1 | /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
|
---|
2 | /* ***** BEGIN LICENSE BLOCK *****
|
---|
3 | * Version: MPL 1.1/GPL 2.0/LGPL 2.1
|
---|
4 | *
|
---|
5 | * The contents of this file are subject to the Mozilla Public License Version
|
---|
6 | * 1.1 (the "License"); you may not use this file except in compliance with
|
---|
7 | * the License. You may obtain a copy of the License at
|
---|
8 | * http://www.mozilla.org/MPL/
|
---|
9 | *
|
---|
10 | * Software distributed under the License is distributed on an "AS IS" basis,
|
---|
11 | * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
|
---|
12 | * for the specific language governing rights and limitations under the
|
---|
13 | * License.
|
---|
14 | *
|
---|
15 | * The Original Code is mozilla.org code.
|
---|
16 | *
|
---|
17 | * The Initial Developer of the Original Code is
|
---|
18 | * Netscape Communications Corporation.
|
---|
19 | * Portions created by the Initial Developer are Copyright (C) 2000
|
---|
20 | * the Initial Developer. All Rights Reserved.
|
---|
21 | *
|
---|
22 | * Contributor(s):
|
---|
23 | * Scott Collins <[email protected]> (original author)
|
---|
24 | * Johnny Stenbeck <[email protected]>
|
---|
25 | *
|
---|
26 | * Alternatively, the contents of this file may be used under the terms of
|
---|
27 | * either of the GNU General Public License Version 2 or later (the "GPL"),
|
---|
28 | * or the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
|
---|
29 | * in which case the provisions of the GPL or the LGPL are applicable instead
|
---|
30 | * of those above. If you wish to allow use of your version of this file only
|
---|
31 | * under the terms of either the GPL or the LGPL, and not to allow others to
|
---|
32 | * use your version of this file under the terms of the MPL, indicate your
|
---|
33 | * decision by deleting the provisions above and replace them with the notice
|
---|
34 | * and other provisions required by the GPL or the LGPL. If you do not delete
|
---|
35 | * the provisions above, a recipient may use your version of this file under
|
---|
36 | * the terms of any one of the MPL, the GPL or the LGPL.
|
---|
37 | *
|
---|
38 | * ***** END LICENSE BLOCK ***** */
|
---|
39 |
|
---|
40 | #ifndef nsReadableUtils_h___
|
---|
41 | #define nsReadableUtils_h___
|
---|
42 |
|
---|
43 | /**
|
---|
44 | * I guess all the routines in this file are all mis-named.
|
---|
45 | * According to our conventions, they should be |NS_xxx|.
|
---|
46 | */
|
---|
47 |
|
---|
48 | #ifndef nsAString_h___
|
---|
49 | #include "nsAString.h"
|
---|
50 | #endif
|
---|
51 |
|
---|
52 | inline size_t Distance( const nsReadingIterator<PRUnichar>& start, const nsReadingIterator<PRUnichar>& end )
|
---|
53 | {
|
---|
54 | return end.get() - start.get();
|
---|
55 | }
|
---|
56 | inline size_t Distance( const nsReadingIterator<char>& start, const nsReadingIterator<char>& end )
|
---|
57 | {
|
---|
58 | return end.get() - start.get();
|
---|
59 | }
|
---|
60 |
|
---|
61 | NS_COM void LossyCopyUTF16toASCII( const nsAString& aSource, nsACString& aDest );
|
---|
62 | NS_COM void CopyASCIItoUTF16( const nsACString& aSource, nsAString& aDest );
|
---|
63 |
|
---|
64 | NS_COM void LossyCopyUTF16toASCII( const PRUnichar* aSource, nsACString& aDest );
|
---|
65 | NS_COM void CopyASCIItoUTF16( const char* aSource, nsAString& aDest );
|
---|
66 |
|
---|
67 | NS_COM void CopyUTF16toUTF8( const nsAString& aSource, nsACString& aDest );
|
---|
68 | NS_COM void CopyUTF8toUTF16( const nsACString& aSource, nsAString& aDest );
|
---|
69 |
|
---|
70 | NS_COM void CopyUTF16toUTF8( const PRUnichar* aSource, nsACString& aDest );
|
---|
71 | NS_COM void CopyUTF8toUTF16( const char* aSource, nsAString& aDest );
|
---|
72 |
|
---|
73 | NS_COM void LossyAppendUTF16toASCII( const nsAString& aSource, nsACString& aDest );
|
---|
74 | NS_COM void AppendASCIItoUTF16( const nsACString& aSource, nsAString& aDest );
|
---|
75 |
|
---|
76 | NS_COM void LossyAppendUTF16toASCII( const PRUnichar* aSource, nsACString& aDest );
|
---|
77 | NS_COM void AppendASCIItoUTF16( const char* aSource, nsAString& aDest );
|
---|
78 |
|
---|
79 | NS_COM void AppendUTF16toUTF8( const nsAString& aSource, nsACString& aDest );
|
---|
80 | NS_COM void AppendUTF8toUTF16( const nsACString& aSource, nsAString& aDest );
|
---|
81 |
|
---|
82 | NS_COM void AppendUTF16toUTF8( const PRUnichar* aSource, nsACString& aDest );
|
---|
83 | NS_COM void AppendUTF8toUTF16( const char* aSource, nsAString& aDest );
|
---|
84 |
|
---|
85 | // Backward compatibility
|
---|
86 | inline
|
---|
87 | NS_COM void CopyUCS2toASCII( const nsAString& aSource, nsACString& aDest )
|
---|
88 | { LossyCopyUTF16toASCII(aSource, aDest); }
|
---|
89 | inline
|
---|
90 | NS_COM void CopyASCIItoUCS2( const nsACString& aSource, nsAString& aDest )
|
---|
91 | { CopyASCIItoUTF16(aSource, aDest); }
|
---|
92 |
|
---|
93 | /**
|
---|
94 | * Returns a new |char| buffer containing a zero-terminated copy of |aSource|.
|
---|
95 | *
|
---|
96 | * Allocates and returns a new |char| buffer which you must free with |nsMemory::Free|.
|
---|
97 | * Performs a lossy encoding conversion by chopping 16-bit wide characters down to 8-bits wide while copying |aSource| to your new buffer.
|
---|
98 | * This conversion is not well defined; but it reproduces legacy string behavior.
|
---|
99 | * The new buffer is zero-terminated, but that may not help you if |aSource| contains embedded nulls.
|
---|
100 | *
|
---|
101 | * @param aSource a 16-bit wide string
|
---|
102 | * @return a new |char| buffer you must free with |nsMemory::Free|.
|
---|
103 | */
|
---|
104 | NS_COM char* ToNewCString( const nsAString& aSource );
|
---|
105 |
|
---|
106 |
|
---|
107 | /**
|
---|
108 | * Returns a new |char| buffer containing a zero-terminated copy of |aSource|.
|
---|
109 | *
|
---|
110 | * Allocates and returns a new |char| buffer which you must free with |nsMemory::Free|.
|
---|
111 | * The new buffer is zero-terminated, but that may not help you if |aSource| contains embedded nulls.
|
---|
112 | *
|
---|
113 | * @param aSource an 8-bit wide string
|
---|
114 | * @return a new |char| buffer you must free with |nsMemory::Free|.
|
---|
115 | */
|
---|
116 | NS_COM char* ToNewCString( const nsACString& aSource );
|
---|
117 |
|
---|
118 | /**
|
---|
119 | * Returns a new |char| buffer containing a zero-terminated copy of |aSource|.
|
---|
120 | *
|
---|
121 | * Allocates and returns a new |char| buffer which you must free with
|
---|
122 | * |nsMemory::Free|.
|
---|
123 | * Performs an encoding conversion from a UTF-16 string to a UTF-8 string
|
---|
124 | * copying |aSource| to your new buffer.
|
---|
125 | * The new buffer is zero-terminated, but that may not help you if |aSource|
|
---|
126 | * contains embedded nulls.
|
---|
127 | *
|
---|
128 | * @param aSource a UTF-16 string (made of PRUnichar's)
|
---|
129 | * @param aUTF8Count the number of 8-bit units that was returned
|
---|
130 | * @return a new |char| buffer you must free with |nsMemory::Free|.
|
---|
131 | */
|
---|
132 |
|
---|
133 | NS_COM char* ToNewUTF8String( const nsAString& aSource, PRUint32 *aUTF8Count = nsnull );
|
---|
134 |
|
---|
135 |
|
---|
136 | /**
|
---|
137 | * Returns a new |PRUnichar| buffer containing a zero-terminated copy of
|
---|
138 | * |aSource|.
|
---|
139 | *
|
---|
140 | * Allocates and returns a new |PRUnichar| buffer which you must free with
|
---|
141 | * |nsMemory::Free|.
|
---|
142 | * The new buffer is zero-terminated, but that may not help you if |aSource|
|
---|
143 | * contains embedded nulls.
|
---|
144 | *
|
---|
145 | * @param aSource a UTF-16 string
|
---|
146 | * @return a new |PRUnichar| buffer you must free with |nsMemory::Free|.
|
---|
147 | */
|
---|
148 | NS_COM PRUnichar* ToNewUnicode( const nsAString& aSource );
|
---|
149 |
|
---|
150 |
|
---|
151 | /**
|
---|
152 | * Returns a new |PRUnichar| buffer containing a zero-terminated copy of |aSource|.
|
---|
153 | *
|
---|
154 | * Allocates and returns a new |PRUnichar| buffer which you must free with |nsMemory::Free|.
|
---|
155 | * Performs an encoding conversion by 0-padding 8-bit wide characters up to 16-bits wide while copying |aSource| to your new buffer.
|
---|
156 | * This conversion is not well defined; but it reproduces legacy string behavior.
|
---|
157 | * The new buffer is zero-terminated, but that may not help you if |aSource| contains embedded nulls.
|
---|
158 | *
|
---|
159 | * @param aSource an 8-bit wide string (a C-string, NOT UTF-8)
|
---|
160 | * @return a new |PRUnichar| buffer you must free with |nsMemory::Free|.
|
---|
161 | */
|
---|
162 | NS_COM PRUnichar* ToNewUnicode( const nsACString& aSource );
|
---|
163 |
|
---|
164 | /**
|
---|
165 | * Returns a new |PRUnichar| buffer containing a zero-terminated copy
|
---|
166 | * of |aSource|.
|
---|
167 | *
|
---|
168 | * Allocates and returns a new |char| buffer which you must free with
|
---|
169 | * |nsMemory::Free|. Performs an encoding conversion from UTF-8 to UTF-16
|
---|
170 | * while copying |aSource| to your new buffer. This conversion is well defined
|
---|
171 | * for a valid UTF-8 string. The new buffer is zero-terminated, but that
|
---|
172 | * may not help you if |aSource| contains embedded nulls.
|
---|
173 | *
|
---|
174 | * @param aSource an 8-bit wide string, UTF-8 encoded
|
---|
175 | * @param aUTF16Count the number of 16-bit units that was returned
|
---|
176 | * @return a new |PRUnichar| buffer you must free with |nsMemory::Free|.
|
---|
177 | * (UTF-16 encoded)
|
---|
178 | */
|
---|
179 | NS_COM PRUnichar* UTF8ToNewUnicode( const nsACString& aSource, PRUint32 *aUTF16Count = nsnull );
|
---|
180 |
|
---|
181 | /**
|
---|
182 | * Copies |aLength| 16-bit code units from the start of |aSource| to the
|
---|
183 | * |PRUnichar| buffer |aDest|.
|
---|
184 | *
|
---|
185 | * After this operation |aDest| is not null terminated.
|
---|
186 | *
|
---|
187 | * @param aSource a UTF-16 string
|
---|
188 | * @param aSrcOffset start offset in the source string
|
---|
189 | * @param aDest a |PRUnichar| buffer
|
---|
190 | * @param aLength the number of 16-bit code units to copy
|
---|
191 | * @return pointer to destination buffer - identical to |aDest|
|
---|
192 | */
|
---|
193 | NS_COM PRUnichar* CopyUnicodeTo( const nsAString& aSource,
|
---|
194 | PRUint32 aSrcOffset,
|
---|
195 | PRUnichar* aDest,
|
---|
196 | PRUint32 aLength );
|
---|
197 |
|
---|
198 |
|
---|
199 | /**
|
---|
200 | * Copies 16-bit characters between iterators |aSrcStart| and
|
---|
201 | * |aSrcEnd| to the writable string |aDest|. Similar to the
|
---|
202 | * |nsString::Mid| method.
|
---|
203 | *
|
---|
204 | * After this operation |aDest| is not null terminated.
|
---|
205 | *
|
---|
206 | * @param aSrcStart start source iterator
|
---|
207 | * @param aSrcEnd end source iterator
|
---|
208 | * @param aDest destination for the copy
|
---|
209 | */
|
---|
210 | NS_COM void CopyUnicodeTo( const nsAString::const_iterator& aSrcStart,
|
---|
211 | const nsAString::const_iterator& aSrcEnd,
|
---|
212 | nsAString& aDest );
|
---|
213 |
|
---|
214 | /**
|
---|
215 | * Appends 16-bit characters between iterators |aSrcStart| and
|
---|
216 | * |aSrcEnd| to the writable string |aDest|.
|
---|
217 | *
|
---|
218 | * After this operation |aDest| is not null terminated.
|
---|
219 | *
|
---|
220 | * @param aSrcStart start source iterator
|
---|
221 | * @param aSrcEnd end source iterator
|
---|
222 | * @param aDest destination for the copy
|
---|
223 | */
|
---|
224 | NS_COM void AppendUnicodeTo( const nsAString::const_iterator& aSrcStart,
|
---|
225 | const nsAString::const_iterator& aSrcEnd,
|
---|
226 | nsAString& aDest );
|
---|
227 |
|
---|
228 | /**
|
---|
229 | * Returns |PR_TRUE| if |aString| contains only ASCII characters, that is, characters in the range (0x00, 0x7F).
|
---|
230 | *
|
---|
231 | * @param aString a 16-bit wide string to scan
|
---|
232 | */
|
---|
233 | NS_COM PRBool IsASCII( const nsAString& aString );
|
---|
234 |
|
---|
235 | /**
|
---|
236 | * Returns |PR_TRUE| if |aString| contains only ASCII characters, that is, characters in the range (0x00, 0x7F).
|
---|
237 | *
|
---|
238 | * @param aString a 8-bit wide string to scan
|
---|
239 | */
|
---|
240 | NS_COM PRBool IsASCII( const nsACString& aString );
|
---|
241 |
|
---|
242 |
|
---|
243 | /**
|
---|
244 | * Returns |PR_TRUE| if |aString| is a valid UTF-8 string.
|
---|
245 | * XXX This is not bullet-proof and nor an all-purpose UTF-8 validator.
|
---|
246 | * It is mainly written to replace and roughly equivalent to
|
---|
247 | *
|
---|
248 | * str.Equals(NS_ConvertUTF16toUTF8(NS_ConvertUTF8toUTF16(str)))
|
---|
249 | *
|
---|
250 | * (see bug 191541)
|
---|
251 | * As such, it does not check for non-UTF-8 7bit encodings such as
|
---|
252 | * ISO-2022-JP and HZ. However, it filters out UTF-8 representation
|
---|
253 | * of surrogate codepoints and non-characters ( 0xFFFE and 0xFFFF
|
---|
254 | * in planes 0 through 16.) as well as overlong UTF-8 sequences.
|
---|
255 | * Also note that it regards UTF-8 sequences corresponding to
|
---|
256 | * codepoints above 0x10FFFF as invalid in accordance with
|
---|
257 | * http://www.ietf.org/internet-drafts/draft-yergeau-rfc2279bis-04.txt
|
---|
258 | *
|
---|
259 | * @param aString an 8-bit wide string to scan
|
---|
260 | */
|
---|
261 | NS_COM PRBool IsUTF8( const nsACString& aString );
|
---|
262 |
|
---|
263 |
|
---|
264 | /**
|
---|
265 | * Converts case in place in the argument string.
|
---|
266 | */
|
---|
267 | NS_COM void ToUpperCase( nsACString& );
|
---|
268 |
|
---|
269 | NS_COM void ToLowerCase( nsACString& );
|
---|
270 |
|
---|
271 | NS_COM void ToUpperCase( nsCSubstring& );
|
---|
272 |
|
---|
273 | NS_COM void ToLowerCase( nsCSubstring& );
|
---|
274 |
|
---|
275 | /**
|
---|
276 | * Converts case from string aSource to aDest.
|
---|
277 | */
|
---|
278 | NS_COM void ToUpperCase( const nsACString& aSource, nsACString& aDest );
|
---|
279 |
|
---|
280 | NS_COM void ToLowerCase( const nsACString& aSource, nsACString& aDest );
|
---|
281 |
|
---|
282 | /**
|
---|
283 | * Finds the leftmost occurance of |aPattern|, if any in the range |aSearchStart|..|aSearchEnd|.
|
---|
284 | *
|
---|
285 | * Returns |PR_TRUE| if a match was found, and adjusts |aSearchStart| and |aSearchEnd| to
|
---|
286 | * point to the match. If no match was found, returns |PR_FALSE| and makes |aSearchStart == aSearchEnd|.
|
---|
287 | *
|
---|
288 | * Currently, this is equivalent to the O(m*n) implementation previously on |ns[C]String|.
|
---|
289 | * If we need something faster, then we can implement that later.
|
---|
290 | */
|
---|
291 |
|
---|
292 | NS_COM PRBool FindInReadable( const nsAString& aPattern, nsAString::const_iterator&, nsAString::const_iterator&, const nsStringComparator& = nsDefaultStringComparator() );
|
---|
293 | NS_COM PRBool FindInReadable( const nsACString& aPattern, nsACString::const_iterator&, nsACString::const_iterator&, const nsCStringComparator& = nsDefaultCStringComparator() );
|
---|
294 |
|
---|
295 | /* sometimes we don't care about where the string was, just that we
|
---|
296 | * found it or not */
|
---|
297 | inline PRBool FindInReadable( const nsAString& aPattern, const nsAString& aSource, const nsStringComparator& compare = nsDefaultStringComparator() )
|
---|
298 | {
|
---|
299 | nsAString::const_iterator start, end;
|
---|
300 | aSource.BeginReading(start);
|
---|
301 | aSource.EndReading(end);
|
---|
302 | return FindInReadable(aPattern, start, end, compare);
|
---|
303 | }
|
---|
304 |
|
---|
305 | inline PRBool FindInReadable( const nsACString& aPattern, const nsACString& aSource, const nsCStringComparator& compare = nsDefaultCStringComparator() )
|
---|
306 | {
|
---|
307 | nsACString::const_iterator start, end;
|
---|
308 | aSource.BeginReading(start);
|
---|
309 | aSource.EndReading(end);
|
---|
310 | return FindInReadable(aPattern, start, end, compare);
|
---|
311 | }
|
---|
312 |
|
---|
313 |
|
---|
314 | NS_COM PRBool CaseInsensitiveFindInReadable( const nsACString& aPattern, nsACString::const_iterator&, nsACString::const_iterator& );
|
---|
315 |
|
---|
316 | /**
|
---|
317 | * Finds the rightmost occurance of |aPattern|
|
---|
318 | * Returns |PR_TRUE| if a match was found, and adjusts |aSearchStart| and |aSearchEnd| to
|
---|
319 | * point to the match. If no match was found, returns |PR_FALSE| and makes |aSearchStart == aSearchEnd|.
|
---|
320 | *
|
---|
321 | * Currently, this is equivalent to the O(m*n) implementation previously on |ns[C]String|.
|
---|
322 | * If we need something faster, then we can implement that later.
|
---|
323 | */
|
---|
324 | NS_COM PRBool RFindInReadable( const nsAString& aPattern, nsAString::const_iterator&, nsAString::const_iterator&, const nsStringComparator& = nsDefaultStringComparator() );
|
---|
325 | NS_COM PRBool RFindInReadable( const nsACString& aPattern, nsACString::const_iterator&, nsACString::const_iterator&, const nsCStringComparator& = nsDefaultCStringComparator() );
|
---|
326 |
|
---|
327 | /**
|
---|
328 | * Finds the leftmost occurance of |aChar|, if any in the range
|
---|
329 | * |aSearchStart|..|aSearchEnd|.
|
---|
330 | *
|
---|
331 | * Returns |PR_TRUE| if a match was found, and adjusts |aSearchStart| to
|
---|
332 | * point to the match. If no match was found, returns |PR_FALSE| and
|
---|
333 | * makes |aSearchStart == aSearchEnd|.
|
---|
334 | */
|
---|
335 | NS_COM PRBool FindCharInReadable( PRUnichar aChar, nsAString::const_iterator& aSearchStart, const nsAString::const_iterator& aSearchEnd );
|
---|
336 | NS_COM PRBool FindCharInReadable( char aChar, nsACString::const_iterator& aSearchStart, const nsACString::const_iterator& aSearchEnd );
|
---|
337 |
|
---|
338 | /**
|
---|
339 | * Finds the number of occurences of |aChar| in the string |aStr|
|
---|
340 | */
|
---|
341 | NS_COM PRUint32 CountCharInReadable( const nsAString& aStr,
|
---|
342 | PRUnichar aChar );
|
---|
343 | NS_COM PRUint32 CountCharInReadable( const nsACString& aStr,
|
---|
344 | char aChar );
|
---|
345 |
|
---|
346 | NS_COM PRBool
|
---|
347 | StringBeginsWith( const nsAString& aSource, const nsAString& aSubstring,
|
---|
348 | const nsStringComparator& aComparator =
|
---|
349 | nsDefaultStringComparator() );
|
---|
350 | NS_COM PRBool
|
---|
351 | StringBeginsWith( const nsACString& aSource, const nsACString& aSubstring,
|
---|
352 | const nsCStringComparator& aComparator =
|
---|
353 | nsDefaultCStringComparator() );
|
---|
354 | NS_COM PRBool
|
---|
355 | StringEndsWith( const nsAString& aSource, const nsAString& aSubstring,
|
---|
356 | const nsStringComparator& aComparator =
|
---|
357 | nsDefaultStringComparator() );
|
---|
358 | NS_COM PRBool
|
---|
359 | StringEndsWith( const nsACString& aSource, const nsACString& aSubstring,
|
---|
360 | const nsCStringComparator& aComparator =
|
---|
361 | nsDefaultCStringComparator() );
|
---|
362 |
|
---|
363 | NS_COM PRUint32 HashString( const nsAString& aStr );
|
---|
364 | NS_COM PRUint32 HashString( const nsACString& aStr );
|
---|
365 |
|
---|
366 | NS_COM const nsAFlatString& EmptyString();
|
---|
367 | NS_COM const nsAFlatCString& EmptyCString();
|
---|
368 |
|
---|
369 |
|
---|
370 | #endif // !defined(nsReadableUtils_h___)
|
---|