summaryrefslogtreecommitdiff
path: root/src/basic/escape.c
diff options
context:
space:
mode:
Diffstat (limited to 'src/basic/escape.c')
-rw-r--r--src/basic/escape.c476
1 files changed, 476 insertions, 0 deletions
diff --git a/src/basic/escape.c b/src/basic/escape.c
new file mode 100644
index 0000000000..5661f36813
--- /dev/null
+++ b/src/basic/escape.c
@@ -0,0 +1,476 @@
+/*-*- Mode: C; c-basic-offset: 8; indent-tabs-mode: nil -*-*/
+
+/***
+ This file is part of systemd.
+
+ Copyright 2010 Lennart Poettering
+
+ systemd is free software; you can redistribute it and/or modify it
+ under the terms of the GNU Lesser General Public License as published by
+ the Free Software Foundation; either version 2.1 of the License, or
+ (at your option) any later version.
+
+ systemd is distributed in the hope that it will be useful, but
+ WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public License
+ along with systemd; If not, see <http://www.gnu.org/licenses/>.
+***/
+
+#include <errno.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "alloc-util.h"
+#include "escape.h"
+#include "hexdecoct.h"
+#include "macro.h"
+#include "utf8.h"
+
+size_t cescape_char(char c, char *buf) {
+ char * buf_old = buf;
+
+ 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;
+}
+
+char *cescape_length(const char *s, size_t n) {
+ const char *f;
+ char *r, *t;
+
+ assert(s || n == 0);
+
+ /* Does C style string escaping. May be reversed with
+ * cunescape(). */
+
+ r = new(char, n*4 + 1);
+ if (!r)
+ return NULL;
+
+ for (f = s, t = r; f < s + n; f++)
+ t += cescape_char(*f, t);
+
+ *t = 0;
+
+ return r;
+}
+
+char *cescape(const char *s) {
+ assert(s);
+
+ return cescape_length(s, strlen(s));
+}
+
+int cunescape_one(const char *p, size_t length, uint32_t *ret, bool *eight_bit) {
+ int r = 1;
+
+ assert(p);
+ assert(*p);
+ assert(ret);
+
+ /* 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.
+ */
+
+ 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;
+
+ *ret = (a << 4U) | b;
+ *eight_bit = true;
+ r = 3;
+ break;
+ }
+
+ case 'u': {
+ /* C++11 style 16bit unicode */
+
+ int a[4];
+ unsigned i;
+ 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;
+
+ *ret = c;
+ r = 5;
+ break;
+ }
+
+ case 'U': {
+ /* C++11 style 32bit unicode */
+
+ int a[8];
+ unsigned i;
+ uint32_t c;
+
+ 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;
+
+ *ret = c;
+ 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;
+ uint32_t m;
+
+ 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;
+ *eight_bit = true;
+ 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. */
+
+ pl = prefix ? strlen(prefix) : 0;
+
+ 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;
+ uint32_t u;
+ bool eight_bit = false;
+ int k;
+
+ remaining = s + length - f;
+ assert(remaining > 0);
+
+ if (*f != '\\') {
+ /* A literal literal, copy verbatim */
+ *(t++) = *f;
+ continue;
+ }
+
+ if (remaining == 1) {
+ if (flags & UNESCAPE_RELAX) {
+ /* A trailing backslash, copy verbatim */
+ *(t++) = *f;
+ continue;
+ }
+
+ free(r);
+ return -EINVAL;
+ }
+
+ k = cunescape_one(f + 1, remaining - 1, &u, &eight_bit);
+ if (k < 0) {
+ if (flags & UNESCAPE_RELAX) {
+ /* Invalid escape code, let's take it literal then */
+ *(t++) = '\\';
+ continue;
+ }
+
+ free(r);
+ return k;
+ }
+
+ f += k;
+ if (eight_bit)
+ /* One byte? Set directly as specified */
+ *(t++) = u;
+ else
+ /* Otherwise encode as multi-byte UTF-8 */
+ 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;
+}
+
+static char *strcpy_backslash_escaped(char *t, const char *s, const char *bad) {
+ assert(bad);
+
+ for (; *s; s++) {
+ 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;
+
+ t = strcpy_backslash_escaped(r, s, bad);
+ *t = 0;
+
+ return r;
+}
+
+char *shell_maybe_quote(const char *s) {
+ const char *p;
+ char *r, *t;
+
+ assert(s);
+
+ /* Encloses a string in double quotes if necessary to make it
+ * OK as shell string. */
+
+ for (p = s; *p; p++)
+ if (*p <= ' ' ||
+ *p >= 127 ||
+ strchr(SHELL_NEED_QUOTES, *p))
+ break;
+
+ if (!*p)
+ return strdup(s);
+
+ r = new(char, 1+strlen(s)*2+1+1);
+ if (!r)
+ return NULL;
+
+ t = r;
+ *(t++) = '"';
+ t = mempcpy(t, s, p - s);
+
+ t = strcpy_backslash_escaped(t, p, SHELL_NEED_ESCAPE);
+
+ *(t++)= '"';
+ *t = 0;
+
+ return r;
+}
generated by cgit v1.2.3-54-g00ecf (git 2.45.2) at 2024-12-12 02:50:45 +0000