this repo has no description
Add vim into build, add extra APIs into CoreServices (#314) - sadly vim doesn't run yet
+388
-1
+1
src/CMakeLists.txt
+1
src/CMakeLists.txt
+2
-1
src/CoreServices/CMakeLists.txt
+2
-1
src/CoreServices/CMakeLists.txt
+2
src/CoreServices/CoreServices.h
+2
src/CoreServices/CoreServices.h
+31
src/CoreServices/TextCommon.cpp
+31
src/CoreServices/TextCommon.cpp
···
1
+
/*
2
+
This file is part of Darling.
3
+
4
+
Copyright (C) 2017 Lubos Dolezel
5
+
6
+
Darling is free software: you can redistribute it and/or modify
7
+
it under the terms of the GNU General Public License as published by
8
+
the Free Software Foundation, either version 3 of the License, or
9
+
(at your option) any later version.
10
+
11
+
Darling is distributed in the hope that it will be useful,
12
+
but WITHOUT ANY WARRANTY; without even the implied warranty of
13
+
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14
+
GNU General Public License for more details.
15
+
16
+
You should have received a copy of the GNU General Public License
17
+
along with Darling. If not, see <http://www.gnu.org/licenses/>.
18
+
*/
19
+
20
+
#include "TextCommon.h"
21
+
22
+
TextEncoding CreateTextEncoding(TextEncodingBase encodingBase, TextEncodingVariant encodingVariant, TextEncodingFormat encodingFormat)
23
+
{
24
+
TextEncoding rv = encodingBase & 0xff;
25
+
rv |= (encodingVariant << 8) & 0xff00;
26
+
rv |= (encodingFormat << 16) & 0xff0000;
27
+
return rv;
28
+
}
29
+
30
+
31
+
+75
src/CoreServices/TextCommon.h
+75
src/CoreServices/TextCommon.h
···
1
+
/*
2
+
This file is part of Darling.
3
+
4
+
Copyright (C) 2017 Lubos Dolezel
5
+
6
+
Darling is free software: you can redistribute it and/or modify
7
+
it under the terms of the GNU General Public License as published by
8
+
the Free Software Foundation, either version 3 of the License, or
9
+
(at your option) any later version.
10
+
11
+
Darling is distributed in the hope that it will be useful,
12
+
but WITHOUT ANY WARRANTY; without even the implied warranty of
13
+
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14
+
GNU General Public License for more details.
15
+
16
+
You should have received a copy of the GNU General Public License
17
+
along with Darling. If not, see <http://www.gnu.org/licenses/>.
18
+
*/
19
+
20
+
#ifndef _TEXTCOMMON_H
21
+
#define _TEXTCOMMON_H
22
+
#include <MacTypes.h>
23
+
24
+
#ifdef __cplusplus
25
+
extern "C" {
26
+
#endif
27
+
28
+
typedef UInt8* TextPtr;
29
+
typedef const UInt8* ConstTextPtr;
30
+
31
+
enum {
32
+
kTextEncodingUnicodeDefault = 0x100
33
+
};
34
+
35
+
enum {
36
+
kTextEncodingDefaultVariant = 0
37
+
};
38
+
39
+
enum {
40
+
kTextEncodingDefaultFormat = 0,
41
+
kUnicodeUTF16Format = 0,
42
+
kUnicodeUTF7Format = 1,
43
+
kUnicodeUTF8Format = 2,
44
+
kUnicodeUTF32Format = 3,
45
+
kUnicodeUTF16BEFormat = 4,
46
+
kUnicodeUTF16LEFormat = 5,
47
+
kUnicodeUTF32BEFormat = 6,
48
+
kUnicodeUTF32LEFormat = 7,
49
+
kUnicodeSCSUFormat = 8,
50
+
kUnicode16BitFormat = 0,
51
+
kUnicode32BitFormat = 3,
52
+
};
53
+
54
+
enum {
55
+
kUnicodeNoSubset = 0,
56
+
kUnicodeNormalizationFormD = 5,
57
+
kUnicodeNormalizationFormC = 3,
58
+
kUnicodeCanonicalCompVariant = kUnicodeNormalizationFormC,
59
+
kUnicodeHFSPlusDecompVariant = 8,
60
+
kUnicodeHFSPlusCompVariant = 9,
61
+
};
62
+
63
+
typedef UInt32 TextEncoding;
64
+
typedef UInt32 TextEncodingBase;
65
+
typedef UInt32 TextEncodingFormat;
66
+
typedef UInt32 TextEncodingVariant;
67
+
68
+
TextEncoding CreateTextEncoding(TextEncodingBase encodingBase, TextEncodingVariant encodingVariant, TextEncodingFormat encodingFormat);
69
+
70
+
#ifdef __cplusplus
71
+
}
72
+
#endif
73
+
74
+
#endif
75
+
+233
src/CoreServices/TextEncodingConverter.cpp
+233
src/CoreServices/TextEncodingConverter.cpp
···
1
+
/*
2
+
This file is part of Darling.
3
+
4
+
Copyright (C) 2017 Lubos Dolezel
5
+
6
+
Darling is free software: you can redistribute it and/or modify
7
+
it under the terms of the GNU General Public License as published by
8
+
the Free Software Foundation, either version 3 of the License, or
9
+
(at your option) any later version.
10
+
11
+
Darling is distributed in the hope that it will be useful,
12
+
but WITHOUT ANY WARRANTY; without even the implied warranty of
13
+
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14
+
GNU General Public License for more details.
15
+
16
+
You should have received a copy of the GNU General Public License
17
+
along with Darling. If not, see <http://www.gnu.org/licenses/>.
18
+
*/
19
+
20
+
#include "TextEncodingConverter.h"
21
+
#include <unicode/ucnv.h>
22
+
#include <unicode/normalizer2.h>
23
+
#include "MacErrors.h"
24
+
25
+
struct OpaqueTECObjectRef
26
+
{
27
+
UConverter* inputConverter;
28
+
UConverter* outputConverter;
29
+
const UNormalizer2* normalizer;
30
+
UChar buffer[64];
31
+
size_t bufferUsed;
32
+
};
33
+
34
+
static UConverter* createConverter(TextEncodingBase base, TextEncodingFormat format)
35
+
{
36
+
switch (base)
37
+
{
38
+
case kTextEncodingUnicodeDefault:
39
+
{
40
+
const char* enc;
41
+
UErrorCode error = U_ZERO_ERROR;
42
+
43
+
switch (format)
44
+
{
45
+
case kUnicodeUTF16Format:
46
+
enc = "UTF-16";
47
+
break;
48
+
case kUnicodeUTF7Format:
49
+
enc = "UTF-7";
50
+
break;
51
+
case kUnicodeUTF8Format:
52
+
enc = "UTF-8";
53
+
break;
54
+
case kUnicodeUTF32Format:
55
+
enc = "UTF-32";
56
+
break;
57
+
case kUnicodeUTF16BEFormat:
58
+
enc = "UTF-16BE";
59
+
break;
60
+
case kUnicodeUTF16LEFormat:
61
+
enc = "UTF-16LE";
62
+
break;
63
+
case kUnicodeUTF32BEFormat:
64
+
enc = "UTF-32BE";
65
+
break;
66
+
case kUnicodeUTF32LEFormat:
67
+
enc = "UTF-32LE";
68
+
break;
69
+
default:
70
+
return NULL;
71
+
}
72
+
73
+
return ucnv_open(enc, &error);
74
+
}
75
+
default:
76
+
return NULL;
77
+
}
78
+
}
79
+
80
+
OSStatus TECCreateConverter(TECObjectRef *newEncodingConverter, TextEncoding inputEncoding, TextEncoding outputEncoding)
81
+
{
82
+
TextEncodingFormat format;
83
+
TextEncodingBase base;
84
+
TextEncodingVariant variant;
85
+
OpaqueTECObjectRef* obj = new OpaqueTECObjectRef;
86
+
87
+
obj->inputConverter = obj->outputConverter = NULL;
88
+
obj->normalizer = NULL;
89
+
obj->bufferUsed = 0;
90
+
91
+
format = (inputEncoding >> 16) & 0xff;
92
+
base = (inputEncoding >> 0) & 0xff;
93
+
variant = (inputEncoding >> 8) & 0xff;
94
+
95
+
obj->inputConverter = createConverter(base, format);
96
+
if (!obj->inputConverter)
97
+
{
98
+
TECDisposeConverter(obj);
99
+
*newEncodingConverter = NULL;
100
+
return unimpErr;
101
+
}
102
+
103
+
format = (outputEncoding >> 16) & 0xff;
104
+
base = (outputEncoding >> 0) & 0xff;
105
+
variant = (outputEncoding >> 8) & 0xff;
106
+
107
+
obj->outputConverter = createConverter(base, format);
108
+
if (!obj->outputConverter)
109
+
{
110
+
TECDisposeConverter(obj);
111
+
*newEncodingConverter = NULL;
112
+
return unimpErr;
113
+
}
114
+
115
+
if (base == kTextEncodingUnicodeDefault)
116
+
{
117
+
switch (variant)
118
+
{
119
+
case kUnicodeNoSubset:
120
+
break;
121
+
case kUnicodeNormalizationFormD:
122
+
{
123
+
UErrorCode error = U_ZERO_ERROR;
124
+
obj->normalizer = unorm2_getNFDInstance(&error);
125
+
break;
126
+
}
127
+
case kUnicodeNormalizationFormC:
128
+
{
129
+
UErrorCode error = U_ZERO_ERROR;
130
+
obj->normalizer = unorm2_getNFCInstance(&error);
131
+
break;
132
+
}
133
+
case kUnicodeHFSPlusDecompVariant:
134
+
case kUnicodeHFSPlusCompVariant:
135
+
{
136
+
UErrorCode error = U_ZERO_ERROR;
137
+
obj->normalizer = unorm2_getNFKDInstance(&error);
138
+
break;
139
+
}
140
+
}
141
+
}
142
+
143
+
*newEncodingConverter = obj;
144
+
return noErr;
145
+
}
146
+
147
+
OSStatus TECConvertText(TECObjectRef encodingConverter, ConstTextPtr inputBuffer,
148
+
ByteCount inputBufferLength, ByteCount *actualInputLength,
149
+
TextPtr outputBuffer, ByteCount outputBufferLength, ByteCount *actualOutputLength)
150
+
{
151
+
if (actualInputLength != NULL)
152
+
*actualInputLength = 0;
153
+
*actualOutputLength = 0;
154
+
155
+
while (outputBufferLength > 0)
156
+
{
157
+
if (encodingConverter->bufferUsed > 0)
158
+
{
159
+
// flush buffer
160
+
UErrorCode error = U_ZERO_ERROR;
161
+
char* target = (char*) outputBuffer;
162
+
const UChar* source = encodingConverter->buffer;
163
+
ByteCount inputUsed;
164
+
165
+
166
+
ucnv_fromUnicode(encodingConverter->outputConverter,
167
+
&target, target + outputBufferLength,
168
+
&source, source + encodingConverter->bufferUsed,
169
+
NULL, false, &error);
170
+
171
+
if (error != U_ZERO_ERROR && error != U_BUFFER_OVERFLOW_ERROR)
172
+
return paramErr;
173
+
174
+
*actualOutputLength = target - ((char*)outputBuffer);
175
+
176
+
inputUsed = source - encodingConverter->buffer;
177
+
if (inputUsed < encodingConverter->bufferUsed)
178
+
{
179
+
memmove(encodingConverter->buffer,
180
+
encodingConverter->buffer + inputUsed,
181
+
encodingConverter->bufferUsed - inputUsed);
182
+
}
183
+
encodingConverter->bufferUsed -= inputUsed;
184
+
outputBuffer += *actualOutputLength;
185
+
outputBufferLength -= *actualOutputLength;
186
+
187
+
if (error == U_BUFFER_OVERFLOW_ERROR)
188
+
break;
189
+
}
190
+
191
+
if (inputBufferLength <= 0)
192
+
break;
193
+
194
+
// Consume input
195
+
{
196
+
UChar* target = encodingConverter->buffer + encodingConverter->bufferUsed;
197
+
const char* source = (const char*) inputBuffer;
198
+
UErrorCode error = U_ZERO_ERROR;
199
+
200
+
ucnv_toUnicode(encodingConverter->inputConverter,
201
+
&target, encodingConverter->buffer + (sizeof(encodingConverter->buffer) / sizeof(encodingConverter->buffer[0])),
202
+
&source, source + inputBufferLength,
203
+
NULL, false, &error);
204
+
205
+
if (error != U_ZERO_ERROR && error != U_BUFFER_OVERFLOW_ERROR)
206
+
return paramErr;
207
+
208
+
encodingConverter->bufferUsed += target - (encodingConverter->buffer + encodingConverter->bufferUsed);
209
+
*actualInputLength += source - ((const char*)inputBuffer);
210
+
inputBufferLength -= source - ((const char*)inputBuffer);
211
+
inputBuffer = (ConstTextPtr) source;
212
+
}
213
+
214
+
// TODO: normalize
215
+
// Normalization may cause the data to no longer fit into our internal buffer :-/
216
+
}
217
+
218
+
return noErr;
219
+
}
220
+
221
+
OSStatus TECFlushText(TECObjectRef encodingConverter, TextPtr outputBuffer, ByteCount outputBufferLength, ByteCount *actualOutputLength)
222
+
{
223
+
return TECConvertText(encodingConverter, NULL, 0, NULL, outputBuffer, outputBufferLength, actualOutputLength);
224
+
}
225
+
226
+
OSStatus TECDisposeConverter(TECObjectRef conv)
227
+
{
228
+
ucnv_close(conv->inputConverter);
229
+
ucnv_close(conv->outputConverter);
230
+
delete conv;
231
+
return noErr;
232
+
}
233
+
+44
src/CoreServices/TextEncodingConverter.h
+44
src/CoreServices/TextEncodingConverter.h
···
1
+
/*
2
+
This file is part of Darling.
3
+
4
+
Copyright (C) 2017 Lubos Dolezel
5
+
6
+
Darling is free software: you can redistribute it and/or modify
7
+
it under the terms of the GNU General Public License as published by
8
+
the Free Software Foundation, either version 3 of the License, or
9
+
(at your option) any later version.
10
+
11
+
Darling is distributed in the hope that it will be useful,
12
+
but WITHOUT ANY WARRANTY; without even the implied warranty of
13
+
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14
+
GNU General Public License for more details.
15
+
16
+
You should have received a copy of the GNU General Public License
17
+
along with Darling. If not, see <http://www.gnu.org/licenses/>.
18
+
*/
19
+
20
+
#ifndef _TEXTENCODINGCONVERTER_H
21
+
#define _TEXTENCODINGCONVERTER_H
22
+
#include <MacTypes.h>
23
+
#include "TextCommon.h"
24
+
25
+
#ifdef __cplusplus
26
+
extern "C" {
27
+
#endif
28
+
29
+
typedef struct OpaqueTECObjectRef* TECObjectRef;
30
+
31
+
OSStatus TECCreateConverter(TECObjectRef *newEncodingConverter, TextEncoding inputEncoding, TextEncoding outputEncoding);
32
+
33
+
OSStatus TECConvertText(TECObjectRef encodingConverter, ConstTextPtr inputBuffer, ByteCount inputBufferLength, ByteCount *actualInputLength, TextPtr outputBuffer, ByteCount outputBufferLength, ByteCount *actualOutputLength);
34
+
35
+
OSStatus TECFlushText(TECObjectRef encodingConverter, TextPtr outputBuffer, ByteCount outputBufferLength, ByteCount *actualOutputLength);
36
+
37
+
OSStatus TECDisposeConverter(TECObjectRef newEncodingConverter);
38
+
39
+
#ifdef __cplusplus
40
+
}
41
+
#endif
42
+
43
+
#endif
44
+