1// © 2016 and later: Unicode, Inc. and others.
2// License & terms of use: http://www.unicode.org/copyright.html
3/*
4**********************************************************************
5* Copyright (C) 1999-2014, International Business Machines
6* Corporation and others. All Rights Reserved.
7**********************************************************************
8 * ucnv.h:
9 * External APIs for the ICU's codeset conversion library
10 * Bertrand A. Damiba
11 *
12 * Modification History:
13 *
14 * Date Name Description
15 * 04/04/99 helena Fixed internal header inclusion.
16 * 05/11/00 helena Added setFallback and usesFallback APIs.
17 * 06/29/2000 helena Major rewrite of the callback APIs.
18 * 12/07/2000 srl Update of documentation
19 */
20
21/**
22 * \file
23 * \brief C API: Character conversion
24 *
25 * <h2>Character Conversion C API</h2>
26 *
27 * <p>This API is used to convert codepage or character encoded data to and
28 * from UTF-16. You can open a converter with {@link ucnv_open() }. With that
29 * converter, you can get its properties, set options, convert your data and
30 * close the converter.</p>
31 *
32 * <p>Since many software programs recognize different converter names for
33 * different types of converters, there are other functions in this API to
34 * iterate over the converter aliases. The functions {@link ucnv_getAvailableName() },
35 * {@link ucnv_getAlias() } and {@link ucnv_getStandardName() } are some of the
36 * more frequently used alias functions to get this information.</p>
37 *
38 * <p>When a converter encounters an illegal, irregular, invalid or unmappable character
39 * its default behavior is to use a substitution character to replace the
40 * bad byte sequence. This behavior can be changed by using {@link ucnv_setFromUCallBack() }
41 * or {@link ucnv_setToUCallBack() } on the converter. The header ucnv_err.h defines
42 * many other callback actions that can be used instead of a character substitution.</p>
43 *
44 * <p>More information about this API can be found in our
45 * <a href="https://unicode-org.github.io/icu/userguide/conversion/">User Guide</a>.</p>
46 */
47
48#ifndef UCNV_H
49#define UCNV_H
50
51#include "unicode/ucnv_err.h"
52#include "unicode/uenum.h"
53
54#if U_SHOW_CPLUSPLUS_API
55#include "unicode/localpointer.h"
56#endif // U_SHOW_CPLUSPLUS_API
57
58#if !defined(USET_DEFINED) && !defined(U_IN_DOXYGEN)
59
60#define USET_DEFINED
61
62/**
63 * USet is the C API type corresponding to C++ class UnicodeSet.
64 * It is forward-declared here to avoid including unicode/uset.h file if related
65 * conversion APIs are not used.
66 *
67 * @see ucnv_getUnicodeSet
68 * @stable ICU 2.4
69 */
70typedef struct USet USet;
71
72#endif
73
74#if !UCONFIG_NO_CONVERSION
75
76U_CDECL_BEGIN
77
78/** Maximum length of a converter name including the terminating NULL @stable ICU 2.0 */
79#define UCNV_MAX_CONVERTER_NAME_LENGTH 60
80/** Maximum length of a converter name including path and terminating NULL @stable ICU 2.0 */
81#define UCNV_MAX_FULL_FILE_NAME_LENGTH (600+UCNV_MAX_CONVERTER_NAME_LENGTH)
82
83/** Shift in for EBDCDIC_STATEFUL and iso2022 states @stable ICU 2.0 */
84#define UCNV_SI 0x0F
85/** Shift out for EBDCDIC_STATEFUL and iso2022 states @stable ICU 2.0 */
86#define UCNV_SO 0x0E
87
88/**
89 * Enum for specifying basic types of converters
90 * @see ucnv_getType
91 * @stable ICU 2.0
92 */
93typedef enum {
94 /** @stable ICU 2.0 */
95 UCNV_UNSUPPORTED_CONVERTER = -1,
96 /** @stable ICU 2.0 */
97 UCNV_SBCS = 0,
98 /** @stable ICU 2.0 */
99 UCNV_DBCS = 1,
100 /** @stable ICU 2.0 */
101 UCNV_MBCS = 2,
102 /** @stable ICU 2.0 */
103 UCNV_LATIN_1 = 3,
104 /** @stable ICU 2.0 */
105 UCNV_UTF8 = 4,
106 /** @stable ICU 2.0 */
107 UCNV_UTF16_BigEndian = 5,
108 /** @stable ICU 2.0 */
109 UCNV_UTF16_LittleEndian = 6,
110 /** @stable ICU 2.0 */
111 UCNV_UTF32_BigEndian = 7,
112 /** @stable ICU 2.0 */
113 UCNV_UTF32_LittleEndian = 8,
114 /** @stable ICU 2.0 */
115 UCNV_EBCDIC_STATEFUL = 9,
116 /** @stable ICU 2.0 */
117 UCNV_ISO_2022 = 10,
118
119 /** @stable ICU 2.0 */
120 UCNV_LMBCS_1 = 11,
121 /** @stable ICU 2.0 */
122 UCNV_LMBCS_2,
123 /** @stable ICU 2.0 */
124 UCNV_LMBCS_3,
125 /** @stable ICU 2.0 */
126 UCNV_LMBCS_4,
127 /** @stable ICU 2.0 */
128 UCNV_LMBCS_5,
129 /** @stable ICU 2.0 */
130 UCNV_LMBCS_6,
131 /** @stable ICU 2.0 */
132 UCNV_LMBCS_8,
133 /** @stable ICU 2.0 */
134 UCNV_LMBCS_11,
135 /** @stable ICU 2.0 */
136 UCNV_LMBCS_16,
137 /** @stable ICU 2.0 */
138 UCNV_LMBCS_17,
139 /** @stable ICU 2.0 */
140 UCNV_LMBCS_18,
141 /** @stable ICU 2.0 */
142 UCNV_LMBCS_19,
143 /** @stable ICU 2.0 */
144 UCNV_LMBCS_LAST = UCNV_LMBCS_19,
145 /** @stable ICU 2.0 */
146 UCNV_HZ,
147 /** @stable ICU 2.0 */
148 UCNV_SCSU,
149 /** @stable ICU 2.0 */
150 UCNV_ISCII,
151 /** @stable ICU 2.0 */
152 UCNV_US_ASCII,
153 /** @stable ICU 2.0 */
154 UCNV_UTF7,
155 /** @stable ICU 2.2 */
156 UCNV_BOCU1,
157 /** @stable ICU 2.2 */
158 UCNV_UTF16,
159 /** @stable ICU 2.2 */
160 UCNV_UTF32,
161 /** @stable ICU 2.2 */
162 UCNV_CESU8,
163 /** @stable ICU 2.4 */
164 UCNV_IMAP_MAILBOX,
165 /** @stable ICU 4.8 */
166 UCNV_COMPOUND_TEXT,
167
168 /* Number of converter types for which we have conversion routines. */
169 UCNV_NUMBER_OF_SUPPORTED_CONVERTER_TYPES
170} UConverterType;
171
172/**
173 * Enum for specifying which platform a converter ID refers to.
174 * The use of platform/CCSID is not recommended. See ucnv_openCCSID().
175 *
176 * @see ucnv_getPlatform
177 * @see ucnv_openCCSID
178 * @see ucnv_getCCSID
179 * @stable ICU 2.0
180 */
181typedef enum {
182 UCNV_UNKNOWN = -1,
183 UCNV_IBM = 0
184} UConverterPlatform;
185
186/**
187 * Function pointer for error callback in the codepage to unicode direction.
188 * Called when an error has occurred in conversion to unicode, or on open/close of the callback (see reason).
189 * @param context Pointer to the callback's private data
190 * @param args Information about the conversion in progress
191 * @param codeUnits Points to 'length' bytes of the concerned codepage sequence
192 * @param length Size (in bytes) of the concerned codepage sequence
193 * @param reason Defines the reason the callback was invoked
194 * @param pErrorCode ICU error code in/out parameter.
195 * For converter callback functions, set to a conversion error
196 * before the call, and the callback may reset it to U_ZERO_ERROR.
197 * @see ucnv_setToUCallBack
198 * @see UConverterToUnicodeArgs
199 * @stable ICU 2.0
200 */
201typedef void (U_EXPORT2 *UConverterToUCallback) (
202 const void* context,
203 UConverterToUnicodeArgs *args,
204 const char *codeUnits,
205 int32_t length,
206 UConverterCallbackReason reason,
207 UErrorCode *pErrorCode);
208
209/**
210 * Function pointer for error callback in the unicode to codepage direction.
211 * Called when an error has occurred in conversion from unicode, or on open/close of the callback (see reason).
212 * @param context Pointer to the callback's private data
213 * @param args Information about the conversion in progress
214 * @param codeUnits Points to 'length' UChars of the concerned Unicode sequence
215 * @param length Size (in bytes) of the concerned codepage sequence
216 * @param codePoint Single UChar32 (UTF-32) containing the concerend Unicode codepoint.
217 * @param reason Defines the reason the callback was invoked
218 * @param pErrorCode ICU error code in/out parameter.
219 * For converter callback functions, set to a conversion error
220 * before the call, and the callback may reset it to U_ZERO_ERROR.
221 * @see ucnv_setFromUCallBack
222 * @stable ICU 2.0
223 */
224typedef void (U_EXPORT2 *UConverterFromUCallback) (
225 const void* context,
226 UConverterFromUnicodeArgs *args,
227 const UChar* codeUnits,
228 int32_t length,
229 UChar32 codePoint,
230 UConverterCallbackReason reason,
231 UErrorCode *pErrorCode);
232
233U_CDECL_END
234
235/**
236 * Character that separates converter names from options and options from each other.
237 * @see ucnv_open
238 * @stable ICU 2.0
239 */
240#define UCNV_OPTION_SEP_CHAR ','
241
242/**
243 * String version of UCNV_OPTION_SEP_CHAR.
244 * @see ucnv_open
245 * @stable ICU 2.0
246 */
247#define UCNV_OPTION_SEP_STRING ","
248
249/**
250 * Character that separates a converter option from its value.
251 * @see ucnv_open
252 * @stable ICU 2.0
253 */
254#define UCNV_VALUE_SEP_CHAR '='
255
256/**
257 * String version of UCNV_VALUE_SEP_CHAR.
258 * @see ucnv_open
259 * @stable ICU 2.0
260 */
261#define UCNV_VALUE_SEP_STRING "="
262
263/**
264 * Converter option for specifying a locale.
265 * For example, ucnv_open("SCSU,locale=ja", &errorCode);
266 * See convrtrs.txt.
267 *
268 * @see ucnv_open
269 * @stable ICU 2.0
270 */
271#define UCNV_LOCALE_OPTION_STRING ",locale="
272
273/**
274 * Converter option for specifying a version selector (0..9) for some converters.
275 * For example,
276 * \code
277 * ucnv_open("UTF-7,version=1", &errorCode);
278 * \endcode
279 * See convrtrs.txt.
280 *
281 * @see ucnv_open
282 * @stable ICU 2.4
283 */
284#define UCNV_VERSION_OPTION_STRING ",version="
285
286/**
287 * Converter option for EBCDIC SBCS or mixed-SBCS/DBCS (stateful) codepages.
288 * Swaps Unicode mappings for EBCDIC LF and NL codes, as used on
289 * S/390 (z/OS) Unix System Services (Open Edition).
290 * For example, ucnv_open("ibm-1047,swaplfnl", &errorCode);
291 * See convrtrs.txt.
292 *
293 * @see ucnv_open
294 * @stable ICU 2.4
295 */
296#define UCNV_SWAP_LFNL_OPTION_STRING ",swaplfnl"
297
298/**
299 * Do a fuzzy compare of two converter/alias names.
300 * The comparison is case-insensitive, ignores leading zeroes if they are not
301 * followed by further digits, and ignores all but letters and digits.
302 * Thus the strings "UTF-8", "utf_8", "u*T@f08" and "Utf 8" are exactly equivalent.
303 * See section 1.4, Charset Alias Matching in Unicode Technical Standard #22
304 * at http://www.unicode.org/reports/tr22/
305 *
306 * @param name1 a converter name or alias, zero-terminated
307 * @param name2 a converter name or alias, zero-terminated
308 * @return 0 if the names match, or a negative value if the name1
309 * lexically precedes name2, or a positive value if the name1
310 * lexically follows name2.
311 * @stable ICU 2.0
312 */
313U_CAPI int U_EXPORT2
314ucnv_compareNames(const char *name1, const char *name2);
315
316
317/**
318 * Creates a UConverter object with the name of a coded character set specified as a C string.
319 * The actual name will be resolved with the alias file
320 * using a case-insensitive string comparison that ignores
321 * leading zeroes and all non-alphanumeric characters.
322 * E.g., the names "UTF8", "utf-8", "u*T@f08" and "Utf 8" are all equivalent.
323 * (See also ucnv_compareNames().)
324 * If <code>NULL</code> is passed for the converter name, it will create one with the
325 * getDefaultName return value.
326 *
327 * <p>A converter name for ICU 1.5 and above may contain options
328 * like a locale specification to control the specific behavior of
329 * the newly instantiated converter.
330 * The meaning of the options depends on the particular converter.
331 * If an option is not defined for or recognized by a given converter, then it is ignored.</p>
332 *
333 * <p>Options are appended to the converter name string, with a
334 * <code>UCNV_OPTION_SEP_CHAR</code> between the name and the first option and
335 * also between adjacent options.</p>
336 *
337 * <p>If the alias is ambiguous, then the preferred converter is used
338 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.</p>
339 *
340 * <p>The conversion behavior and names can vary between platforms. ICU may
341 * convert some characters differently from other platforms. Details on this topic
342 * are in the <a href="https://unicode-org.github.io/icu/userguide/conversion/">User
343 * Guide</a>. Aliases starting with a "cp" prefix have no specific meaning
344 * other than its an alias starting with the letters "cp". Please do not
345 * associate any meaning to these aliases.</p>
346 *
347 * \snippet samples/ucnv/convsamp.cpp ucnv_open
348 *
349 * @param converterName Name of the coded character set table.
350 * This may have options appended to the string.
351 * IANA alias character set names, IBM CCSIDs starting with "ibm-",
352 * Windows codepage numbers starting with "windows-" are frequently
353 * used for this parameter. See ucnv_getAvailableName and
354 * ucnv_getAlias for a complete list that is available.
355 * If this parameter is NULL, the default converter will be used.
356 * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
357 * @return the created Unicode converter object, or <TT>NULL</TT> if an error occurred
358 * @see ucnv_openU
359 * @see ucnv_openCCSID
360 * @see ucnv_getAvailableName
361 * @see ucnv_getAlias
362 * @see ucnv_getDefaultName
363 * @see ucnv_close
364 * @see ucnv_compareNames
365 * @stable ICU 2.0
366 */
367U_CAPI UConverter* U_EXPORT2
368ucnv_open(const char *converterName, UErrorCode *err);
369
370
371/**
372 * Creates a Unicode converter with the names specified as unicode string.
373 * The name should be limited to the ASCII-7 alphanumerics range.
374 * The actual name will be resolved with the alias file
375 * using a case-insensitive string comparison that ignores
376 * leading zeroes and all non-alphanumeric characters.
377 * E.g., the names "UTF8", "utf-8", "u*T@f08" and "Utf 8" are all equivalent.
378 * (See also ucnv_compareNames().)
379 * If <TT>NULL</TT> is passed for the converter name, it will create
380 * one with the ucnv_getDefaultName() return value.
381 * If the alias is ambiguous, then the preferred converter is used
382 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
383 *
384 * <p>See ucnv_open for the complete details</p>
385 * @param name Name of the UConverter table in a zero terminated
386 * Unicode string
387 * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR,
388 * U_FILE_ACCESS_ERROR</TT>
389 * @return the created Unicode converter object, or <TT>NULL</TT> if an
390 * error occurred
391 * @see ucnv_open
392 * @see ucnv_openCCSID
393 * @see ucnv_close
394 * @see ucnv_compareNames
395 * @stable ICU 2.0
396 */
397U_CAPI UConverter* U_EXPORT2
398ucnv_openU(const UChar *name,
399 UErrorCode *err);
400
401/**
402 * Creates a UConverter object from a CCSID number and platform pair.
403 * Note that the usefulness of this function is limited to platforms with numeric
404 * encoding IDs. Only IBM and Microsoft platforms use numeric (16-bit) identifiers for
405 * encodings.
406 *
407 * In addition, IBM CCSIDs and Unicode conversion tables are not 1:1 related.
408 * For many IBM CCSIDs there are multiple (up to six) Unicode conversion tables, and
409 * for some Unicode conversion tables there are multiple CCSIDs.
410 * Some "alternate" Unicode conversion tables are provided by the
411 * IBM CDRA conversion table registry.
412 * The most prominent example of a systematic modification of conversion tables that is
413 * not provided in the form of conversion table files in the repository is
414 * that S/390 Unix System Services swaps the codes for Line Feed and New Line in all
415 * EBCDIC codepages, which requires such a swap in the Unicode conversion tables as well.
416 *
417 * Only IBM default conversion tables are accessible with ucnv_openCCSID().
418 * ucnv_getCCSID() will return the same CCSID for all conversion tables that are associated
419 * with that CCSID.
420 *
421 * Currently, the only "platform" supported in the ICU converter API is UCNV_IBM.
422 *
423 * In summary, the use of CCSIDs and the associated API functions is not recommended.
424 *
425 * In order to open a converter with the default IBM CDRA Unicode conversion table,
426 * you can use this function or use the prefix "ibm-":
427 * \code
428 * char name[20];
429 * sprintf(name, "ibm-%hu", ccsid);
430 * cnv=ucnv_open(name, &errorCode);
431 * \endcode
432 *
433 * In order to open a converter with the IBM S/390 Unix System Services variant
434 * of a Unicode/EBCDIC conversion table,
435 * you can use the prefix "ibm-" together with the option string UCNV_SWAP_LFNL_OPTION_STRING:
436 * \code
437 * char name[20];
438 * sprintf(name, "ibm-%hu" UCNV_SWAP_LFNL_OPTION_STRING, ccsid);
439 * cnv=ucnv_open(name, &errorCode);
440 * \endcode
441 *
442 * In order to open a converter from a Microsoft codepage number, use the prefix "cp":
443 * \code
444 * char name[20];
445 * sprintf(name, "cp%hu", codepageID);
446 * cnv=ucnv_open(name, &errorCode);
447 * \endcode
448 *
449 * If the alias is ambiguous, then the preferred converter is used
450 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
451 *
452 * @param codepage codepage number to create
453 * @param platform the platform in which the codepage number exists
454 * @param err error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
455 * @return the created Unicode converter object, or <TT>NULL</TT> if an error
456 * occurred.
457 * @see ucnv_open
458 * @see ucnv_openU
459 * @see ucnv_close
460 * @see ucnv_getCCSID
461 * @see ucnv_getPlatform
462 * @see UConverterPlatform
463 * @stable ICU 2.0
464 */
465U_CAPI UConverter* U_EXPORT2
466ucnv_openCCSID(int32_t codepage,
467 UConverterPlatform platform,
468 UErrorCode * err);
469
470/**
471 * <p>Creates a UConverter object specified from a packageName and a converterName.</p>
472 *
473 * <p>The packageName and converterName must point to an ICU udata object, as defined by
474 * <code> udata_open( packageName, "cnv", converterName, err) </code> or equivalent.
475 * Typically, packageName will refer to a (.dat) file, or to a package registered with
476 * udata_setAppData(). Using a full file or directory pathname for packageName is deprecated.</p>
477 *
478 * <p>The name will NOT be looked up in the alias mechanism, nor will the converter be
479 * stored in the converter cache or the alias table. The only way to open further converters
480 * is call this function multiple times, or use the ucnv_clone() function to clone a
481 * 'primary' converter.</p>
482 *
483 * <p>A future version of ICU may add alias table lookups and/or caching
484 * to this function.</p>
485 *
486 * <p>Example Use:
487 * <code>cnv = ucnv_openPackage("myapp", "myconverter", &err);</code>
488 * </p>
489 *
490 * @param packageName name of the package (equivalent to 'path' in udata_open() call)
491 * @param converterName name of the data item to be used, without suffix.
492 * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
493 * @return the created Unicode converter object, or <TT>NULL</TT> if an error occurred
494 * @see udata_open
495 * @see ucnv_open
496 * @see ucnv_clone
497 * @see ucnv_close
498 * @stable ICU 2.2
499 */
500U_CAPI UConverter* U_EXPORT2
501ucnv_openPackage(const char *packageName, const char *converterName, UErrorCode *err);
502
503/**
504 * Thread safe converter cloning operation.
505 *
506 * You must ucnv_close() the clone.
507 *
508 * @param cnv converter to be cloned
509 * @param status to indicate whether the operation went on smoothly or there were errors
510 * @return pointer to the new clone
511 * @stable ICU 71
512 */
513U_CAPI UConverter* U_EXPORT2 ucnv_clone(const UConverter *cnv, UErrorCode *status);
514
515#ifndef U_HIDE_DEPRECATED_API
516
517/**
518 * Thread safe converter cloning operation.
519 * For most efficient operation, pass in a stackBuffer (and a *pBufferSize)
520 * with at least U_CNV_SAFECLONE_BUFFERSIZE bytes of space.
521 * If the buffer size is sufficient, then the clone will use the stack buffer;
522 * otherwise, it will be allocated, and *pBufferSize will indicate
523 * the actual size. (This should not occur with U_CNV_SAFECLONE_BUFFERSIZE.)
524 *
525 * You must ucnv_close() the clone in any case.
526 *
527 * If *pBufferSize==0, (regardless of whether stackBuffer==NULL or not)
528 * then *pBufferSize will be changed to a sufficient size
529 * for cloning this converter,
530 * without actually cloning the converter ("pure pre-flighting").
531 *
532 * If *pBufferSize is greater than zero but not large enough for a stack-based
533 * clone, then the converter is cloned using newly allocated memory
534 * and *pBufferSize is changed to the necessary size.
535 *
536 * If the converter clone fits into the stack buffer but the stack buffer is not
537 * sufficiently aligned for the clone, then the clone will use an
538 * adjusted pointer and use an accordingly smaller buffer size.
539 *
540 * @param cnv converter to be cloned
541 * @param stackBuffer <em>Deprecated functionality as of ICU 52, use NULL.</em><br>
542 * user allocated space for the new clone. If NULL new memory will be allocated.
543 * If buffer is not large enough, new memory will be allocated.
544 * Clients can use the U_CNV_SAFECLONE_BUFFERSIZE. This will probably be enough to avoid memory allocations.
545 * @param pBufferSize <em>Deprecated functionality as of ICU 52, use NULL or 1.</em><br>
546 * pointer to size of allocated space.
547 * @param status to indicate whether the operation went on smoothly or there were errors
548 * An informational status value, U_SAFECLONE_ALLOCATED_WARNING,
549 * is used if pBufferSize != NULL and any allocations were necessary
550 * However, it is better to check if *pBufferSize grew for checking for
551 * allocations because warning codes can be overridden by subsequent
552 * function calls.
553 * @return pointer to the new clone
554 * @deprecated ICU 71 Use ucnv_clone() instead.
555 */
556U_DEPRECATED UConverter * U_EXPORT2
557ucnv_safeClone(const UConverter *cnv,
558 void *stackBuffer,
559 int32_t *pBufferSize,
560 UErrorCode *status);
561
562/**
563 * \def U_CNV_SAFECLONE_BUFFERSIZE
564 * Definition of a buffer size that is designed to be large enough for
565 * converters to be cloned with ucnv_safeClone().
566 * @deprecated ICU 52. Do not rely on ucnv_safeClone() cloning into any provided buffer.
567 */
568#define U_CNV_SAFECLONE_BUFFERSIZE 1024
569
570#endif /* U_HIDE_DEPRECATED_API */
571
572/**
573 * Deletes the unicode converter and releases resources associated
574 * with just this instance.
575 * Does not free up shared converter tables.
576 *
577 * @param converter the converter object to be deleted
578 * @see ucnv_open
579 * @see ucnv_openU
580 * @see ucnv_openCCSID
581 * @stable ICU 2.0
582 */
583U_CAPI void U_EXPORT2
584ucnv_close(UConverter * converter);
585
586#if U_SHOW_CPLUSPLUS_API
587
588U_NAMESPACE_BEGIN
589
590/**
591 * \class LocalUConverterPointer
592 * "Smart pointer" class, closes a UConverter via ucnv_close().
593 * For most methods see the LocalPointerBase base class.
594 *
595 * @see LocalPointerBase
596 * @see LocalPointer
597 * @stable ICU 4.4
598 */
599U_DEFINE_LOCAL_OPEN_POINTER(LocalUConverterPointer, UConverter, ucnv_close);
600
601U_NAMESPACE_END
602
603#endif
604
605/**
606 * Fills in the output parameter, subChars, with the substitution characters
607 * as multiple bytes.
608 * If ucnv_setSubstString() set a Unicode string because the converter is
609 * stateful, then subChars will be an empty string.
610 *
611 * @param converter the Unicode converter
612 * @param subChars the substitution characters
613 * @param len on input the capacity of subChars, on output the number
614 * of bytes copied to it
615 * @param err the outgoing error status code.
616 * If the substitution character array is too small, an
617 * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
618 * @see ucnv_setSubstString
619 * @see ucnv_setSubstChars
620 * @stable ICU 2.0
621 */
622U_CAPI void U_EXPORT2
623ucnv_getSubstChars(const UConverter *converter,
624 char *subChars,
625 int8_t *len,
626 UErrorCode *err);
627
628/**
629 * Sets the substitution chars when converting from unicode to a codepage. The
630 * substitution is specified as a string of 1-4 bytes, and may contain
631 * <TT>NULL</TT> bytes.
632 * The subChars must represent a single character. The caller needs to know the
633 * byte sequence of a valid character in the converter's charset.
634 * For some converters, for example some ISO 2022 variants, only single-byte
635 * substitution characters may be supported.
636 * The newer ucnv_setSubstString() function relaxes these limitations.
637 *
638 * @param converter the Unicode converter
639 * @param subChars the substitution character byte sequence we want set
640 * @param len the number of bytes in subChars
641 * @param err the error status code. <TT>U_INDEX_OUTOFBOUNDS_ERROR </TT> if
642 * len is bigger than the maximum number of bytes allowed in subchars
643 * @see ucnv_setSubstString
644 * @see ucnv_getSubstChars
645 * @stable ICU 2.0
646 */
647U_CAPI void U_EXPORT2
648ucnv_setSubstChars(UConverter *converter,
649 const char *subChars,
650 int8_t len,
651 UErrorCode *err);
652
653/**
654 * Set a substitution string for converting from Unicode to a charset.
655 * The caller need not know the charset byte sequence for each charset.
656 *
657 * Unlike ucnv_setSubstChars() which is designed to set a charset byte sequence
658 * for a single character, this function takes a Unicode string with
659 * zero, one or more characters, and immediately verifies that the string can be
660 * converted to the charset.
661 * If not, or if the result is too long (more than 32 bytes as of ICU 3.6),
662 * then the function returns with an error accordingly.
663 *
664 * Also unlike ucnv_setSubstChars(), this function works for stateful charsets
665 * by converting on the fly at the point of substitution rather than setting
666 * a fixed byte sequence.
667 *
668 * @param cnv The UConverter object.
669 * @param s The Unicode string.
670 * @param length The number of UChars in s, or -1 for a NUL-terminated string.
671 * @param err Pointer to a standard ICU error code. Its input value must
672 * pass the U_SUCCESS() test, or else the function returns
673 * immediately. Check for U_FAILURE() on output or use with
674 * function chaining. (See User Guide for details.)
675 *
676 * @see ucnv_setSubstChars
677 * @see ucnv_getSubstChars
678 * @stable ICU 3.6
679 */
680U_CAPI void U_EXPORT2
681ucnv_setSubstString(UConverter *cnv,
682 const UChar *s,
683 int32_t length,
684 UErrorCode *err);
685
686/**
687 * Fills in the output parameter, errBytes, with the error characters from the
688 * last failing conversion.
689 *
690 * @param converter the Unicode converter
691 * @param errBytes the codepage bytes which were in error
692 * @param len on input the capacity of errBytes, on output the number of
693 * bytes which were copied to it
694 * @param err the error status code.
695 * If the substitution character array is too small, an
696 * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
697 * @stable ICU 2.0
698 */
699U_CAPI void U_EXPORT2
700ucnv_getInvalidChars(const UConverter *converter,
701 char *errBytes,
702 int8_t *len,
703 UErrorCode *err);
704
705/**
706 * Fills in the output parameter, errChars, with the error characters from the
707 * last failing conversion.
708 *
709 * @param converter the Unicode converter
710 * @param errUChars the UChars which were in error
711 * @param len on input the capacity of errUChars, on output the number of
712 * UChars which were copied to it
713 * @param err the error status code.
714 * If the substitution character array is too small, an
715 * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
716 * @stable ICU 2.0
717 */
718U_CAPI void U_EXPORT2
719ucnv_getInvalidUChars(const UConverter *converter,
720 UChar *errUChars,
721 int8_t *len,
722 UErrorCode *err);
723
724/**
725 * Resets the state of a converter to the default state. This is used
726 * in the case of an error, to restart a conversion from a known default state.
727 * It will also empty the internal output buffers.
728 * @param converter the Unicode converter
729 * @stable ICU 2.0
730 */
731U_CAPI void U_EXPORT2
732ucnv_reset(UConverter *converter);
733
734/**
735 * Resets the to-Unicode part of a converter state to the default state.
736 * This is used in the case of an error to restart a conversion to
737 * Unicode to a known default state. It will also empty the internal
738 * output buffers used for the conversion to Unicode codepoints.
739 * @param converter the Unicode converter
740 * @stable ICU 2.0
741 */
742U_CAPI void U_EXPORT2
743ucnv_resetToUnicode(UConverter *converter);
744
745/**
746 * Resets the from-Unicode part of a converter state to the default state.
747 * This is used in the case of an error to restart a conversion from
748 * Unicode to a known default state. It will also empty the internal output
749 * buffers used for the conversion from Unicode codepoints.
750 * @param converter the Unicode converter
751 * @stable ICU 2.0
752 */
753U_CAPI void U_EXPORT2
754ucnv_resetFromUnicode(UConverter *converter);
755
756/**
757 * Returns the maximum number of bytes that are output per UChar in conversion
758 * from Unicode using this converter.
759 * The returned number can be used with UCNV_GET_MAX_BYTES_FOR_STRING
760 * to calculate the size of a target buffer for conversion from Unicode.
761 *
762 * Note: Before ICU 2.8, this function did not return reliable numbers for
763 * some stateful converters (EBCDIC_STATEFUL, ISO-2022) and LMBCS.
764 *
765 * This number may not be the same as the maximum number of bytes per
766 * "conversion unit". In other words, it may not be the intuitively expected
767 * number of bytes per character that would be published for a charset,
768 * and may not fulfill any other purpose than the allocation of an output
769 * buffer of guaranteed sufficient size for a given input length and converter.
770 *
771 * Examples for special cases that are taken into account:
772 * - Supplementary code points may convert to more bytes than BMP code points.
773 * This function returns bytes per UChar (UTF-16 code unit), not per
774 * Unicode code point, for efficient buffer allocation.
775 * - State-shifting output (SI/SO, escapes, etc.) from stateful converters.
776 * - When m input UChars are converted to n output bytes, then the maximum m/n
777 * is taken into account.
778 *
779 * The number returned here does not take into account
780 * (see UCNV_GET_MAX_BYTES_FOR_STRING):
781 * - callbacks which output more than one charset character sequence per call,
782 * like escape callbacks
783 * - initial and final non-character bytes that are output by some converters
784 * (automatic BOMs, initial escape sequence, final SI, etc.)
785 *
786 * Examples for returned values:
787 * - SBCS charsets: 1
788 * - Shift-JIS: 2
789 * - UTF-16: 2 (2 per BMP, 4 per surrogate _pair_, BOM not counted)
790 * - UTF-8: 3 (3 per BMP, 4 per surrogate _pair_)
791 * - EBCDIC_STATEFUL (EBCDIC mixed SBCS/DBCS): 3 (SO + DBCS)
792 * - ISO-2022: 3 (always outputs UTF-8)
793 * - ISO-2022-JP: 6 (4-byte escape sequences + DBCS)
794 * - ISO-2022-CN: 8 (4-byte designator sequences + 2-byte SS2/SS3 + DBCS)
795 *
796 * @param converter The Unicode converter.
797 * @return The maximum number of bytes per UChar (16 bit code unit)
798 * that are output by ucnv_fromUnicode(),
799 * to be used together with UCNV_GET_MAX_BYTES_FOR_STRING
800 * for buffer allocation.
801 *
802 * @see UCNV_GET_MAX_BYTES_FOR_STRING
803 * @see ucnv_getMinCharSize
804 * @stable ICU 2.0
805 */
806U_CAPI int8_t U_EXPORT2
807ucnv_getMaxCharSize(const UConverter *converter);
808
809/**
810 * Calculates the size of a buffer for conversion from Unicode to a charset.
811 * The calculated size is guaranteed to be sufficient for this conversion.
812 *
813 * It takes into account initial and final non-character bytes that are output
814 * by some converters.
815 * It does not take into account callbacks which output more than one charset
816 * character sequence per call, like escape callbacks.
817 * The default (substitution) callback only outputs one charset character sequence.
818 *
819 * @param length Number of UChars to be converted.
820 * @param maxCharSize Return value from ucnv_getMaxCharSize() for the converter
821 * that will be used.
822 * @return Size of a buffer that will be large enough to hold the output bytes of
823 * converting length UChars with the converter that returned the maxCharSize.
824 *
825 * @see ucnv_getMaxCharSize
826 * @stable ICU 2.8
827 */
828#define UCNV_GET_MAX_BYTES_FOR_STRING(length, maxCharSize) \
829 (((int32_t)(length)+10)*(int32_t)(maxCharSize))
830
831/**
832 * Returns the minimum byte length (per codepoint) for characters in this codepage.
833 * This is usually either 1 or 2.
834 * @param converter the Unicode converter
835 * @return the minimum number of bytes per codepoint allowed by this particular converter
836 * @see ucnv_getMaxCharSize
837 * @stable ICU 2.0
838 */
839U_CAPI int8_t U_EXPORT2
840ucnv_getMinCharSize(const UConverter *converter);
841
842/**
843 * Returns the display name of the converter passed in based on the Locale
844 * passed in. If the locale contains no display name, the internal ASCII
845 * name will be filled in.
846 *
847 * @param converter the Unicode converter.
848 * @param displayLocale is the specific Locale we want to localized for
849 * @param displayName user provided buffer to be filled in
850 * @param displayNameCapacity size of displayName Buffer
851 * @param err error status code
852 * @return displayNameLength number of UChar needed in displayName
853 * @see ucnv_getName
854 * @stable ICU 2.0
855 */
856U_CAPI int32_t U_EXPORT2
857ucnv_getDisplayName(const UConverter *converter,
858 const char *displayLocale,
859 UChar *displayName,
860 int32_t displayNameCapacity,
861 UErrorCode *err);
862
863/**
864 * Gets the internal, canonical name of the converter (zero-terminated).
865 * The lifetime of the returned string will be that of the converter
866 * passed to this function.
867 * @param converter the Unicode converter
868 * @param err UErrorCode status
869 * @return the internal name of the converter
870 * @see ucnv_getDisplayName
871 * @stable ICU 2.0
872 */
873U_CAPI const char * U_EXPORT2
874ucnv_getName(const UConverter *converter, UErrorCode *err);
875
876/**
877 * Gets a codepage number associated with the converter. This is not guaranteed
878 * to be the one used to create the converter. Some converters do not represent
879 * platform registered codepages and return zero for the codepage number.
880 * The error code fill-in parameter indicates if the codepage number
881 * is available.
882 * Does not check if the converter is <TT>NULL</TT> or if converter's data
883 * table is <TT>NULL</TT>.
884 *
885 * Important: The use of CCSIDs is not recommended because it is limited
886 * to only two platforms in principle and only one (UCNV_IBM) in the current
887 * ICU converter API.
888 * Also, CCSIDs are insufficient to identify IBM Unicode conversion tables precisely.
889 * For more details see ucnv_openCCSID().
890 *
891 * @param converter the Unicode converter
892 * @param err the error status code.
893 * @return If any error occurs, -1 will be returned otherwise, the codepage number
894 * will be returned
895 * @see ucnv_openCCSID
896 * @see ucnv_getPlatform
897 * @stable ICU 2.0
898 */
899U_CAPI int32_t U_EXPORT2
900ucnv_getCCSID(const UConverter *converter,
901 UErrorCode *err);
902
903/**
904 * Gets a codepage platform associated with the converter. Currently,
905 * only <TT>UCNV_IBM</TT> will be returned.
906 * Does not test if the converter is <TT>NULL</TT> or if converter's data
907 * table is <TT>NULL</TT>.
908 * @param converter the Unicode converter
909 * @param err the error status code.
910 * @return The codepage platform
911 * @stable ICU 2.0
912 */
913U_CAPI UConverterPlatform U_EXPORT2
914ucnv_getPlatform(const UConverter *converter,
915 UErrorCode *err);
916
917/**
918 * Gets the type of the converter
919 * e.g. SBCS, MBCS, DBCS, UTF8, UTF16_BE, UTF16_LE, ISO_2022,
920 * EBCDIC_STATEFUL, LATIN_1
921 * @param converter a valid, opened converter
922 * @return the type of the converter
923 * @stable ICU 2.0
924 */
925U_CAPI UConverterType U_EXPORT2
926ucnv_getType(const UConverter * converter);
927
928/**
929 * Gets the "starter" (lead) bytes for converters of type MBCS.
930 * Will fill in an <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> if converter passed in
931 * is not MBCS. Fills in an array of type UBool, with the value of the byte
932 * as offset to the array. For example, if (starters[0x20] == true) at return,
933 * it means that the byte 0x20 is a starter byte in this converter.
934 * Context pointers are always owned by the caller.
935 *
936 * @param converter a valid, opened converter of type MBCS
937 * @param starters an array of size 256 to be filled in
938 * @param err error status, <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> if the
939 * converter is not a type which can return starters.
940 * @see ucnv_getType
941 * @stable ICU 2.0
942 */
943U_CAPI void U_EXPORT2
944ucnv_getStarters(const UConverter* converter,
945 UBool starters[256],
946 UErrorCode* err);
947
948
949/**
950 * Selectors for Unicode sets that can be returned by ucnv_getUnicodeSet().
951 * @see ucnv_getUnicodeSet
952 * @stable ICU 2.6
953 */
954typedef enum UConverterUnicodeSet {
955 /** Select the set of roundtrippable Unicode code points. @stable ICU 2.6 */
956 UCNV_ROUNDTRIP_SET,
957 /** Select the set of Unicode code points with roundtrip or fallback mappings. @stable ICU 4.0 */
958 UCNV_ROUNDTRIP_AND_FALLBACK_SET,
959#ifndef U_HIDE_DEPRECATED_API
960 /**
961 * Number of UConverterUnicodeSet selectors.
962 * @deprecated ICU 58 The numeric value may change over time, see ICU ticket #12420.
963 */
964 UCNV_SET_COUNT
965#endif // U_HIDE_DEPRECATED_API
966} UConverterUnicodeSet;
967
968
969/**
970 * Returns the set of Unicode code points that can be converted by an ICU converter.
971 *
972 * Returns one of several kinds of set:
973 *
974 * 1. UCNV_ROUNDTRIP_SET
975 *
976 * The set of all Unicode code points that can be roundtrip-converted
977 * (converted without any data loss) with the converter (ucnv_fromUnicode()).
978 * This set will not include code points that have fallback mappings
979 * or are only the result of reverse fallback mappings.
980 * This set will also not include PUA code points with fallbacks, although
981 * ucnv_fromUnicode() will always uses those mappings despite ucnv_setFallback().
982 * See UTR #22 "Character Mapping Markup Language"
983 * at http://www.unicode.org/reports/tr22/
984 *
985 * This is useful for example for
986 * - checking that a string or document can be roundtrip-converted with a converter,
987 * without/before actually performing the conversion
988 * - testing if a converter can be used for text for typical text for a certain locale,
989 * by comparing its roundtrip set with the set of ExemplarCharacters from
990 * ICU's locale data or other sources
991 *
992 * 2. UCNV_ROUNDTRIP_AND_FALLBACK_SET
993 *
994 * The set of all Unicode code points that can be converted with the converter (ucnv_fromUnicode())
995 * when fallbacks are turned on (see ucnv_setFallback()).
996 * This set includes all code points with roundtrips and fallbacks (but not reverse fallbacks).
997 *
998 * In the future, there may be more UConverterUnicodeSet choices to select
999 * sets with different properties.
1000 *
1001 * @param cnv The converter for which a set is requested.
1002 * @param setFillIn A valid USet *. It will be cleared by this function before
1003 * the converter's specific set is filled into the USet.
1004 * @param whichSet A UConverterUnicodeSet selector;
1005 * currently UCNV_ROUNDTRIP_SET is the only supported value.
1006 * @param pErrorCode ICU error code in/out parameter.
1007 * Must fulfill U_SUCCESS before the function call.
1008 *
1009 * @see UConverterUnicodeSet
1010 * @see uset_open
1011 * @see uset_close
1012 * @stable ICU 2.6
1013 */
1014U_CAPI void U_EXPORT2
1015ucnv_getUnicodeSet(const UConverter *cnv,
1016 USet *setFillIn,
1017 UConverterUnicodeSet whichSet,
1018 UErrorCode *pErrorCode);
1019
1020/**
1021 * Gets the current callback function used by the converter when an illegal
1022 * or invalid codepage sequence is found.
1023 * Context pointers are always owned by the caller.
1024 *
1025 * @param converter the unicode converter
1026 * @param action fillin: returns the callback function pointer
1027 * @param context fillin: returns the callback's private void* context
1028 * @see ucnv_setToUCallBack
1029 * @stable ICU 2.0
1030 */
1031U_CAPI void U_EXPORT2
1032ucnv_getToUCallBack (const UConverter * converter,
1033 UConverterToUCallback *action,
1034 const void **context);
1035
1036/**
1037 * Gets the current callback function used by the converter when illegal
1038 * or invalid Unicode sequence is found.
1039 * Context pointers are always owned by the caller.
1040 *
1041 * @param converter the unicode converter
1042 * @param action fillin: returns the callback function pointer
1043 * @param context fillin: returns the callback's private void* context
1044 * @see ucnv_setFromUCallBack
1045 * @stable ICU 2.0
1046 */
1047U_CAPI void U_EXPORT2
1048ucnv_getFromUCallBack (const UConverter * converter,
1049 UConverterFromUCallback *action,
1050 const void **context);
1051
1052/**
1053 * Changes the callback function used by the converter when
1054 * an illegal or invalid sequence is found.
1055 * Context pointers are always owned by the caller.
1056 * Predefined actions and contexts can be found in the ucnv_err.h header.
1057 *
1058 * @param converter the unicode converter
1059 * @param newAction the new callback function
1060 * @param newContext the new toUnicode callback context pointer. This can be NULL.
1061 * @param oldAction fillin: returns the old callback function pointer. This can be NULL.
1062 * @param oldContext fillin: returns the old callback's private void* context. This can be NULL.
1063 * @param err The error code status
1064 * @see ucnv_getToUCallBack
1065 * @stable ICU 2.0
1066 */
1067U_CAPI void U_EXPORT2
1068ucnv_setToUCallBack (UConverter * converter,
1069 UConverterToUCallback newAction,
1070 const void* newContext,
1071 UConverterToUCallback *oldAction,
1072 const void** oldContext,
1073 UErrorCode * err);
1074
1075/**
1076 * Changes the current callback function used by the converter when
1077 * an illegal or invalid sequence is found.
1078 * Context pointers are always owned by the caller.
1079 * Predefined actions and contexts can be found in the ucnv_err.h header.
1080 *
1081 * @param converter the unicode converter
1082 * @param newAction the new callback function
1083 * @param newContext the new fromUnicode callback context pointer. This can be NULL.
1084 * @param oldAction fillin: returns the old callback function pointer. This can be NULL.
1085 * @param oldContext fillin: returns the old callback's private void* context. This can be NULL.
1086 * @param err The error code status
1087 * @see ucnv_getFromUCallBack
1088 * @stable ICU 2.0
1089 */
1090U_CAPI void U_EXPORT2
1091ucnv_setFromUCallBack (UConverter * converter,
1092 UConverterFromUCallback newAction,
1093 const void *newContext,
1094 UConverterFromUCallback *oldAction,
1095 const void **oldContext,
1096 UErrorCode * err);
1097
1098/**
1099 * Converts an array of unicode characters to an array of codepage
1100 * characters. This function is optimized for converting a continuous
1101 * stream of data in buffer-sized chunks, where the entire source and
1102 * target does not fit in available buffers.
1103 *
1104 * The source pointer is an in/out parameter. It starts out pointing where the
1105 * conversion is to begin, and ends up pointing after the last UChar consumed.
1106 *
1107 * Target similarly starts out pointer at the first available byte in the output
1108 * buffer, and ends up pointing after the last byte written to the output.
1109 *
1110 * The converter always attempts to consume the entire source buffer, unless
1111 * (1.) the target buffer is full, or (2.) a failing error is returned from the
1112 * current callback function. When a successful error status has been
1113 * returned, it means that all of the source buffer has been
1114 * consumed. At that point, the caller should reset the source and
1115 * sourceLimit pointers to point to the next chunk.
1116 *
1117 * At the end of the stream (flush==true), the input is completely consumed
1118 * when *source==sourceLimit and no error code is set.
1119 * The converter object is then automatically reset by this function.
1120 * (This means that a converter need not be reset explicitly between data
1121 * streams if it finishes the previous stream without errors.)
1122 *
1123 * This is a <I>stateful</I> conversion. Additionally, even when all source data has
1124 * been consumed, some data may be in the converters' internal state.
1125 * Call this function repeatedly, updating the target pointers with
1126 * the next empty chunk of target in case of a
1127 * <TT>U_BUFFER_OVERFLOW_ERROR</TT>, and updating the source pointers
1128 * with the next chunk of source when a successful error status is
1129 * returned, until there are no more chunks of source data.
1130 * @param converter the Unicode converter
1131 * @param target I/O parameter. Input : Points to the beginning of the buffer to copy
1132 * codepage characters to. Output : points to after the last codepage character copied
1133 * to <TT>target</TT>.
1134 * @param targetLimit the pointer just after last of the <TT>target</TT> buffer
1135 * @param source I/O parameter, pointer to pointer to the source Unicode character buffer.
1136 * @param sourceLimit the pointer just after the last of the source buffer
1137 * @param offsets if NULL is passed, nothing will happen to it, otherwise it needs to have the same number
1138 * of allocated cells as <TT>target</TT>. Will fill in offsets from target to source pointer
1139 * e.g: <TT>offsets[3]</TT> is equal to 6, it means that the <TT>target[3]</TT> was a result of transcoding <TT>source[6]</TT>
1140 * For output data carried across calls, and other data without a specific source character
1141 * (such as from escape sequences or callbacks) -1 will be placed for offsets.
1142 * @param flush set to <TT>true</TT> if the current source buffer is the last available
1143 * chunk of the source, <TT>false</TT> otherwise. Note that if a failing status is returned,
1144 * this function may have to be called multiple times with flush set to <TT>true</TT> until
1145 * the source buffer is consumed.
1146 * @param err the error status. <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> will be set if the
1147 * converter is <TT>NULL</TT>.
1148 * <code>U_BUFFER_OVERFLOW_ERROR</code> will be set if the target is full and there is
1149 * still data to be written to the target.
1150 * @see ucnv_fromUChars
1151 * @see ucnv_convert
1152 * @see ucnv_getMinCharSize
1153 * @see ucnv_setToUCallBack
1154 * @stable ICU 2.0
1155 */
1156U_CAPI void U_EXPORT2
1157ucnv_fromUnicode (UConverter * converter,
1158 char **target,
1159 const char *targetLimit,
1160 const UChar ** source,
1161 const UChar * sourceLimit,
1162 int32_t* offsets,
1163 UBool flush,
1164 UErrorCode * err);
1165
1166/**
1167 * Converts a buffer of codepage bytes into an array of unicode UChars
1168 * characters. This function is optimized for converting a continuous
1169 * stream of data in buffer-sized chunks, where the entire source and
1170 * target does not fit in available buffers.
1171 *
1172 * The source pointer is an in/out parameter. It starts out pointing where the
1173 * conversion is to begin, and ends up pointing after the last byte of source consumed.
1174 *
1175 * Target similarly starts out pointer at the first available UChar in the output
1176 * buffer, and ends up pointing after the last UChar written to the output.
1177 * It does NOT necessarily keep UChar sequences together.
1178 *
1179 * The converter always attempts to consume the entire source buffer, unless
1180 * (1.) the target buffer is full, or (2.) a failing error is returned from the
1181 * current callback function. When a successful error status has been
1182 * returned, it means that all of the source buffer has been
1183 * consumed. At that point, the caller should reset the source and
1184 * sourceLimit pointers to point to the next chunk.
1185 *
1186 * At the end of the stream (flush==true), the input is completely consumed
1187 * when *source==sourceLimit and no error code is set
1188 * The converter object is then automatically reset by this function.
1189 * (This means that a converter need not be reset explicitly between data
1190 * streams if it finishes the previous stream without errors.)
1191 *
1192 * This is a <I>stateful</I> conversion. Additionally, even when all source data has
1193 * been consumed, some data may be in the converters' internal state.
1194 * Call this function repeatedly, updating the target pointers with
1195 * the next empty chunk of target in case of a
1196 * <TT>U_BUFFER_OVERFLOW_ERROR</TT>, and updating the source pointers
1197 * with the next chunk of source when a successful error status is
1198 * returned, until there are no more chunks of source data.
1199 * @param converter the Unicode converter
1200 * @param target I/O parameter. Input : Points to the beginning of the buffer to copy
1201 * UChars into. Output : points to after the last UChar copied.
1202 * @param targetLimit the pointer just after the end of the <TT>target</TT> buffer
1203 * @param source I/O parameter, pointer to pointer to the source codepage buffer.
1204 * @param sourceLimit the pointer to the byte after the end of the source buffer
1205 * @param offsets if NULL is passed, nothing will happen to it, otherwise it needs to have the same number
1206 * of allocated cells as <TT>target</TT>. Will fill in offsets from target to source pointer
1207 * e.g: <TT>offsets[3]</TT> is equal to 6, it means that the <TT>target[3]</TT> was a result of transcoding <TT>source[6]</TT>
1208 * For output data carried across calls, and other data without a specific source character
1209 * (such as from escape sequences or callbacks) -1 will be placed for offsets.
1210 * @param flush set to <TT>true</TT> if the current source buffer is the last available
1211 * chunk of the source, <TT>false</TT> otherwise. Note that if a failing status is returned,
1212 * this function may have to be called multiple times with flush set to <TT>true</TT> until
1213 * the source buffer is consumed.
1214 * @param err the error status. <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> will be set if the
1215 * converter is <TT>NULL</TT>.
1216 * <code>U_BUFFER_OVERFLOW_ERROR</code> will be set if the target is full and there is
1217 * still data to be written to the target.
1218 * @see ucnv_fromUChars
1219 * @see ucnv_convert
1220 * @see ucnv_getMinCharSize
1221 * @see ucnv_setFromUCallBack
1222 * @see ucnv_getNextUChar
1223 * @stable ICU 2.0
1224 */
1225U_CAPI void U_EXPORT2
1226ucnv_toUnicode(UConverter *converter,
1227 UChar **target,
1228 const UChar *targetLimit,
1229 const char **source,
1230 const char *sourceLimit,
1231 int32_t *offsets,
1232 UBool flush,
1233 UErrorCode *err);
1234
1235/**
1236 * Convert the Unicode string into a codepage string using an existing UConverter.
1237 * The output string is NUL-terminated if possible.
1238 *
1239 * This function is a more convenient but less powerful version of ucnv_fromUnicode().
1240 * It is only useful for whole strings, not for streaming conversion.
1241 *
1242 * The maximum output buffer capacity required (barring output from callbacks) will be
1243 * UCNV_GET_MAX_BYTES_FOR_STRING(srcLength, ucnv_getMaxCharSize(cnv)).
1244 *
1245 * @param cnv the converter object to be used (ucnv_resetFromUnicode() will be called)
1246 * @param src the input Unicode string
1247 * @param srcLength the input string length, or -1 if NUL-terminated
1248 * @param dest destination string buffer, can be NULL if destCapacity==0
1249 * @param destCapacity the number of chars available at dest
1250 * @param pErrorCode normal ICU error code;
1251 * common error codes that may be set by this function include
1252 * U_BUFFER_OVERFLOW_ERROR, U_STRING_NOT_TERMINATED_WARNING,
1253 * U_ILLEGAL_ARGUMENT_ERROR, and conversion errors
1254 * @return the length of the output string, not counting the terminating NUL;
1255 * if the length is greater than destCapacity, then the string will not fit
1256 * and a buffer of the indicated length would need to be passed in
1257 * @see ucnv_fromUnicode
1258 * @see ucnv_convert
1259 * @see UCNV_GET_MAX_BYTES_FOR_STRING
1260 * @stable ICU 2.0
1261 */
1262U_CAPI int32_t U_EXPORT2
1263ucnv_fromUChars(UConverter *cnv,
1264 char *dest, int32_t destCapacity,
1265 const UChar *src, int32_t srcLength,
1266 UErrorCode *pErrorCode);
1267
1268/**
1269 * Convert the codepage string into a Unicode string using an existing UConverter.
1270 * The output string is NUL-terminated if possible.
1271 *
1272 * This function is a more convenient but less powerful version of ucnv_toUnicode().
1273 * It is only useful for whole strings, not for streaming conversion.
1274 *
1275 * The maximum output buffer capacity required (barring output from callbacks) will be
1276 * 2*srcLength (each char may be converted into a surrogate pair).
1277 *
1278 * @param cnv the converter object to be used (ucnv_resetToUnicode() will be called)
1279 * @param src the input codepage string
1280 * @param srcLength the input string length, or -1 if NUL-terminated
1281 * @param dest destination string buffer, can be NULL if destCapacity==0
1282 * @param destCapacity the number of UChars available at dest
1283 * @param pErrorCode normal ICU error code;
1284 * common error codes that may be set by this function include
1285 * U_BUFFER_OVERFLOW_ERROR, U_STRING_NOT_TERMINATED_WARNING,
1286 * U_ILLEGAL_ARGUMENT_ERROR, and conversion errors
1287 * @return the length of the output string, not counting the terminating NUL;
1288 * if the length is greater than destCapacity, then the string will not fit
1289 * and a buffer of the indicated length would need to be passed in
1290 * @see ucnv_toUnicode
1291 * @see ucnv_convert
1292 * @stable ICU 2.0
1293 */
1294U_CAPI int32_t U_EXPORT2
1295ucnv_toUChars(UConverter *cnv,
1296 UChar *dest, int32_t destCapacity,
1297 const char *src, int32_t srcLength,
1298 UErrorCode *pErrorCode);
1299
1300/**
1301 * Convert a codepage buffer into Unicode one character at a time.
1302 * The input is completely consumed when the U_INDEX_OUTOFBOUNDS_ERROR is set.
1303 *
1304 * Advantage compared to ucnv_toUnicode() or ucnv_toUChars():
1305 * - Faster for small amounts of data, for most converters, e.g.,
1306 * US-ASCII, ISO-8859-1, UTF-8/16/32, and most "normal" charsets.
1307 * (For complex converters, e.g., SCSU, UTF-7 and ISO 2022 variants,
1308 * it uses ucnv_toUnicode() internally.)
1309 * - Convenient.
1310 *
1311 * Limitations compared to ucnv_toUnicode():
1312 * - Always assumes flush=true.
1313 * This makes ucnv_getNextUChar() unsuitable for "streaming" conversion,
1314 * that is, for where the input is supplied in multiple buffers,
1315 * because ucnv_getNextUChar() will assume the end of the input at the end
1316 * of the first buffer.
1317 * - Does not provide offset output.
1318 *
1319 * It is possible to "mix" ucnv_getNextUChar() and ucnv_toUnicode() because
1320 * ucnv_getNextUChar() uses the current state of the converter
1321 * (unlike ucnv_toUChars() which always resets first).
1322 * However, if ucnv_getNextUChar() is called after ucnv_toUnicode()
1323 * stopped in the middle of a character sequence (with flush=false),
1324 * then ucnv_getNextUChar() will always use the slower ucnv_toUnicode()
1325 * internally until the next character boundary.
1326 * (This is new in ICU 2.6. In earlier releases, ucnv_getNextUChar() had to
1327 * start at a character boundary.)
1328 *
1329 * Instead of using ucnv_getNextUChar(), it is recommended
1330 * to convert using ucnv_toUnicode() or ucnv_toUChars()
1331 * and then iterate over the text using U16_NEXT() or a UCharIterator (uiter.h)
1332 * or a C++ CharacterIterator or similar.
1333 * This allows streaming conversion and offset output, for example.
1334 *
1335 * <p>Handling of surrogate pairs and supplementary-plane code points:<br>
1336 * There are two different kinds of codepages that provide mappings for surrogate characters:
1337 * <ul>
1338 * <li>Codepages like UTF-8, UTF-32, and GB 18030 provide direct representations for Unicode
1339 * code points U+10000-U+10ffff as well as for single surrogates U+d800-U+dfff.
1340 * Each valid sequence will result in exactly one returned code point.
1341 * If a sequence results in a single surrogate, then that will be returned
1342 * by itself, even if a neighboring sequence encodes the matching surrogate.</li>
1343 * <li>Codepages like SCSU and LMBCS (and UTF-16) provide direct representations only for BMP code points
1344 * including surrogates. Code points in supplementary planes are represented with
1345 * two sequences, each encoding a surrogate.
1346 * For these codepages, matching pairs of surrogates will be combined into single
1347 * code points for returning from this function.
1348 * (Note that SCSU is actually a mix of these codepage types.)</li>
1349 * </ul></p>
1350 *
1351 * @param converter an open UConverter
1352 * @param source the address of a pointer to the codepage buffer, will be
1353 * updated to point after the bytes consumed in the conversion call.
1354 * @param sourceLimit points to the end of the input buffer
1355 * @param err fills in error status (see ucnv_toUnicode)
1356 * <code>U_INDEX_OUTOFBOUNDS_ERROR</code> will be set if the input
1357 * is empty or does not convert to any output (e.g.: pure state-change
1358 * codes SI/SO, escape sequences for ISO 2022,
1359 * or if the callback did not output anything, ...).
1360 * This function will not set a <code>U_BUFFER_OVERFLOW_ERROR</code> because
1361 * the "buffer" is the return code. However, there might be subsequent output
1362 * stored in the converter object
1363 * that will be returned in following calls to this function.
1364 * @return a UChar32 resulting from the partial conversion of source
1365 * @see ucnv_toUnicode
1366 * @see ucnv_toUChars
1367 * @see ucnv_convert
1368 * @stable ICU 2.0
1369 */
1370U_CAPI UChar32 U_EXPORT2
1371ucnv_getNextUChar(UConverter * converter,
1372 const char **source,
1373 const char * sourceLimit,
1374 UErrorCode * err);
1375
1376/**
1377 * Convert from one external charset to another using two existing UConverters.
1378 * Internally, two conversions - ucnv_toUnicode() and ucnv_fromUnicode() -
1379 * are used, "pivoting" through 16-bit Unicode.
1380 *
1381 * Important: For streaming conversion (multiple function calls for successive
1382 * parts of a text stream), the caller must provide a pivot buffer explicitly,
1383 * and must preserve the pivot buffer and associated pointers from one
1384 * call to another. (The buffer may be moved if its contents and the relative
1385 * pointer positions are preserved.)
1386 *
1387 * There is a similar function, ucnv_convert(),
1388 * which has the following limitations:
1389 * - it takes charset names, not converter objects, so that
1390 * - two converters are opened for each call
1391 * - only single-string conversion is possible, not streaming operation
1392 * - it does not provide enough information to find out,
1393 * in case of failure, whether the toUnicode or
1394 * the fromUnicode conversion failed
1395 *
1396 * By contrast, ucnv_convertEx()
1397 * - takes UConverter parameters instead of charset names
1398 * - fully exposes the pivot buffer for streaming conversion and complete error handling
1399 *
1400 * ucnv_convertEx() also provides further convenience:
1401 * - an option to reset the converters at the beginning
1402 * (if reset==true, see parameters;
1403 * also sets *pivotTarget=*pivotSource=pivotStart)
1404 * - allow NUL-terminated input
1405 * (only a single NUL byte, will not work for charsets with multi-byte NULs)
1406 * (if sourceLimit==NULL, see parameters)
1407 * - terminate with a NUL on output
1408 * (only a single NUL byte, not useful for charsets with multi-byte NULs),
1409 * or set U_STRING_NOT_TERMINATED_WARNING if the output exactly fills
1410 * the target buffer
1411 * - the pivot buffer can be provided internally;
1412 * possible only for whole-string conversion, not streaming conversion;
1413 * in this case, the caller will not be able to get details about where an
1414 * error occurred
1415 * (if pivotStart==NULL, see below)
1416 *
1417 * The function returns when one of the following is true:
1418 * - the entire source text has been converted successfully to the target buffer
1419 * - a target buffer overflow occurred (U_BUFFER_OVERFLOW_ERROR)
1420 * - a conversion error occurred
1421 * (other U_FAILURE(), see description of pErrorCode)
1422 *
1423 * Limitation compared to the direct use of
1424 * ucnv_fromUnicode() and ucnv_toUnicode():
1425 * ucnv_convertEx() does not provide offset information.
1426 *
1427 * Limitation compared to ucnv_fromUChars() and ucnv_toUChars():
1428 * ucnv_convertEx() does not support preflighting directly.
1429 *
1430 * Sample code for converting a single string from
1431 * one external charset to UTF-8, ignoring the location of errors:
1432 *
1433 * \code
1434 * int32_t
1435 * myToUTF8(UConverter *cnv,
1436 * const char *s, int32_t length,
1437 * char *u8, int32_t capacity,
1438 * UErrorCode *pErrorCode) {
1439 * UConverter *utf8Cnv;
1440 * char *target;
1441 *
1442 * if(U_FAILURE(*pErrorCode)) {
1443 * return 0;
1444 * }
1445 *
1446 * utf8Cnv=myGetCachedUTF8Converter(pErrorCode);
1447 * if(U_FAILURE(*pErrorCode)) {
1448 * return 0;
1449 * }
1450 *
1451 * if(length<0) {
1452 * length=strlen(s);
1453 * }
1454 * target=u8;
1455 * ucnv_convertEx(utf8Cnv, cnv,
1456 * &target, u8+capacity,
1457 * &s, s+length,
1458 * NULL, NULL, NULL, NULL,
1459 * true, true,
1460 * pErrorCode);
1461 *
1462 * myReleaseCachedUTF8Converter(utf8Cnv);
1463 *
1464 * // return the output string length, but without preflighting
1465 * return (int32_t)(target-u8);
1466 * }
1467 * \endcode
1468 *
1469 * @param targetCnv Output converter, used to convert from the UTF-16 pivot
1470 * to the target using ucnv_fromUnicode().
1471 * @param sourceCnv Input converter, used to convert from the source to
1472 * the UTF-16 pivot using ucnv_toUnicode().
1473 * @param target I/O parameter, same as for ucnv_fromUChars().
1474 * Input: *target points to the beginning of the target buffer.
1475 * Output: *target points to the first unit after the last char written.
1476 * @param targetLimit Pointer to the first unit after the target buffer.
1477 * @param source I/O parameter, same as for ucnv_toUChars().
1478 * Input: *source points to the beginning of the source buffer.
1479 * Output: *source points to the first unit after the last char read.
1480 * @param sourceLimit Pointer to the first unit after the source buffer.
1481 * @param pivotStart Pointer to the UTF-16 pivot buffer. If pivotStart==NULL,
1482 * then an internal buffer is used and the other pivot
1483 * arguments are ignored and can be NULL as well.
1484 * @param pivotSource I/O parameter, same as source in ucnv_fromUChars() for
1485 * conversion from the pivot buffer to the target buffer.
1486 * @param pivotTarget I/O parameter, same as target in ucnv_toUChars() for
1487 * conversion from the source buffer to the pivot buffer.
1488 * It must be pivotStart<=*pivotSource<=*pivotTarget<=pivotLimit
1489 * and pivotStart<pivotLimit (unless pivotStart==NULL).
1490 * @param pivotLimit Pointer to the first unit after the pivot buffer.
1491 * @param reset If true, then ucnv_resetToUnicode(sourceCnv) and
1492 * ucnv_resetFromUnicode(targetCnv) are called, and the
1493 * pivot pointers are reset (*pivotTarget=*pivotSource=pivotStart).
1494 * @param flush If true, indicates the end of the input.
1495 * Passed directly to ucnv_toUnicode(), and carried over to
1496 * ucnv_fromUnicode() when the source is empty as well.
1497 * @param pErrorCode ICU error code in/out parameter.
1498 * Must fulfill U_SUCCESS before the function call.
1499 * U_BUFFER_OVERFLOW_ERROR always refers to the target buffer
1500 * because overflows into the pivot buffer are handled internally.
1501 * Other conversion errors are from the source-to-pivot
1502 * conversion if *pivotSource==pivotStart, otherwise from
1503 * the pivot-to-target conversion.
1504 *
1505 * @see ucnv_convert
1506 * @see ucnv_fromAlgorithmic
1507 * @see ucnv_toAlgorithmic
1508 * @see ucnv_fromUnicode
1509 * @see ucnv_toUnicode
1510 * @see ucnv_fromUChars
1511 * @see ucnv_toUChars
1512 * @stable ICU 2.6
1513 */
1514U_CAPI void U_EXPORT2
1515ucnv_convertEx(UConverter *targetCnv, UConverter *sourceCnv,
1516 char **target, const char *targetLimit,
1517 const char **source, const char *sourceLimit,
1518 UChar *pivotStart, UChar **pivotSource,
1519 UChar **pivotTarget, const UChar *pivotLimit,
1520 UBool reset, UBool flush,
1521 UErrorCode *pErrorCode);
1522
1523/**
1524 * Convert from one external charset to another.
1525 * Internally, two converters are opened according to the name arguments,
1526 * then the text is converted to and from the 16-bit Unicode "pivot"
1527 * using ucnv_convertEx(), then the converters are closed again.
1528 *
1529 * This is a convenience function, not an efficient way to convert a lot of text:
1530 * ucnv_convert()
1531 * - takes charset names, not converter objects, so that
1532 * - two converters are opened for each call
1533 * - only single-string conversion is possible, not streaming operation
1534 * - does not provide enough information to find out,
1535 * in case of failure, whether the toUnicode or
1536 * the fromUnicode conversion failed
1537 * - allows NUL-terminated input
1538 * (only a single NUL byte, will not work for charsets with multi-byte NULs)
1539 * (if sourceLength==-1, see parameters)
1540 * - terminate with a NUL on output
1541 * (only a single NUL byte, not useful for charsets with multi-byte NULs),
1542 * or set U_STRING_NOT_TERMINATED_WARNING if the output exactly fills
1543 * the target buffer
1544 * - a pivot buffer is provided internally
1545 *
1546 * The function returns when one of the following is true:
1547 * - the entire source text has been converted successfully to the target buffer
1548 * and either the target buffer is terminated with a single NUL byte
1549 * or the error code is set to U_STRING_NOT_TERMINATED_WARNING
1550 * - a target buffer overflow occurred (U_BUFFER_OVERFLOW_ERROR)
1551 * and the full output string length is returned ("preflighting")
1552 * - a conversion error occurred
1553 * (other U_FAILURE(), see description of pErrorCode)
1554 *
1555 * @param toConverterName The name of the converter that is used to convert
1556 * from the UTF-16 pivot buffer to the target.
1557 * @param fromConverterName The name of the converter that is used to convert
1558 * from the source to the UTF-16 pivot buffer.
1559 * @param target Pointer to the output buffer.
1560 * @param targetCapacity Capacity of the target, in bytes.
1561 * @param source Pointer to the input buffer.
1562 * @param sourceLength Length of the input text, in bytes, or -1 for NUL-terminated input.
1563 * @param pErrorCode ICU error code in/out parameter.
1564 * Must fulfill U_SUCCESS before the function call.
1565 * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
1566 * and a U_BUFFER_OVERFLOW_ERROR is set.
1567 *
1568 * @see ucnv_convertEx
1569 * @see ucnv_fromAlgorithmic
1570 * @see ucnv_toAlgorithmic
1571 * @see ucnv_fromUnicode
1572 * @see ucnv_toUnicode
1573 * @see ucnv_fromUChars
1574 * @see ucnv_toUChars
1575 * @see ucnv_getNextUChar
1576 * @stable ICU 2.0
1577 */
1578U_CAPI int32_t U_EXPORT2
1579ucnv_convert(const char *toConverterName,
1580 const char *fromConverterName,
1581 char *target,
1582 int32_t targetCapacity,
1583 const char *source,
1584 int32_t sourceLength,
1585 UErrorCode *pErrorCode);
1586
1587/**
1588 * Convert from one external charset to another.
1589 * Internally, the text is converted to and from the 16-bit Unicode "pivot"
1590 * using ucnv_convertEx(). ucnv_toAlgorithmic() works exactly like ucnv_convert()
1591 * except that the two converters need not be looked up and opened completely.
1592 *
1593 * The source-to-pivot conversion uses the cnv converter parameter.
1594 * The pivot-to-target conversion uses a purely algorithmic converter
1595 * according to the specified type, e.g., UCNV_UTF8 for a UTF-8 converter.
1596 *
1597 * Internally, the algorithmic converter is opened and closed for each
1598 * function call, which is more efficient than using the public ucnv_open()
1599 * but somewhat less efficient than only resetting an existing converter
1600 * and using ucnv_convertEx().
1601 *
1602 * This function is more convenient than ucnv_convertEx() for single-string
1603 * conversions, especially when "preflighting" is desired (returning the length
1604 * of the complete output even if it does not fit into the target buffer;
1605 * see the User Guide Strings chapter). See ucnv_convert() for details.
1606 *
1607 * @param algorithmicType UConverterType constant identifying the desired target
1608 * charset as a purely algorithmic converter.
1609 * Those are converters for Unicode charsets like
1610 * UTF-8, BOCU-1, SCSU, UTF-7, IMAP-mailbox-name, etc.,
1611 * as well as US-ASCII and ISO-8859-1.
1612 * @param cnv The converter that is used to convert
1613 * from the source to the UTF-16 pivot buffer.
1614 * @param target Pointer to the output buffer.
1615 * @param targetCapacity Capacity of the target, in bytes.
1616 * @param source Pointer to the input buffer.
1617 * @param sourceLength Length of the input text, in bytes
1618 * @param pErrorCode ICU error code in/out parameter.
1619 * Must fulfill U_SUCCESS before the function call.
1620 * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
1621 * and a U_BUFFER_OVERFLOW_ERROR is set.
1622 *
1623 * @see ucnv_fromAlgorithmic
1624 * @see ucnv_convert
1625 * @see ucnv_convertEx
1626 * @see ucnv_fromUnicode
1627 * @see ucnv_toUnicode
1628 * @see ucnv_fromUChars
1629 * @see ucnv_toUChars
1630 * @stable ICU 2.6
1631 */
1632U_CAPI int32_t U_EXPORT2
1633ucnv_toAlgorithmic(UConverterType algorithmicType,
1634 UConverter *cnv,
1635 char *target, int32_t targetCapacity,
1636 const char *source, int32_t sourceLength,
1637 UErrorCode *pErrorCode);
1638
1639/**
1640 * Convert from one external charset to another.
1641 * Internally, the text is converted to and from the 16-bit Unicode "pivot"
1642 * using ucnv_convertEx(). ucnv_fromAlgorithmic() works exactly like ucnv_convert()
1643 * except that the two converters need not be looked up and opened completely.
1644 *
1645 * The source-to-pivot conversion uses a purely algorithmic converter
1646 * according to the specified type, e.g., UCNV_UTF8 for a UTF-8 converter.
1647 * The pivot-to-target conversion uses the cnv converter parameter.
1648 *
1649 * Internally, the algorithmic converter is opened and closed for each
1650 * function call, which is more efficient than using the public ucnv_open()
1651 * but somewhat less efficient than only resetting an existing converter
1652 * and using ucnv_convertEx().
1653 *
1654 * This function is more convenient than ucnv_convertEx() for single-string
1655 * conversions, especially when "preflighting" is desired (returning the length
1656 * of the complete output even if it does not fit into the target buffer;
1657 * see the User Guide Strings chapter). See ucnv_convert() for details.
1658 *
1659 * @param cnv The converter that is used to convert
1660 * from the UTF-16 pivot buffer to the target.
1661 * @param algorithmicType UConverterType constant identifying the desired source
1662 * charset as a purely algorithmic converter.
1663 * Those are converters for Unicode charsets like
1664 * UTF-8, BOCU-1, SCSU, UTF-7, IMAP-mailbox-name, etc.,
1665 * as well as US-ASCII and ISO-8859-1.
1666 * @param target Pointer to the output buffer.
1667 * @param targetCapacity Capacity of the target, in bytes.
1668 * @param source Pointer to the input buffer.
1669 * @param sourceLength Length of the input text, in bytes
1670 * @param pErrorCode ICU error code in/out parameter.
1671 * Must fulfill U_SUCCESS before the function call.
1672 * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
1673 * and a U_BUFFER_OVERFLOW_ERROR is set.
1674 *
1675 * @see ucnv_fromAlgorithmic
1676 * @see ucnv_convert
1677 * @see ucnv_convertEx
1678 * @see ucnv_fromUnicode
1679 * @see ucnv_toUnicode
1680 * @see ucnv_fromUChars
1681 * @see ucnv_toUChars
1682 * @stable ICU 2.6
1683 */
1684U_CAPI int32_t U_EXPORT2
1685ucnv_fromAlgorithmic(UConverter *cnv,
1686 UConverterType algorithmicType,
1687 char *target, int32_t targetCapacity,
1688 const char *source, int32_t sourceLength,
1689 UErrorCode *pErrorCode);
1690
1691/**
1692 * Frees up memory occupied by unused, cached converter shared data.
1693 *
1694 * @return the number of cached converters successfully deleted
1695 * @see ucnv_close
1696 * @stable ICU 2.0
1697 */
1698U_CAPI int32_t U_EXPORT2
1699ucnv_flushCache(void);
1700
1701/**
1702 * Returns the number of available converters, as per the alias file.
1703 *
1704 * @return the number of available converters
1705 * @see ucnv_getAvailableName
1706 * @stable ICU 2.0
1707 */
1708U_CAPI int32_t U_EXPORT2
1709ucnv_countAvailable(void);
1710
1711/**
1712 * Gets the canonical converter name of the specified converter from a list of
1713 * all available converters contained in the alias file. All converters
1714 * in this list can be opened.
1715 *
1716 * @param n the index to a converter available on the system (in the range <TT>[0..ucnv_countAvailable()]</TT>)
1717 * @return a pointer a string (library owned), or <TT>NULL</TT> if the index is out of bounds.
1718 * @see ucnv_countAvailable
1719 * @stable ICU 2.0
1720 */
1721U_CAPI const char* U_EXPORT2
1722ucnv_getAvailableName(int32_t n);
1723
1724/**
1725 * Returns a UEnumeration to enumerate all of the canonical converter
1726 * names, as per the alias file, regardless of the ability to open each
1727 * converter.
1728 *
1729 * @return A UEnumeration object for getting all the recognized canonical
1730 * converter names.
1731 * @see ucnv_getAvailableName
1732 * @see uenum_close
1733 * @see uenum_next
1734 * @stable ICU 2.4
1735 */
1736U_CAPI UEnumeration * U_EXPORT2
1737ucnv_openAllNames(UErrorCode *pErrorCode);
1738
1739/**
1740 * Gives the number of aliases for a given converter or alias name.
1741 * If the alias is ambiguous, then the preferred converter is used
1742 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
1743 * This method only enumerates the listed entries in the alias file.
1744 * @param alias alias name
1745 * @param pErrorCode error status
1746 * @return number of names on alias list for given alias
1747 * @stable ICU 2.0
1748 */
1749U_CAPI uint16_t U_EXPORT2
1750ucnv_countAliases(const char *alias, UErrorCode *pErrorCode);
1751
1752/**
1753 * Gives the name of the alias at given index of alias list.
1754 * This method only enumerates the listed entries in the alias file.
1755 * If the alias is ambiguous, then the preferred converter is used
1756 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
1757 * @param alias alias name
1758 * @param n index in alias list
1759 * @param pErrorCode result of operation
1760 * @return returns the name of the alias at given index
1761 * @see ucnv_countAliases
1762 * @stable ICU 2.0
1763 */
1764U_CAPI const char * U_EXPORT2
1765ucnv_getAlias(const char *alias, uint16_t n, UErrorCode *pErrorCode);
1766
1767/**
1768 * Fill-up the list of alias names for the given alias.
1769 * This method only enumerates the listed entries in the alias file.
1770 * If the alias is ambiguous, then the preferred converter is used
1771 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
1772 * @param alias alias name
1773 * @param aliases fill-in list, aliases is a pointer to an array of
1774 * <code>ucnv_countAliases()</code> string-pointers
1775 * (<code>const char *</code>) that will be filled in.
1776 * The strings themselves are owned by the library.
1777 * @param pErrorCode result of operation
1778 * @stable ICU 2.0
1779 */
1780U_CAPI void U_EXPORT2
1781ucnv_getAliases(const char *alias, const char **aliases, UErrorCode *pErrorCode);
1782
1783/**
1784 * Return a new UEnumeration object for enumerating all the
1785 * alias names for a given converter that are recognized by a standard.
1786 * This method only enumerates the listed entries in the alias file.
1787 * The convrtrs.txt file can be modified to change the results of
1788 * this function.
1789 * The first result in this list is the same result given by
1790 * <code>ucnv_getStandardName</code>, which is the default alias for
1791 * the specified standard name. The returned object must be closed with
1792 * <code>uenum_close</code> when you are done with the object.
1793 *
1794 * @param convName original converter name
1795 * @param standard name of the standard governing the names; MIME and IANA
1796 * are such standards
1797 * @param pErrorCode The error code
1798 * @return A UEnumeration object for getting all aliases that are recognized
1799 * by a standard. If any of the parameters are invalid, NULL
1800 * is returned.
1801 * @see ucnv_getStandardName
1802 * @see uenum_close
1803 * @see uenum_next
1804 * @stable ICU 2.2
1805 */
1806U_CAPI UEnumeration * U_EXPORT2
1807ucnv_openStandardNames(const char *convName,
1808 const char *standard,
1809 UErrorCode *pErrorCode);
1810
1811/**
1812 * Gives the number of standards associated to converter names.
1813 * @return number of standards
1814 * @stable ICU 2.0
1815 */
1816U_CAPI uint16_t U_EXPORT2
1817ucnv_countStandards(void);
1818
1819/**
1820 * Gives the name of the standard at given index of standard list.
1821 * @param n index in standard list
1822 * @param pErrorCode result of operation
1823 * @return returns the name of the standard at given index. Owned by the library.
1824 * @stable ICU 2.0
1825 */
1826U_CAPI const char * U_EXPORT2
1827ucnv_getStandard(uint16_t n, UErrorCode *pErrorCode);
1828
1829/**
1830 * Returns a standard name for a given converter name.
1831 * <p>
1832 * Example alias table:<br>
1833 * conv alias1 { STANDARD1 } alias2 { STANDARD1* }
1834 * <p>
1835 * Result of ucnv_getStandardName("conv", "STANDARD1") from example
1836 * alias table:<br>
1837 * <b>"alias2"</b>
1838 *
1839 * @param name original converter name
1840 * @param standard name of the standard governing the names; MIME and IANA
1841 * are such standards
1842 * @param pErrorCode result of operation
1843 * @return returns the standard converter name;
1844 * if a standard converter name cannot be determined,
1845 * then <code>NULL</code> is returned. Owned by the library.
1846 * @stable ICU 2.0
1847 */
1848U_CAPI const char * U_EXPORT2
1849ucnv_getStandardName(const char *name, const char *standard, UErrorCode *pErrorCode);
1850
1851/**
1852 * This function will return the internal canonical converter name of the
1853 * tagged alias. This is the opposite of ucnv_openStandardNames, which
1854 * returns the tagged alias given the canonical name.
1855 * <p>
1856 * Example alias table:<br>
1857 * conv alias1 { STANDARD1 } alias2 { STANDARD1* }
1858 * <p>
1859 * Result of ucnv_getStandardName("alias1", "STANDARD1") from example
1860 * alias table:<br>
1861 * <b>"conv"</b>
1862 *
1863 * @return returns the canonical converter name;
1864 * if a standard or alias name cannot be determined,
1865 * then <code>NULL</code> is returned. The returned string is
1866 * owned by the library.
1867 * @see ucnv_getStandardName
1868 * @stable ICU 2.4
1869 */
1870U_CAPI const char * U_EXPORT2
1871ucnv_getCanonicalName(const char *alias, const char *standard, UErrorCode *pErrorCode);
1872
1873/**
1874 * Returns the current default converter name. If you want to open
1875 * a default converter, you do not need to use this function.
1876 * It is faster if you pass a NULL argument to ucnv_open the
1877 * default converter.
1878 *
1879 * If U_CHARSET_IS_UTF8 is defined to 1 in utypes.h then this function
1880 * always returns "UTF-8".
1881 *
1882 * @return returns the current default converter name.
1883 * Storage owned by the library
1884 * @see ucnv_setDefaultName
1885 * @stable ICU 2.0
1886 */
1887U_CAPI const char * U_EXPORT2
1888ucnv_getDefaultName(void);
1889
1890#ifndef U_HIDE_SYSTEM_API
1891/**
1892 * This function is not thread safe. DO NOT call this function when ANY ICU
1893 * function is being used from more than one thread! This function sets the
1894 * current default converter name. If this function needs to be called, it
1895 * should be called during application initialization. Most of the time, the
1896 * results from ucnv_getDefaultName() or ucnv_open with a NULL string argument
1897 * is sufficient for your application.
1898 *
1899 * If U_CHARSET_IS_UTF8 is defined to 1 in utypes.h then this function
1900 * does nothing.
1901 *
1902 * @param name the converter name to be the default (must be known by ICU).
1903 * @see ucnv_getDefaultName
1904 * @system
1905 * @stable ICU 2.0
1906 */
1907U_CAPI void U_EXPORT2
1908ucnv_setDefaultName(const char *name);
1909#endif /* U_HIDE_SYSTEM_API */
1910
1911/**
1912 * Fixes the backslash character mismapping. For example, in SJIS, the backslash
1913 * character in the ASCII portion is also used to represent the yen currency sign.
1914 * When mapping from Unicode character 0x005C, it's unclear whether to map the
1915 * character back to yen or backslash in SJIS. This function will take the input
1916 * buffer and replace all the yen sign characters with backslash. This is necessary
1917 * when the user tries to open a file with the input buffer on Windows.
1918 * This function will test the converter to see whether such mapping is
1919 * required. You can sometimes avoid using this function by using the correct version
1920 * of Shift-JIS.
1921 *
1922 * @param cnv The converter representing the target codepage.
1923 * @param source the input buffer to be fixed
1924 * @param sourceLen the length of the input buffer
1925 * @see ucnv_isAmbiguous
1926 * @stable ICU 2.0
1927 */
1928U_CAPI void U_EXPORT2
1929ucnv_fixFileSeparator(const UConverter *cnv, UChar *source, int32_t sourceLen);
1930
1931/**
1932 * Determines if the converter contains ambiguous mappings of the same
1933 * character or not.
1934 * @param cnv the converter to be tested
1935 * @return true if the converter contains ambiguous mapping of the same
1936 * character, false otherwise.
1937 * @stable ICU 2.0
1938 */
1939U_CAPI UBool U_EXPORT2
1940ucnv_isAmbiguous(const UConverter *cnv);
1941
1942/**
1943 * Sets the converter to use fallback mappings or not.
1944 * Regardless of this flag, the converter will always use
1945 * fallbacks from Unicode Private Use code points, as well as
1946 * reverse fallbacks (to Unicode).
1947 * For details see ".ucm File Format"
1948 * in the Conversion Data chapter of the ICU User Guide:
1949 * https://unicode-org.github.io/icu/userguide/conversion/data.html#ucm-file-format
1950 *
1951 * @param cnv The converter to set the fallback mapping usage on.
1952 * @param usesFallback true if the user wants the converter to take advantage of the fallback
1953 * mapping, false otherwise.
1954 * @stable ICU 2.0
1955 * @see ucnv_usesFallback
1956 */
1957U_CAPI void U_EXPORT2
1958ucnv_setFallback(UConverter *cnv, UBool usesFallback);
1959
1960/**
1961 * Determines if the converter uses fallback mappings or not.
1962 * This flag has restrictions, see ucnv_setFallback().
1963 *
1964 * @param cnv The converter to be tested
1965 * @return true if the converter uses fallback, false otherwise.
1966 * @stable ICU 2.0
1967 * @see ucnv_setFallback
1968 */
1969U_CAPI UBool U_EXPORT2
1970ucnv_usesFallback(const UConverter *cnv);
1971
1972/**
1973 * Detects Unicode signature byte sequences at the start of the byte stream
1974 * and returns the charset name of the indicated Unicode charset.
1975 * NULL is returned when no Unicode signature is recognized.
1976 * The number of bytes in the signature is output as well.
1977 *
1978 * The caller can ucnv_open() a converter using the charset name.
1979 * The first code unit (UChar) from the start of the stream will be U+FEFF
1980 * (the Unicode BOM/signature character) and can usually be ignored.
1981 *
1982 * For most Unicode charsets it is also possible to ignore the indicated
1983 * number of initial stream bytes and start converting after them.
1984 * However, there are stateful Unicode charsets (UTF-7 and BOCU-1) for which
1985 * this will not work. Therefore, it is best to ignore the first output UChar
1986 * instead of the input signature bytes.
1987 * <p>
1988 * Usage:
1989 * \snippet samples/ucnv/convsamp.cpp ucnv_detectUnicodeSignature
1990 *
1991 * @param source The source string in which the signature should be detected.
1992 * @param sourceLength Length of the input string, or -1 if terminated with a NUL byte.
1993 * @param signatureLength A pointer to int32_t to receive the number of bytes that make up the signature
1994 * of the detected UTF. 0 if not detected.
1995 * Can be a NULL pointer.
1996 * @param pErrorCode ICU error code in/out parameter.
1997 * Must fulfill U_SUCCESS before the function call.
1998 * @return The name of the encoding detected. NULL if encoding is not detected.
1999 * @stable ICU 2.4
2000 */
2001U_CAPI const char* U_EXPORT2
2002ucnv_detectUnicodeSignature(const char* source,
2003 int32_t sourceLength,
2004 int32_t *signatureLength,
2005 UErrorCode *pErrorCode);
2006
2007/**
2008 * Returns the number of UChars held in the converter's internal state
2009 * because more input is needed for completing the conversion. This function is
2010 * useful for mapping semantics of ICU's converter interface to those of iconv,
2011 * and this information is not needed for normal conversion.
2012 * @param cnv The converter in which the input is held
2013 * @param status ICU error code in/out parameter.
2014 * Must fulfill U_SUCCESS before the function call.
2015 * @return The number of UChars in the state. -1 if an error is encountered.
2016 * @stable ICU 3.4
2017 */
2018U_CAPI int32_t U_EXPORT2
2019ucnv_fromUCountPending(const UConverter* cnv, UErrorCode* status);
2020
2021/**
2022 * Returns the number of chars held in the converter's internal state
2023 * because more input is needed for completing the conversion. This function is
2024 * useful for mapping semantics of ICU's converter interface to those of iconv,
2025 * and this information is not needed for normal conversion.
2026 * @param cnv The converter in which the input is held as internal state
2027 * @param status ICU error code in/out parameter.
2028 * Must fulfill U_SUCCESS before the function call.
2029 * @return The number of chars in the state. -1 if an error is encountered.
2030 * @stable ICU 3.4
2031 */
2032U_CAPI int32_t U_EXPORT2
2033ucnv_toUCountPending(const UConverter* cnv, UErrorCode* status);
2034
2035/**
2036 * Returns whether or not the charset of the converter has a fixed number of bytes
2037 * per charset character.
2038 * An example of this are converters that are of the type UCNV_SBCS or UCNV_DBCS.
2039 * Another example is UTF-32 which is always 4 bytes per character.
2040 * A Unicode code point may be represented by more than one UTF-8 or UTF-16 code unit
2041 * but a UTF-32 converter encodes each code point with 4 bytes.
2042 * Note: This method is not intended to be used to determine whether the charset has a
2043 * fixed ratio of bytes to Unicode codes <i>units</i> for any particular Unicode encoding form.
2044 * false is returned with the UErrorCode if error occurs or cnv is NULL.
2045 * @param cnv The converter to be tested
2046 * @param status ICU error code in/out parameter
2047 * @return true if the converter is fixed-width
2048 * @stable ICU 4.8
2049 */
2050U_CAPI UBool U_EXPORT2
2051ucnv_isFixedWidth(UConverter *cnv, UErrorCode *status);
2052
2053#endif
2054
2055#endif
2056/*_UCNV*/
2057