utf-16.cpp@ 25000

Last change on this file since 25000 was 21740, checked in by vboxsync, 15 years ago
IPRT: Some latin-1 cleanup and some preditions in the recoding loops.
Property svn:eol-style set to `native` Property svn:keywords set to `Id`
File size: 29.8 KB

Line
1	/* $Id: utf-16.cpp 21740 2009-07-21 12:06:38Z vboxsync $ */
2	/** @file
3	* IPRT - UTF-16.
4	*/
5
6	/*
7	* Copyright (C) 2006-2007 Sun Microsystems, Inc.
8	*
9	* This file is part of VirtualBox Open Source Edition (OSE), as
10	* available from http://www.virtualbox.org. This file is free software;
11	* you can redistribute it and/or modify it under the terms of the GNU
12	* General Public License (GPL) as published by the Free Software
13	* Foundation, in version 2 as it comes in the "COPYING" file of the
14	* VirtualBox OSE distribution. VirtualBox OSE is distributed in the
15	* hope that it will be useful, but WITHOUT ANY WARRANTY of any kind.
16	*
17	* The contents of this file may alternatively be used under the terms
18	* of the Common Development and Distribution License Version 1.0
19	* (CDDL) only, as it comes in the "COPYING.CDDL" file of the
20	* VirtualBox OSE distribution, in which case the provisions of the
21	* CDDL are applicable instead of those of the GPL.
22	*
23	* You may elect to license modified versions of this file under the
24	* terms and conditions of either the GPL or the CDDL or both.
25	*
26	* Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa
27	* Clara, CA 95054 USA or visit http://www.sun.com if you need
28	* additional information or have any questions.
29	*/
30
31
32	/*******************************************************************************
33	* Header Files *
34	*******************************************************************************/
35	#include <iprt/string.h>
36	#include "internal/iprt.h"
37
38	#include <iprt/uni.h>
39	#include <iprt/alloc.h>
40	#include <iprt/assert.h>
41	#include <iprt/err.h>
42	#include "internal/string.h"
43
44
45
46	RTDECL(void) RTUtf16Free(PRTUTF16 pwszString)
47	{
48	if (pwszString)
49	RTMemTmpFree(pwszString);
50	}
51	RT_EXPORT_SYMBOL(RTUtf16Free);
52
53
54	RTDECL(PRTUTF16) RTUtf16Dup(PCRTUTF16 pwszString)
55	{
56	Assert(pwszString);
57	size_t cb = (RTUtf16Len(pwszString) + 1) * sizeof(RTUTF16);
58	PRTUTF16 pwsz = (PRTUTF16)RTMemAlloc(cb);
59	if (pwsz)
60	memcpy(pwsz, pwszString, cb);
61	return pwsz;
62	}
63	RT_EXPORT_SYMBOL(RTUtf16Dup);
64
65
66	RTDECL(int) RTUtf16DupEx(PRTUTF16 *ppwszString, PCRTUTF16 pwszString, size_t cwcExtra)
67	{
68	Assert(pwszString);
69	size_t cb = (RTUtf16Len(pwszString) + 1) * sizeof(RTUTF16);
70	PRTUTF16 pwsz = (PRTUTF16)RTMemAlloc(cb + cwcExtra * sizeof(RTUTF16));
71	if (pwsz)
72	{
73	memcpy(pwsz, pwszString, cb);
74	*ppwszString = pwsz;
75	return VINF_SUCCESS;
76	}
77	return VERR_NO_MEMORY;
78	}
79	RT_EXPORT_SYMBOL(RTUtf16DupEx);
80
81
82	RTDECL(size_t) RTUtf16Len(PCRTUTF16 pwszString)
83	{
84	if (!pwszString)
85	return 0;
86
87	PCRTUTF16 pwsz = pwszString;
88	while (*pwsz)
89	pwsz++;
90	return pwsz - pwszString;
91	}
92	RT_EXPORT_SYMBOL(RTUtf16Len);
93
94
95	RTDECL(int) RTUtf16Cmp(register PCRTUTF16 pwsz1, register PCRTUTF16 pwsz2)
96	{
97	if (pwsz1 == pwsz2)
98	return 0;
99	if (!pwsz1)
100	return -1;
101	if (!pwsz2)
102	return 1;
103
104	for (;;)
105	{
106	register RTUTF16 wcs = *pwsz1;
107	register int iDiff = wcs - *pwsz2;
108	if (iDiff \|\| !wcs)
109	return iDiff;
110	pwsz1++;
111	pwsz2++;
112	}
113	}
114	RT_EXPORT_SYMBOL(RTUtf16Cmp);
115
116
117	RTDECL(int) RTUtf16ICmp(register PCRTUTF16 pwsz1, register PCRTUTF16 pwsz2)
118	{
119	if (pwsz1 == pwsz2)
120	return 0;
121	if (!pwsz1)
122	return -1;
123	if (!pwsz2)
124	return 1;
125
126	PCRTUTF16 pwsz1Start = pwsz1; /* keep it around in case we have to backtrack on a surrogate pair */
127	for (;;)
128	{
129	register RTUTF16 wc1 = *pwsz1;
130	register RTUTF16 wc2 = *pwsz2;
131	register int iDiff = wc1 - wc2;
132	if (iDiff)
133	{
134	/* unless they are both surrogate pairs, there is no chance they'll be identical. */
135	if ( wc1 < 0xd800
136	\|\| wc2 < 0xd800
137	\|\| wc1 > 0xdfff
138	\|\| wc2 > 0xdfff)
139	{
140	/* simple UCS-2 char */
141	iDiff = RTUniCpToUpper(wc1) - RTUniCpToUpper(wc2);
142	if (iDiff)
143	iDiff = RTUniCpToLower(wc1) - RTUniCpToLower(wc2);
144	}
145	else
146	{
147	/* a damned pair */
148	RTUNICP uc1;
149	RTUNICP uc2;
150	if (wc1 >= 0xdc00)
151	{
152	if (pwsz1Start == pwsz1)
153	return iDiff;
154	uc1 = pwsz1[-1];
155	if (uc1 < 0xd800 \|\| uc1 >= 0xdc00)
156	return iDiff;
157	uc1 = 0x10000 + (((uc1 & 0x3ff) << 10) \| (wc1 & 0x3ff));
158	uc2 = 0x10000 + (((pwsz2[-1] & 0x3ff) << 10) \| (wc2 & 0x3ff));
159	}
160	else
161	{
162	uc1 = *++pwsz1;
163	if (uc1 < 0xdc00 \|\| uc1 >= 0xe000)
164	return iDiff;
165	uc1 = 0x10000 + (((wc1 & 0x3ff) << 10) \| (uc1 & 0x3ff));
166	uc2 = 0x10000 + (((wc2 & 0x3ff) << 10) \| (*++pwsz2 & 0x3ff));
167	}
168	iDiff = RTUniCpToUpper(uc1) - RTUniCpToUpper(uc2);
169	if (iDiff)
170	iDiff = RTUniCpToLower(uc1) - RTUniCpToLower(uc2); /* serious paranoia! */
171	}
172	if (iDiff)
173	return iDiff;
174	}
175	if (!wc1)
176	return 0;
177	pwsz1++;
178	pwsz2++;
179	}
180	}
181	RT_EXPORT_SYMBOL(RTUtf16ICmp);
182
183
184	RTDECL(PRTUTF16) RTUtf16ToLower(PRTUTF16 pwsz)
185	{
186	PRTUTF16 pwc = pwsz;
187	for (;;)
188	{
189	RTUTF16 wc = *pwc;
190	if (!wc)
191	break;
192	if (wc < 0xd800 \|\| wc >= 0xdc00)
193	{
194	RTUNICP ucFolded = RTUniCpToLower(wc);
195	if (ucFolded < 0x10000)
196	*pwc++ = RTUniCpToLower(wc);
197	}
198	else
199	{
200	/* surrogate */
201	RTUTF16 wc2 = pwc[1];
202	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
203	{
204	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
205	RTUNICP ucFolded = RTUniCpToLower(uc);
206	if (uc != ucFolded && ucFolded >= 0x10000) /* we don't support shrinking the string */
207	{
208	uc -= 0x10000;
209	*pwc++ = 0xd800 \| (uc >> 10);
210	*pwc++ = 0xdc00 \| (uc & 0x3ff);
211	}
212	}
213	else /* invalid encoding. */
214	pwc++;
215	}
216	}
217	return pwsz;
218	}
219	RT_EXPORT_SYMBOL(RTUtf16ToLower);
220
221
222	RTDECL(PRTUTF16) RTUtf16ToUpper(PRTUTF16 pwsz)
223	{
224	PRTUTF16 pwc = pwsz;
225	for (;;)
226	{
227	RTUTF16 wc = *pwc;
228	if (!wc)
229	break;
230	if (wc < 0xd800 \|\| wc >= 0xdc00)
231	*pwc++ = RTUniCpToUpper(wc);
232	else
233	{
234	/* surrogate */
235	RTUTF16 wc2 = pwc[1];
236	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
237	{
238	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
239	RTUNICP ucFolded = RTUniCpToUpper(uc);
240	if (uc != ucFolded && ucFolded >= 0x10000) /* we don't support shrinking the string */
241	{
242	uc -= 0x10000;
243	*pwc++ = 0xd800 \| (uc >> 10);
244	*pwc++ = 0xdc00 \| (uc & 0x3ff);
245	}
246	}
247	else /* invalid encoding. */
248	pwc++;
249	}
250	}
251	return pwsz;
252	}
253	RT_EXPORT_SYMBOL(RTUtf16ToUpper);
254
255
256	/**
257	* Validate the UTF-16 encoding and calculates the length of an UTF-8 encoding.
258	*
259	* @returns iprt status code.
260	* @param pwsz The UTF-16 string.
261	* @param cwc The max length of the UTF-16 string to consider.
262	* @param pcch Where to store the length (excluding '\\0') of the UTF-8 string. (cch == cb, btw)
263	*/
264	static int rtUtf16CalcUtf8Length(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
265	{
266	int rc = VINF_SUCCESS;
267	size_t cch = 0;
268	while (cwc > 0)
269	{
270	RTUTF16 wc = *pwsz++; cwc--;
271	if (!wc)
272	break;
273	else if (wc < 0xd800 \|\| wc > 0xdfff)
274	{
275	if (wc < 0x80)
276	cch++;
277	else if (wc < 0x800)
278	cch += 2;
279	else if (wc < 0xfffe)
280	cch += 3;
281	else
282	{
283	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
284	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
285	break;
286	}
287	}
288	else
289	{
290	if (wc >= 0xdc00)
291	{
292	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
293	rc = VERR_INVALID_UTF16_ENCODING;
294	break;
295	}
296	if (cwc <= 0)
297	{
298	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
299	rc = VERR_INVALID_UTF16_ENCODING;
300	break;
301	}
302	wc = *pwsz++; cwc--;
303	if (wc < 0xdc00 \|\| wc > 0xdfff)
304	{
305	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
306	rc = VERR_INVALID_UTF16_ENCODING;
307	break;
308	}
309	cch += 4;
310	}
311	}
312
313
314	/* done */
315	*pcch = cch;
316	return rc;
317	}
318
319
320	/**
321	* Recodes an valid UTF-16 string as UTF-8.
322	*
323	* @returns iprt status code.
324	* @param pwsz The UTF-16 string.
325	* @param cwc The number of RTUTF16 characters to process from pwsz. The recoding
326	* will stop when cwc or '\\0' is reached.
327	* @param psz Where to store the UTF-8 string.
328	* @param cch The size of the UTF-8 buffer, excluding the terminator.
329	* @param pcch Where to store the number of octets actually encoded.
330	*/
331	static int rtUtf16RecodeAsUtf8(PCRTUTF16 pwsz, size_t cwc, char psz, size_t cch, size_t pcch)
332	{
333	unsigned char pwch = (unsigned char )psz;
334	int rc = VINF_SUCCESS;
335	while (cwc > 0)
336	{
337	RTUTF16 wc = *pwsz++; cwc--;
338	if (!wc)
339	break;
340	else if (wc < 0xd800 \|\| wc > 0xdfff)
341	{
342	if (wc < 0x80)
343	{
344	if (RT_UNLIKELY(cch < 1))
345	{
346	RTStrAssertMsgFailed(("Buffer overflow! 1\n"));
347	rc = VERR_BUFFER_OVERFLOW;
348	break;
349	}
350	cch--;
351	*pwch++ = (unsigned char)wc;
352	}
353	else if (wc < 0x800)
354	{
355	if (RT_UNLIKELY(cch < 2))
356	{
357	RTStrAssertMsgFailed(("Buffer overflow! 2\n"));
358	rc = VERR_BUFFER_OVERFLOW;
359	break;
360	}
361	cch -= 2;
362	*pwch++ = 0xc0 \| (wc >> 6);
363	*pwch++ = 0x80 \| (wc & 0x3f);
364	}
365	else if (wc < 0xfffe)
366	{
367	if (RT_UNLIKELY(cch < 3))
368	{
369	RTStrAssertMsgFailed(("Buffer overflow! 3\n"));
370	rc = VERR_BUFFER_OVERFLOW;
371	break;
372	}
373	cch -= 3;
374	*pwch++ = 0xe0 \| (wc >> 12);
375	*pwch++ = 0x80 \| ((wc >> 6) & 0x3f);
376	*pwch++ = 0x80 \| (wc & 0x3f);
377	}
378	else
379	{
380	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
381	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
382	break;
383	}
384	}
385	else
386	{
387	if (wc >= 0xdc00)
388	{
389	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
390	rc = VERR_INVALID_UTF16_ENCODING;
391	break;
392	}
393	if (cwc <= 0)
394	{
395	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
396	rc = VERR_INVALID_UTF16_ENCODING;
397	break;
398	}
399	RTUTF16 wc2 = *pwsz++; cwc--;
400	if (wc2 < 0xdc00 \|\| wc2 > 0xdfff)
401	{
402	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
403	rc = VERR_INVALID_UTF16_ENCODING;
404	break;
405	}
406	uint32_t CodePoint = 0x10000
407	+ ( ((wc & 0x3ff) << 10)
408	\| (wc2 & 0x3ff));
409	if (RT_UNLIKELY(cch < 4))
410	{
411	RTStrAssertMsgFailed(("Buffer overflow! 4\n"));
412	rc = VERR_BUFFER_OVERFLOW;
413	break;
414	}
415	cch -= 4;
416	*pwch++ = 0xf0 \| (CodePoint >> 18);
417	*pwch++ = 0x80 \| ((CodePoint >> 12) & 0x3f);
418	*pwch++ = 0x80 \| ((CodePoint >> 6) & 0x3f);
419	*pwch++ = 0x80 \| (CodePoint & 0x3f);
420	}
421	}
422
423	/* done */
424	*pwch = '\0';
425	pcch = (char )pwch - psz;
426	return rc;
427	}
428
429
430
431	RTDECL(int) RTUtf16ToUtf8(PCRTUTF16 pwszString, char **ppszString)
432	{
433	/*
434	* Validate input.
435	*/
436	Assert(VALID_PTR(ppszString));
437	Assert(VALID_PTR(pwszString));
438	*ppszString = NULL;
439
440	/*
441	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
442	*/
443	size_t cch;
444	int rc = rtUtf16CalcUtf8Length(pwszString, RTSTR_MAX, &cch);
445	if (RT_SUCCESS(rc))
446	{
447	/*
448	* Allocate buffer and recode it.
449	*/
450	char pszResult = (char )RTMemAlloc(cch + 1);
451	if (pszResult)
452	{
453	rc = rtUtf16RecodeAsUtf8(pwszString, RTSTR_MAX, pszResult, cch, &cch);
454	if (RT_SUCCESS(rc))
455	{
456	*ppszString = pszResult;
457	return rc;
458	}
459
460	RTMemFree(pszResult);
461	}
462	else
463	rc = VERR_NO_STR_MEMORY;
464	}
465	return rc;
466	}
467	RT_EXPORT_SYMBOL(RTUtf16ToUtf8);
468
469
470	RTDECL(int) RTUtf16ToUtf8Ex(PCRTUTF16 pwszString, size_t cwcString, char *ppsz, size_t cch, size_t pcch)
471	{
472	/*
473	* Validate input.
474	*/
475	Assert(VALID_PTR(pwszString));
476	Assert(VALID_PTR(ppsz));
477	Assert(!pcch \|\| VALID_PTR(pcch));
478
479	/*
480	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
481	*/
482	size_t cchResult;
483	int rc = rtUtf16CalcUtf8Length(pwszString, cwcString, &cchResult);
484	if (RT_SUCCESS(rc))
485	{
486	if (pcch)
487	*pcch = cchResult;
488
489	/*
490	* Check buffer size / Allocate buffer and recode it.
491	*/
492	bool fShouldFree;
493	char *pszResult;
494	if (cch > 0 && *ppsz)
495	{
496	fShouldFree = false;
497	if (RT_UNLIKELY(cch <= cchResult))
498	return VERR_BUFFER_OVERFLOW;
499	pszResult = *ppsz;
500	}
501	else
502	{
503	*ppsz = NULL;
504	fShouldFree = true;
505	cch = RT_MAX(cch, cchResult + 1);
506	pszResult = (char *)RTMemAlloc(cch);
507	}
508	if (pszResult)
509	{
510	rc = rtUtf16RecodeAsUtf8(pwszString, cwcString, pszResult, cch - 1, &cch);
511	if (RT_SUCCESS(rc))
512	{
513	*ppsz = pszResult;
514	return rc;
515	}
516
517	if (fShouldFree)
518	RTMemFree(pszResult);
519	}
520	else
521	rc = VERR_NO_STR_MEMORY;
522	}
523	return rc;
524	}
525	RT_EXPORT_SYMBOL(RTUtf16ToUtf8Ex);
526
527
528	RTDECL(size_t) RTUtf16CalcUtf8Len(PCRTUTF16 pwsz)
529	{
530	size_t cch;
531	int rc = rtUtf16CalcUtf8Length(pwsz, RTSTR_MAX, &cch);
532	return RT_SUCCESS(rc) ? cch : 0;
533	}
534	RT_EXPORT_SYMBOL(RTUtf16CalcUtf8Len);
535
536
537	RTDECL(int) RTUtf16CalcUtf8LenEx(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
538	{
539	size_t cch;
540	int rc = rtUtf16CalcUtf8Length(pwsz, cwc, &cch);
541	if (pcch)
542	*pcch = RT_SUCCESS(rc) ? cch : ~(size_t)0;
543	return rc;
544	}
545	RT_EXPORT_SYMBOL(RTUtf16CalcUtf8LenEx);
546
547
548	RTDECL(RTUNICP) RTUtf16GetCpInternal(PCRTUTF16 pwsz)
549	{
550	const RTUTF16 wc = *pwsz;
551
552	/* simple */
553	if (wc < 0xd800 \|\| (wc > 0xdfff && wc < 0xfffe))
554	return wc;
555	if (wc < 0xfffe)
556	{
557	/* surrogate pair */
558	if (wc < 0xdc00)
559	{
560	const RTUTF16 wc2 = pwsz[1];
561	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
562	{
563	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
564	return uc;
565	}
566
567	RTStrAssertMsgFailed(("wc=%#08x wc2=%#08x - invalid 2nd char in surrogate pair\n", wc, wc2));
568	}
569	else
570	RTStrAssertMsgFailed(("wc=%#08x - invalid surrogate pair order\n", wc));
571	}
572	else
573	RTStrAssertMsgFailed(("wc=%#08x - endian indicator\n", wc));
574	return RTUNICP_INVALID;
575	}
576	RT_EXPORT_SYMBOL(RTUtf16GetCpInternal);
577
578
579	RTDECL(int) RTUtf16GetCpExInternal(PCRTUTF16 *ppwsz, PRTUNICP pCp)
580	{
581	const RTUTF16 wc = **ppwsz;
582
583	/* simple */
584	if (wc < 0xd800 \|\| (wc > 0xdfff && wc < 0xfffe))
585	{
586	(*ppwsz)++;
587	*pCp = wc;
588	return VINF_SUCCESS;
589	}
590
591	int rc;
592	if (wc < 0xfffe)
593	{
594	/* surrogate pair */
595	if (wc < 0xdc00)
596	{
597	const RTUTF16 wc2 = (*ppwsz)[1];
598	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
599	{
600	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
601	*pCp = uc;
602	(*ppwsz) += 2;
603	return VINF_SUCCESS;
604	}
605
606	RTStrAssertMsgFailed(("wc=%#08x wc2=%#08x - invalid 2nd char in surrogate pair\n", wc, wc2));
607	}
608	else
609	RTStrAssertMsgFailed(("wc=%#08x - invalid surrogate pair order\n", wc));
610	rc = VERR_INVALID_UTF16_ENCODING;
611	}
612	else
613	{
614	RTStrAssertMsgFailed(("wc=%#08x - endian indicator\n", wc));
615	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
616	}
617	*pCp = RTUNICP_INVALID;
618	(*ppwsz)++;
619	return rc;
620	}
621	RT_EXPORT_SYMBOL(RTUtf16GetCpExInternal);
622
623
624	RTDECL(PRTUTF16) RTUtf16PutCpInternal(PRTUTF16 pwsz, RTUNICP CodePoint)
625	{
626	/* simple */
627	if ( CodePoint < 0xd800
628	\|\| ( CodePoint > 0xdfff
629	&& CodePoint < 0xfffe))
630	{
631	*pwsz++ = (RTUTF16)CodePoint;
632	return pwsz;
633	}
634
635	/* surrogate pair */
636	if (CodePoint >= 0x10000 && CodePoint <= 0x0010ffff)
637	{
638	CodePoint -= 0x10000;
639	*pwsz++ = 0xd800 \| (CodePoint >> 10);
640	*pwsz++ = 0xdc00 \| (CodePoint & 0x3ff);
641	return pwsz;
642	}
643
644	/* invalid code point. */
645	RTStrAssertMsgFailed(("Invalid codepoint %#x\n", CodePoint));
646	*pwsz++ = 0x7f;
647	return pwsz;
648	}
649	RT_EXPORT_SYMBOL(RTUtf16PutCpInternal);
650
651
652	/**
653	* Validate the UTF-16 encoding and calculates the length of a Latin1 encoding.
654	*
655	* @returns iprt status code.
656	* @param pwsz The UTF-16 string.
657	* @param cwc The max length of the UTF-16 string to consider.
658	* @param pcch Where to store the length (excluding '\\0') of the Latin1 string. (cch == cb, btw)
659	*/
660	static int rtUtf16CalcLatin1Length(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
661	{
662	int rc = VINF_SUCCESS;
663	size_t cch = 0;
664	while (cwc > 0)
665	{
666	RTUTF16 wc = *pwsz++; cwc--;
667	if (!wc)
668	break;
669	else if (RT_LIKELY(wc < 0x100))
670	++cch;
671	else
672	{
673	if (wc < 0xd800 \|\| wc > 0xdfff)
674	{
675	if (wc >= 0xfffe)
676	{
677	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
678	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
679	break;
680	}
681	}
682	else
683	{
684	if (wc >= 0xdc00)
685	{
686	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
687	rc = VERR_INVALID_UTF16_ENCODING;
688	break;
689	}
690	if (cwc <= 0)
691	{
692	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
693	rc = VERR_INVALID_UTF16_ENCODING;
694	break;
695	}
696	wc = *pwsz++; cwc--;
697	if (wc < 0xdc00 \|\| wc > 0xdfff)
698	{
699	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
700	rc = VERR_INVALID_UTF16_ENCODING;
701	break;
702	}
703	}
704
705	rc = VERR_NO_TRANSLATION;
706	break;
707	}
708	}
709
710	/* done */
711	*pcch = cch;
712	return rc;
713	}
714
715
716	/**
717	* Recodes an valid UTF-16 string as Latin1.
718	*
719	* @returns iprt status code.
720	* @param pwsz The UTF-16 string.
721	* @param cwc The number of RTUTF16 characters to process from pwsz. The recoding
722	* will stop when cwc or '\\0' is reached.
723	* @param psz Where to store the Latin1 string.
724	* @param cch The size of the Latin1 buffer, excluding the terminator.
725	*/
726	static int rtUtf16RecodeAsLatin1(PCRTUTF16 pwsz, size_t cwc, char *psz, size_t cch)
727	{
728	unsigned char pch = (unsigned char )psz;
729	int rc = VINF_SUCCESS;
730	while (cwc > 0)
731	{
732	RTUTF16 wc = *pwsz++; cwc--;
733	if (!wc)
734	break;
735	if (RT_LIKELY(wc < 0x100))
736	{
737	if (RT_UNLIKELY(cch < 1))
738	{
739	RTStrAssertMsgFailed(("Buffer overflow! 1\n"));
740	rc = VERR_BUFFER_OVERFLOW;
741	break;
742	}
743	cch--;
744	*pch++ = (unsigned char)wc;
745	}
746	else
747	{
748	if (wc < 0xd800 \|\| wc > 0xdfff)
749	{
750	if (wc >= 0xfffe)
751	{
752	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
753	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
754	break;
755	}
756	}
757	else
758	{
759	if (wc >= 0xdc00)
760	{
761	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
762	rc = VERR_INVALID_UTF16_ENCODING;
763	break;
764	}
765	if (cwc <= 0)
766	{
767	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
768	rc = VERR_INVALID_UTF16_ENCODING;
769	break;
770	}
771	RTUTF16 wc2 = *pwsz++; cwc--;
772	if (wc2 < 0xdc00 \|\| wc2 > 0xdfff)
773	{
774	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
775	rc = VERR_INVALID_UTF16_ENCODING;
776	break;
777	}
778	}
779
780	rc = VERR_NO_TRANSLATION;
781	break;
782	}
783	}
784
785	/* done */
786	*pch = '\0';
787	return rc;
788	}
789
790
791	RTDECL(int) RTUtf16ToLatin1(PCRTUTF16 pwszString, char **ppszString)
792	{
793	/*
794	* Validate input.
795	*/
796	Assert(VALID_PTR(ppszString));
797	Assert(VALID_PTR(pwszString));
798	*ppszString = NULL;
799
800	/*
801	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
802	*/
803	size_t cch;
804	int rc = rtUtf16CalcLatin1Length(pwszString, RTSTR_MAX, &cch);
805	if (RT_SUCCESS(rc))
806	{
807	/*
808	* Allocate buffer and recode it.
809	*/
810	char pszResult = (char )RTMemAlloc(cch + 1);
811	if (pszResult)
812	{
813	rc = rtUtf16RecodeAsLatin1(pwszString, RTSTR_MAX, pszResult, cch);
814	if (RT_SUCCESS(rc))
815	{
816	*ppszString = pszResult;
817	return rc;
818	}
819
820	RTMemFree(pszResult);
821	}
822	else
823	rc = VERR_NO_STR_MEMORY;
824	}
825	return rc;
826	}
827	RT_EXPORT_SYMBOL(RTUtf16ToLatin1);
828
829
830	RTDECL(int) RTUtf16ToLatin1Ex(PCRTUTF16 pwszString, size_t cwcString, char *ppsz, size_t cch, size_t pcch)
831	{
832	/*
833	* Validate input.
834	*/
835	AssertPtr(pwszString);
836	AssertPtr(ppsz);
837	AssertPtrNull(pcch);
838
839	/*
840	* Validate the UTF-16 string and calculate the length of the Latin1 encoding of it.
841	*/
842	size_t cchResult;
843	int rc = rtUtf16CalcLatin1Length(pwszString, cwcString, &cchResult);
844	if (RT_SUCCESS(rc))
845	{
846	if (pcch)
847	*pcch = cchResult;
848
849	/*
850	* Check buffer size / Allocate buffer and recode it.
851	*/
852	bool fShouldFree;
853	char *pszResult;
854	if (cch > 0 && *ppsz)
855	{
856	fShouldFree = false;
857	if (cch <= cchResult)
858	return VERR_BUFFER_OVERFLOW;
859	pszResult = *ppsz;
860	}
861	else
862	{
863	*ppsz = NULL;
864	fShouldFree = true;
865	cch = RT_MAX(cch, cchResult + 1);
866	pszResult = (char *)RTMemAlloc(cch);
867	}
868	if (pszResult)
869	{
870	rc = rtUtf16RecodeAsLatin1(pwszString, cwcString, pszResult, cch - 1);
871	if (RT_SUCCESS(rc))
872	{
873	*ppsz = pszResult;
874	return rc;
875	}
876
877	if (fShouldFree)
878	RTMemFree(pszResult);
879	}
880	else
881	rc = VERR_NO_STR_MEMORY;
882	}
883	return rc;
884	}
885	RT_EXPORT_SYMBOL(RTUtf16ToLatin1Ex);
886
887
888	RTDECL(size_t) RTUtf16CalcLatin1Len(PCRTUTF16 pwsz)
889	{
890	size_t cch;
891	int rc = rtUtf16CalcLatin1Length(pwsz, RTSTR_MAX, &cch);
892	return RT_SUCCESS(rc) ? cch : 0;
893	}
894	RT_EXPORT_SYMBOL(RTUtf16CalcLatin1Len);
895
896
897	RTDECL(int) RTUtf16CalcLatin1LenEx(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
898	{
899	size_t cch;
900	int rc = rtUtf16CalcLatin1Length(pwsz, cwc, &cch);
901	if (pcch)
902	*pcch = RT_SUCCESS(rc) ? cch : ~(size_t)0;
903	return rc;
904	}
905	RT_EXPORT_SYMBOL(RTUtf16CalcLatin1LenEx);
906
907
908	/**
909	* Calculates the UTF-16 length of a Latin1 string. In fact this is just the
910	* original length, but the function saves us nasty comments to that effect
911	* all over the place.
912	*
913	* @returns IPRT status code.
914	* @param psz Pointer to the Latin1 string.
915	* @param cch The max length of the string. (btw cch = cb)
916	* Use RTSTR_MAX if all of the string is to be examined.s
917	* @param pcwc Where to store the length of the UTF-16 string as a number of RTUTF16 characters.
918	*/
919	static int rtLatin1CalcUtf16Length(const char psz, size_t cch, size_t pcwc)
920	{
921	*pcwc = RTStrNLen(psz, cch);
922	return VINF_SUCCESS;
923	}
924
925
926	/**
927	* Recodes a Latin1 string as UTF-16. This is just a case of expanding it to
928	* sixteen bits, as Unicode is a superset of Latin1.
929	*
930	* Since we know the input is valid, we do not perform length checks.
931	*
932	* @returns iprt status code.
933	* @param psz The Latin1 string to recode.
934	* @param cch The number of chars (the type char, so bytes if you like) to process of the Latin1 string.
935	* The recoding will stop when cch or '\\0' is reached. Pass RTSTR_MAX to process up to '\\0'.
936	* @param pwsz Where to store the UTF-16 string.
937	* @param cwc The number of RTUTF16 items the pwsz buffer can hold, excluding the terminator ('\\0').
938	*/
939	static int rtLatin1RecodeAsUtf16(const char *psz, size_t cch, PRTUTF16 pwsz, size_t cwc)
940	{
941	int rc = VINF_SUCCESS;
942	const unsigned char puch = (const unsigned char )psz;
943	PRTUTF16 pwc = pwsz;
944	while (cch-- > 0)
945	{
946	/* read the next char and check for terminator. */
947	const unsigned char uch = *puch;
948	if (!uch)
949	break;
950
951	/* check for output overflow */
952	if (RT_UNLIKELY(cwc < 1))
953	{
954	rc = VERR_BUFFER_OVERFLOW;
955	break;
956	}
957
958	/* expand the code point */
959	*pwc++ = uch;
960	cwc--;
961	puch++;
962	}
963
964	/* done */
965	*pwc = '\0';
966	return rc;
967	}
968
969
970	RTDECL(int) RTLatin1ToUtf16(const char pszString, PRTUTF16 ppwszString)
971	{
972	/*
973	* Validate input.
974	*/
975	Assert(VALID_PTR(ppwszString));
976	Assert(VALID_PTR(pszString));
977	*ppwszString = NULL;
978
979	/*
980	* Validate the input and calculate the length of the UTF-16 string.
981	*/
982	size_t cwc;
983	int rc = rtLatin1CalcUtf16Length(pszString, RTSTR_MAX, &cwc);
984	if (RT_SUCCESS(rc))
985	{
986	/*
987	* Allocate buffer.
988	*/
989	PRTUTF16 pwsz = (PRTUTF16)RTMemAlloc((cwc + 1) * sizeof(RTUTF16));
990	if (pwsz)
991	{
992	/*
993	* Encode the UTF-16 string.
994	*/
995	rc = rtLatin1RecodeAsUtf16(pszString, RTSTR_MAX, pwsz, cwc);
996	if (RT_SUCCESS(rc))
997	{
998	*ppwszString = pwsz;
999	return rc;
1000	}
1001	RTMemFree(pwsz);
1002	}
1003	else
1004	rc = VERR_NO_UTF16_MEMORY;
1005	}
1006	return rc;
1007	}
1008	RT_EXPORT_SYMBOL(RTLatin1ToUtf16);
1009
1010
1011	RTDECL(int) RTLatin1ToUtf16Ex(const char pszString, size_t cchString, PRTUTF16 ppwsz, size_t cwc, size_t *pcwc)
1012	{
1013	/*
1014	* Validate input.
1015	*/
1016	Assert(VALID_PTR(pszString));
1017	Assert(VALID_PTR(ppwsz));
1018	Assert(!pcwc \|\| VALID_PTR(pcwc));
1019
1020	/*
1021	* Validate the input and calculate the length of the UTF-16 string.
1022	*/
1023	size_t cwcResult;
1024	int rc = rtLatin1CalcUtf16Length(pszString, cchString, &cwcResult);
1025	if (RT_SUCCESS(rc))
1026	{
1027	if (pcwc)
1028	*pcwc = cwcResult;
1029
1030	/*
1031	* Check buffer size / Allocate buffer.
1032	*/
1033	bool fShouldFree;
1034	PRTUTF16 pwszResult;
1035	if (cwc > 0 && *ppwsz)
1036	{
1037	fShouldFree = false;
1038	if (cwc <= cwcResult)
1039	return VERR_BUFFER_OVERFLOW;
1040	pwszResult = *ppwsz;
1041	}
1042	else
1043	{
1044	*ppwsz = NULL;
1045	fShouldFree = true;
1046	cwc = RT_MAX(cwcResult + 1, cwc);
1047	pwszResult = (PRTUTF16)RTMemAlloc(cwc * sizeof(RTUTF16));
1048	}
1049	if (pwszResult)
1050	{
1051	/*
1052	* Encode the UTF-16 string.
1053	*/
1054	rc = rtLatin1RecodeAsUtf16(pszString, cchString, pwszResult, cwc - 1);
1055	if (RT_SUCCESS(rc))
1056	{
1057	*ppwsz = pwszResult;
1058	return rc;
1059	}
1060	if (fShouldFree)
1061	RTMemFree(pwszResult);
1062	}
1063	else
1064	rc = VERR_NO_UTF16_MEMORY;
1065	}
1066	return rc;
1067	}
1068	RT_EXPORT_SYMBOL(RTLatin1ToUtf16Ex);
1069
1070
1071	RTDECL(size_t) RTLatin1CalcUtf16Len(const char *psz)
1072	{
1073	size_t cwc;
1074	int rc = rtLatin1CalcUtf16Length(psz, RTSTR_MAX, &cwc);
1075	return RT_SUCCESS(rc) ? cwc : 0;
1076	}
1077	RT_EXPORT_SYMBOL(RTLatin1CalcUtf16Len);
1078
1079
1080	RTDECL(int) RTLatin1CalcUtf16LenEx(const char psz, size_t cch, size_t pcwc)
1081	{
1082	size_t cwc;
1083	int rc = rtLatin1CalcUtf16Length(psz, cch, &cwc);
1084	if (pcwc)
1085	*pcwc = RT_SUCCESS(rc) ? cwc : ~(size_t)0;
1086	return rc;
1087	}
1088	RT_EXPORT_SYMBOL(RTLatin1CalcUtf16LenEx);

Note: See TracBrowser for help on using the repository browser.

source: vbox/trunk/src/VBox/Runtime/common/string/utf-16.cpp@ 25000

Download in other formats: