utf-16.cpp@ 34079

Last change on this file since 34079 was 31157, checked in by vboxsync, 14 years ago
iprt,++: Tag allocation in all builds with a string, defaulting to FILE.
Property svn:eol-style set to `native` Property svn:keywords set to `Id`
File size: 29.9 KB

Line
1	/* $Id: utf-16.cpp 31157 2010-07-28 03:15:35Z vboxsync $ */
2	/** @file
3	* IPRT - UTF-16.
4	*/
5
6	/*
7	* Copyright (C) 2006-2010 Oracle Corporation
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
27
28	/*******************************************************************************
29	* Header Files *
30	*******************************************************************************/
31	#include <iprt/string.h>
32	#include "internal/iprt.h"
33
34	#include <iprt/uni.h>
35	#include <iprt/alloc.h>
36	#include <iprt/assert.h>
37	#include <iprt/err.h>
38	#include "internal/string.h"
39
40
41
42	RTDECL(void) RTUtf16Free(PRTUTF16 pwszString)
43	{
44	if (pwszString)
45	RTMemTmpFree(pwszString);
46	}
47	RT_EXPORT_SYMBOL(RTUtf16Free);
48
49
50	RTDECL(PRTUTF16) RTUtf16DupTag(PCRTUTF16 pwszString, const char *pszTag)
51	{
52	Assert(pwszString);
53	size_t cb = (RTUtf16Len(pwszString) + 1) * sizeof(RTUTF16);
54	PRTUTF16 pwsz = (PRTUTF16)RTMemAllocTag(cb, pszTag);
55	if (pwsz)
56	memcpy(pwsz, pwszString, cb);
57	return pwsz;
58	}
59	RT_EXPORT_SYMBOL(RTUtf16DupTag);
60
61
62	RTDECL(int) RTUtf16DupExTag(PRTUTF16 ppwszString, PCRTUTF16 pwszString, size_t cwcExtra, const char pszTag)
63	{
64	Assert(pwszString);
65	size_t cb = (RTUtf16Len(pwszString) + 1) * sizeof(RTUTF16);
66	PRTUTF16 pwsz = (PRTUTF16)RTMemAllocTag(cb + cwcExtra * sizeof(RTUTF16), pszTag);
67	if (pwsz)
68	{
69	memcpy(pwsz, pwszString, cb);
70	*ppwszString = pwsz;
71	return VINF_SUCCESS;
72	}
73	return VERR_NO_MEMORY;
74	}
75	RT_EXPORT_SYMBOL(RTUtf16DupExTag);
76
77
78	RTDECL(size_t) RTUtf16Len(PCRTUTF16 pwszString)
79	{
80	if (!pwszString)
81	return 0;
82
83	PCRTUTF16 pwsz = pwszString;
84	while (*pwsz)
85	pwsz++;
86	return pwsz - pwszString;
87	}
88	RT_EXPORT_SYMBOL(RTUtf16Len);
89
90
91	RTDECL(int) RTUtf16Cmp(register PCRTUTF16 pwsz1, register PCRTUTF16 pwsz2)
92	{
93	if (pwsz1 == pwsz2)
94	return 0;
95	if (!pwsz1)
96	return -1;
97	if (!pwsz2)
98	return 1;
99
100	for (;;)
101	{
102	register RTUTF16 wcs = *pwsz1;
103	register int iDiff = wcs - *pwsz2;
104	if (iDiff \|\| !wcs)
105	return iDiff;
106	pwsz1++;
107	pwsz2++;
108	}
109	}
110	RT_EXPORT_SYMBOL(RTUtf16Cmp);
111
112
113	RTDECL(int) RTUtf16ICmp(register PCRTUTF16 pwsz1, register PCRTUTF16 pwsz2)
114	{
115	if (pwsz1 == pwsz2)
116	return 0;
117	if (!pwsz1)
118	return -1;
119	if (!pwsz2)
120	return 1;
121
122	PCRTUTF16 pwsz1Start = pwsz1; /* keep it around in case we have to backtrack on a surrogate pair */
123	for (;;)
124	{
125	register RTUTF16 wc1 = *pwsz1;
126	register RTUTF16 wc2 = *pwsz2;
127	register int iDiff = wc1 - wc2;
128	if (iDiff)
129	{
130	/* unless they are both surrogate pairs, there is no chance they'll be identical. */
131	if ( wc1 < 0xd800
132	\|\| wc2 < 0xd800
133	\|\| wc1 > 0xdfff
134	\|\| wc2 > 0xdfff)
135	{
136	/* simple UCS-2 char */
137	iDiff = RTUniCpToUpper(wc1) - RTUniCpToUpper(wc2);
138	if (iDiff)
139	iDiff = RTUniCpToLower(wc1) - RTUniCpToLower(wc2);
140	}
141	else
142	{
143	/* a damned pair */
144	RTUNICP uc1;
145	RTUNICP uc2;
146	if (wc1 >= 0xdc00)
147	{
148	if (pwsz1Start == pwsz1)
149	return iDiff;
150	uc1 = pwsz1[-1];
151	if (uc1 < 0xd800 \|\| uc1 >= 0xdc00)
152	return iDiff;
153	uc1 = 0x10000 + (((uc1 & 0x3ff) << 10) \| (wc1 & 0x3ff));
154	uc2 = 0x10000 + (((pwsz2[-1] & 0x3ff) << 10) \| (wc2 & 0x3ff));
155	}
156	else
157	{
158	uc1 = *++pwsz1;
159	if (uc1 < 0xdc00 \|\| uc1 >= 0xe000)
160	return iDiff;
161	uc1 = 0x10000 + (((wc1 & 0x3ff) << 10) \| (uc1 & 0x3ff));
162	uc2 = 0x10000 + (((wc2 & 0x3ff) << 10) \| (*++pwsz2 & 0x3ff));
163	}
164	iDiff = RTUniCpToUpper(uc1) - RTUniCpToUpper(uc2);
165	if (iDiff)
166	iDiff = RTUniCpToLower(uc1) - RTUniCpToLower(uc2); /* serious paranoia! */
167	}
168	if (iDiff)
169	return iDiff;
170	}
171	if (!wc1)
172	return 0;
173	pwsz1++;
174	pwsz2++;
175	}
176	}
177	RT_EXPORT_SYMBOL(RTUtf16ICmp);
178
179
180	RTDECL(PRTUTF16) RTUtf16ToLower(PRTUTF16 pwsz)
181	{
182	PRTUTF16 pwc = pwsz;
183	for (;;)
184	{
185	RTUTF16 wc = *pwc;
186	if (!wc)
187	break;
188	if (wc < 0xd800 \|\| wc >= 0xdc00)
189	{
190	RTUNICP ucFolded = RTUniCpToLower(wc);
191	if (ucFolded < 0x10000)
192	*pwc++ = RTUniCpToLower(wc);
193	}
194	else
195	{
196	/* surrogate */
197	RTUTF16 wc2 = pwc[1];
198	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
199	{
200	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
201	RTUNICP ucFolded = RTUniCpToLower(uc);
202	if (uc != ucFolded && ucFolded >= 0x10000) /* we don't support shrinking the string */
203	{
204	uc -= 0x10000;
205	*pwc++ = 0xd800 \| (uc >> 10);
206	*pwc++ = 0xdc00 \| (uc & 0x3ff);
207	}
208	}
209	else /* invalid encoding. */
210	pwc++;
211	}
212	}
213	return pwsz;
214	}
215	RT_EXPORT_SYMBOL(RTUtf16ToLower);
216
217
218	RTDECL(PRTUTF16) RTUtf16ToUpper(PRTUTF16 pwsz)
219	{
220	PRTUTF16 pwc = pwsz;
221	for (;;)
222	{
223	RTUTF16 wc = *pwc;
224	if (!wc)
225	break;
226	if (wc < 0xd800 \|\| wc >= 0xdc00)
227	*pwc++ = RTUniCpToUpper(wc);
228	else
229	{
230	/* surrogate */
231	RTUTF16 wc2 = pwc[1];
232	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
233	{
234	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
235	RTUNICP ucFolded = RTUniCpToUpper(uc);
236	if (uc != ucFolded && ucFolded >= 0x10000) /* we don't support shrinking the string */
237	{
238	uc -= 0x10000;
239	*pwc++ = 0xd800 \| (uc >> 10);
240	*pwc++ = 0xdc00 \| (uc & 0x3ff);
241	}
242	}
243	else /* invalid encoding. */
244	pwc++;
245	}
246	}
247	return pwsz;
248	}
249	RT_EXPORT_SYMBOL(RTUtf16ToUpper);
250
251
252	/**
253	* Validate the UTF-16 encoding and calculates the length of an UTF-8 encoding.
254	*
255	* @returns iprt status code.
256	* @param pwsz The UTF-16 string.
257	* @param cwc The max length of the UTF-16 string to consider.
258	* @param pcch Where to store the length (excluding '\\0') of the UTF-8 string. (cch == cb, btw)
259	*/
260	static int rtUtf16CalcUtf8Length(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
261	{
262	int rc = VINF_SUCCESS;
263	size_t cch = 0;
264	while (cwc > 0)
265	{
266	RTUTF16 wc = *pwsz++; cwc--;
267	if (!wc)
268	break;
269	else if (wc < 0xd800 \|\| wc > 0xdfff)
270	{
271	if (wc < 0x80)
272	cch++;
273	else if (wc < 0x800)
274	cch += 2;
275	else if (wc < 0xfffe)
276	cch += 3;
277	else
278	{
279	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
280	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
281	break;
282	}
283	}
284	else
285	{
286	if (wc >= 0xdc00)
287	{
288	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
289	rc = VERR_INVALID_UTF16_ENCODING;
290	break;
291	}
292	if (cwc <= 0)
293	{
294	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
295	rc = VERR_INVALID_UTF16_ENCODING;
296	break;
297	}
298	wc = *pwsz++; cwc--;
299	if (wc < 0xdc00 \|\| wc > 0xdfff)
300	{
301	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
302	rc = VERR_INVALID_UTF16_ENCODING;
303	break;
304	}
305	cch += 4;
306	}
307	}
308
309
310	/* done */
311	*pcch = cch;
312	return rc;
313	}
314
315
316	/**
317	* Recodes an valid UTF-16 string as UTF-8.
318	*
319	* @returns iprt status code.
320	* @param pwsz The UTF-16 string.
321	* @param cwc The number of RTUTF16 characters to process from pwsz. The recoding
322	* will stop when cwc or '\\0' is reached.
323	* @param psz Where to store the UTF-8 string.
324	* @param cch The size of the UTF-8 buffer, excluding the terminator.
325	* @param pcch Where to store the number of octets actually encoded.
326	*/
327	static int rtUtf16RecodeAsUtf8(PCRTUTF16 pwsz, size_t cwc, char psz, size_t cch, size_t pcch)
328	{
329	unsigned char pwch = (unsigned char )psz;
330	int rc = VINF_SUCCESS;
331	while (cwc > 0)
332	{
333	RTUTF16 wc = *pwsz++; cwc--;
334	if (!wc)
335	break;
336	else if (wc < 0xd800 \|\| wc > 0xdfff)
337	{
338	if (wc < 0x80)
339	{
340	if (RT_UNLIKELY(cch < 1))
341	{
342	RTStrAssertMsgFailed(("Buffer overflow! 1\n"));
343	rc = VERR_BUFFER_OVERFLOW;
344	break;
345	}
346	cch--;
347	*pwch++ = (unsigned char)wc;
348	}
349	else if (wc < 0x800)
350	{
351	if (RT_UNLIKELY(cch < 2))
352	{
353	RTStrAssertMsgFailed(("Buffer overflow! 2\n"));
354	rc = VERR_BUFFER_OVERFLOW;
355	break;
356	}
357	cch -= 2;
358	*pwch++ = 0xc0 \| (wc >> 6);
359	*pwch++ = 0x80 \| (wc & 0x3f);
360	}
361	else if (wc < 0xfffe)
362	{
363	if (RT_UNLIKELY(cch < 3))
364	{
365	RTStrAssertMsgFailed(("Buffer overflow! 3\n"));
366	rc = VERR_BUFFER_OVERFLOW;
367	break;
368	}
369	cch -= 3;
370	*pwch++ = 0xe0 \| (wc >> 12);
371	*pwch++ = 0x80 \| ((wc >> 6) & 0x3f);
372	*pwch++ = 0x80 \| (wc & 0x3f);
373	}
374	else
375	{
376	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
377	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
378	break;
379	}
380	}
381	else
382	{
383	if (wc >= 0xdc00)
384	{
385	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
386	rc = VERR_INVALID_UTF16_ENCODING;
387	break;
388	}
389	if (cwc <= 0)
390	{
391	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
392	rc = VERR_INVALID_UTF16_ENCODING;
393	break;
394	}
395	RTUTF16 wc2 = *pwsz++; cwc--;
396	if (wc2 < 0xdc00 \|\| wc2 > 0xdfff)
397	{
398	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
399	rc = VERR_INVALID_UTF16_ENCODING;
400	break;
401	}
402	uint32_t CodePoint = 0x10000
403	+ ( ((wc & 0x3ff) << 10)
404	\| (wc2 & 0x3ff));
405	if (RT_UNLIKELY(cch < 4))
406	{
407	RTStrAssertMsgFailed(("Buffer overflow! 4\n"));
408	rc = VERR_BUFFER_OVERFLOW;
409	break;
410	}
411	cch -= 4;
412	*pwch++ = 0xf0 \| (CodePoint >> 18);
413	*pwch++ = 0x80 \| ((CodePoint >> 12) & 0x3f);
414	*pwch++ = 0x80 \| ((CodePoint >> 6) & 0x3f);
415	*pwch++ = 0x80 \| (CodePoint & 0x3f);
416	}
417	}
418
419	/* done */
420	*pwch = '\0';
421	pcch = (char )pwch - psz;
422	return rc;
423	}
424
425
426
427	RTDECL(int) RTUtf16ToUtf8Tag(PCRTUTF16 pwszString, char *ppszString, const char pszTag)
428	{
429	/*
430	* Validate input.
431	*/
432	Assert(VALID_PTR(ppszString));
433	Assert(VALID_PTR(pwszString));
434	*ppszString = NULL;
435
436	/*
437	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
438	*/
439	size_t cch;
440	int rc = rtUtf16CalcUtf8Length(pwszString, RTSTR_MAX, &cch);
441	if (RT_SUCCESS(rc))
442	{
443	/*
444	* Allocate buffer and recode it.
445	*/
446	char pszResult = (char )RTMemAllocTag(cch + 1, pszTag);
447	if (pszResult)
448	{
449	rc = rtUtf16RecodeAsUtf8(pwszString, RTSTR_MAX, pszResult, cch, &cch);
450	if (RT_SUCCESS(rc))
451	{
452	*ppszString = pszResult;
453	return rc;
454	}
455
456	RTMemFree(pszResult);
457	}
458	else
459	rc = VERR_NO_STR_MEMORY;
460	}
461	return rc;
462	}
463	RT_EXPORT_SYMBOL(RTUtf16ToUtf8Tag);
464
465
466	RTDECL(int) RTUtf16ToUtf8ExTag(PCRTUTF16 pwszString, size_t cwcString, char *ppsz, size_t cch, size_t pcch, const char *pszTag)
467	{
468	/*
469	* Validate input.
470	*/
471	Assert(VALID_PTR(pwszString));
472	Assert(VALID_PTR(ppsz));
473	Assert(!pcch \|\| VALID_PTR(pcch));
474
475	/*
476	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
477	*/
478	size_t cchResult;
479	int rc = rtUtf16CalcUtf8Length(pwszString, cwcString, &cchResult);
480	if (RT_SUCCESS(rc))
481	{
482	if (pcch)
483	*pcch = cchResult;
484
485	/*
486	* Check buffer size / Allocate buffer and recode it.
487	*/
488	bool fShouldFree;
489	char *pszResult;
490	if (cch > 0 && *ppsz)
491	{
492	fShouldFree = false;
493	if (RT_UNLIKELY(cch <= cchResult))
494	return VERR_BUFFER_OVERFLOW;
495	pszResult = *ppsz;
496	}
497	else
498	{
499	*ppsz = NULL;
500	fShouldFree = true;
501	cch = RT_MAX(cch, cchResult + 1);
502	pszResult = (char *)RTStrAllocTag(cch, pszTag);
503	}
504	if (pszResult)
505	{
506	rc = rtUtf16RecodeAsUtf8(pwszString, cwcString, pszResult, cch - 1, &cch);
507	if (RT_SUCCESS(rc))
508	{
509	*ppsz = pszResult;
510	return rc;
511	}
512
513	if (fShouldFree)
514	RTStrFree(pszResult);
515	}
516	else
517	rc = VERR_NO_STR_MEMORY;
518	}
519	return rc;
520	}
521	RT_EXPORT_SYMBOL(RTUtf16ToUtf8ExTag);
522
523
524	RTDECL(size_t) RTUtf16CalcUtf8Len(PCRTUTF16 pwsz)
525	{
526	size_t cch;
527	int rc = rtUtf16CalcUtf8Length(pwsz, RTSTR_MAX, &cch);
528	return RT_SUCCESS(rc) ? cch : 0;
529	}
530	RT_EXPORT_SYMBOL(RTUtf16CalcUtf8Len);
531
532
533	RTDECL(int) RTUtf16CalcUtf8LenEx(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
534	{
535	size_t cch;
536	int rc = rtUtf16CalcUtf8Length(pwsz, cwc, &cch);
537	if (pcch)
538	*pcch = RT_SUCCESS(rc) ? cch : ~(size_t)0;
539	return rc;
540	}
541	RT_EXPORT_SYMBOL(RTUtf16CalcUtf8LenEx);
542
543
544	RTDECL(RTUNICP) RTUtf16GetCpInternal(PCRTUTF16 pwsz)
545	{
546	const RTUTF16 wc = *pwsz;
547
548	/* simple */
549	if (wc < 0xd800 \|\| (wc > 0xdfff && wc < 0xfffe))
550	return wc;
551	if (wc < 0xfffe)
552	{
553	/* surrogate pair */
554	if (wc < 0xdc00)
555	{
556	const RTUTF16 wc2 = pwsz[1];
557	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
558	{
559	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
560	return uc;
561	}
562
563	RTStrAssertMsgFailed(("wc=%#08x wc2=%#08x - invalid 2nd char in surrogate pair\n", wc, wc2));
564	}
565	else
566	RTStrAssertMsgFailed(("wc=%#08x - invalid surrogate pair order\n", wc));
567	}
568	else
569	RTStrAssertMsgFailed(("wc=%#08x - endian indicator\n", wc));
570	return RTUNICP_INVALID;
571	}
572	RT_EXPORT_SYMBOL(RTUtf16GetCpInternal);
573
574
575	RTDECL(int) RTUtf16GetCpExInternal(PCRTUTF16 *ppwsz, PRTUNICP pCp)
576	{
577	const RTUTF16 wc = **ppwsz;
578
579	/* simple */
580	if (wc < 0xd800 \|\| (wc > 0xdfff && wc < 0xfffe))
581	{
582	(*ppwsz)++;
583	*pCp = wc;
584	return VINF_SUCCESS;
585	}
586
587	int rc;
588	if (wc < 0xfffe)
589	{
590	/* surrogate pair */
591	if (wc < 0xdc00)
592	{
593	const RTUTF16 wc2 = (*ppwsz)[1];
594	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
595	{
596	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
597	*pCp = uc;
598	(*ppwsz) += 2;
599	return VINF_SUCCESS;
600	}
601
602	RTStrAssertMsgFailed(("wc=%#08x wc2=%#08x - invalid 2nd char in surrogate pair\n", wc, wc2));
603	}
604	else
605	RTStrAssertMsgFailed(("wc=%#08x - invalid surrogate pair order\n", wc));
606	rc = VERR_INVALID_UTF16_ENCODING;
607	}
608	else
609	{
610	RTStrAssertMsgFailed(("wc=%#08x - endian indicator\n", wc));
611	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
612	}
613	*pCp = RTUNICP_INVALID;
614	(*ppwsz)++;
615	return rc;
616	}
617	RT_EXPORT_SYMBOL(RTUtf16GetCpExInternal);
618
619
620	RTDECL(PRTUTF16) RTUtf16PutCpInternal(PRTUTF16 pwsz, RTUNICP CodePoint)
621	{
622	/* simple */
623	if ( CodePoint < 0xd800
624	\|\| ( CodePoint > 0xdfff
625	&& CodePoint < 0xfffe))
626	{
627	*pwsz++ = (RTUTF16)CodePoint;
628	return pwsz;
629	}
630
631	/* surrogate pair */
632	if (CodePoint >= 0x10000 && CodePoint <= 0x0010ffff)
633	{
634	CodePoint -= 0x10000;
635	*pwsz++ = 0xd800 \| (CodePoint >> 10);
636	*pwsz++ = 0xdc00 \| (CodePoint & 0x3ff);
637	return pwsz;
638	}
639
640	/* invalid code point. */
641	RTStrAssertMsgFailed(("Invalid codepoint %#x\n", CodePoint));
642	*pwsz++ = 0x7f;
643	return pwsz;
644	}
645	RT_EXPORT_SYMBOL(RTUtf16PutCpInternal);
646
647
648	/**
649	* Validate the UTF-16 encoding and calculates the length of a Latin1 encoding.
650	*
651	* @returns iprt status code.
652	* @param pwsz The UTF-16 string.
653	* @param cwc The max length of the UTF-16 string to consider.
654	* @param pcch Where to store the length (excluding '\\0') of the Latin1 string. (cch == cb, btw)
655	*/
656	static int rtUtf16CalcLatin1Length(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
657	{
658	int rc = VINF_SUCCESS;
659	size_t cch = 0;
660	while (cwc > 0)
661	{
662	RTUTF16 wc = *pwsz++; cwc--;
663	if (!wc)
664	break;
665	else if (RT_LIKELY(wc < 0x100))
666	++cch;
667	else
668	{
669	if (wc < 0xd800 \|\| wc > 0xdfff)
670	{
671	if (wc >= 0xfffe)
672	{
673	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
674	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
675	break;
676	}
677	}
678	else
679	{
680	if (wc >= 0xdc00)
681	{
682	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
683	rc = VERR_INVALID_UTF16_ENCODING;
684	break;
685	}
686	if (cwc <= 0)
687	{
688	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
689	rc = VERR_INVALID_UTF16_ENCODING;
690	break;
691	}
692	wc = *pwsz++; cwc--;
693	if (wc < 0xdc00 \|\| wc > 0xdfff)
694	{
695	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
696	rc = VERR_INVALID_UTF16_ENCODING;
697	break;
698	}
699	}
700
701	rc = VERR_NO_TRANSLATION;
702	break;
703	}
704	}
705
706	/* done */
707	*pcch = cch;
708	return rc;
709	}
710
711
712	/**
713	* Recodes an valid UTF-16 string as Latin1.
714	*
715	* @returns iprt status code.
716	* @param pwsz The UTF-16 string.
717	* @param cwc The number of RTUTF16 characters to process from pwsz. The recoding
718	* will stop when cwc or '\\0' is reached.
719	* @param psz Where to store the Latin1 string.
720	* @param cch The size of the Latin1 buffer, excluding the terminator.
721	*/
722	static int rtUtf16RecodeAsLatin1(PCRTUTF16 pwsz, size_t cwc, char *psz, size_t cch)
723	{
724	unsigned char pch = (unsigned char )psz;
725	int rc = VINF_SUCCESS;
726	while (cwc > 0)
727	{
728	RTUTF16 wc = *pwsz++; cwc--;
729	if (!wc)
730	break;
731	if (RT_LIKELY(wc < 0x100))
732	{
733	if (RT_UNLIKELY(cch < 1))
734	{
735	RTStrAssertMsgFailed(("Buffer overflow! 1\n"));
736	rc = VERR_BUFFER_OVERFLOW;
737	break;
738	}
739	cch--;
740	*pch++ = (unsigned char)wc;
741	}
742	else
743	{
744	if (wc < 0xd800 \|\| wc > 0xdfff)
745	{
746	if (wc >= 0xfffe)
747	{
748	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
749	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
750	break;
751	}
752	}
753	else
754	{
755	if (wc >= 0xdc00)
756	{
757	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
758	rc = VERR_INVALID_UTF16_ENCODING;
759	break;
760	}
761	if (cwc <= 0)
762	{
763	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
764	rc = VERR_INVALID_UTF16_ENCODING;
765	break;
766	}
767	RTUTF16 wc2 = *pwsz++; cwc--;
768	if (wc2 < 0xdc00 \|\| wc2 > 0xdfff)
769	{
770	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
771	rc = VERR_INVALID_UTF16_ENCODING;
772	break;
773	}
774	}
775
776	rc = VERR_NO_TRANSLATION;
777	break;
778	}
779	}
780
781	/* done */
782	*pch = '\0';
783	return rc;
784	}
785
786
787	RTDECL(int) RTUtf16ToLatin1Tag(PCRTUTF16 pwszString, char *ppszString, const char pszTag)
788	{
789	/*
790	* Validate input.
791	*/
792	Assert(VALID_PTR(ppszString));
793	Assert(VALID_PTR(pwszString));
794	*ppszString = NULL;
795
796	/*
797	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
798	*/
799	size_t cch;
800	int rc = rtUtf16CalcLatin1Length(pwszString, RTSTR_MAX, &cch);
801	if (RT_SUCCESS(rc))
802	{
803	/*
804	* Allocate buffer and recode it.
805	*/
806	char pszResult = (char )RTMemAllocTag(cch + 1, pszTag);
807	if (pszResult)
808	{
809	rc = rtUtf16RecodeAsLatin1(pwszString, RTSTR_MAX, pszResult, cch);
810	if (RT_SUCCESS(rc))
811	{
812	*ppszString = pszResult;
813	return rc;
814	}
815
816	RTMemFree(pszResult);
817	}
818	else
819	rc = VERR_NO_STR_MEMORY;
820	}
821	return rc;
822	}
823	RT_EXPORT_SYMBOL(RTUtf16ToLatin1Tag);
824
825
826	RTDECL(int) RTUtf16ToLatin1ExTag(PCRTUTF16 pwszString, size_t cwcString, char *ppsz, size_t cch, size_t pcch, const char *pszTag)
827	{
828	/*
829	* Validate input.
830	*/
831	AssertPtr(pwszString);
832	AssertPtr(ppsz);
833	AssertPtrNull(pcch);
834
835	/*
836	* Validate the UTF-16 string and calculate the length of the Latin1 encoding of it.
837	*/
838	size_t cchResult;
839	int rc = rtUtf16CalcLatin1Length(pwszString, cwcString, &cchResult);
840	if (RT_SUCCESS(rc))
841	{
842	if (pcch)
843	*pcch = cchResult;
844
845	/*
846	* Check buffer size / Allocate buffer and recode it.
847	*/
848	bool fShouldFree;
849	char *pszResult;
850	if (cch > 0 && *ppsz)
851	{
852	fShouldFree = false;
853	if (cch <= cchResult)
854	return VERR_BUFFER_OVERFLOW;
855	pszResult = *ppsz;
856	}
857	else
858	{
859	*ppsz = NULL;
860	fShouldFree = true;
861	cch = RT_MAX(cch, cchResult + 1);
862	pszResult = (char *)RTMemAllocTag(cch, pszTag);
863	}
864	if (pszResult)
865	{
866	rc = rtUtf16RecodeAsLatin1(pwszString, cwcString, pszResult, cch - 1);
867	if (RT_SUCCESS(rc))
868	{
869	*ppsz = pszResult;
870	return rc;
871	}
872
873	if (fShouldFree)
874	RTMemFree(pszResult);
875	}
876	else
877	rc = VERR_NO_STR_MEMORY;
878	}
879	return rc;
880	}
881	RT_EXPORT_SYMBOL(RTUtf16ToLatin1ExTag);
882
883
884	RTDECL(size_t) RTUtf16CalcLatin1Len(PCRTUTF16 pwsz)
885	{
886	size_t cch;
887	int rc = rtUtf16CalcLatin1Length(pwsz, RTSTR_MAX, &cch);
888	return RT_SUCCESS(rc) ? cch : 0;
889	}
890	RT_EXPORT_SYMBOL(RTUtf16CalcLatin1Len);
891
892
893	RTDECL(int) RTUtf16CalcLatin1LenEx(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
894	{
895	size_t cch;
896	int rc = rtUtf16CalcLatin1Length(pwsz, cwc, &cch);
897	if (pcch)
898	*pcch = RT_SUCCESS(rc) ? cch : ~(size_t)0;
899	return rc;
900	}
901	RT_EXPORT_SYMBOL(RTUtf16CalcLatin1LenEx);
902
903
904	/**
905	* Calculates the UTF-16 length of a Latin1 string. In fact this is just the
906	* original length, but the function saves us nasty comments to that effect
907	* all over the place.
908	*
909	* @returns IPRT status code.
910	* @param psz Pointer to the Latin1 string.
911	* @param cch The max length of the string. (btw cch = cb)
912	* Use RTSTR_MAX if all of the string is to be examined.s
913	* @param pcwc Where to store the length of the UTF-16 string as a number of RTUTF16 characters.
914	*/
915	static int rtLatin1CalcUtf16Length(const char psz, size_t cch, size_t pcwc)
916	{
917	*pcwc = RTStrNLen(psz, cch);
918	return VINF_SUCCESS;
919	}
920
921
922	/**
923	* Recodes a Latin1 string as UTF-16. This is just a case of expanding it to
924	* sixteen bits, as Unicode is a superset of Latin1.
925	*
926	* Since we know the input is valid, we do not perform length checks.
927	*
928	* @returns iprt status code.
929	* @param psz The Latin1 string to recode.
930	* @param cch The number of chars (the type char, so bytes if you like) to process of the Latin1 string.
931	* The recoding will stop when cch or '\\0' is reached. Pass RTSTR_MAX to process up to '\\0'.
932	* @param pwsz Where to store the UTF-16 string.
933	* @param cwc The number of RTUTF16 items the pwsz buffer can hold, excluding the terminator ('\\0').
934	*/
935	static int rtLatin1RecodeAsUtf16(const char *psz, size_t cch, PRTUTF16 pwsz, size_t cwc)
936	{
937	int rc = VINF_SUCCESS;
938	const unsigned char puch = (const unsigned char )psz;
939	PRTUTF16 pwc = pwsz;
940	while (cch-- > 0)
941	{
942	/* read the next char and check for terminator. */
943	const unsigned char uch = *puch;
944	if (!uch)
945	break;
946
947	/* check for output overflow */
948	if (RT_UNLIKELY(cwc < 1))
949	{
950	rc = VERR_BUFFER_OVERFLOW;
951	break;
952	}
953
954	/* expand the code point */
955	*pwc++ = uch;
956	cwc--;
957	puch++;
958	}
959
960	/* done */
961	*pwc = '\0';
962	return rc;
963	}
964
965
966	RTDECL(int) RTLatin1ToUtf16Tag(const char pszString, PRTUTF16 ppwszString, const char *pszTag)
967	{
968	/*
969	* Validate input.
970	*/
971	Assert(VALID_PTR(ppwszString));
972	Assert(VALID_PTR(pszString));
973	*ppwszString = NULL;
974
975	/*
976	* Validate the input and calculate the length of the UTF-16 string.
977	*/
978	size_t cwc;
979	int rc = rtLatin1CalcUtf16Length(pszString, RTSTR_MAX, &cwc);
980	if (RT_SUCCESS(rc))
981	{
982	/*
983	* Allocate buffer.
984	*/
985	PRTUTF16 pwsz = (PRTUTF16)RTMemAllocTag((cwc + 1) * sizeof(RTUTF16), pszTag);
986	if (pwsz)
987	{
988	/*
989	* Encode the UTF-16 string.
990	*/
991	rc = rtLatin1RecodeAsUtf16(pszString, RTSTR_MAX, pwsz, cwc);
992	if (RT_SUCCESS(rc))
993	{
994	*ppwszString = pwsz;
995	return rc;
996	}
997	RTMemFree(pwsz);
998	}
999	else
1000	rc = VERR_NO_UTF16_MEMORY;
1001	}
1002	return rc;
1003	}
1004	RT_EXPORT_SYMBOL(RTLatin1ToUtf16Tag);
1005
1006
1007	RTDECL(int) RTLatin1ToUtf16ExTag(const char *pszString, size_t cchString,
1008	PRTUTF16 ppwsz, size_t cwc, size_t pcwc, const char *pszTag)
1009	{
1010	/*
1011	* Validate input.
1012	*/
1013	Assert(VALID_PTR(pszString));
1014	Assert(VALID_PTR(ppwsz));
1015	Assert(!pcwc \|\| VALID_PTR(pcwc));
1016
1017	/*
1018	* Validate the input and calculate the length of the UTF-16 string.
1019	*/
1020	size_t cwcResult;
1021	int rc = rtLatin1CalcUtf16Length(pszString, cchString, &cwcResult);
1022	if (RT_SUCCESS(rc))
1023	{
1024	if (pcwc)
1025	*pcwc = cwcResult;
1026
1027	/*
1028	* Check buffer size / Allocate buffer.
1029	*/
1030	bool fShouldFree;
1031	PRTUTF16 pwszResult;
1032	if (cwc > 0 && *ppwsz)
1033	{
1034	fShouldFree = false;
1035	if (cwc <= cwcResult)
1036	return VERR_BUFFER_OVERFLOW;
1037	pwszResult = *ppwsz;
1038	}
1039	else
1040	{
1041	*ppwsz = NULL;
1042	fShouldFree = true;
1043	cwc = RT_MAX(cwcResult + 1, cwc);
1044	pwszResult = (PRTUTF16)RTMemAllocTag(cwc * sizeof(RTUTF16), pszTag);
1045	}
1046	if (pwszResult)
1047	{
1048	/*
1049	* Encode the UTF-16 string.
1050	*/
1051	rc = rtLatin1RecodeAsUtf16(pszString, cchString, pwszResult, cwc - 1);
1052	if (RT_SUCCESS(rc))
1053	{
1054	*ppwsz = pwszResult;
1055	return rc;
1056	}
1057	if (fShouldFree)
1058	RTMemFree(pwszResult);
1059	}
1060	else
1061	rc = VERR_NO_UTF16_MEMORY;
1062	}
1063	return rc;
1064	}
1065	RT_EXPORT_SYMBOL(RTLatin1ToUtf16ExTag);
1066
1067
1068	RTDECL(size_t) RTLatin1CalcUtf16Len(const char *psz)
1069	{
1070	size_t cwc;
1071	int rc = rtLatin1CalcUtf16Length(psz, RTSTR_MAX, &cwc);
1072	return RT_SUCCESS(rc) ? cwc : 0;
1073	}
1074	RT_EXPORT_SYMBOL(RTLatin1CalcUtf16Len);
1075
1076
1077	RTDECL(int) RTLatin1CalcUtf16LenEx(const char psz, size_t cch, size_t pcwc)
1078	{
1079	size_t cwc;
1080	int rc = rtLatin1CalcUtf16Length(psz, cch, &cwc);
1081	if (pcwc)
1082	*pcwc = RT_SUCCESS(rc) ? cwc : ~(size_t)0;
1083	return rc;
1084	}
1085	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@ 34079

Download in other formats: