escape.c 13.6 KB
Newer Older
1
/* SPDX-License-Identifier: LGPL-2.1+ */
2

3 4 5 6
#include <errno.h>
#include <stdlib.h>
#include <string.h>

7 8 9
#include "alloc-util.h"
#include "escape.h"
#include "hexdecoct.h"
10
#include "macro.h"
11 12
#include "utf8.h"

13 14 15 16
int cescape_char(char c, char *buf) {
        char *buf_old = buf;

        /* Needs space for 4 characters in the buffer */
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77

        switch (c) {

                case '\a':
                        *(buf++) = '\\';
                        *(buf++) = 'a';
                        break;
                case '\b':
                        *(buf++) = '\\';
                        *(buf++) = 'b';
                        break;
                case '\f':
                        *(buf++) = '\\';
                        *(buf++) = 'f';
                        break;
                case '\n':
                        *(buf++) = '\\';
                        *(buf++) = 'n';
                        break;
                case '\r':
                        *(buf++) = '\\';
                        *(buf++) = 'r';
                        break;
                case '\t':
                        *(buf++) = '\\';
                        *(buf++) = 't';
                        break;
                case '\v':
                        *(buf++) = '\\';
                        *(buf++) = 'v';
                        break;
                case '\\':
                        *(buf++) = '\\';
                        *(buf++) = '\\';
                        break;
                case '"':
                        *(buf++) = '\\';
                        *(buf++) = '"';
                        break;
                case '\'':
                        *(buf++) = '\\';
                        *(buf++) = '\'';
                        break;

                default:
                        /* For special chars we prefer octal over
                         * hexadecimal encoding, simply because glib's
                         * g_strescape() does the same */
                        if ((c < ' ') || (c >= 127)) {
                                *(buf++) = '\\';
                                *(buf++) = octchar((unsigned char) c >> 6);
                                *(buf++) = octchar((unsigned char) c >> 3);
                                *(buf++) = octchar((unsigned char) c);
                        } else
                                *(buf++) = c;
                        break;
        }

        return buf - buf_old;
}

78
char *cescape_length(const char *s, size_t n) {
79
        const char *f;
80
        char *r, *t;
81

82
        assert(s || n == 0);
83 84 85 86

        /* Does C style string escaping. May be reversed with
         * cunescape(). */

87
        r = new(char, n*4 + 1);
88 89 90
        if (!r)
                return NULL;

91
        for (f = s, t = r; f < s + n; f++)
92 93 94 95 96 97 98
                t += cescape_char(*f, t);

        *t = 0;

        return r;
}

99 100 101 102 103 104
char *cescape(const char *s) {
        assert(s);

        return cescape_length(s, strlen(s));
}

105
int cunescape_one(const char *p, size_t length, char32_t *ret, bool *eight_bit) {
106 107 108 109 110 111
        int r = 1;

        assert(p);
        assert(*p);
        assert(ret);

112 113 114 115 116
        /* Unescapes C style. Returns the unescaped character in ret.
         * Sets *eight_bit to true if the escaped sequence either fits in
         * one byte in UTF-8 or is a non-unicode literal byte and should
         * instead be copied directly.
         */
117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177

        if (length != (size_t) -1 && length < 1)
                return -EINVAL;

        switch (p[0]) {

        case 'a':
                *ret = '\a';
                break;
        case 'b':
                *ret = '\b';
                break;
        case 'f':
                *ret = '\f';
                break;
        case 'n':
                *ret = '\n';
                break;
        case 'r':
                *ret = '\r';
                break;
        case 't':
                *ret = '\t';
                break;
        case 'v':
                *ret = '\v';
                break;
        case '\\':
                *ret = '\\';
                break;
        case '"':
                *ret = '"';
                break;
        case '\'':
                *ret = '\'';
                break;

        case 's':
                /* This is an extension of the XDG syntax files */
                *ret = ' ';
                break;

        case 'x': {
                /* hexadecimal encoding */
                int a, b;

                if (length != (size_t) -1 && length < 3)
                        return -EINVAL;

                a = unhexchar(p[1]);
                if (a < 0)
                        return -EINVAL;

                b = unhexchar(p[2]);
                if (b < 0)
                        return -EINVAL;

                /* Don't allow NUL bytes */
                if (a == 0 && b == 0)
                        return -EINVAL;

178 179
                *ret = (a << 4U) | b;
                *eight_bit = true;
180 181 182 183 184 185 186 187
                r = 3;
                break;
        }

        case 'u': {
                /* C++11 style 16bit unicode */

                int a[4];
188
                size_t i;
189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205
                uint32_t c;

                if (length != (size_t) -1 && length < 5)
                        return -EINVAL;

                for (i = 0; i < 4; i++) {
                        a[i] = unhexchar(p[1 + i]);
                        if (a[i] < 0)
                                return a[i];
                }

                c = ((uint32_t) a[0] << 12U) | ((uint32_t) a[1] << 8U) | ((uint32_t) a[2] << 4U) | (uint32_t) a[3];

                /* Don't allow 0 chars */
                if (c == 0)
                        return -EINVAL;

206
                *ret = c;
207 208 209 210 211 212 213 214
                r = 5;
                break;
        }

        case 'U': {
                /* C++11 style 32bit unicode */

                int a[8];
215
                size_t i;
216
                char32_t c;
217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237

                if (length != (size_t) -1 && length < 9)
                        return -EINVAL;

                for (i = 0; i < 8; i++) {
                        a[i] = unhexchar(p[1 + i]);
                        if (a[i] < 0)
                                return a[i];
                }

                c = ((uint32_t) a[0] << 28U) | ((uint32_t) a[1] << 24U) | ((uint32_t) a[2] << 20U) | ((uint32_t) a[3] << 16U) |
                    ((uint32_t) a[4] << 12U) | ((uint32_t) a[5] <<  8U) | ((uint32_t) a[6] <<  4U) |  (uint32_t) a[7];

                /* Don't allow 0 chars */
                if (c == 0)
                        return -EINVAL;

                /* Don't allow invalid code points */
                if (!unichar_is_valid(c))
                        return -EINVAL;

238
                *ret = c;
239 240 241 242 243 244 245 246 247 248 249 250 251 252
                r = 9;
                break;
        }

        case '0':
        case '1':
        case '2':
        case '3':
        case '4':
        case '5':
        case '6':
        case '7': {
                /* octal encoding */
                int a, b, c;
253
                char32_t m;
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279

                if (length != (size_t) -1 && length < 3)
                        return -EINVAL;

                a = unoctchar(p[0]);
                if (a < 0)
                        return -EINVAL;

                b = unoctchar(p[1]);
                if (b < 0)
                        return -EINVAL;

                c = unoctchar(p[2]);
                if (c < 0)
                        return -EINVAL;

                /* don't allow NUL bytes */
                if (a == 0 && b == 0 && c == 0)
                        return -EINVAL;

                /* Don't allow bytes above 255 */
                m = ((uint32_t) a << 6U) | ((uint32_t) b << 3U) | (uint32_t) c;
                if (m > 255)
                        return -EINVAL;

                *ret = m;
280
                *eight_bit = true;
281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301
                r = 3;
                break;
        }

        default:
                return -EINVAL;
        }

        return r;
}

int cunescape_length_with_prefix(const char *s, size_t length, const char *prefix, UnescapeFlags flags, char **ret) {
        char *r, *t;
        const char *f;
        size_t pl;

        assert(s);
        assert(ret);

        /* Undoes C style string escaping, and optionally prefixes it. */

302
        pl = strlen_ptr(prefix);
303 304 305 306 307 308 309 310 311 312

        r = new(char, pl+length+1);
        if (!r)
                return -ENOMEM;

        if (prefix)
                memcpy(r, prefix, pl);

        for (f = s, t = r + pl; f < s + length; f++) {
                size_t remaining;
313
                bool eight_bit = false;
314
                char32_t u;
315 316 317 318 319 320
                int k;

                remaining = s + length - f;
                assert(remaining > 0);

                if (*f != '\\') {
321
                        /* A literal, copy verbatim */
322 323 324 325 326 327 328 329 330 331 332 333 334 335 336
                        *(t++) = *f;
                        continue;
                }

                if (remaining == 1) {
                        if (flags & UNESCAPE_RELAX) {
                                /* A trailing backslash, copy verbatim */
                                *(t++) = *f;
                                continue;
                        }

                        free(r);
                        return -EINVAL;
                }

337
                k = cunescape_one(f + 1, remaining - 1, &u, &eight_bit);
338 339 340 341 342 343 344 345 346 347 348
                if (k < 0) {
                        if (flags & UNESCAPE_RELAX) {
                                /* Invalid escape code, let's take it literal then */
                                *(t++) = '\\';
                                continue;
                        }

                        free(r);
                        return k;
                }

349 350 351 352
                f += k;
                if (eight_bit)
                        /* One byte? Set directly as specified */
                        *(t++) = u;
353
                else
354
                        /* Otherwise encode as multi-byte UTF-8 */
355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400
                        t += utf8_encode_unichar(t, u);
        }

        *t = 0;

        *ret = r;
        return t - r;
}

int cunescape_length(const char *s, size_t length, UnescapeFlags flags, char **ret) {
        return cunescape_length_with_prefix(s, length, NULL, flags, ret);
}

int cunescape(const char *s, UnescapeFlags flags, char **ret) {
        return cunescape_length(s, strlen(s), flags, ret);
}

char *xescape(const char *s, const char *bad) {
        char *r, *t;
        const char *f;

        /* Escapes all chars in bad, in addition to \ and all special
         * chars, in \xFF style escaping. May be reversed with
         * cunescape(). */

        r = new(char, strlen(s) * 4 + 1);
        if (!r)
                return NULL;

        for (f = s, t = r; *f; f++) {

                if ((*f < ' ') || (*f >= 127) ||
                    (*f == '\\') || strchr(bad, *f)) {
                        *(t++) = '\\';
                        *(t++) = 'x';
                        *(t++) = hexchar(*f >> 4);
                        *(t++) = hexchar(*f);
                } else
                        *(t++) = *f;
        }

        *t = 0;

        return r;
}

401 402 403 404 405 406 407 408 409 410 411 412 413
char *octescape(const char *s, size_t len) {
        char *r, *t;
        const char *f;

        /* Escapes all chars in bad, in addition to \ and " chars,
         * in \nnn style escaping. */

        r = new(char, len * 4 + 1);
        if (!r)
                return NULL;

        for (f = s, t = r; f < s + len; f++) {

414
                if (*f < ' ' || *f >= 127 || IN_SET(*f, '\\', '"')) {
415 416 417 418 419 420 421 422 423 424 425 426 427 428
                        *(t++) = '\\';
                        *(t++) = '0' + (*f >> 6);
                        *(t++) = '0' + ((*f >> 3) & 8);
                        *(t++) = '0' + (*f & 8);
                } else
                        *(t++) = *f;
        }

        *t = 0;

        return r;

}

429
static char *strcpy_backslash_escaped(char *t, const char *s, const char *bad, bool escape_tab_nl) {
430 431 432
        assert(bad);

        for (; *s; s++) {
433 434 435 436 437 438
                if (escape_tab_nl && IN_SET(*s, '\n', '\t')) {
                        *(t++) = '\\';
                        *(t++) = *s == '\n' ? 'n' : 't';
                        continue;
                }

439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454
                if (*s == '\\' || strchr(bad, *s))
                        *(t++) = '\\';

                *(t++) = *s;
        }

        return t;
}

char *shell_escape(const char *s, const char *bad) {
        char *r, *t;

        r = new(char, strlen(s)*2+1);
        if (!r)
                return NULL;

455
        t = strcpy_backslash_escaped(r, s, bad, false);
456 457 458 459 460
        *t = 0;

        return r;
}

461
char* shell_maybe_quote(const char *s, EscapeStyle style) {
462 463 464 465 466
        const char *p;
        char *r, *t;

        assert(s);

467 468 469
        /* Encloses a string in quotes if necessary to make it OK as a shell
         * string. Note that we treat benign UTF-8 characters as needing
         * escaping too, but that should be OK. */
470 471 472 473 474 475 476 477 478 479

        for (p = s; *p; p++)
                if (*p <= ' ' ||
                    *p >= 127 ||
                    strchr(SHELL_NEED_QUOTES, *p))
                        break;

        if (!*p)
                return strdup(s);

480
        r = new(char, (style == ESCAPE_POSIX) + 1 + strlen(s)*2 + 1 + 1);
481 482 483 484
        if (!r)
                return NULL;

        t = r;
485 486 487 488 489 490 491 492
        if (style == ESCAPE_BACKSLASH)
                *(t++) = '"';
        else if (style == ESCAPE_POSIX) {
                *(t++) = '$';
                *(t++) = '\'';
        } else
                assert_not_reached("Bad EscapeStyle");

493 494
        t = mempcpy(t, s, p - s);

495 496 497 498
        if (style == ESCAPE_BACKSLASH)
                t = strcpy_backslash_escaped(t, p, SHELL_NEED_ESCAPE, false);
        else
                t = strcpy_backslash_escaped(t, p, SHELL_NEED_ESCAPE_POSIX, true);
499

500 501 502 503
        if (style == ESCAPE_BACKSLASH)
                *(t++) = '"';
        else
                *(t++) = '\'';
504 505 506 507
        *t = 0;

        return r;
}