136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
|
buf[i] = 0x80 | (c & 0x3F);
return 4;
}
return 0;
}
of_unichar_t
of_string_utf8_to_unicode(const char *buf_, size_t len)
if (!(*buf & 0x80))
return buf[0];
if ((*buf & 0xE0) == 0xC0) {
if (OF_UNLIKELY(len < 2))
return OF_INVALID_UNICHAR;
return ((buf[0] & 0x1F) << 6) | (buf[1] & 0x3F);
}
if ((*buf & 0xF0) == 0xE0) {
if (OF_UNLIKELY(len < 3))
return OF_INVALID_UNICHAR;
return ((buf[0] & 0x0F) << 12) | ((buf[1] & 0x3F) << 6) |
(buf[2] & 0x3F);
}
if ((*buf & 0xF8) == 0xF0) {
if (OF_UNLIKELY(len < 4))
return OF_INVALID_UNICHAR;
return ((buf[0] & 0x07) << 18) | ((buf[1] & 0x3F) << 12) |
((buf[2] & 0x3F) << 6) | (buf[3] & 0x3F);
}
return OF_INVALID_UNICHAR;
}
size_t
of_string_position_to_index(const char *str, size_t pos)
{
size_t i, idx = pos;
|
|
|
|
>
|
>
|
|
>
|
|
>
|
|
>
|
|
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
|
buf[i] = 0x80 | (c & 0x3F);
return 4;
}
return 0;
}
size_t
of_string_utf8_to_unicode(const char *buf_, size_t len, of_unichar_t *ret)
{
const uint8_t *buf = (const uint8_t*)buf_;
if (!(*buf & 0x80)) {
*ret = buf[0];
return 1;
}
if ((*buf & 0xE0) == 0xC0) {
if (OF_UNLIKELY(len < 2))
return 0;
*ret = ((buf[0] & 0x1F) << 6) | (buf[1] & 0x3F);
return 2;
}
if ((*buf & 0xF0) == 0xE0) {
if (OF_UNLIKELY(len < 3))
return 0;
*ret = ((buf[0] & 0x0F) << 12) | ((buf[1] & 0x3F) << 6) |
(buf[2] & 0x3F);
return 3;
}
if ((*buf & 0xF8) == 0xF0) {
if (OF_UNLIKELY(len < 4))
return 0;
*ret = ((buf[0] & 0x07) << 18) | ((buf[1] & 0x3F) << 12) |
((buf[2] & 0x3F) << 6) | (buf[3] & 0x3F);
return 4;
}
return 0;
}
size_t
of_string_position_to_index(const char *str, size_t pos)
{
size_t i, idx = pos;
|
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
|
of_unichar_t c;
index = of_string_index_to_position(string, index, length);
if (index >= length)
@throw [OFOutOfRangeException newWithClass: isa];
if ((c = of_string_utf8_to_unicode(string + index, length - index)) ==
OF_INVALID_UNICHAR)
@throw [OFInvalidEncodingException newWithClass: isa];
return c;
}
- (size_t)indexOfFirstOccurrenceOfString: (OFString*)str
{
|
|
<
|
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
|
of_unichar_t c;
index = of_string_index_to_position(string, index, length);
if (index >= length)
@throw [OFOutOfRangeException newWithClass: isa];
if (!of_string_utf8_to_unicode(string + index, length - index, &c))
@throw [OFInvalidEncodingException newWithClass: isa];
return c;
}
- (size_t)indexOfFirstOccurrenceOfString: (OFString*)str
{
|