Commit | Line | Data |
---|---|---|
3c9f3681 JB |
1 | /* |
2 | * Helpers for formatting and printing strings | |
3 | * | |
4 | * Copyright 31 August 2008 James Bottomley | |
16c7fa05 | 5 | * Copyright (C) 2013, Intel Corporation |
3c9f3681 JB |
6 | */ |
7 | #include <linux/kernel.h> | |
8 | #include <linux/math64.h> | |
8bc3bcc9 | 9 | #include <linux/export.h> |
16c7fa05 | 10 | #include <linux/ctype.h> |
c8250381 AS |
11 | #include <linux/errno.h> |
12 | #include <linux/string.h> | |
3c9f3681 JB |
13 | #include <linux/string_helpers.h> |
14 | ||
15 | /** | |
16 | * string_get_size - get the size in the specified units | |
17 | * @size: The size to be converted | |
18 | * @units: units to use (powers of 1000 or 1024) | |
19 | * @buf: buffer to format to | |
20 | * @len: length of buffer | |
21 | * | |
22 | * This function returns a string formatted to 3 significant figures | |
d1214c65 RV |
23 | * giving the size in the required units. @buf should have room for |
24 | * at least 9 bytes and will always be zero terminated. | |
3c9f3681 JB |
25 | * |
26 | */ | |
d1214c65 RV |
27 | void string_get_size(u64 size, const enum string_size_units units, |
28 | char *buf, int len) | |
3c9f3681 | 29 | { |
142cda5d | 30 | static const char *const units_10[] = { |
7eed8fde | 31 | "B", "kB", "MB", "GB", "TB", "PB", "EB" |
142cda5d MK |
32 | }; |
33 | static const char *const units_2[] = { | |
7eed8fde | 34 | "B", "KiB", "MiB", "GiB", "TiB", "PiB", "EiB" |
142cda5d MK |
35 | }; |
36 | static const char *const *const units_str[] = { | |
37 | [STRING_UNITS_10] = units_10, | |
3c9f3681 JB |
38 | [STRING_UNITS_2] = units_2, |
39 | }; | |
68aecfb9 | 40 | static const unsigned int divisor[] = { |
3c9f3681 JB |
41 | [STRING_UNITS_10] = 1000, |
42 | [STRING_UNITS_2] = 1024, | |
43 | }; | |
44 | int i, j; | |
84b9fbed | 45 | u32 remainder = 0, sf_cap; |
3c9f3681 JB |
46 | char tmp[8]; |
47 | ||
48 | tmp[0] = '\0'; | |
a8659597 PA |
49 | i = 0; |
50 | if (size >= divisor[units]) { | |
7eed8fde | 51 | while (size >= divisor[units]) { |
a8659597 PA |
52 | remainder = do_div(size, divisor[units]); |
53 | i++; | |
54 | } | |
3c9f3681 | 55 | |
a8659597 PA |
56 | sf_cap = size; |
57 | for (j = 0; sf_cap*10 < 1000; j++) | |
58 | sf_cap *= 10; | |
3c9f3681 | 59 | |
a8659597 PA |
60 | if (j) { |
61 | remainder *= 1000; | |
84b9fbed RV |
62 | remainder /= divisor[units]; |
63 | snprintf(tmp, sizeof(tmp), ".%03u", remainder); | |
a8659597 PA |
64 | tmp[j+1] = '\0'; |
65 | } | |
3c9f3681 JB |
66 | } |
67 | ||
84b9fbed | 68 | snprintf(buf, len, "%u%s %s", (u32)size, |
3c9f3681 | 69 | tmp, units_str[units][i]); |
3c9f3681 JB |
70 | } |
71 | EXPORT_SYMBOL(string_get_size); | |
16c7fa05 AS |
72 | |
73 | static bool unescape_space(char **src, char **dst) | |
74 | { | |
75 | char *p = *dst, *q = *src; | |
76 | ||
77 | switch (*q) { | |
78 | case 'n': | |
79 | *p = '\n'; | |
80 | break; | |
81 | case 'r': | |
82 | *p = '\r'; | |
83 | break; | |
84 | case 't': | |
85 | *p = '\t'; | |
86 | break; | |
87 | case 'v': | |
88 | *p = '\v'; | |
89 | break; | |
90 | case 'f': | |
91 | *p = '\f'; | |
92 | break; | |
93 | default: | |
94 | return false; | |
95 | } | |
96 | *dst += 1; | |
97 | *src += 1; | |
98 | return true; | |
99 | } | |
100 | ||
101 | static bool unescape_octal(char **src, char **dst) | |
102 | { | |
103 | char *p = *dst, *q = *src; | |
104 | u8 num; | |
105 | ||
106 | if (isodigit(*q) == 0) | |
107 | return false; | |
108 | ||
109 | num = (*q++) & 7; | |
110 | while (num < 32 && isodigit(*q) && (q - *src < 3)) { | |
111 | num <<= 3; | |
112 | num += (*q++) & 7; | |
113 | } | |
114 | *p = num; | |
115 | *dst += 1; | |
116 | *src = q; | |
117 | return true; | |
118 | } | |
119 | ||
120 | static bool unescape_hex(char **src, char **dst) | |
121 | { | |
122 | char *p = *dst, *q = *src; | |
123 | int digit; | |
124 | u8 num; | |
125 | ||
126 | if (*q++ != 'x') | |
127 | return false; | |
128 | ||
129 | num = digit = hex_to_bin(*q++); | |
130 | if (digit < 0) | |
131 | return false; | |
132 | ||
133 | digit = hex_to_bin(*q); | |
134 | if (digit >= 0) { | |
135 | q++; | |
136 | num = (num << 4) | digit; | |
137 | } | |
138 | *p = num; | |
139 | *dst += 1; | |
140 | *src = q; | |
141 | return true; | |
142 | } | |
143 | ||
144 | static bool unescape_special(char **src, char **dst) | |
145 | { | |
146 | char *p = *dst, *q = *src; | |
147 | ||
148 | switch (*q) { | |
149 | case '\"': | |
150 | *p = '\"'; | |
151 | break; | |
152 | case '\\': | |
153 | *p = '\\'; | |
154 | break; | |
155 | case 'a': | |
156 | *p = '\a'; | |
157 | break; | |
158 | case 'e': | |
159 | *p = '\e'; | |
160 | break; | |
161 | default: | |
162 | return false; | |
163 | } | |
164 | *dst += 1; | |
165 | *src += 1; | |
166 | return true; | |
167 | } | |
168 | ||
d295634e AS |
169 | /** |
170 | * string_unescape - unquote characters in the given string | |
171 | * @src: source buffer (escaped) | |
172 | * @dst: destination buffer (unescaped) | |
173 | * @size: size of the destination buffer (0 to unlimit) | |
174 | * @flags: combination of the flags (bitwise OR): | |
175 | * %UNESCAPE_SPACE: | |
176 | * '\f' - form feed | |
177 | * '\n' - new line | |
178 | * '\r' - carriage return | |
179 | * '\t' - horizontal tab | |
180 | * '\v' - vertical tab | |
181 | * %UNESCAPE_OCTAL: | |
182 | * '\NNN' - byte with octal value NNN (1 to 3 digits) | |
183 | * %UNESCAPE_HEX: | |
184 | * '\xHH' - byte with hexadecimal value HH (1 to 2 digits) | |
185 | * %UNESCAPE_SPECIAL: | |
186 | * '\"' - double quote | |
187 | * '\\' - backslash | |
188 | * '\a' - alert (BEL) | |
189 | * '\e' - escape | |
190 | * %UNESCAPE_ANY: | |
191 | * all previous together | |
192 | * | |
193 | * Description: | |
194 | * The function unquotes characters in the given string. | |
195 | * | |
196 | * Because the size of the output will be the same as or less than the size of | |
197 | * the input, the transformation may be performed in place. | |
198 | * | |
199 | * Caller must provide valid source and destination pointers. Be aware that | |
200 | * destination buffer will always be NULL-terminated. Source string must be | |
201 | * NULL-terminated as well. | |
202 | * | |
203 | * Return: | |
204 | * The amount of the characters processed to the destination buffer excluding | |
205 | * trailing '\0' is returned. | |
206 | */ | |
16c7fa05 AS |
207 | int string_unescape(char *src, char *dst, size_t size, unsigned int flags) |
208 | { | |
209 | char *out = dst; | |
210 | ||
211 | while (*src && --size) { | |
212 | if (src[0] == '\\' && src[1] != '\0' && size > 1) { | |
213 | src++; | |
214 | size--; | |
215 | ||
216 | if (flags & UNESCAPE_SPACE && | |
217 | unescape_space(&src, &out)) | |
218 | continue; | |
219 | ||
220 | if (flags & UNESCAPE_OCTAL && | |
221 | unescape_octal(&src, &out)) | |
222 | continue; | |
223 | ||
224 | if (flags & UNESCAPE_HEX && | |
225 | unescape_hex(&src, &out)) | |
226 | continue; | |
227 | ||
228 | if (flags & UNESCAPE_SPECIAL && | |
229 | unescape_special(&src, &out)) | |
230 | continue; | |
231 | ||
232 | *out++ = '\\'; | |
233 | } | |
234 | *out++ = *src++; | |
235 | } | |
236 | *out = '\0'; | |
237 | ||
238 | return out - dst; | |
239 | } | |
240 | EXPORT_SYMBOL(string_unescape); | |
c8250381 AS |
241 | |
242 | static int escape_passthrough(unsigned char c, char **dst, size_t *osz) | |
243 | { | |
244 | char *out = *dst; | |
245 | ||
246 | if (*osz < 1) | |
247 | return -ENOMEM; | |
248 | ||
249 | *out++ = c; | |
250 | ||
251 | *dst = out; | |
252 | *osz -= 1; | |
253 | ||
254 | return 1; | |
255 | } | |
256 | ||
257 | static int escape_space(unsigned char c, char **dst, size_t *osz) | |
258 | { | |
259 | char *out = *dst; | |
260 | unsigned char to; | |
261 | ||
262 | if (*osz < 2) | |
263 | return -ENOMEM; | |
264 | ||
265 | switch (c) { | |
266 | case '\n': | |
267 | to = 'n'; | |
268 | break; | |
269 | case '\r': | |
270 | to = 'r'; | |
271 | break; | |
272 | case '\t': | |
273 | to = 't'; | |
274 | break; | |
275 | case '\v': | |
276 | to = 'v'; | |
277 | break; | |
278 | case '\f': | |
279 | to = 'f'; | |
280 | break; | |
281 | default: | |
282 | return 0; | |
283 | } | |
284 | ||
285 | *out++ = '\\'; | |
286 | *out++ = to; | |
287 | ||
288 | *dst = out; | |
289 | *osz -= 2; | |
290 | ||
291 | return 1; | |
292 | } | |
293 | ||
294 | static int escape_special(unsigned char c, char **dst, size_t *osz) | |
295 | { | |
296 | char *out = *dst; | |
297 | unsigned char to; | |
298 | ||
299 | if (*osz < 2) | |
300 | return -ENOMEM; | |
301 | ||
302 | switch (c) { | |
303 | case '\\': | |
304 | to = '\\'; | |
305 | break; | |
306 | case '\a': | |
307 | to = 'a'; | |
308 | break; | |
309 | case '\e': | |
310 | to = 'e'; | |
311 | break; | |
312 | default: | |
313 | return 0; | |
314 | } | |
315 | ||
316 | *out++ = '\\'; | |
317 | *out++ = to; | |
318 | ||
319 | *dst = out; | |
320 | *osz -= 2; | |
321 | ||
322 | return 1; | |
323 | } | |
324 | ||
325 | static int escape_null(unsigned char c, char **dst, size_t *osz) | |
326 | { | |
327 | char *out = *dst; | |
328 | ||
329 | if (*osz < 2) | |
330 | return -ENOMEM; | |
331 | ||
332 | if (c) | |
333 | return 0; | |
334 | ||
335 | *out++ = '\\'; | |
336 | *out++ = '0'; | |
337 | ||
338 | *dst = out; | |
339 | *osz -= 2; | |
340 | ||
341 | return 1; | |
342 | } | |
343 | ||
344 | static int escape_octal(unsigned char c, char **dst, size_t *osz) | |
345 | { | |
346 | char *out = *dst; | |
347 | ||
348 | if (*osz < 4) | |
349 | return -ENOMEM; | |
350 | ||
351 | *out++ = '\\'; | |
352 | *out++ = ((c >> 6) & 0x07) + '0'; | |
353 | *out++ = ((c >> 3) & 0x07) + '0'; | |
354 | *out++ = ((c >> 0) & 0x07) + '0'; | |
355 | ||
356 | *dst = out; | |
357 | *osz -= 4; | |
358 | ||
359 | return 1; | |
360 | } | |
361 | ||
362 | static int escape_hex(unsigned char c, char **dst, size_t *osz) | |
363 | { | |
364 | char *out = *dst; | |
365 | ||
366 | if (*osz < 4) | |
367 | return -ENOMEM; | |
368 | ||
369 | *out++ = '\\'; | |
370 | *out++ = 'x'; | |
371 | *out++ = hex_asc_hi(c); | |
372 | *out++ = hex_asc_lo(c); | |
373 | ||
374 | *dst = out; | |
375 | *osz -= 4; | |
376 | ||
377 | return 1; | |
378 | } | |
379 | ||
380 | /** | |
381 | * string_escape_mem - quote characters in the given memory buffer | |
382 | * @src: source buffer (unescaped) | |
383 | * @isz: source buffer size | |
384 | * @dst: destination buffer (escaped) | |
385 | * @osz: destination buffer size | |
386 | * @flags: combination of the flags (bitwise OR): | |
387 | * %ESCAPE_SPACE: | |
388 | * '\f' - form feed | |
389 | * '\n' - new line | |
390 | * '\r' - carriage return | |
391 | * '\t' - horizontal tab | |
392 | * '\v' - vertical tab | |
393 | * %ESCAPE_SPECIAL: | |
394 | * '\\' - backslash | |
395 | * '\a' - alert (BEL) | |
396 | * '\e' - escape | |
397 | * %ESCAPE_NULL: | |
398 | * '\0' - null | |
399 | * %ESCAPE_OCTAL: | |
400 | * '\NNN' - byte with octal value NNN (3 digits) | |
401 | * %ESCAPE_ANY: | |
402 | * all previous together | |
403 | * %ESCAPE_NP: | |
404 | * escape only non-printable characters (checked by isprint) | |
405 | * %ESCAPE_ANY_NP: | |
406 | * all previous together | |
407 | * %ESCAPE_HEX: | |
408 | * '\xHH' - byte with hexadecimal value HH (2 digits) | |
409 | * @esc: NULL-terminated string of characters any of which, if found in | |
410 | * the source, has to be escaped | |
411 | * | |
412 | * Description: | |
413 | * The process of escaping byte buffer includes several parts. They are applied | |
414 | * in the following sequence. | |
415 | * 1. The character is matched to the printable class, if asked, and in | |
416 | * case of match it passes through to the output. | |
417 | * 2. The character is not matched to the one from @esc string and thus | |
418 | * must go as is to the output. | |
419 | * 3. The character is checked if it falls into the class given by @flags. | |
420 | * %ESCAPE_OCTAL and %ESCAPE_HEX are going last since they cover any | |
421 | * character. Note that they actually can't go together, otherwise | |
422 | * %ESCAPE_HEX will be ignored. | |
423 | * | |
424 | * Caller must provide valid source and destination pointers. Be aware that | |
425 | * destination buffer will not be NULL-terminated, thus caller have to append | |
426 | * it if needs. | |
427 | * | |
428 | * Return: | |
429 | * The amount of the characters processed to the destination buffer, or | |
430 | * %-ENOMEM if the size of buffer is not enough to put an escaped character is | |
431 | * returned. | |
432 | * | |
433 | * Even in the case of error @dst pointer will be updated to point to the byte | |
434 | * after the last processed character. | |
435 | */ | |
436 | int string_escape_mem(const char *src, size_t isz, char **dst, size_t osz, | |
437 | unsigned int flags, const char *esc) | |
438 | { | |
439 | char *out = *dst, *p = out; | |
440 | bool is_dict = esc && *esc; | |
441 | int ret = 0; | |
442 | ||
443 | while (isz--) { | |
444 | unsigned char c = *src++; | |
445 | ||
446 | /* | |
447 | * Apply rules in the following sequence: | |
448 | * - the character is printable, when @flags has | |
449 | * %ESCAPE_NP bit set | |
450 | * - the @esc string is supplied and does not contain a | |
451 | * character under question | |
452 | * - the character doesn't fall into a class of symbols | |
453 | * defined by given @flags | |
454 | * In these cases we just pass through a character to the | |
455 | * output buffer. | |
456 | */ | |
457 | if ((flags & ESCAPE_NP && isprint(c)) || | |
458 | (is_dict && !strchr(esc, c))) { | |
459 | /* do nothing */ | |
460 | } else { | |
461 | if (flags & ESCAPE_SPACE) { | |
462 | ret = escape_space(c, &p, &osz); | |
463 | if (ret < 0) | |
464 | break; | |
465 | if (ret > 0) | |
466 | continue; | |
467 | } | |
468 | ||
469 | if (flags & ESCAPE_SPECIAL) { | |
470 | ret = escape_special(c, &p, &osz); | |
471 | if (ret < 0) | |
472 | break; | |
473 | if (ret > 0) | |
474 | continue; | |
475 | } | |
476 | ||
477 | if (flags & ESCAPE_NULL) { | |
478 | ret = escape_null(c, &p, &osz); | |
479 | if (ret < 0) | |
480 | break; | |
481 | if (ret > 0) | |
482 | continue; | |
483 | } | |
484 | ||
485 | /* ESCAPE_OCTAL and ESCAPE_HEX always go last */ | |
486 | if (flags & ESCAPE_OCTAL) { | |
487 | ret = escape_octal(c, &p, &osz); | |
488 | if (ret < 0) | |
489 | break; | |
490 | continue; | |
491 | } | |
492 | if (flags & ESCAPE_HEX) { | |
493 | ret = escape_hex(c, &p, &osz); | |
494 | if (ret < 0) | |
495 | break; | |
496 | continue; | |
497 | } | |
498 | } | |
499 | ||
500 | ret = escape_passthrough(c, &p, &osz); | |
501 | if (ret < 0) | |
502 | break; | |
503 | } | |
504 | ||
505 | *dst = p; | |
506 | ||
507 | if (ret < 0) | |
508 | return ret; | |
509 | ||
510 | return p - out; | |
511 | } | |
512 | EXPORT_SYMBOL(string_escape_mem); |