-/*-*- 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/>.
-***/
+/* SPDX-License-Identifier: LGPL-2.1+ */
#include <errno.h>
#include <stdlib.h>
#include "macro.h"
#include "utf8.h"
-size_t cescape_char(char c, char *buf) {
- char * buf_old = buf;
+int cescape_char(char c, char *buf) {
+ char *buf_old = buf;
+
+ /* Needs space for 4 characters in the buffer */
switch (c) {
return cescape_length(s, strlen(s));
}
-int cunescape_one(const char *p, size_t length, uint32_t *ret, bool *eight_bit) {
+int cunescape_one(const char *p, size_t length, char32_t *ret, bool *eight_bit) {
int r = 1;
assert(p);
- assert(*p);
assert(ret);
/* Unescapes C style. Returns the unescaped character in ret.
/* C++11 style 16bit unicode */
int a[4];
- unsigned i;
+ size_t i;
uint32_t c;
if (length != (size_t) -1 && length < 5)
/* C++11 style 32bit unicode */
int a[8];
- unsigned i;
- uint32_t c;
+ size_t i;
+ char32_t c;
if (length != (size_t) -1 && length < 9)
return -EINVAL;
case '7': {
/* octal encoding */
int a, b, c;
- uint32_t m;
+ char32_t m;
if (length != (size_t) -1 && length < 3)
return -EINVAL;
/* Undoes C style string escaping, and optionally prefixes it. */
- pl = prefix ? strlen(prefix) : 0;
+ pl = strlen_ptr(prefix);
r = new(char, pl+length+1);
if (!r)
for (f = s, t = r + pl; f < s + length; f++) {
size_t remaining;
- uint32_t u;
bool eight_bit = false;
+ char32_t u;
int k;
remaining = s + length - f;
assert(remaining > 0);
if (*f != '\\') {
- /* A literal literal, copy verbatim */
+ /* A literal, copy verbatim */
*(t++) = *f;
continue;
}
return cunescape_length(s, strlen(s), flags, ret);
}
-char *xescape(const char *s, const char *bad) {
- char *r, *t;
+char *xescape_full(const char *s, const char *bad, size_t console_width, bool eight_bits) {
+ char *ans, *t, *prev, *prev2;
const char *f;
- /* Escapes all chars in bad, in addition to \ and all special
- * chars, in \xFF style escaping. May be reversed with
- * cunescape(). */
+ /* Escapes all chars in bad, in addition to \ and all special chars, in \xFF style escaping. May be
+ * reversed with cunescape(). If eight_bits is true, characters >= 127 are let through unchanged.
+ * This corresponds to non-ASCII printable characters in pre-unicode encodings.
+ *
+ * If console_width is reached, output is truncated and "..." is appended. */
- r = new(char, strlen(s) * 4 + 1);
- if (!r)
+ if (console_width == 0)
+ return strdup("");
+
+ ans = new(char, MIN(strlen(s), console_width) * 4 + 1);
+ if (!ans)
return NULL;
- for (f = s, t = r; *f; f++) {
+ memset(ans, '_', MIN(strlen(s), console_width) * 4);
+ ans[MIN(strlen(s), console_width) * 4] = 0;
+
+ for (f = s, t = prev = prev2 = ans; ; f++) {
+ char *tmp_t = t;
+
+ if (!*f) {
+ *t = 0;
+ return ans;
+ }
+
+ if ((unsigned char) *f < ' ' || (!eight_bits && (unsigned char) *f >= 127) ||
+ *f == '\\' || strchr(bad, *f)) {
+ if ((size_t) (t - ans) + 4 > console_width)
+ break;
- if ((*f < ' ') || (*f >= 127) ||
- (*f == '\\') || strchr(bad, *f)) {
*(t++) = '\\';
*(t++) = 'x';
*(t++) = hexchar(*f >> 4);
*(t++) = hexchar(*f);
+ } else {
+ if ((size_t) (t - ans) + 1 > console_width)
+ break;
+
+ *(t++) = *f;
+ }
+
+ /* We might need to go back two cycles to fit three dots, so remember two positions */
+ prev2 = prev;
+ prev = tmp_t;
+ }
+
+ /* We can just write where we want, since chars are one-byte */
+ size_t c = MIN(console_width, 3u); /* If the console is too narrow, write fewer dots */
+ size_t off;
+ if (console_width - c >= (size_t) (t - ans))
+ off = (size_t) (t - ans);
+ else if (console_width - c >= (size_t) (prev - ans))
+ off = (size_t) (prev - ans);
+ else if (console_width - c >= (size_t) (prev2 - ans))
+ off = (size_t) (prev2 - ans);
+ else
+ off = console_width - c;
+ assert(off <= (size_t) (t - ans));
+
+ memcpy(ans + off, "...", c);
+ ans[off + c] = '\0';
+ return ans;
+}
+
+char *escape_non_printable_full(const char *str, size_t console_width, bool eight_bit) {
+ if (eight_bit)
+ return xescape_full(str, "", console_width, true);
+ else
+ return utf8_escape_non_printable_full(str, console_width);
+}
+
+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++) {
+
+ if (*f < ' ' || *f >= 127 || IN_SET(*f, '\\', '"')) {
+ *(t++) = '\\';
+ *(t++) = '0' + (*f >> 6);
+ *(t++) = '0' + ((*f >> 3) & 8);
+ *(t++) = '0' + (*f & 8);
} else
*(t++) = *f;
}
*t = 0;
return r;
+
}
-static char *strcpy_backslash_escaped(char *t, const char *s, const char *bad) {
+static char *strcpy_backslash_escaped(char *t, const char *s, const char *bad, bool escape_tab_nl) {
assert(bad);
for (; *s; s++) {
+ if (escape_tab_nl && IN_SET(*s, '\n', '\t')) {
+ *(t++) = '\\';
+ *(t++) = *s == '\n' ? 'n' : 't';
+ continue;
+ }
+
if (*s == '\\' || strchr(bad, *s))
*(t++) = '\\';
if (!r)
return NULL;
- t = strcpy_backslash_escaped(r, s, bad);
+ t = strcpy_backslash_escaped(r, s, bad, false);
*t = 0;
return r;
}
-char *shell_maybe_quote(const char *s) {
+char* shell_maybe_quote(const char *s, EscapeStyle style) {
const char *p;
char *r, *t;
assert(s);
- /* Encloses a string in double quotes if necessary to make it
- * OK as shell string. */
+ /* 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. */
for (p = s; *p; p++)
if (*p <= ' ' ||
if (!*p)
return strdup(s);
- r = new(char, 1+strlen(s)*2+1+1);
+ r = new(char, (style == ESCAPE_POSIX) + 1 + strlen(s)*2 + 1 + 1);
if (!r)
return NULL;
t = r;
- *(t++) = '"';
+ if (style == ESCAPE_BACKSLASH)
+ *(t++) = '"';
+ else if (style == ESCAPE_POSIX) {
+ *(t++) = '$';
+ *(t++) = '\'';
+ } else
+ assert_not_reached("Bad EscapeStyle");
+
t = mempcpy(t, s, p - s);
- t = strcpy_backslash_escaped(t, p, SHELL_NEED_ESCAPE);
+ 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);
- *(t++)= '"';
+ if (style == ESCAPE_BACKSLASH)
+ *(t++) = '"';
+ else
+ *(t++) = '\'';
*t = 0;
return r;