utf-16.cpp@ 67391

Last change on this file since 67391 was 67391, checked in by vboxsync, 8 years ago
IPRT: More ISO maker code.
Property svn:eol-style set to `native` Property svn:keywords set to `Id Revision`
File size: 30.2 KB

Line
1	/* $Id: utf-16.cpp 67391 2017-06-14 12:13:48Z vboxsync $ */
2	/** @file
3	* IPRT - UTF-16.
4	*/
5
6	/*
7	* Copyright (C) 2006-2016 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/asm.h>
36	#include <iprt/mem.h>
37	#include <iprt/assert.h>
38	#include <iprt/err.h>
39	#include "internal/string.h"
40
41
42	/**
43	* Get get length in code points of an UTF-16 encoded string, validating the
44	* string while doing so.
45	*
46	* @returns IPRT status code.
47	* @param pwsz Pointer to the UTF-16 string.
48	* @param cwc The max length of the string in UTF-16 units. Use
49	* RTSTR_MAX if all of the string is to be examined.
50	* @param pcuc Where to store the length in unicode code points.
51	* @param pcwcActual Where to store the actual size of the UTF-16 string
52	* on success. Optional.
53	*/
54	static int rtUtf16Length(PCRTUTF16 pwsz, size_t cwc, size_t pcuc, size_t pcwcActual)
55	{
56	PCRTUTF16 pwszStart = pwsz;
57	size_t cCodePoints = 0;
58	while (cwc > 0)
59	{
60	RTUTF16 wc = *pwsz;
61	if (!wc)
62	break;
63	if (wc < 0xd800 \|\| wc > 0xdfff)
64	{
65	cCodePoints++;
66	pwsz++;
67	cwc--;
68	}
69	/* Surrogate pair: */
70	else if (wc >= 0xdc00)
71	{
72	RTStrAssertMsgFailed(("Lone UTF-16 trail surrogate: %#x (%.Rhxs)\n", wc, RT_MIN(cwc 2, 10), pwsz));
73	return VERR_INVALID_UTF16_ENCODING;
74	}
75	else if (cwc < 2)
76	{
77	RTStrAssertMsgFailed(("Lone UTF-16 lead surrogate: %#x\n", wc));
78	return VERR_INVALID_UTF16_ENCODING;
79	}
80	else
81	{
82	RTUTF16 wcTrail = pwsz[1];
83	if (wcTrail < 0xdc00 \|\| wcTrail > 0xdfff)
84	{
85	RTStrAssertMsgFailed(("Invalid UTF-16 trail surrogate: %#x (lead %#x)\n", wcTrail, wc));
86	return VERR_INVALID_UTF16_ENCODING;
87	}
88
89	cCodePoints++;
90	pwsz += 2;
91	cwc -= 2;
92	}
93	}
94
95	/* done */
96	*pcuc = cCodePoints;
97	if (pcwcActual)
98	*pcwcActual = pwsz - pwszStart;
99	return VINF_SUCCESS;
100	}
101
102
103	RTDECL(PRTUTF16) RTUtf16AllocTag(size_t cb, const char *pszTag)
104	{
105	if (cb > sizeof(RTUTF16))
106	cb = RT_ALIGN_Z(cb, sizeof(RTUTF16));
107	else
108	cb = sizeof(RTUTF16);
109	PRTUTF16 pwsz = (PRTUTF16)RTMemAllocTag(cb, pszTag);
110	if (pwsz)
111	*pwsz = '\0';
112	return pwsz;
113	}
114	RT_EXPORT_SYMBOL(RTUtf16AllocTag);
115
116
117	RTDECL(int) RTUtf16ReallocTag(PRTUTF16 ppwsz, size_t cbNew, const char pszTag)
118	{
119	PRTUTF16 pwszOld = *ppwsz;
120	cbNew = RT_ALIGN_Z(cbNew, sizeof(RTUTF16));
121	if (!cbNew)
122	{
123	RTMemFree(pwszOld);
124	*ppwsz = NULL;
125	}
126	else if (pwszOld)
127	{
128	PRTUTF16 pwszNew = (PRTUTF16)RTMemReallocTag(pwszOld, cbNew, pszTag);
129	if (!pwszNew)
130	return VERR_NO_STR_MEMORY;
131	pwszNew[cbNew / sizeof(RTUTF16) - 1] = '\0';
132	*ppwsz = pwszNew;
133	}
134	else
135	{
136	PRTUTF16 pwszNew = (PRTUTF16)RTMemAllocTag(cbNew, pszTag);
137	if (!pwszNew)
138	return VERR_NO_UTF16_MEMORY;
139	pwszNew[0] = '\0';
140	pwszNew[cbNew / sizeof(RTUTF16) - 1] = '\0';
141	*ppwsz = pwszNew;
142	}
143	return VINF_SUCCESS;
144	}
145	RT_EXPORT_SYMBOL(RTUtf16ReallocTag);
146
147
148	RTDECL(void) RTUtf16Free(PRTUTF16 pwszString)
149	{
150	if (pwszString)
151	RTMemTmpFree(pwszString);
152	}
153	RT_EXPORT_SYMBOL(RTUtf16Free);
154
155
156	RTDECL(PRTUTF16) RTUtf16DupTag(PCRTUTF16 pwszString, const char *pszTag)
157	{
158	Assert(pwszString);
159	size_t cb = (RTUtf16Len(pwszString) + 1) * sizeof(RTUTF16);
160	PRTUTF16 pwsz = (PRTUTF16)RTMemAllocTag(cb, pszTag);
161	if (pwsz)
162	memcpy(pwsz, pwszString, cb);
163	return pwsz;
164	}
165	RT_EXPORT_SYMBOL(RTUtf16DupTag);
166
167
168	RTDECL(int) RTUtf16DupExTag(PRTUTF16 ppwszString, PCRTUTF16 pwszString, size_t cwcExtra, const char pszTag)
169	{
170	Assert(pwszString);
171	size_t cb = (RTUtf16Len(pwszString) + 1) * sizeof(RTUTF16);
172	PRTUTF16 pwsz = (PRTUTF16)RTMemAllocTag(cb + cwcExtra * sizeof(RTUTF16), pszTag);
173	if (pwsz)
174	{
175	memcpy(pwsz, pwszString, cb);
176	*ppwszString = pwsz;
177	return VINF_SUCCESS;
178	}
179	return VERR_NO_MEMORY;
180	}
181	RT_EXPORT_SYMBOL(RTUtf16DupExTag);
182
183
184	RTDECL(size_t) RTUtf16Len(PCRTUTF16 pwszString)
185	{
186	if (!pwszString)
187	return 0;
188
189	PCRTUTF16 pwsz = pwszString;
190	while (*pwsz)
191	pwsz++;
192	return pwsz - pwszString;
193	}
194	RT_EXPORT_SYMBOL(RTUtf16Len);
195
196
197	RTDECL(int) RTUtf16Cmp(register PCRTUTF16 pwsz1, register PCRTUTF16 pwsz2)
198	{
199	if (pwsz1 == pwsz2)
200	return 0;
201	if (!pwsz1)
202	return -1;
203	if (!pwsz2)
204	return 1;
205
206	for (;;)
207	{
208	register RTUTF16 wcs = *pwsz1;
209	register int iDiff = wcs - *pwsz2;
210	if (iDiff \|\| !wcs)
211	return iDiff;
212	pwsz1++;
213	pwsz2++;
214	}
215	}
216	RT_EXPORT_SYMBOL(RTUtf16Cmp);
217
218
219	RTDECL(int) RTUtf16CmpUtf8(PCRTUTF16 pwsz1, const char *psz2)
220	{
221	/*
222	* NULL and empty strings are all the same.
223	*/
224	if (!pwsz1)
225	return !psz2 \|\| !*psz2 ? 0 : -1;
226	if (!psz2)
227	return !*pwsz1 ? 0 : 1;
228
229	/*
230	* Compare with a UTF-8 string by enumerating them char by char.
231	*/
232	for (;;)
233	{
234	RTUNICP uc1;
235	int rc = RTUtf16GetCpEx(&pwsz1, &uc1);
236	AssertRCReturn(rc, 1);
237
238	RTUNICP uc2;
239	rc = RTStrGetCpEx(&psz2, &uc2);
240	AssertRCReturn(rc, -1);
241	if (uc1 == uc2)
242	{
243	if (uc1)
244	continue;
245	return 0;
246	}
247	return uc1 < uc2 ? -1 : 1;
248	}
249	}
250	RT_EXPORT_SYMBOL(RTUtf16CmpUtf8);
251
252
253	RTDECL(int) RTUtf16ValidateEncoding(PCRTUTF16 pwsz)
254	{
255	return RTUtf16ValidateEncodingEx(pwsz, RTSTR_MAX, 0);
256	}
257	RT_EXPORT_SYMBOL(RTUtf16ValidateEncoding);
258
259
260	RTDECL(int) RTUtf16ValidateEncodingEx(PCRTUTF16 pwsz, size_t cwc, uint32_t fFlags)
261	{
262	AssertReturn(!(fFlags & ~(RTSTR_VALIDATE_ENCODING_ZERO_TERMINATED \| RTSTR_VALIDATE_ENCODING_EXACT_LENGTH)),
263	VERR_INVALID_PARAMETER);
264	AssertPtr(pwsz);
265
266	/*
267	* Use rtUtf16Length for the job.
268	*/
269	size_t cwcActual = 0; /* Shut up cc1plus. */
270	size_t cCpsIgnored;
271	int rc = rtUtf16Length(pwsz, cwc, &cCpsIgnored, &cwcActual);
272	if (RT_SUCCESS(rc))
273	{
274	if (fFlags & RTSTR_VALIDATE_ENCODING_EXACT_LENGTH)
275	{
276	if (fFlags & RTSTR_VALIDATE_ENCODING_ZERO_TERMINATED)
277	cwcActual++;
278	if (cwcActual == cwc)
279	rc = VINF_SUCCESS;
280	else if (cwcActual < cwc)
281	rc = VERR_BUFFER_UNDERFLOW;
282	else
283	rc = VERR_BUFFER_OVERFLOW;
284	}
285	else if ( (fFlags & RTSTR_VALIDATE_ENCODING_ZERO_TERMINATED)
286	&& cwcActual >= cwc)
287	rc = VERR_BUFFER_OVERFLOW;
288	}
289	return rc;
290	}
291	RT_EXPORT_SYMBOL(RTUtf16ValidateEncodingEx);
292
293
294	RTDECL(bool) RTUtf16IsValidEncoding(PCRTUTF16 pwsz)
295	{
296	int rc = RTUtf16ValidateEncodingEx(pwsz, RTSTR_MAX, 0);
297	return RT_SUCCESS(rc);
298	}
299	RT_EXPORT_SYMBOL(RTUtf16IsValidEncoding);
300
301
302	/**
303	* Helper for RTUtf16PurgeComplementSet.
304	*
305	* @returns true if @a Cp is valid, false if not.
306	* @param Cp The code point to validate.
307	* @param puszValidPairs Pair of valid code point sets.
308	* @param cValidPairs Number of pairs.
309	*/
310	DECLINLINE(bool) rtUtf16PurgeIsInSet(RTUNICP Cp, PCRTUNICP puszValidPairs, uint32_t cValidPairs)
311	{
312	while (cValidPairs-- > 0)
313	{
314	if ( Cp >= puszValidPairs[0]
315	&& Cp <= puszValidPairs[1])
316	return true;
317	puszValidPairs += 2;
318	}
319	return false;
320	}
321
322
323	RTDECL(ssize_t) RTUtf16PurgeComplementSet(PRTUTF16 pwsz, PCRTUNICP puszValidPairs, char chReplacement)
324	{
325	AssertReturn(chReplacement && (unsigned)chReplacement < 128, -1);
326
327	/*
328	* Calc valid pairs and check that we've got an even number.
329	*/
330	uint32_t cValidPairs = 0;
331	while (puszValidPairs[cValidPairs * 2])
332	{
333	AssertReturn(puszValidPairs[cValidPairs * 2 + 1], -1);
334	AssertMsg(puszValidPairs[cValidPairs * 2] <= puszValidPairs[cValidPairs * 2 + 1],
335	("%#x vs %#x\n", puszValidPairs[cValidPairs * 2], puszValidPairs[cValidPairs * 2 + 1]));
336	cValidPairs++;
337	}
338
339	/*
340	* Do the replacing.
341	*/
342	ssize_t cReplacements = 0;
343	for (;;)
344	{
345	PRTUTF16 pwszCur = pwsz;
346	RTUNICP Cp;
347	int rc = RTUtf16GetCpEx((PCRTUTF16 *)&pwsz, &Cp);
348	if (RT_SUCCESS(rc))
349	{
350	if (Cp)
351	{
352	if (!rtUtf16PurgeIsInSet(Cp, puszValidPairs, cValidPairs))
353	{
354	for (; pwszCur != pwsz; ++pwszCur)
355	*pwszCur = chReplacement;
356	++cReplacements;
357	}
358	}
359	else
360	break;
361	}
362	else
363	return -1;
364	}
365	return cReplacements;
366	}
367	RT_EXPORT_SYMBOL(RTUtf16PurgeComplementSet);
368
369
370	/**
371	* Validate the UTF-16 encoding and calculates the length of an UTF-8 encoding.
372	*
373	* @returns iprt status code.
374	* @param pwsz The UTF-16 string.
375	* @param cwc The max length of the UTF-16 string to consider.
376	* @param pcch Where to store the length (excluding '\\0') of the UTF-8 string. (cch == cb, btw)
377	*
378	* @note rtUtf16BigCalcUtf8Length is a copy of this.
379	*/
380	static int rtUtf16CalcUtf8Length(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
381	{
382	int rc = VINF_SUCCESS;
383	size_t cch = 0;
384	while (cwc > 0)
385	{
386	RTUTF16 wc = *pwsz++; cwc--;
387	if (!wc)
388	break;
389	if (wc < 0xd800 \|\| wc > 0xdfff)
390	{
391	if (wc < 0x80)
392	cch++;
393	else if (wc < 0x800)
394	cch += 2;
395	else if (wc < 0xfffe)
396	cch += 3;
397	else
398	{
399	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
400	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
401	break;
402	}
403	}
404	else
405	{
406	if (wc >= 0xdc00)
407	{
408	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
409	rc = VERR_INVALID_UTF16_ENCODING;
410	break;
411	}
412	if (cwc <= 0)
413	{
414	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
415	rc = VERR_INVALID_UTF16_ENCODING;
416	break;
417	}
418	wc = *pwsz++; cwc--;
419	if (wc < 0xdc00 \|\| wc > 0xdfff)
420	{
421	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
422	rc = VERR_INVALID_UTF16_ENCODING;
423	break;
424	}
425	cch += 4;
426	}
427	}
428
429
430	/* done */
431	*pcch = cch;
432	return rc;
433	}
434
435
436	/**
437	* Validate the UTF-16BE encoding and calculates the length of an UTF-8
438	* encoding.
439	*
440	* @returns iprt status code.
441	* @param pwsz The UTF-16 string.
442	* @param cwc The max length of the UTF-16BE string to consider.
443	* @param pcch Where to store the length (excluding '\\0') of the UTF-8 string. (cch == cb, btw)
444	*
445	* @note Code is a copy of rtUtf16CalcUtf8Length, but with two RT_BE2H_U16
446	* invocations inserted.
447	*/
448	static int rtUtf16BigCalcUtf8Length(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
449	{
450	int rc = VINF_SUCCESS;
451	size_t cch = 0;
452	while (cwc > 0)
453	{
454	RTUTF16 wc = *pwsz++; cwc--;
455	if (!wc)
456	break;
457	wc = RT_BE2H_U16(wc);
458	if (wc < 0xd800 \|\| wc > 0xdfff)
459	{
460	if (wc < 0x80)
461	cch++;
462	else if (wc < 0x800)
463	cch += 2;
464	else if (wc < 0xfffe)
465	cch += 3;
466	else
467	{
468	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
469	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
470	break;
471	}
472	}
473	else
474	{
475	if (wc >= 0xdc00)
476	{
477	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
478	rc = VERR_INVALID_UTF16_ENCODING;
479	break;
480	}
481	if (cwc <= 0)
482	{
483	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
484	rc = VERR_INVALID_UTF16_ENCODING;
485	break;
486	}
487	wc = *pwsz++; cwc--;
488	wc = RT_BE2H_U16(wc);
489	if (wc < 0xdc00 \|\| wc > 0xdfff)
490	{
491	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
492	rc = VERR_INVALID_UTF16_ENCODING;
493	break;
494	}
495	cch += 4;
496	}
497	}
498
499
500	/* done */
501	*pcch = cch;
502	return rc;
503	}
504
505
506	/**
507	* Recodes an valid UTF-16 string as UTF-8.
508	*
509	* @returns iprt status code.
510	* @param pwsz The UTF-16 string.
511	* @param cwc The number of RTUTF16 characters to process from pwsz. The recoding
512	* will stop when cwc or '\\0' is reached.
513	* @param psz Where to store the UTF-8 string.
514	* @param cch The size of the UTF-8 buffer, excluding the terminator.
515	* @param pcch Where to store the number of octets actually encoded.
516	* @note rtUtf16BigRecodeAsUtf8 is a copy of this.
517	*/
518	static int rtUtf16RecodeAsUtf8(PCRTUTF16 pwsz, size_t cwc, char psz, size_t cch, size_t pcch)
519	{
520	unsigned char pwch = (unsigned char )psz;
521	int rc = VINF_SUCCESS;
522	while (cwc > 0)
523	{
524	RTUTF16 wc = *pwsz++; cwc--;
525	if (!wc)
526	break;
527	if (wc < 0xd800 \|\| wc > 0xdfff)
528	{
529	if (wc < 0x80)
530	{
531	if (RT_UNLIKELY(cch < 1))
532	{
533	RTStrAssertMsgFailed(("Buffer overflow! 1\n"));
534	rc = VERR_BUFFER_OVERFLOW;
535	break;
536	}
537	cch--;
538	*pwch++ = (unsigned char)wc;
539	}
540	else if (wc < 0x800)
541	{
542	if (RT_UNLIKELY(cch < 2))
543	{
544	RTStrAssertMsgFailed(("Buffer overflow! 2\n"));
545	rc = VERR_BUFFER_OVERFLOW;
546	break;
547	}
548	cch -= 2;
549	*pwch++ = 0xc0 \| (wc >> 6);
550	*pwch++ = 0x80 \| (wc & 0x3f);
551	}
552	else if (wc < 0xfffe)
553	{
554	if (RT_UNLIKELY(cch < 3))
555	{
556	RTStrAssertMsgFailed(("Buffer overflow! 3\n"));
557	rc = VERR_BUFFER_OVERFLOW;
558	break;
559	}
560	cch -= 3;
561	*pwch++ = 0xe0 \| (wc >> 12);
562	*pwch++ = 0x80 \| ((wc >> 6) & 0x3f);
563	*pwch++ = 0x80 \| (wc & 0x3f);
564	}
565	else
566	{
567	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
568	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
569	break;
570	}
571	}
572	else
573	{
574	if (wc >= 0xdc00)
575	{
576	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
577	rc = VERR_INVALID_UTF16_ENCODING;
578	break;
579	}
580	if (cwc <= 0)
581	{
582	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
583	rc = VERR_INVALID_UTF16_ENCODING;
584	break;
585	}
586	RTUTF16 wc2 = *pwsz++; cwc--;
587	if (wc2 < 0xdc00 \|\| wc2 > 0xdfff)
588	{
589	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
590	rc = VERR_INVALID_UTF16_ENCODING;
591	break;
592	}
593	uint32_t CodePoint = 0x10000
594	+ ( ((wc & 0x3ff) << 10)
595	\| (wc2 & 0x3ff));
596	if (RT_UNLIKELY(cch < 4))
597	{
598	RTStrAssertMsgFailed(("Buffer overflow! 4\n"));
599	rc = VERR_BUFFER_OVERFLOW;
600	break;
601	}
602	cch -= 4;
603	*pwch++ = 0xf0 \| (CodePoint >> 18);
604	*pwch++ = 0x80 \| ((CodePoint >> 12) & 0x3f);
605	*pwch++ = 0x80 \| ((CodePoint >> 6) & 0x3f);
606	*pwch++ = 0x80 \| (CodePoint & 0x3f);
607	}
608	}
609
610	/* done */
611	*pwch = '\0';
612	pcch = (char )pwch - psz;
613	return rc;
614	}
615
616
617	/**
618	* Recodes an valid UTF-16BE string as UTF-8.
619	*
620	* @returns iprt status code.
621	* @param pwsz The UTF-16BE string.
622	* @param cwc The number of RTUTF16 characters to process from pwsz. The recoding
623	* will stop when cwc or '\\0' is reached.
624	* @param psz Where to store the UTF-8 string.
625	* @param cch The size of the UTF-8 buffer, excluding the terminator.
626	* @param pcch Where to store the number of octets actually encoded.
627	*
628	* @note Copy of rtUtf16RecodeAsUtf8 with a few RT_BE2H_U16 invocations
629	* insterted.
630	*/
631	static int rtUtf16BigRecodeAsUtf8(PCRTUTF16 pwsz, size_t cwc, char psz, size_t cch, size_t pcch)
632	{
633	unsigned char pwch = (unsigned char )psz;
634	int rc = VINF_SUCCESS;
635	while (cwc > 0)
636	{
637	RTUTF16 wc = *pwsz++; cwc--;
638	if (!wc)
639	break;
640	wc = RT_BE2H_U16(wc);
641	if (wc < 0xd800 \|\| wc > 0xdfff)
642	{
643	if (wc < 0x80)
644	{
645	if (RT_UNLIKELY(cch < 1))
646	{
647	RTStrAssertMsgFailed(("Buffer overflow! 1\n"));
648	rc = VERR_BUFFER_OVERFLOW;
649	break;
650	}
651	cch--;
652	*pwch++ = (unsigned char)wc;
653	}
654	else if (wc < 0x800)
655	{
656	if (RT_UNLIKELY(cch < 2))
657	{
658	RTStrAssertMsgFailed(("Buffer overflow! 2\n"));
659	rc = VERR_BUFFER_OVERFLOW;
660	break;
661	}
662	cch -= 2;
663	*pwch++ = 0xc0 \| (wc >> 6);
664	*pwch++ = 0x80 \| (wc & 0x3f);
665	}
666	else if (wc < 0xfffe)
667	{
668	if (RT_UNLIKELY(cch < 3))
669	{
670	RTStrAssertMsgFailed(("Buffer overflow! 3\n"));
671	rc = VERR_BUFFER_OVERFLOW;
672	break;
673	}
674	cch -= 3;
675	*pwch++ = 0xe0 \| (wc >> 12);
676	*pwch++ = 0x80 \| ((wc >> 6) & 0x3f);
677	*pwch++ = 0x80 \| (wc & 0x3f);
678	}
679	else
680	{
681	RTStrAssertMsgFailed(("endian indicator! wc=%#x\n", wc));
682	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
683	break;
684	}
685	}
686	else
687	{
688	if (wc >= 0xdc00)
689	{
690	RTStrAssertMsgFailed(("Wrong 1st char in surrogate! wc=%#x\n", wc));
691	rc = VERR_INVALID_UTF16_ENCODING;
692	break;
693	}
694	if (cwc <= 0)
695	{
696	RTStrAssertMsgFailed(("Invalid length! wc=%#x\n", wc));
697	rc = VERR_INVALID_UTF16_ENCODING;
698	break;
699	}
700	RTUTF16 wc2 = *pwsz++; cwc--;
701	wc2 = RT_BE2H_U16(wc2);
702	if (wc2 < 0xdc00 \|\| wc2 > 0xdfff)
703	{
704	RTStrAssertMsgFailed(("Wrong 2nd char in surrogate! wc=%#x\n", wc));
705	rc = VERR_INVALID_UTF16_ENCODING;
706	break;
707	}
708	uint32_t CodePoint = 0x10000
709	+ ( ((wc & 0x3ff) << 10)
710	\| (wc2 & 0x3ff));
711	if (RT_UNLIKELY(cch < 4))
712	{
713	RTStrAssertMsgFailed(("Buffer overflow! 4\n"));
714	rc = VERR_BUFFER_OVERFLOW;
715	break;
716	}
717	cch -= 4;
718	*pwch++ = 0xf0 \| (CodePoint >> 18);
719	*pwch++ = 0x80 \| ((CodePoint >> 12) & 0x3f);
720	*pwch++ = 0x80 \| ((CodePoint >> 6) & 0x3f);
721	*pwch++ = 0x80 \| (CodePoint & 0x3f);
722	}
723	}
724
725	/* done */
726	*pwch = '\0';
727	pcch = (char )pwch - psz;
728	return rc;
729	}
730
731
732
733	RTDECL(int) RTUtf16ToUtf8Tag(PCRTUTF16 pwszString, char *ppszString, const char pszTag)
734	{
735	/*
736	* Validate input.
737	*/
738	Assert(VALID_PTR(ppszString));
739	Assert(VALID_PTR(pwszString));
740	*ppszString = NULL;
741
742	/*
743	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
744	*/
745	size_t cch;
746	int rc = rtUtf16CalcUtf8Length(pwszString, RTSTR_MAX, &cch);
747	if (RT_SUCCESS(rc))
748	{
749	/*
750	* Allocate buffer and recode it.
751	*/
752	char pszResult = (char )RTMemAllocTag(cch + 1, pszTag);
753	if (pszResult)
754	{
755	rc = rtUtf16RecodeAsUtf8(pwszString, RTSTR_MAX, pszResult, cch, &cch);
756	if (RT_SUCCESS(rc))
757	{
758	*ppszString = pszResult;
759	return rc;
760	}
761
762	RTMemFree(pszResult);
763	}
764	else
765	rc = VERR_NO_STR_MEMORY;
766	}
767	return rc;
768	}
769	RT_EXPORT_SYMBOL(RTUtf16ToUtf8Tag);
770
771
772	RTDECL(int) RTUtf16BigToUtf8Tag(PCRTUTF16 pwszString, char *ppszString, const char pszTag)
773	{
774	/*
775	* Validate input.
776	*/
777	Assert(VALID_PTR(ppszString));
778	Assert(VALID_PTR(pwszString));
779	*ppszString = NULL;
780
781	/*
782	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
783	*/
784	size_t cch;
785	int rc = rtUtf16BigCalcUtf8Length(pwszString, RTSTR_MAX, &cch);
786	if (RT_SUCCESS(rc))
787	{
788	/*
789	* Allocate buffer and recode it.
790	*/
791	char pszResult = (char )RTMemAllocTag(cch + 1, pszTag);
792	if (pszResult)
793	{
794	rc = rtUtf16BigRecodeAsUtf8(pwszString, RTSTR_MAX, pszResult, cch, &cch);
795	if (RT_SUCCESS(rc))
796	{
797	*ppszString = pszResult;
798	return rc;
799	}
800
801	RTMemFree(pszResult);
802	}
803	else
804	rc = VERR_NO_STR_MEMORY;
805	}
806	return rc;
807	}
808	RT_EXPORT_SYMBOL(RTUtf16BigToUtf8Tag);
809
810
811	RTDECL(int) RTUtf16ToUtf8ExTag(PCRTUTF16 pwszString, size_t cwcString, char *ppsz, size_t cch, size_t pcch, const char *pszTag)
812	{
813	/*
814	* Validate input.
815	*/
816	AssertPtr(pwszString);
817	AssertPtr(ppsz);
818	AssertPtrNull(pcch);
819
820	/*
821	* Validate the UTF-16 string and calculate the length of the UTF-8 encoding of it.
822	*/
823	size_t cchResult;
824	int rc = rtUtf16CalcUtf8Length(pwszString, cwcString, &cchResult);
825	if (RT_SUCCESS(rc))
826	{
827	if (pcch)
828	*pcch = cchResult;
829
830	/*
831	* Check buffer size / Allocate buffer and recode it.
832	*/
833	bool fShouldFree;
834	char *pszResult;
835	if (cch > 0 && *ppsz)
836	{
837	fShouldFree = false;
838	if (RT_UNLIKELY(cch <= cchResult))
839	return VERR_BUFFER_OVERFLOW;
840	pszResult = *ppsz;
841	}
842	else
843	{
844	*ppsz = NULL;
845	fShouldFree = true;
846	cch = RT_MAX(cch, cchResult + 1);
847	pszResult = (char *)RTStrAllocTag(cch, pszTag);
848	}
849	if (pszResult)
850	{
851	rc = rtUtf16RecodeAsUtf8(pwszString, cwcString, pszResult, cch - 1, &cch);
852	if (RT_SUCCESS(rc))
853	{
854	*ppsz = pszResult;
855	return rc;
856	}
857
858	if (fShouldFree)
859	RTStrFree(pszResult);
860	}
861	else
862	rc = VERR_NO_STR_MEMORY;
863	}
864	return rc;
865	}
866	RT_EXPORT_SYMBOL(RTUtf16ToUtf8ExTag);
867
868
869	RTDECL(int) RTUtf16BigToUtf8ExTag(PCRTUTF16 pwszString, size_t cwcString, char *ppsz, size_t cch, size_t pcch, const char *pszTag)
870	{
871	/*
872	* Validate input.
873	*/
874	AssertPtr(pwszString);
875	AssertPtr(ppsz);
876	AssertPtrNull(pcch);
877
878	/*
879	* Validate the UTF-16BE string and calculate the length of the UTF-8 encoding of it.
880	*/
881	size_t cchResult;
882	int rc = rtUtf16BigCalcUtf8Length(pwszString, cwcString, &cchResult);
883	if (RT_SUCCESS(rc))
884	{
885	if (pcch)
886	*pcch = cchResult;
887
888	/*
889	* Check buffer size / Allocate buffer and recode it.
890	*/
891	bool fShouldFree;
892	char *pszResult;
893	if (cch > 0 && *ppsz)
894	{
895	fShouldFree = false;
896	if (RT_UNLIKELY(cch <= cchResult))
897	return VERR_BUFFER_OVERFLOW;
898	pszResult = *ppsz;
899	}
900	else
901	{
902	*ppsz = NULL;
903	fShouldFree = true;
904	cch = RT_MAX(cch, cchResult + 1);
905	pszResult = (char *)RTStrAllocTag(cch, pszTag);
906	}
907	if (pszResult)
908	{
909	rc = rtUtf16BigRecodeAsUtf8(pwszString, cwcString, pszResult, cch - 1, &cch);
910	if (RT_SUCCESS(rc))
911	{
912	*ppsz = pszResult;
913	return rc;
914	}
915
916	if (fShouldFree)
917	RTStrFree(pszResult);
918	}
919	else
920	rc = VERR_NO_STR_MEMORY;
921	}
922	return rc;
923	}
924	RT_EXPORT_SYMBOL(RTUtf16BigToUtf8ExTag);
925
926
927	RTDECL(size_t) RTUtf16CalcUtf8Len(PCRTUTF16 pwsz)
928	{
929	size_t cch;
930	int rc = rtUtf16CalcUtf8Length(pwsz, RTSTR_MAX, &cch);
931	return RT_SUCCESS(rc) ? cch : 0;
932	}
933	RT_EXPORT_SYMBOL(RTUtf16CalcUtf8Len);
934
935
936	RTDECL(int) RTUtf16CalcUtf8LenEx(PCRTUTF16 pwsz, size_t cwc, size_t *pcch)
937	{
938	size_t cch;
939	int rc = rtUtf16CalcUtf8Length(pwsz, cwc, &cch);
940	if (pcch)
941	*pcch = RT_SUCCESS(rc) ? cch : ~(size_t)0;
942	return rc;
943	}
944	RT_EXPORT_SYMBOL(RTUtf16CalcUtf8LenEx);
945
946
947	RTDECL(RTUNICP) RTUtf16GetCpInternal(PCRTUTF16 pwsz)
948	{
949	const RTUTF16 wc = *pwsz;
950
951	/* simple */
952	if (wc < 0xd800 \|\| (wc > 0xdfff && wc < 0xfffe))
953	return wc;
954	if (wc < 0xfffe)
955	{
956	/* surrogate pair */
957	if (wc < 0xdc00)
958	{
959	const RTUTF16 wc2 = pwsz[1];
960	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
961	{
962	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
963	return uc;
964	}
965
966	RTStrAssertMsgFailed(("wc=%#08x wc2=%#08x - invalid 2nd char in surrogate pair\n", wc, wc2));
967	}
968	else
969	RTStrAssertMsgFailed(("wc=%#08x - invalid surrogate pair order\n", wc));
970	}
971	else
972	RTStrAssertMsgFailed(("wc=%#08x - endian indicator\n", wc));
973	return RTUNICP_INVALID;
974	}
975	RT_EXPORT_SYMBOL(RTUtf16GetCpInternal);
976
977
978	RTDECL(int) RTUtf16GetCpExInternal(PCRTUTF16 *ppwsz, PRTUNICP pCp)
979	{
980	const RTUTF16 wc = **ppwsz;
981
982	/* simple */
983	if (wc < 0xd800 \|\| (wc > 0xdfff && wc < 0xfffe))
984	{
985	(*ppwsz)++;
986	*pCp = wc;
987	return VINF_SUCCESS;
988	}
989
990	int rc;
991	if (wc < 0xfffe)
992	{
993	/* surrogate pair */
994	if (wc < 0xdc00)
995	{
996	const RTUTF16 wc2 = (*ppwsz)[1];
997	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
998	{
999	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
1000	*pCp = uc;
1001	(*ppwsz) += 2;
1002	return VINF_SUCCESS;
1003	}
1004
1005	RTStrAssertMsgFailed(("wc=%#08x wc2=%#08x - invalid 2nd char in surrogate pair\n", wc, wc2));
1006	}
1007	else
1008	RTStrAssertMsgFailed(("wc=%#08x - invalid surrogate pair order\n", wc));
1009	rc = VERR_INVALID_UTF16_ENCODING;
1010	}
1011	else
1012	{
1013	RTStrAssertMsgFailed(("wc=%#08x - endian indicator\n", wc));
1014	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
1015	}
1016	*pCp = RTUNICP_INVALID;
1017	(*ppwsz)++;
1018	return rc;
1019	}
1020	RT_EXPORT_SYMBOL(RTUtf16GetCpExInternal);
1021
1022
1023	RTDECL(int) RTUtf16BigGetCpExInternal(PCRTUTF16 *ppwsz, PRTUNICP pCp)
1024	{
1025	const RTUTF16 wc = RT_BE2H_U16(**ppwsz);
1026
1027	/* simple */
1028	if (wc < 0xd800 \|\| (wc > 0xdfff && wc < 0xfffe))
1029	{
1030	(*ppwsz)++;
1031	*pCp = wc;
1032	return VINF_SUCCESS;
1033	}
1034
1035	int rc;
1036	if (wc < 0xfffe)
1037	{
1038	/* surrogate pair */
1039	if (wc < 0xdc00)
1040	{
1041	const RTUTF16 wc2 = RT_BE2H_U16((*ppwsz)[1]);
1042	if (wc2 >= 0xdc00 && wc2 <= 0xdfff)
1043	{
1044	RTUNICP uc = 0x10000 + (((wc & 0x3ff) << 10) \| (wc2 & 0x3ff));
1045	*pCp = uc;
1046	(*ppwsz) += 2;
1047	return VINF_SUCCESS;
1048	}
1049
1050	RTStrAssertMsgFailed(("wc=%#08x wc2=%#08x - invalid 2nd char in surrogate pair\n", wc, wc2));
1051	}
1052	else
1053	RTStrAssertMsgFailed(("wc=%#08x - invalid surrogate pair order\n", wc));
1054	rc = VERR_INVALID_UTF16_ENCODING;
1055	}
1056	else
1057	{
1058	RTStrAssertMsgFailed(("wc=%#08x - endian indicator\n", wc));
1059	rc = VERR_CODE_POINT_ENDIAN_INDICATOR;
1060	}
1061	*pCp = RTUNICP_INVALID;
1062	(*ppwsz)++;
1063	return rc;
1064	}
1065	RT_EXPORT_SYMBOL(RTUtf16GetCpExInternal);
1066
1067
1068	RTDECL(PRTUTF16) RTUtf16PutCpInternal(PRTUTF16 pwsz, RTUNICP CodePoint)
1069	{
1070	/* simple */
1071	if ( CodePoint < 0xd800
1072	\|\| ( CodePoint > 0xdfff
1073	&& CodePoint < 0xfffe))
1074	{
1075	*pwsz++ = (RTUTF16)CodePoint;
1076	return pwsz;
1077	}
1078
1079	/* surrogate pair */
1080	if (CodePoint >= 0x10000 && CodePoint <= 0x0010ffff)
1081	{
1082	CodePoint -= 0x10000;
1083	*pwsz++ = 0xd800 \| (CodePoint >> 10);
1084	*pwsz++ = 0xdc00 \| (CodePoint & 0x3ff);
1085	return pwsz;
1086	}
1087
1088	/* invalid code point. */
1089	RTStrAssertMsgFailed(("Invalid codepoint %#x\n", CodePoint));
1090	*pwsz++ = 0x7f;
1091	return pwsz;
1092	}
1093	RT_EXPORT_SYMBOL(RTUtf16PutCpInternal);
1094

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

source: vbox/trunk/src/VBox/Runtime/common/string/utf-16.cpp@ 67391

Download in other formats: