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