5 $ |
---|
| 5 | // |
---|
| 6 | // Library: Foundation |
---|
| 7 | // Package: Text |
---|
| 8 | // Module: TextConverter |
---|
| 9 | // |
---|
| 10 | // Copyright (c) 2004-2006, Applied Informatics Software Engineering GmbH. |
---|
| 11 | // and Contributors. |
---|
| 12 | // |
---|
| 13 | // Permission is hereby granted, free of charge, to any person or organization |
---|
| 14 | // obtaining a copy of the software and accompanying documentation covered by |
---|
| 15 | // this license (the "Software") to use, reproduce, display, distribute, |
---|
| 16 | // execute, and transmit the Software, and to prepare derivative works of the |
---|
| 17 | // Software, and to permit third-parties to whom the Software is furnished to |
---|
| 18 | // do so, all subject to the following: |
---|
| 19 | // |
---|
| 20 | // The copyright notices in the Software and this entire statement, including |
---|
| 21 | // the above license grant, this restriction and the following disclaimer, |
---|
| 22 | // must be included in all copies of the Software, in whole or in part, and |
---|
| 23 | // all derivative works of the Software, unless such copies or derivative |
---|
| 24 | // works are solely in the form of machine-executable object code generated by |
---|
| 25 | // a source language processor. |
---|
| 26 | // |
---|
| 27 | // THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR |
---|
| 28 | // IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
---|
| 29 | // FITNESS FOR A PARTICULAR PURPOSE, TITLE AND NON-INFRINGEMENT. IN NO EVENT |
---|
| 30 | // SHALL THE COPYRIGHT HOLDERS OR ANYONE DISTRIBUTING THE SOFTWARE BE LIABLE |
---|
| 31 | // FOR ANY DAMAGES OR OTHER LIABILITY, WHETHER IN CONTRACT, TORT OR OTHERWISE, |
---|
| 32 | // ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER |
---|
| 33 | // DEALINGS IN THE SOFTWARE. |
---|
| 34 | // |
---|
| 35 | |
---|
| 36 | |
---|
| 37 | #include "Poco/TextConverter.h" |
---|
| 38 | #include "Poco/TextIterator.h" |
---|
| 39 | #include "Poco/TextEncoding.h" |
---|
| 40 | |
---|
| 41 | |
---|
| 42 | namespace { |
---|
| 43 | int nullTransform(int ch) |
---|
| 44 | { |
---|
| 45 | return ch; |
---|
| 46 | } |
---|
| 47 | } |
---|
| 48 | |
---|
| 49 | |
---|
| 50 | namespace Poco { |
---|
| 51 | |
---|
| 52 | |
---|
| 53 | TextConverter::TextConverter(const TextEncoding& inEncoding, const TextEncoding& outEncoding, int defaultChar): |
---|
| 54 | _inEncoding(inEncoding), |
---|
| 55 | _outEncoding(outEncoding), |
---|
| 56 | _defaultChar(defaultChar) |
---|
| 57 | { |
---|
| 58 | } |
---|
| 59 | |
---|
| 60 | |
---|
| 61 | TextConverter::~TextConverter() |
---|
| 62 | { |
---|
| 63 | } |
---|
| 64 | |
---|
| 65 | |
---|
| 66 | int TextConverter::convert(const std::string& source, std::string& destination, Transform trans) |
---|
| 67 | { |
---|
| 68 | int errors = 0; |
---|
| 69 | TextIterator it(source, _inEncoding); |
---|
| 70 | TextIterator end(source); |
---|
| 71 | unsigned char buffer[TextEncoding::MAX_SEQUENCE_LENGTH]; |
---|
| 72 | |
---|
| 73 | while (it != end) |
---|
| 74 | { |
---|
| 75 | int c = *it; |
---|
| 76 | if (c == -1) { ++errors; c = _defaultChar; } |
---|
| 77 | c = trans(c); |
---|
| 78 | int n = _outEncoding.convert(c, buffer, sizeof(buffer)); |
---|
| 79 | if (n == 0) n = _outEncoding.convert(_defaultChar, buffer, sizeof(buffer)); |
---|
| 80 | poco_assert (n <= sizeof(buffer)); |
---|
| 81 | destination.append((const char*) buffer, n); |
---|
| 82 | ++it; |
---|
| 83 | } |
---|
| 84 | return errors; |
---|
| 85 | } |
---|
| 86 | |
---|
| 87 | |
---|
| 88 | int TextConverter::convert(const void* source, int length, std::string& destination, Transform trans) |
---|
| 89 | { |
---|
| 90 | poco_check_ptr (source); |
---|
| 91 | |
---|
| 92 | int errors = 0; |
---|
| 93 | const unsigned char* it = (const unsigned char*) source; |
---|
| 94 | const unsigned char* end = (const unsigned char*) source + length; |
---|
| 95 | unsigned char buffer[TextEncoding::MAX_SEQUENCE_LENGTH]; |
---|
| 96 | |
---|
| 97 | while (it < end) |
---|
| 98 | { |
---|
| 99 | int n = _inEncoding.queryConvert(it, 1); |
---|
| 100 | int uc; |
---|
| 101 | int read = 1; |
---|
| 102 | |
---|
| 103 | while (-1 > n && (end - it) >= -n) |
---|
| 104 | { |
---|
| 105 | read = -n; |
---|
| 106 | n = _inEncoding.queryConvert(it, read); |
---|
| 107 | } |
---|
| 108 | |
---|
| 109 | if (-1 > n) |
---|
| 110 | { |
---|
| 111 | it = end; |
---|
| 112 | } |
---|
| 113 | else |
---|
| 114 | { |
---|
| 115 | it += read; |
---|
| 116 | } |
---|
| 117 | |
---|
| 118 | if (-1 >= n) |
---|
| 119 | { |
---|
| 120 | uc = _defaultChar; |
---|
| 121 | ++errors; |
---|
| 122 | } |
---|
| 123 | else |
---|
| 124 | { |
---|
| 125 | uc = n; |
---|
| 126 | } |
---|
| 127 | |
---|
| 128 | uc = trans(uc); |
---|
| 129 | n = _outEncoding.convert(uc, buffer, sizeof(buffer)); |
---|
| 130 | if (n == 0) n = _outEncoding.convert(_defaultChar, buffer, sizeof(buffer)); |
---|
| 131 | poco_assert (n <= sizeof(buffer)); |
---|
| 132 | destination.append((const char*) buffer, n); |
---|
| 133 | } |
---|
| 134 | return errors; |
---|
| 135 | } |
---|
| 136 | |
---|
| 137 | |
---|
| 138 | int TextConverter::convert(const std::string& source, std::string& destination) |
---|
| 139 | { |
---|
| 140 | return convert(source, destination, nullTransform); |
---|
| 141 | } |
---|
| 142 | |
---|
| 143 | |
---|
| 144 | int TextConverter::convert(const void* source, int length, std::string& destination) |
---|
| 145 | { |
---|
| 146 | return convert(source, length, destination, nullTransform); |
---|
| 147 | } |
---|
| 148 | |
---|
| 149 | |
---|
| 150 | } // namespace Poco |
---|