utf-16.cpp@ 44300

Last change on this file since 44300 was 40938, checked in by vboxsync, 13 years ago
runtime: backed out r77481,r77482,r77483,r77484,r77485
Property svn:eol-style set to `native` Property svn:keywords set to `Id`
File size: 30.8 KB

Line
1	/* $Id: utf-16.cpp 40938 2012-04-16 11:58:26Z 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	RTDECL(ssize_t) RTUtf16PurgeComplementSet(PRTUTF16 pwsz, PCRTUNICP puszValidSet, char chReplacement)
253	{
254	size_t cReplacements = 0;
255	AssertReturn(chReplacement && (unsigned)chReplacement < 128, -1);
256	/* Validate the encoding. */
257	for (;;)
258	{
259	RTUNICP Cp;
260	PCRTUNICP pCp;
261	PRTUTF16 pwszOld = pwsz;
262	if (RT_FAILURE(RTUtf16GetCpEx((PCRTUTF16 *)&pwsz, &Cp)))
263	return -1;
264	if (!Cp)
265	break;
266	for (pCp = puszValidSet; *pCp; pCp += 2)
267	{
268	AssertReturn(*(pCp + 1), -1);
269	if (pCp <= Cp && (pCp + 1) >= Cp) /* No, I won't do * and ++. */
270	break;
271	}
272	if (!*pCp)
273	{
274	for (; pwszOld != pwsz; ++pwszOld)
275	*pwszOld = chReplacement;
276	++cReplacements;
277	}
278	}
279	return cReplacements;
280	}
281	RT_EXPORT_SYMBOL(RTUtf16PurgeComplementSet);
282
283
284	/**
285	* Validate the UTF-16 encoding and calculates the length of an UTF-8 encoding.
286	*
287	* @returns iprt status code.
288	* @param pwsz The UTF-16 string.
289	* @param cwc The max length of the UTF-16 string to consider.
290	* @param pcch Where to store the length (excluding '\\0') of the UTF-8 string. (cch == cb, btw)
291	*/
292	static int rtUtf16CalcUtf8Length(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
293	{
294	int rc = VINF_SUCCESS;
295	size_t cch = 0;
296	while (cwc > 0)
297	{
298	RTUTF16 wc = *pwsz++; cwc--;
299	if (!wc)
300	break;
301	else if (wc < 0xd800 \|\| wc > 0xdfff)
302	{
303	if (wc < 0x80)
304	cch++;
305	else if (wc < 0x800)
306	cch += 2;
307	else if (wc < 0xfffe)
308	cch += 3;
309	else
310	{
311	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
312	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
313	break;
314	}
315	}
316	else
317	{
318	if (wc >= 0xdc00)
319	{
320	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
321	rc = VERR_INVALID_UTF16_ENCODING;
322	break;
323	}
324	if (cwc <= 0)
325	{
326	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
327	rc = VERR_INVALID_UTF16_ENCODING;
328	break;
329	}
330	wc = *pwsz++; cwc--;
331	if (wc < 0xdc00 \|\| wc > 0xdfff)
332	{
333	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
334	rc = VERR_INVALID_UTF16_ENCODING;
335	break;
336	}
337	cch += 4;
338	}
339	}
340
341
342	/* done */
343	*pcch = cch;
344	return rc;
345	}
346
347
348	/**
349	* Recodes an valid UTF-16 string as UTF-8.
350	*
351	* @returns iprt status code.
352	* @param pwsz The UTF-16 string.
353	* @param cwc The number of RTUTF16 characters to process from pwsz. The recoding
354	* will stop when cwc or '\\0' is reached.
355	* @param psz Where to store the UTF-8 string.
356	* @param cch The size of the UTF-8 buffer, excluding the terminator.
357	* @param pcch Where to store the number of octets actually encoded.
358	*/
359	static int rtUtf16RecodeAsUtf8(PCRTUTF16 pwsz, size_t cwc, char psz, size_t cch, size_t pcch)
360	{
361	unsigned char pwch = (unsigned char )psz;
362	int rc = VINF_SUCCESS;
363	while (cwc > 0)
364	{
365	RTUTF16 wc = *pwsz++; cwc--;
366	if (!wc)
367	break;
368	else if (wc < 0xd800 \|\| wc > 0xdfff)
369	{
370	if (wc < 0x80)
371	{
372	if (RT_UNLIKELY(cch < 1))
373	{
374	RTStrAssertMsgFailed(("Buffer overflow! 1\n"));
375	rc = VERR_BUFFER_OVERFLOW;
376	break;
377	}
378	cch--;
379	*pwch++ = (unsigned char)wc;
380	}
381	else if (wc < 0x800)
382	{
383	if (RT_UNLIKELY(cch < 2))
384	{
385	RTStrAssertMsgFailed(("Buffer overflow! 2\n"));
386	rc = VERR_BUFFER_OVERFLOW;
387	break;
388	}
389	cch -= 2;
390	*pwch++ = 0xc0 \| (wc >> 6);
391	*pwch++ = 0x80 \| (wc & 0x3f);
392	}
393	else if (wc < 0xfffe)
394	{
395	if (RT_UNLIKELY(cch < 3))
396	{
397	RTStrAssertMsgFailed(("Buffer overflow! 3\n"));
398	rc = VERR_BUFFER_OVERFLOW;
399	break;
400	}
401	cch -= 3;
402	*pwch++ = 0xe0 \| (wc >> 12);
403	*pwch++ = 0x80 \| ((wc >> 6) & 0x3f);
404	*pwch++ = 0x80 \| (wc & 0x3f);
405	}
406	else
407	{
408	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
409	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
410	break;
411	}
412	}
413	else
414	{
415	if (wc >= 0xdc00)
416	{
417	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
418	rc = VERR_INVALID_UTF16_ENCODING;
419	break;
420	}
421	if (cwc <= 0)
422	{
423	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
424	rc = VERR_INVALID_UTF16_ENCODING;
425	break;
426	}
427	RTUTF16 wc2 = *pwsz++; cwc--;
428	if (wc2 < 0xdc00 \|\| wc2 > 0xdfff)
429	{
430	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
431	rc = VERR_INVALID_UTF16_ENCODING;
432	break;
433	}
434	uint32_t CodePoint = 0x10000
435	+ ( ((wc & 0x3ff) << 10)
436	\| (wc2 & 0x3ff));
437	if (RT_UNLIKELY(cch < 4))
438	{
439	RTStrAssertMsgFailed(("Buffer overflow! 4\n"));
440	rc = VERR_BUFFER_OVERFLOW;
441	break;
442	}
443	cch -= 4;
444	*pwch++ = 0xf0 \| (CodePoint >> 18);
445	*pwch++ = 0x80 \| ((CodePoint >> 12) & 0x3f);
446	*pwch++ = 0x80 \| ((CodePoint >> 6) & 0x3f);
447	*pwch++ = 0x80 \| (CodePoint & 0x3f);
448	}
449	}
450
451	/* done */
452	*pwch = '\0';
453	pcch = (char )pwch - psz;
454	return rc;
455	}
456
457
458
459	RTDECL(int) RTUtf16ToUtf8Tag(PCRTUTF16 pwszString, char *ppszString, const char pszTag)
460	{
461	/*
462	* Validate input.
463	*/
464	Assert(VALID_PTR(ppszString));
465	Assert(VALID_PTR(pwszString));
466	*ppszString = NULL;
467
468	/*
469	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
470	*/
471	size_t cch;
472	int rc = rtUtf16CalcUtf8Length(pwszString, RTSTR_MAX, &cch);
473	if (RT_SUCCESS(rc))
474	{
475	/*
476	* Allocate buffer and recode it.
477	*/
478	char pszResult = (char )RTMemAllocTag(cch + 1, pszTag);
479	if (pszResult)
480	{
481	rc = rtUtf16RecodeAsUtf8(pwszString, RTSTR_MAX, pszResult, cch, &cch);
482	if (RT_SUCCESS(rc))
483	{
484	*ppszString = pszResult;
485	return rc;
486	}
487
488	RTMemFree(pszResult);
489	}
490	else
491	rc = VERR_NO_STR_MEMORY;
492	}
493	return rc;
494	}
495	RT_EXPORT_SYMBOL(RTUtf16ToUtf8Tag);
496
497
498	RTDECL(int) RTUtf16ToUtf8ExTag(PCRTUTF16 pwszString, size_t cwcString, char *ppsz, size_t cch, size_t pcch, const char *pszTag)
499	{
500	/*
501	* Validate input.
502	*/
503	Assert(VALID_PTR(pwszString));
504	Assert(VALID_PTR(ppsz));
505	Assert(!pcch \|\| VALID_PTR(pcch));
506
507	/*
508	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
509	*/
510	size_t cchResult;
511	int rc = rtUtf16CalcUtf8Length(pwszString, cwcString, &cchResult);
512	if (RT_SUCCESS(rc))
513	{
514	if (pcch)
515	*pcch = cchResult;
516
517	/*
518	* Check buffer size / Allocate buffer and recode it.
519	*/
520	bool fShouldFree;
521	char *pszResult;
522	if (cch > 0 && *ppsz)
523	{
524	fShouldFree = false;
525	if (RT_UNLIKELY(cch <= cchResult))
526	return VERR_BUFFER_OVERFLOW;
527	pszResult = *ppsz;
528	}
529	else
530	{
531	*ppsz = NULL;
532	fShouldFree = true;
533	cch = RT_MAX(cch, cchResult + 1);
534	pszResult = (char *)RTStrAllocTag(cch, pszTag);
535	}
536	if (pszResult)
537	{
538	rc = rtUtf16RecodeAsUtf8(pwszString, cwcString, pszResult, cch - 1, &cch);
539	if (RT_SUCCESS(rc))
540	{
541	*ppsz = pszResult;
542	return rc;
543	}
544
545	if (fShouldFree)
546	RTStrFree(pszResult);
547	}
548	else
549	rc = VERR_NO_STR_MEMORY;
550	}
551	return rc;
552	}
553	RT_EXPORT_SYMBOL(RTUtf16ToUtf8ExTag);
554
555
556	RTDECL(size_t) RTUtf16CalcUtf8Len(PCRTUTF16 pwsz)
557	{
558	size_t cch;
559	int rc = rtUtf16CalcUtf8Length(pwsz, RTSTR_MAX, &cch);
560	return RT_SUCCESS(rc) ? cch : 0;
561	}
562	RT_EXPORT_SYMBOL(RTUtf16CalcUtf8Len);
563
564
565	RTDECL(int) RTUtf16CalcUtf8LenEx(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
566	{
567	size_t cch;
568	int rc = rtUtf16CalcUtf8Length(pwsz, cwc, &cch);
569	if (pcch)
570	*pcch = RT_SUCCESS(rc) ? cch : ~(size_t)0;
571	return rc;
572	}
573	RT_EXPORT_SYMBOL(RTUtf16CalcUtf8LenEx);
574
575
576	RTDECL(RTUNICP) RTUtf16GetCpInternal(PCRTUTF16 pwsz)
577	{
578	const RTUTF16 wc = *pwsz;
579
580	/* simple */
581	if (wc < 0xd800 \|\| (wc > 0xdfff && wc < 0xfffe))
582	return wc;
583	if (wc < 0xfffe)
584	{
585	/* surrogate pair */
586	if (wc < 0xdc00)
587	{
588	const RTUTF16 wc2 = pwsz[1];
589	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
590	{
591	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
592	return uc;
593	}
594
595	RTStrAssertMsgFailed(("wc=%#08x wc2=%#08x - invalid 2nd char in surrogate pair\n", wc, wc2));
596	}
597	else
598	RTStrAssertMsgFailed(("wc=%#08x - invalid surrogate pair order\n", wc));
599	}
600	else
601	RTStrAssertMsgFailed(("wc=%#08x - endian indicator\n", wc));
602	return RTUNICP_INVALID;
603	}
604	RT_EXPORT_SYMBOL(RTUtf16GetCpInternal);
605
606
607	RTDECL(int) RTUtf16GetCpExInternal(PCRTUTF16 *ppwsz, PRTUNICP pCp)
608	{
609	const RTUTF16 wc = **ppwsz;
610
611	/* simple */
612	if (wc < 0xd800 \|\| (wc > 0xdfff && wc < 0xfffe))
613	{
614	(*ppwsz)++;
615	*pCp = wc;
616	return VINF_SUCCESS;
617	}
618
619	int rc;
620	if (wc < 0xfffe)
621	{
622	/* surrogate pair */
623	if (wc < 0xdc00)
624	{
625	const RTUTF16 wc2 = (*ppwsz)[1];
626	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
627	{
628	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
629	*pCp = uc;
630	(*ppwsz) += 2;
631	return VINF_SUCCESS;
632	}
633
634	RTStrAssertMsgFailed(("wc=%#08x wc2=%#08x - invalid 2nd char in surrogate pair\n", wc, wc2));
635	}
636	else
637	RTStrAssertMsgFailed(("wc=%#08x - invalid surrogate pair order\n", wc));
638	rc = VERR_INVALID_UTF16_ENCODING;
639	}
640	else
641	{
642	RTStrAssertMsgFailed(("wc=%#08x - endian indicator\n", wc));
643	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
644	}
645	*pCp = RTUNICP_INVALID;
646	(*ppwsz)++;
647	return rc;
648	}
649	RT_EXPORT_SYMBOL(RTUtf16GetCpExInternal);
650
651
652	RTDECL(PRTUTF16) RTUtf16PutCpInternal(PRTUTF16 pwsz, RTUNICP CodePoint)
653	{
654	/* simple */
655	if ( CodePoint < 0xd800
656	\|\| ( CodePoint > 0xdfff
657	&& CodePoint < 0xfffe))
658	{
659	*pwsz++ = (RTUTF16)CodePoint;
660	return pwsz;
661	}
662
663	/* surrogate pair */
664	if (CodePoint >= 0x10000 && CodePoint <= 0x0010ffff)
665	{
666	CodePoint -= 0x10000;
667	*pwsz++ = 0xd800 \| (CodePoint >> 10);
668	*pwsz++ = 0xdc00 \| (CodePoint & 0x3ff);
669	return pwsz;
670	}
671
672	/* invalid code point. */
673	RTStrAssertMsgFailed(("Invalid codepoint %#x\n", CodePoint));
674	*pwsz++ = 0x7f;
675	return pwsz;
676	}
677	RT_EXPORT_SYMBOL(RTUtf16PutCpInternal);
678
679
680	/**
681	* Validate the UTF-16 encoding and calculates the length of a Latin1 encoding.
682	*
683	* @returns iprt status code.
684	* @param pwsz The UTF-16 string.
685	* @param cwc The max length of the UTF-16 string to consider.
686	* @param pcch Where to store the length (excluding '\\0') of the Latin1 string. (cch == cb, btw)
687	*/
688	static int rtUtf16CalcLatin1Length(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
689	{
690	int rc = VINF_SUCCESS;
691	size_t cch = 0;
692	while (cwc > 0)
693	{
694	RTUTF16 wc = *pwsz++; cwc--;
695	if (!wc)
696	break;
697	else if (RT_LIKELY(wc < 0x100))
698	++cch;
699	else
700	{
701	if (wc < 0xd800 \|\| wc > 0xdfff)
702	{
703	if (wc >= 0xfffe)
704	{
705	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
706	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
707	break;
708	}
709	}
710	else
711	{
712	if (wc >= 0xdc00)
713	{
714	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
715	rc = VERR_INVALID_UTF16_ENCODING;
716	break;
717	}
718	if (cwc <= 0)
719	{
720	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
721	rc = VERR_INVALID_UTF16_ENCODING;
722	break;
723	}
724	wc = *pwsz++; cwc--;
725	if (wc < 0xdc00 \|\| wc > 0xdfff)
726	{
727	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
728	rc = VERR_INVALID_UTF16_ENCODING;
729	break;
730	}
731	}
732
733	rc = VERR_NO_TRANSLATION;
734	break;
735	}
736	}
737
738	/* done */
739	*pcch = cch;
740	return rc;
741	}
742
743
744	/**
745	* Recodes an valid UTF-16 string as Latin1.
746	*
747	* @returns iprt status code.
748	* @param pwsz The UTF-16 string.
749	* @param cwc The number of RTUTF16 characters to process from pwsz. The recoding
750	* will stop when cwc or '\\0' is reached.
751	* @param psz Where to store the Latin1 string.
752	* @param cch The size of the Latin1 buffer, excluding the terminator.
753	*/
754	static int rtUtf16RecodeAsLatin1(PCRTUTF16 pwsz, size_t cwc, char *psz, size_t cch)
755	{
756	unsigned char pch = (unsigned char )psz;
757	int rc = VINF_SUCCESS;
758	while (cwc > 0)
759	{
760	RTUTF16 wc = *pwsz++; cwc--;
761	if (!wc)
762	break;
763	if (RT_LIKELY(wc < 0x100))
764	{
765	if (RT_UNLIKELY(cch < 1))
766	{
767	RTStrAssertMsgFailed(("Buffer overflow! 1\n"));
768	rc = VERR_BUFFER_OVERFLOW;
769	break;
770	}
771	cch--;
772	*pch++ = (unsigned char)wc;
773	}
774	else
775	{
776	if (wc < 0xd800 \|\| wc > 0xdfff)
777	{
778	if (wc >= 0xfffe)
779	{
780	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
781	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
782	break;
783	}
784	}
785	else
786	{
787	if (wc >= 0xdc00)
788	{
789	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
790	rc = VERR_INVALID_UTF16_ENCODING;
791	break;
792	}
793	if (cwc <= 0)
794	{
795	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
796	rc = VERR_INVALID_UTF16_ENCODING;
797	break;
798	}
799	RTUTF16 wc2 = *pwsz++; cwc--;
800	if (wc2 < 0xdc00 \|\| wc2 > 0xdfff)
801	{
802	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
803	rc = VERR_INVALID_UTF16_ENCODING;
804	break;
805	}
806	}
807
808	rc = VERR_NO_TRANSLATION;
809	break;
810	}
811	}
812
813	/* done */
814	*pch = '\0';
815	return rc;
816	}
817
818
819	RTDECL(int) RTUtf16ToLatin1Tag(PCRTUTF16 pwszString, char *ppszString, const char pszTag)
820	{
821	/*
822	* Validate input.
823	*/
824	Assert(VALID_PTR(ppszString));
825	Assert(VALID_PTR(pwszString));
826	*ppszString = NULL;
827
828	/*
829	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
830	*/
831	size_t cch;
832	int rc = rtUtf16CalcLatin1Length(pwszString, RTSTR_MAX, &cch);
833	if (RT_SUCCESS(rc))
834	{
835	/*
836	* Allocate buffer and recode it.
837	*/
838	char pszResult = (char )RTMemAllocTag(cch + 1, pszTag);
839	if (pszResult)
840	{
841	rc = rtUtf16RecodeAsLatin1(pwszString, RTSTR_MAX, pszResult, cch);
842	if (RT_SUCCESS(rc))
843	{
844	*ppszString = pszResult;
845	return rc;
846	}
847
848	RTMemFree(pszResult);
849	}
850	else
851	rc = VERR_NO_STR_MEMORY;
852	}
853	return rc;
854	}
855	RT_EXPORT_SYMBOL(RTUtf16ToLatin1Tag);
856
857
858	RTDECL(int) RTUtf16ToLatin1ExTag(PCRTUTF16 pwszString, size_t cwcString, char *ppsz, size_t cch, size_t pcch, const char *pszTag)
859	{
860	/*
861	* Validate input.
862	*/
863	AssertPtr(pwszString);
864	AssertPtr(ppsz);
865	AssertPtrNull(pcch);
866
867	/*
868	* Validate the UTF-16 string and calculate the length of the Latin1 encoding of it.
869	*/
870	size_t cchResult;
871	int rc = rtUtf16CalcLatin1Length(pwszString, cwcString, &cchResult);
872	if (RT_SUCCESS(rc))
873	{
874	if (pcch)
875	*pcch = cchResult;
876
877	/*
878	* Check buffer size / Allocate buffer and recode it.
879	*/
880	bool fShouldFree;
881	char *pszResult;
882	if (cch > 0 && *ppsz)
883	{
884	fShouldFree = false;
885	if (cch <= cchResult)
886	return VERR_BUFFER_OVERFLOW;
887	pszResult = *ppsz;
888	}
889	else
890	{
891	*ppsz = NULL;
892	fShouldFree = true;
893	cch = RT_MAX(cch, cchResult + 1);
894	pszResult = (char *)RTMemAllocTag(cch, pszTag);
895	}
896	if (pszResult)
897	{
898	rc = rtUtf16RecodeAsLatin1(pwszString, cwcString, pszResult, cch - 1);
899	if (RT_SUCCESS(rc))
900	{
901	*ppsz = pszResult;
902	return rc;
903	}
904
905	if (fShouldFree)
906	RTMemFree(pszResult);
907	}
908	else
909	rc = VERR_NO_STR_MEMORY;
910	}
911	return rc;
912	}
913	RT_EXPORT_SYMBOL(RTUtf16ToLatin1ExTag);
914
915
916	RTDECL(size_t) RTUtf16CalcLatin1Len(PCRTUTF16 pwsz)
917	{
918	size_t cch;
919	int rc = rtUtf16CalcLatin1Length(pwsz, RTSTR_MAX, &cch);
920	return RT_SUCCESS(rc) ? cch : 0;
921	}
922	RT_EXPORT_SYMBOL(RTUtf16CalcLatin1Len);
923
924
925	RTDECL(int) RTUtf16CalcLatin1LenEx(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
926	{
927	size_t cch;
928	int rc = rtUtf16CalcLatin1Length(pwsz, cwc, &cch);
929	if (pcch)
930	*pcch = RT_SUCCESS(rc) ? cch : ~(size_t)0;
931	return rc;
932	}
933	RT_EXPORT_SYMBOL(RTUtf16CalcLatin1LenEx);
934
935
936	/**
937	* Calculates the UTF-16 length of a Latin1 string. In fact this is just the
938	* original length, but the function saves us nasty comments to that effect
939	* all over the place.
940	*
941	* @returns IPRT status code.
942	* @param psz Pointer to the Latin1 string.
943	* @param cch The max length of the string. (btw cch = cb)
944	* Use RTSTR_MAX if all of the string is to be examined.s
945	* @param pcwc Where to store the length of the UTF-16 string as a number of RTUTF16 characters.
946	*/
947	static int rtLatin1CalcUtf16Length(const char psz, size_t cch, size_t pcwc)
948	{
949	*pcwc = RTStrNLen(psz, cch);
950	return VINF_SUCCESS;
951	}
952
953
954	/**
955	* Recodes a Latin1 string as UTF-16. This is just a case of expanding it to
956	* sixteen bits, as Unicode is a superset of Latin1.
957	*
958	* Since we know the input is valid, we do not perform length checks.
959	*
960	* @returns iprt status code.
961	* @param psz The Latin1 string to recode.
962	* @param cch The number of chars (the type char, so bytes if you like) to process of the Latin1 string.
963	* The recoding will stop when cch or '\\0' is reached. Pass RTSTR_MAX to process up to '\\0'.
964	* @param pwsz Where to store the UTF-16 string.
965	* @param cwc The number of RTUTF16 items the pwsz buffer can hold, excluding the terminator ('\\0').
966	*/
967	static int rtLatin1RecodeAsUtf16(const char *psz, size_t cch, PRTUTF16 pwsz, size_t cwc)
968	{
969	int rc = VINF_SUCCESS;
970	const unsigned char puch = (const unsigned char )psz;
971	PRTUTF16 pwc = pwsz;
972	while (cch-- > 0)
973	{
974	/* read the next char and check for terminator. */
975	const unsigned char uch = *puch;
976	if (!uch)
977	break;
978
979	/* check for output overflow */
980	if (RT_UNLIKELY(cwc < 1))
981	{
982	rc = VERR_BUFFER_OVERFLOW;
983	break;
984	}
985
986	/* expand the code point */
987	*pwc++ = uch;
988	cwc--;
989	puch++;
990	}
991
992	/* done */
993	*pwc = '\0';
994	return rc;
995	}
996
997
998	RTDECL(int) RTLatin1ToUtf16Tag(const char pszString, PRTUTF16 ppwszString, const char *pszTag)
999	{
1000	/*
1001	* Validate input.
1002	*/
1003	Assert(VALID_PTR(ppwszString));
1004	Assert(VALID_PTR(pszString));
1005	*ppwszString = NULL;
1006
1007	/*
1008	* Validate the input and calculate the length of the UTF-16 string.
1009	*/
1010	size_t cwc;
1011	int rc = rtLatin1CalcUtf16Length(pszString, RTSTR_MAX, &cwc);
1012	if (RT_SUCCESS(rc))
1013	{
1014	/*
1015	* Allocate buffer.
1016	*/
1017	PRTUTF16 pwsz = (PRTUTF16)RTMemAllocTag((cwc + 1) * sizeof(RTUTF16), pszTag);
1018	if (pwsz)
1019	{
1020	/*
1021	* Encode the UTF-16 string.
1022	*/
1023	rc = rtLatin1RecodeAsUtf16(pszString, RTSTR_MAX, pwsz, cwc);
1024	if (RT_SUCCESS(rc))
1025	{
1026	*ppwszString = pwsz;
1027	return rc;
1028	}
1029	RTMemFree(pwsz);
1030	}
1031	else
1032	rc = VERR_NO_UTF16_MEMORY;
1033	}
1034	return rc;
1035	}
1036	RT_EXPORT_SYMBOL(RTLatin1ToUtf16Tag);
1037
1038
1039	RTDECL(int) RTLatin1ToUtf16ExTag(const char *pszString, size_t cchString,
1040	PRTUTF16 ppwsz, size_t cwc, size_t pcwc, const char *pszTag)
1041	{
1042	/*
1043	* Validate input.
1044	*/
1045	Assert(VALID_PTR(pszString));
1046	Assert(VALID_PTR(ppwsz));
1047	Assert(!pcwc \|\| VALID_PTR(pcwc));
1048
1049	/*
1050	* Validate the input and calculate the length of the UTF-16 string.
1051	*/
1052	size_t cwcResult;
1053	int rc = rtLatin1CalcUtf16Length(pszString, cchString, &cwcResult);
1054	if (RT_SUCCESS(rc))
1055	{
1056	if (pcwc)
1057	*pcwc = cwcResult;
1058
1059	/*
1060	* Check buffer size / Allocate buffer.
1061	*/
1062	bool fShouldFree;
1063	PRTUTF16 pwszResult;
1064	if (cwc > 0 && *ppwsz)
1065	{
1066	fShouldFree = false;
1067	if (cwc <= cwcResult)
1068	return VERR_BUFFER_OVERFLOW;
1069	pwszResult = *ppwsz;
1070	}
1071	else
1072	{
1073	*ppwsz = NULL;
1074	fShouldFree = true;
1075	cwc = RT_MAX(cwcResult + 1, cwc);
1076	pwszResult = (PRTUTF16)RTMemAllocTag(cwc * sizeof(RTUTF16), pszTag);
1077	}
1078	if (pwszResult)
1079	{
1080	/*
1081	* Encode the UTF-16 string.
1082	*/
1083	rc = rtLatin1RecodeAsUtf16(pszString, cchString, pwszResult, cwc - 1);
1084	if (RT_SUCCESS(rc))
1085	{
1086	*ppwsz = pwszResult;
1087	return rc;
1088	}
1089	if (fShouldFree)
1090	RTMemFree(pwszResult);
1091	}
1092	else
1093	rc = VERR_NO_UTF16_MEMORY;
1094	}
1095	return rc;
1096	}
1097	RT_EXPORT_SYMBOL(RTLatin1ToUtf16ExTag);
1098
1099
1100	RTDECL(size_t) RTLatin1CalcUtf16Len(const char *psz)
1101	{
1102	size_t cwc;
1103	int rc = rtLatin1CalcUtf16Length(psz, RTSTR_MAX, &cwc);
1104	return RT_SUCCESS(rc) ? cwc : 0;
1105	}
1106	RT_EXPORT_SYMBOL(RTLatin1CalcUtf16Len);
1107
1108
1109	RTDECL(int) RTLatin1CalcUtf16LenEx(const char psz, size_t cch, size_t pcwc)
1110	{
1111	size_t cwc;
1112	int rc = rtLatin1CalcUtf16Length(psz, cch, &cwc);
1113	if (pcwc)
1114	*pcwc = RT_SUCCESS(rc) ? cwc : ~(size_t)0;
1115	return rc;
1116	}
1117	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@ 44300

Download in other formats: