1 | // © 2016 and later: Unicode, Inc. and others. |
2 | // License & terms of use: http://www.unicode.org/copyright.html |
3 | /* |
4 | ******************************************************************************* |
5 | * |
6 | * Copyright (C) 1998-2014, International Business Machines |
7 | * Corporation and others. All Rights Reserved. |
8 | * |
9 | ******************************************************************************* |
10 | * file name: ustr_cnv.cpp |
11 | * encoding: UTF-8 |
12 | * tab size: 8 (not used) |
13 | * indentation:4 |
14 | * |
15 | * created on: 2004aug24 |
16 | * created by: Markus W. Scherer |
17 | * |
18 | * Character conversion functions moved here from ustring.c |
19 | */ |
20 | |
21 | #include "unicode/utypes.h" |
22 | |
23 | #if !UCONFIG_NO_CONVERSION |
24 | |
25 | #include "unicode/ustring.h" |
26 | #include "unicode/ucnv.h" |
27 | #include "cstring.h" |
28 | #include "cmemory.h" |
29 | #include "umutex.h" |
30 | #include "ustr_cnv.h" |
31 | #include "ucnv_bld.h" |
32 | |
33 | /* mutexed access to a shared default converter ----------------------------- */ |
34 | |
35 | static UConverter *gDefaultConverter = nullptr; |
36 | |
37 | U_CAPI UConverter* U_EXPORT2 |
38 | u_getDefaultConverter(UErrorCode *status) |
39 | { |
40 | UConverter *converter = nullptr; |
41 | |
42 | if (gDefaultConverter != nullptr) { |
43 | icu::umtx_lock(nullptr); |
44 | |
45 | /* need to check to make sure it wasn't taken out from under us */ |
46 | if (gDefaultConverter != nullptr) { |
47 | converter = gDefaultConverter; |
48 | gDefaultConverter = nullptr; |
49 | } |
50 | icu::umtx_unlock(nullptr); |
51 | } |
52 | |
53 | /* if the cache was empty, create a converter */ |
54 | if(converter == nullptr) { |
55 | converter = ucnv_open(nullptr, status); |
56 | if(U_FAILURE(*status)) { |
57 | ucnv_close(converter); |
58 | converter = nullptr; |
59 | } |
60 | } |
61 | |
62 | return converter; |
63 | } |
64 | |
65 | U_CAPI void U_EXPORT2 |
66 | u_releaseDefaultConverter(UConverter *converter) |
67 | { |
68 | if(gDefaultConverter == nullptr) { |
69 | if (converter != nullptr) { |
70 | ucnv_reset(converter); |
71 | } |
72 | ucnv_enableCleanup(); |
73 | icu::umtx_lock(nullptr); |
74 | if(gDefaultConverter == nullptr) { |
75 | gDefaultConverter = converter; |
76 | converter = nullptr; |
77 | } |
78 | icu::umtx_unlock(nullptr); |
79 | } |
80 | |
81 | if(converter != nullptr) { |
82 | ucnv_close(converter); |
83 | } |
84 | } |
85 | |
86 | U_CAPI void U_EXPORT2 |
87 | u_flushDefaultConverter() |
88 | { |
89 | UConverter *converter = nullptr; |
90 | |
91 | if (gDefaultConverter != nullptr) { |
92 | icu::umtx_lock(nullptr); |
93 | |
94 | /* need to check to make sure it wasn't taken out from under us */ |
95 | if (gDefaultConverter != nullptr) { |
96 | converter = gDefaultConverter; |
97 | gDefaultConverter = nullptr; |
98 | } |
99 | icu::umtx_unlock(nullptr); |
100 | } |
101 | |
102 | /* if the cache was populated, flush it */ |
103 | if(converter != nullptr) { |
104 | ucnv_close(converter); |
105 | } |
106 | } |
107 | |
108 | |
109 | /* conversions between char* and char16_t* ------------------------------------- */ |
110 | |
111 | /* maximum string length for u_uastrcpy() and u_austrcpy() implementations */ |
112 | #define MAX_STRLEN 0x0FFFFFFF |
113 | |
114 | /* |
115 | returns the minimum of (the length of the null-terminated string) and n. |
116 | */ |
117 | static int32_t u_astrnlen(const char *s1, int32_t n) |
118 | { |
119 | int32_t len = 0; |
120 | |
121 | if (s1) |
122 | { |
123 | while (n-- && *(s1++)) |
124 | { |
125 | len++; |
126 | } |
127 | } |
128 | return len; |
129 | } |
130 | |
131 | U_CAPI char16_t* U_EXPORT2 |
132 | u_uastrncpy(char16_t *ucs1, |
133 | const char *s2, |
134 | int32_t n) |
135 | { |
136 | char16_t *target = ucs1; |
137 | UErrorCode err = U_ZERO_ERROR; |
138 | UConverter *cnv = u_getDefaultConverter(&err); |
139 | if(U_SUCCESS(err) && cnv != nullptr) { |
140 | ucnv_reset(cnv); |
141 | ucnv_toUnicode(cnv, |
142 | &target, |
143 | ucs1+n, |
144 | &s2, |
145 | s2+u_astrnlen(s2, n), |
146 | nullptr, |
147 | true, |
148 | &err); |
149 | ucnv_reset(cnv); /* be good citizens */ |
150 | u_releaseDefaultConverter(cnv); |
151 | if(U_FAILURE(err) && (err != U_BUFFER_OVERFLOW_ERROR) ) { |
152 | *ucs1 = 0; /* failure */ |
153 | } |
154 | if(target < (ucs1+n)) { /* U_BUFFER_OVERFLOW_ERROR isn't an err, just means no termination will happen. */ |
155 | *target = 0; /* terminate */ |
156 | } |
157 | } else { |
158 | *ucs1 = 0; |
159 | } |
160 | return ucs1; |
161 | } |
162 | |
163 | U_CAPI char16_t* U_EXPORT2 |
164 | u_uastrcpy(char16_t *ucs1, |
165 | const char *s2 ) |
166 | { |
167 | UErrorCode err = U_ZERO_ERROR; |
168 | UConverter *cnv = u_getDefaultConverter(&err); |
169 | if(U_SUCCESS(err) && cnv != nullptr) { |
170 | ucnv_toUChars(cnv, |
171 | ucs1, |
172 | MAX_STRLEN, |
173 | s2, |
174 | (int32_t)uprv_strlen(s2), |
175 | &err); |
176 | u_releaseDefaultConverter(cnv); |
177 | if(U_FAILURE(err)) { |
178 | *ucs1 = 0; |
179 | } |
180 | } else { |
181 | *ucs1 = 0; |
182 | } |
183 | return ucs1; |
184 | } |
185 | |
186 | /* |
187 | returns the minimum of (the length of the null-terminated string) and n. |
188 | */ |
189 | static int32_t u_ustrnlen(const char16_t *ucs1, int32_t n) |
190 | { |
191 | int32_t len = 0; |
192 | |
193 | if (ucs1) |
194 | { |
195 | while (n-- && *(ucs1++)) |
196 | { |
197 | len++; |
198 | } |
199 | } |
200 | return len; |
201 | } |
202 | |
203 | U_CAPI char* U_EXPORT2 |
204 | u_austrncpy(char *s1, |
205 | const char16_t *ucs2, |
206 | int32_t n) |
207 | { |
208 | char *target = s1; |
209 | UErrorCode err = U_ZERO_ERROR; |
210 | UConverter *cnv = u_getDefaultConverter(&err); |
211 | if(U_SUCCESS(err) && cnv != nullptr) { |
212 | ucnv_reset(cnv); |
213 | ucnv_fromUnicode(cnv, |
214 | &target, |
215 | s1+n, |
216 | &ucs2, |
217 | ucs2+u_ustrnlen(ucs2, n), |
218 | nullptr, |
219 | true, |
220 | &err); |
221 | ucnv_reset(cnv); /* be good citizens */ |
222 | u_releaseDefaultConverter(cnv); |
223 | if(U_FAILURE(err) && (err != U_BUFFER_OVERFLOW_ERROR) ) { |
224 | *s1 = 0; /* failure */ |
225 | } |
226 | if(target < (s1+n)) { /* U_BUFFER_OVERFLOW_ERROR isn't an err, just means no termination will happen. */ |
227 | *target = 0; /* terminate */ |
228 | } |
229 | } else { |
230 | *s1 = 0; |
231 | } |
232 | return s1; |
233 | } |
234 | |
235 | U_CAPI char* U_EXPORT2 |
236 | u_austrcpy(char *s1, |
237 | const char16_t *ucs2 ) |
238 | { |
239 | UErrorCode err = U_ZERO_ERROR; |
240 | UConverter *cnv = u_getDefaultConverter(&err); |
241 | if(U_SUCCESS(err) && cnv != nullptr) { |
242 | int32_t len = ucnv_fromUChars(cnv, |
243 | s1, |
244 | MAX_STRLEN, |
245 | ucs2, |
246 | -1, |
247 | &err); |
248 | u_releaseDefaultConverter(cnv); |
249 | s1[len] = 0; |
250 | } else { |
251 | *s1 = 0; |
252 | } |
253 | return s1; |
254 | } |
255 | |
256 | #endif |
257 | |