1 | /* Copyright (c) 2003, 2012, Oracle and/or its affiliates. All rights reserved. |
2 | |
3 | This program is free software; you can redistribute it and/or modify |
4 | it under the terms of the GNU General Public License as published by |
5 | the Free Software Foundation; version 2 of the License. |
6 | |
7 | This program is distributed in the hope that it will be useful, |
8 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
9 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
10 | GNU General Public License for more details. |
11 | |
12 | You should have received a copy of the GNU General Public License |
13 | along with this program; if not, write to the Free Software |
14 | Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA */ |
15 | |
16 | /* Some useful string utility functions used by the MySQL server */ |
17 | |
18 | #include "mariadb.h" |
19 | #include "sql_priv.h" |
20 | #include "unireg.h" |
21 | #include "strfunc.h" |
22 | #include "sql_class.h" |
23 | #include "typelib.h" // TYPELIB |
24 | #include "m_ctype.h" // my_charset_latin1 |
25 | #include "mysqld.h" // system_charset_info |
26 | |
27 | /* |
28 | Return bitmap for strings used in a set |
29 | |
30 | SYNOPSIS |
31 | find_set() |
32 | lib Strings in set |
33 | str Strings of set-strings separated by ',' |
34 | err_pos If error, set to point to start of wrong set string |
35 | err_len If error, set to the length of wrong set string |
36 | set_warning Set to 1 if some string in set couldn't be used |
37 | |
38 | NOTE |
39 | We delete all end space from str before comparison |
40 | |
41 | RETURN |
42 | bitmap of all sets found in x. |
43 | set_warning is set to 1 if there was any sets that couldn't be set |
44 | */ |
45 | |
46 | static const char field_separator=','; |
47 | |
48 | ulonglong find_set(TYPELIB *lib, const char *str, size_t length, CHARSET_INFO *cs, |
49 | char **err_pos, uint *err_len, bool *set_warning) |
50 | { |
51 | CHARSET_INFO *strip= cs ? cs : &my_charset_latin1; |
52 | const char *end= str + strip->cset->lengthsp(strip, str, length); |
53 | ulonglong found= 0; |
54 | *err_pos= 0; // No error yet |
55 | *err_len= 0; |
56 | if (str != end) |
57 | { |
58 | const char *start= str; |
59 | for (;;) |
60 | { |
61 | const char *pos= start; |
62 | uint var_len; |
63 | int mblen= 1; |
64 | |
65 | if (cs && cs->mbminlen > 1) |
66 | { |
67 | for ( ; pos < end; pos+= mblen) |
68 | { |
69 | my_wc_t wc; |
70 | if ((mblen= cs->cset->mb_wc(cs, &wc, (const uchar *) pos, |
71 | (const uchar *) end)) < 1) |
72 | mblen= 1; // Not to hang on a wrong multibyte sequence |
73 | if (wc == (my_wc_t) field_separator) |
74 | break; |
75 | } |
76 | } |
77 | else |
78 | for (; pos != end && *pos != field_separator; pos++) ; |
79 | var_len= (uint) (pos - start); |
80 | uint find= cs ? find_type2(lib, start, var_len, cs) : |
81 | find_type(lib, start, var_len, (bool) 0); |
82 | if (unlikely(!find && *err_len == 0)) |
83 | { |
84 | // report the first error with length > 0 |
85 | *err_pos= (char*) start; |
86 | *err_len= var_len; |
87 | *set_warning= 1; |
88 | } |
89 | else |
90 | found|= 1ULL << (find - 1); |
91 | if (pos >= end) |
92 | break; |
93 | start= pos + mblen; |
94 | } |
95 | } |
96 | return found; |
97 | } |
98 | |
99 | /* |
100 | Function to find a string in a TYPELIB |
101 | (similar to find_type() of mysys/typelib.c) |
102 | |
103 | SYNOPSIS |
104 | find_type() |
105 | lib TYPELIB (struct of pointer to values + count) |
106 | find String to find |
107 | length Length of string to find |
108 | part_match Allow part matching of value |
109 | |
110 | RETURN |
111 | 0 error |
112 | > 0 position in TYPELIB->type_names +1 |
113 | */ |
114 | |
115 | uint find_type(const TYPELIB *lib, const char *find, size_t length, |
116 | bool part_match) |
117 | { |
118 | uint found_count=0, found_pos=0; |
119 | const char *end= find+length; |
120 | const char *i; |
121 | const char *j; |
122 | for (uint pos=0 ; (j=lib->type_names[pos++]) ; ) |
123 | { |
124 | for (i=find ; i != end && |
125 | my_toupper(system_charset_info,*i) == |
126 | my_toupper(system_charset_info,*j) ; i++, j++) ; |
127 | if (i == end) |
128 | { |
129 | if (! *j) |
130 | return(pos); |
131 | found_count++; |
132 | found_pos= pos; |
133 | } |
134 | } |
135 | return(found_count == 1 && part_match ? found_pos : 0); |
136 | } |
137 | |
138 | |
139 | /* |
140 | Find a string in a list of strings according to collation |
141 | |
142 | SYNOPSIS |
143 | find_type2() |
144 | lib TYPELIB (struct of pointer to values + count) |
145 | x String to find |
146 | length String length |
147 | cs Character set + collation to use for comparison |
148 | |
149 | NOTES |
150 | |
151 | RETURN |
152 | 0 No matching value |
153 | >0 Offset+1 in typelib for matched string |
154 | */ |
155 | |
156 | uint find_type2(const TYPELIB *typelib, const char *x, size_t length, |
157 | CHARSET_INFO *cs) |
158 | { |
159 | int pos; |
160 | const char *j; |
161 | DBUG_ENTER("find_type2" ); |
162 | DBUG_PRINT("enter" ,("x: '%.*s' lib: %p" , (int)length, x, typelib)); |
163 | |
164 | if (!typelib->count) |
165 | { |
166 | DBUG_PRINT("exit" ,("no count" )); |
167 | DBUG_RETURN(0); |
168 | } |
169 | |
170 | for (pos=0 ; (j=typelib->type_names[pos]) ; pos++) |
171 | { |
172 | if (!my_strnncoll(cs, (const uchar*) x, length, |
173 | (const uchar*) j, typelib->type_lengths[pos])) |
174 | DBUG_RETURN(pos+1); |
175 | } |
176 | DBUG_PRINT("exit" ,("Couldn't find type" )); |
177 | DBUG_RETURN(0); |
178 | } /* find_type */ |
179 | |
180 | |
181 | /* |
182 | Un-hex all elements in a typelib |
183 | |
184 | SYNOPSIS |
185 | unhex_type2() |
186 | interval TYPELIB (struct of pointer to values + lengths + count) |
187 | |
188 | NOTES |
189 | |
190 | RETURN |
191 | N/A |
192 | */ |
193 | |
194 | void unhex_type2(TYPELIB *interval) |
195 | { |
196 | for (uint pos= 0; pos < interval->count; pos++) |
197 | { |
198 | char *from, *to; |
199 | for (from= to= (char*) interval->type_names[pos]; *from; ) |
200 | { |
201 | /* |
202 | Note, hexchar_to_int(*from++) doesn't work |
203 | one some compilers, e.g. IRIX. Looks like a compiler |
204 | bug in inline functions in combination with arguments |
205 | that have a side effect. So, let's use from[0] and from[1] |
206 | and increment 'from' by two later. |
207 | */ |
208 | |
209 | *to++= (char) (hexchar_to_int(from[0]) << 4) + |
210 | hexchar_to_int(from[1]); |
211 | from+= 2; |
212 | } |
213 | interval->type_lengths[pos] /= 2; |
214 | } |
215 | } |
216 | |
217 | |
218 | /* |
219 | Check if the first word in a string is one of the ones in TYPELIB |
220 | |
221 | SYNOPSIS |
222 | check_word() |
223 | lib TYPELIB |
224 | val String to check |
225 | end End of input |
226 | end_of_word Store value of last used byte here if we found word |
227 | |
228 | RETURN |
229 | 0 No matching value |
230 | > 1 lib->type_names[#-1] matched |
231 | end_of_word will point to separator character/end in 'val' |
232 | */ |
233 | |
234 | uint check_word(TYPELIB *lib, const char *val, const char *end, |
235 | const char **end_of_word) |
236 | { |
237 | int res; |
238 | const char *ptr; |
239 | |
240 | /* Fiend end of word */ |
241 | for (ptr= val ; ptr < end && my_isalpha(&my_charset_latin1, *ptr) ; ptr++) |
242 | ; |
243 | if ((res=find_type(lib, val, (uint) (ptr - val), 1)) > 0) |
244 | *end_of_word= ptr; |
245 | return res; |
246 | } |
247 | |
248 | |
249 | /* |
250 | Converts a string between character sets |
251 | |
252 | SYNOPSIS |
253 | strconvert() |
254 | from_cs source character set |
255 | from source, a null terminated string |
256 | to destination buffer |
257 | to_length destination buffer length |
258 | |
259 | NOTES |
260 | 'to' is always terminated with a '\0' character. |
261 | If there is no enough space to convert whole string, |
262 | only prefix is converted, and terminated with '\0'. |
263 | |
264 | RETURN VALUES |
265 | result string length |
266 | */ |
267 | |
268 | |
269 | uint strconvert(CHARSET_INFO *from_cs, const char *from, size_t from_length, |
270 | CHARSET_INFO *to_cs, char *to, size_t to_length, uint *errors) |
271 | { |
272 | int cnvres; |
273 | my_wc_t wc; |
274 | char *to_start= to; |
275 | uchar *to_end= (uchar*) to + to_length - 1; |
276 | const uchar *from_end= (const uchar*) from + from_length; |
277 | my_charset_conv_mb_wc mb_wc= from_cs->cset->mb_wc; |
278 | my_charset_conv_wc_mb wc_mb= to_cs->cset->wc_mb; |
279 | uint error_count= 0; |
280 | |
281 | while (1) |
282 | { |
283 | if ((cnvres= (*mb_wc)(from_cs, &wc, |
284 | (uchar*) from, from_end)) > 0) |
285 | { |
286 | if (!wc) |
287 | break; |
288 | from+= cnvres; |
289 | } |
290 | else if (cnvres == MY_CS_ILSEQ) |
291 | { |
292 | error_count++; |
293 | from++; |
294 | wc= '?'; |
295 | } |
296 | else |
297 | break; // Impossible char. |
298 | |
299 | outp: |
300 | |
301 | if ((cnvres= (*wc_mb)(to_cs, wc, (uchar*) to, to_end)) > 0) |
302 | to+= cnvres; |
303 | else if (cnvres == MY_CS_ILUNI && wc != '?') |
304 | { |
305 | error_count++; |
306 | wc= '?'; |
307 | goto outp; |
308 | } |
309 | else |
310 | break; |
311 | } |
312 | *to= '\0'; |
313 | *errors= error_count; |
314 | return (uint32) (to - to_start); |
315 | |
316 | } |
317 | |
318 | |
319 | /* |
320 | Searches for a LEX_STRING in an LEX_STRING array. |
321 | |
322 | SYNOPSIS |
323 | find_string_in_array() |
324 | heap The array |
325 | needle The string to search for |
326 | |
327 | NOTE |
328 | The last LEX_STRING in the array should have str member set to NULL |
329 | |
330 | RETURN VALUES |
331 | -1 Not found |
332 | >=0 Ordinal position |
333 | */ |
334 | |
335 | int find_string_in_array(LEX_CSTRING * const haystack, LEX_CSTRING * const needle, |
336 | CHARSET_INFO * const cs) |
337 | { |
338 | const LEX_CSTRING *pos; |
339 | for (pos= haystack; pos->str; pos++) |
340 | if (!cs->coll->strnncollsp(cs, (uchar *) pos->str, pos->length, |
341 | (uchar *) needle->str, needle->length)) |
342 | { |
343 | return (int)(pos - haystack); |
344 | } |
345 | return -1; |
346 | } |
347 | |
348 | |
349 | const char *set_to_string(THD *thd, LEX_CSTRING *result, ulonglong set, |
350 | const char *lib[]) |
351 | { |
352 | char buff[STRING_BUFFER_USUAL_SIZE*8]; |
353 | String tmp(buff, sizeof(buff), &my_charset_latin1); |
354 | LEX_CSTRING unused; |
355 | |
356 | if (!result) |
357 | result= &unused; |
358 | |
359 | tmp.length(0); |
360 | |
361 | for (uint i= 0; set; i++, set >>= 1) |
362 | if (set & 1) { |
363 | tmp.append(lib[i]); |
364 | tmp.append(','); |
365 | } |
366 | |
367 | if (tmp.length()) |
368 | { |
369 | result->str= thd->strmake(tmp.ptr(), tmp.length()-1); |
370 | result->length= tmp.length()-1; |
371 | } |
372 | else |
373 | { |
374 | result->str= const_cast<char*>("" ); |
375 | result->length= 0; |
376 | } |
377 | return result->str; |
378 | } |
379 | |
380 | const char *flagset_to_string(THD *thd, LEX_CSTRING *result, ulonglong set, |
381 | const char *lib[]) |
382 | { |
383 | char buff[STRING_BUFFER_USUAL_SIZE*8]; |
384 | String tmp(buff, sizeof(buff), &my_charset_latin1); |
385 | LEX_CSTRING unused; |
386 | |
387 | if (!result) result= &unused; |
388 | |
389 | tmp.length(0); |
390 | |
391 | // note that the last element is always "default", and it's ignored below |
392 | for (uint i= 0; lib[i+1]; i++, set >>= 1) |
393 | { |
394 | tmp.append(lib[i]); |
395 | tmp.append(set & 1 ? "=on," : "=off," ); |
396 | } |
397 | |
398 | result->str= thd->strmake(tmp.ptr(), tmp.length()-1); |
399 | result->length= tmp.length()-1; |
400 | |
401 | return result->str; |
402 | } |
403 | |
404 | |