summaryrefslogtreecommitdiff
path: root/src/basic
diff options
context:
space:
mode:
authorLennart Poettering <lennart@poettering.net>2015-10-22 20:12:31 +0200
committerLennart Poettering <lennart@poettering.net>2015-10-24 23:03:49 +0200
commit84ac7bea360cd369df26910e9685a7eed2327088 (patch)
tree7b941067ff1af844404ad7c433473b1d5090c1ec /src/basic
parent0f03c2a4c093e3d44f4072144827e943c05c8904 (diff)
util: split out extract_first_word() and related calls into extract-word.[ch]
This is quite a lot of code these days, hence move it to its own source file.
Diffstat (limited to 'src/basic')
-rw-r--r--src/basic/cgroup-util.c25
-rw-r--r--src/basic/cpu-set-util.c1
-rw-r--r--src/basic/extract-word.c274
-rw-r--r--src/basic/extract-word.h36
-rw-r--r--src/basic/strv.h3
-rw-r--r--src/basic/util.c251
-rw-r--r--src/basic/util.h13
7 files changed, 328 insertions, 275 deletions
diff --git a/src/basic/cgroup-util.c b/src/basic/cgroup-util.c
index 95fc2b9e5d..a3ea512165 100644
--- a/src/basic/cgroup-util.c
+++ b/src/basic/cgroup-util.c
@@ -19,27 +19,28 @@
along with systemd; If not, see <http://www.gnu.org/licenses/>.
***/
+#include <dirent.h>
#include <errno.h>
-#include <unistd.h>
+#include <ftw.h>
#include <signal.h>
-#include <string.h>
#include <stdlib.h>
-#include <dirent.h>
+#include <string.h>
#include <sys/stat.h>
#include <sys/types.h>
-#include <ftw.h>
+#include <unistd.h>
-#include "set.h"
-#include "macro.h"
-#include "util.h"
+#include "extract-word.h"
+#include "fileio.h"
#include "formats-util.h"
-#include "process-util.h"
+#include "login-util.h"
+#include "macro.h"
+#include "mkdir.h"
#include "path-util.h"
-#include "unit-name.h"
-#include "fileio.h"
+#include "process-util.h"
+#include "set.h"
#include "special.h"
-#include "mkdir.h"
-#include "login-util.h"
+#include "unit-name.h"
+#include "util.h"
#include "cgroup-util.h"
int cg_enumerate_processes(const char *controller, const char *path, FILE **_f) {
diff --git a/src/basic/cpu-set-util.c b/src/basic/cpu-set-util.c
index 519583c167..5e064d854f 100644
--- a/src/basic/cpu-set-util.c
+++ b/src/basic/cpu-set-util.c
@@ -20,6 +20,7 @@
along with systemd; If not, see <http://www.gnu.org/licenses/>.
***/
+#include "extract-word.h"
#include "util.h"
#include "cpu-set-util.h"
diff --git a/src/basic/extract-word.c b/src/basic/extract-word.c
new file mode 100644
index 0000000000..474e6fdd57
--- /dev/null
+++ b/src/basic/extract-word.c
@@ -0,0 +1,274 @@
+/*-*- 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 "utf8.h"
+#include "util.h"
+
+#include "extract-word.h"
+
+int extract_first_word(const char **p, char **ret, const char *separators, ExtractFlags flags) {
+ _cleanup_free_ char *s = NULL;
+ size_t allocated = 0, sz = 0;
+ int r;
+
+ char quote = 0; /* 0 or ' or " */
+ bool backslash = false; /* whether we've just seen a backslash */
+ bool separator = false; /* whether we've just seen a separator */
+ bool start = true; /* false means we're looking at a value */
+
+ assert(p);
+ assert(ret);
+
+ if (!separators)
+ separators = WHITESPACE;
+
+ /* Bail early if called after last value or with no input */
+ if (!*p)
+ goto finish_force_terminate;
+
+ /* Parses the first word of a string, and returns it in
+ * *ret. Removes all quotes in the process. When parsing fails
+ * (because of an uneven number of quotes or similar), leaves
+ * the pointer *p at the first invalid character. */
+
+ for (;;) {
+ char c = **p;
+
+ if (start) {
+ if (flags & EXTRACT_DONT_COALESCE_SEPARATORS)
+ if (!GREEDY_REALLOC(s, allocated, sz+1))
+ return -ENOMEM;
+
+ if (c == 0)
+ goto finish_force_terminate;
+ else if (strchr(separators, c)) {
+ (*p) ++;
+ if (flags & EXTRACT_DONT_COALESCE_SEPARATORS)
+ goto finish_force_next;
+ continue;
+ }
+
+ /* We found a non-blank character, so we will always
+ * want to return a string (even if it is empty),
+ * allocate it here. */
+ if (!GREEDY_REALLOC(s, allocated, sz+1))
+ return -ENOMEM;
+
+ start = false;
+ }
+
+ if (backslash) {
+ if (!GREEDY_REALLOC(s, allocated, sz+7))
+ return -ENOMEM;
+
+ if (c == 0) {
+ if ((flags & EXTRACT_CUNESCAPE_RELAX) &&
+ (!quote || flags & EXTRACT_RELAX)) {
+ /* If we find an unquoted trailing backslash and we're in
+ * EXTRACT_CUNESCAPE_RELAX mode, keep it verbatim in the
+ * output.
+ *
+ * Unbalanced quotes will only be allowed in EXTRACT_RELAX
+ * mode, EXTRACT_CUNESCAPE_RELAX mode does not allow them.
+ */
+ s[sz++] = '\\';
+ goto finish_force_terminate;
+ }
+ if (flags & EXTRACT_RELAX)
+ goto finish_force_terminate;
+ return -EINVAL;
+ }
+
+ if (flags & EXTRACT_CUNESCAPE) {
+ uint32_t u;
+
+ r = cunescape_one(*p, (size_t) -1, &c, &u);
+ if (r < 0) {
+ if (flags & EXTRACT_CUNESCAPE_RELAX) {
+ s[sz++] = '\\';
+ s[sz++] = c;
+ goto end_escape;
+ }
+ return -EINVAL;
+ }
+
+ (*p) += r - 1;
+
+ if (c != 0)
+ s[sz++] = c; /* normal explicit char */
+ else
+ sz += utf8_encode_unichar(s + sz, u); /* unicode chars we'll encode as utf8 */
+ } else
+ s[sz++] = c;
+
+end_escape:
+ backslash = false;
+
+ } else if (quote) { /* inside either single or double quotes */
+ if (c == 0) {
+ if (flags & EXTRACT_RELAX)
+ goto finish_force_terminate;
+ return -EINVAL;
+ } else if (c == quote) /* found the end quote */
+ quote = 0;
+ else if (c == '\\')
+ backslash = true;
+ else {
+ if (!GREEDY_REALLOC(s, allocated, sz+2))
+ return -ENOMEM;
+
+ s[sz++] = c;
+ }
+
+ } else if (separator) {
+ if (c == 0)
+ goto finish_force_terminate;
+ if (!strchr(separators, c))
+ goto finish;
+
+ } else {
+ if (c == 0)
+ goto finish_force_terminate;
+ else if ((c == '\'' || c == '"') && (flags & EXTRACT_QUOTES))
+ quote = c;
+ else if (c == '\\')
+ backslash = true;
+ else if (strchr(separators, c)) {
+ if (flags & EXTRACT_DONT_COALESCE_SEPARATORS) {
+ (*p) ++;
+ goto finish_force_next;
+ }
+ separator = true;
+ } else {
+ if (!GREEDY_REALLOC(s, allocated, sz+2))
+ return -ENOMEM;
+
+ s[sz++] = c;
+ }
+ }
+
+ (*p) ++;
+ }
+
+finish_force_terminate:
+ *p = NULL;
+finish:
+ if (!s) {
+ *p = NULL;
+ *ret = NULL;
+ return 0;
+ }
+
+finish_force_next:
+ s[sz] = 0;
+ *ret = s;
+ s = NULL;
+
+ return 1;
+}
+
+int extract_first_word_and_warn(
+ const char **p,
+ char **ret,
+ const char *separators,
+ ExtractFlags flags,
+ const char *unit,
+ const char *filename,
+ unsigned line,
+ const char *rvalue) {
+
+ /* Try to unquote it, if it fails, warn about it and try again but this
+ * time using EXTRACT_CUNESCAPE_RELAX to keep the backslashes verbatim
+ * in invalid escape sequences. */
+ const char *save;
+ int r;
+
+ save = *p;
+ r = extract_first_word(p, ret, separators, flags);
+ if (r < 0 && !(flags & EXTRACT_CUNESCAPE_RELAX)) {
+
+ /* Retry it with EXTRACT_CUNESCAPE_RELAX. */
+ *p = save;
+ r = extract_first_word(p, ret, separators, flags|EXTRACT_CUNESCAPE_RELAX);
+ if (r < 0)
+ log_syntax(unit, LOG_ERR, filename, line, r, "Unbalanced quoting in command line, ignoring: \"%s\"", rvalue);
+ else
+ log_syntax(unit, LOG_WARNING, filename, line, 0, "Invalid escape sequences in command line: \"%s\"", rvalue);
+ }
+
+ return r;
+}
+
+int extract_many_words(const char **p, const char *separators, ExtractFlags flags, ...) {
+ va_list ap;
+ char **l;
+ int n = 0, i, c, r;
+
+ /* Parses a number of words from a string, stripping any
+ * quotes if necessary. */
+
+ assert(p);
+
+ /* Count how many words are expected */
+ va_start(ap, flags);
+ for (;;) {
+ if (!va_arg(ap, char **))
+ break;
+ n++;
+ }
+ va_end(ap);
+
+ if (n <= 0)
+ return 0;
+
+ /* Read all words into a temporary array */
+ l = newa0(char*, n);
+ for (c = 0; c < n; c++) {
+
+ r = extract_first_word(p, &l[c], separators, flags);
+ if (r < 0) {
+ int j;
+
+ for (j = 0; j < c; j++)
+ free(l[j]);
+
+ return r;
+ }
+
+ if (r == 0)
+ break;
+ }
+
+ /* If we managed to parse all words, return them in the passed
+ * in parameters */
+ va_start(ap, flags);
+ for (i = 0; i < n; i++) {
+ char **v;
+
+ v = va_arg(ap, char **);
+ assert(v);
+
+ *v = l[i];
+ }
+ va_end(ap);
+
+ return c;
+}
diff --git a/src/basic/extract-word.h b/src/basic/extract-word.h
new file mode 100644
index 0000000000..ddc1c4f463
--- /dev/null
+++ b/src/basic/extract-word.h
@@ -0,0 +1,36 @@
+/*-*- Mode: C; c-basic-offset: 8; indent-tabs-mode: nil -*-*/
+
+#pragma once
+
+/***
+ 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 "macro.h"
+
+typedef enum ExtractFlags {
+ EXTRACT_RELAX = 1,
+ EXTRACT_CUNESCAPE = 2,
+ EXTRACT_CUNESCAPE_RELAX = 4,
+ EXTRACT_QUOTES = 8,
+ EXTRACT_DONT_COALESCE_SEPARATORS = 16,
+} ExtractFlags;
+
+int extract_first_word(const char **p, char **ret, const char *separators, ExtractFlags flags);
+int extract_first_word_and_warn(const char **p, char **ret, const char *separators, ExtractFlags flags, const char *unit, const char *filename, unsigned line, const char *rvalue);
+int extract_many_words(const char **p, const char *separators, ExtractFlags flags, ...) _sentinel_;
diff --git a/src/basic/strv.h b/src/basic/strv.h
index a5dc696a87..e66794fc34 100644
--- a/src/basic/strv.h
+++ b/src/basic/strv.h
@@ -21,10 +21,11 @@
along with systemd; If not, see <http://www.gnu.org/licenses/>.
***/
+#include <fnmatch.h>
#include <stdarg.h>
#include <stdbool.h>
-#include <fnmatch.h>
+#include "extract-word.h"
#include "util.h"
char *strv_find(char **l, const char *name) _pure_;
diff --git a/src/basic/util.c b/src/basic/util.c
index 63c8abcf82..641e0b4d89 100644
--- a/src/basic/util.c
+++ b/src/basic/util.c
@@ -1587,7 +1587,7 @@ char *cescape(const char *s) {
return r;
}
-static int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode) {
+int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode) {
int r = 1;
assert(p);
@@ -5801,255 +5801,6 @@ int is_device_node(const char *path) {
return !!(S_ISBLK(info.st_mode) || S_ISCHR(info.st_mode));
}
-int extract_first_word(const char **p, char **ret, const char *separators, ExtractFlags flags) {
- _cleanup_free_ char *s = NULL;
- size_t allocated = 0, sz = 0;
- int r;
-
- char quote = 0; /* 0 or ' or " */
- bool backslash = false; /* whether we've just seen a backslash */
- bool separator = false; /* whether we've just seen a separator */
- bool start = true; /* false means we're looking at a value */
-
- assert(p);
- assert(ret);
-
- if (!separators)
- separators = WHITESPACE;
-
- /* Bail early if called after last value or with no input */
- if (!*p)
- goto finish_force_terminate;
-
- /* Parses the first word of a string, and returns it in
- * *ret. Removes all quotes in the process. When parsing fails
- * (because of an uneven number of quotes or similar), leaves
- * the pointer *p at the first invalid character. */
-
- for (;;) {
- char c = **p;
-
- if (start) {
- if (flags & EXTRACT_DONT_COALESCE_SEPARATORS)
- if (!GREEDY_REALLOC(s, allocated, sz+1))
- return -ENOMEM;
-
- if (c == 0)
- goto finish_force_terminate;
- else if (strchr(separators, c)) {
- (*p) ++;
- if (flags & EXTRACT_DONT_COALESCE_SEPARATORS)
- goto finish_force_next;
- continue;
- }
-
- /* We found a non-blank character, so we will always
- * want to return a string (even if it is empty),
- * allocate it here. */
- if (!GREEDY_REALLOC(s, allocated, sz+1))
- return -ENOMEM;
-
- start = false;
- }
-
- if (backslash) {
- if (!GREEDY_REALLOC(s, allocated, sz+7))
- return -ENOMEM;
-
- if (c == 0) {
- if ((flags & EXTRACT_CUNESCAPE_RELAX) &&
- (!quote || flags & EXTRACT_RELAX)) {
- /* If we find an unquoted trailing backslash and we're in
- * EXTRACT_CUNESCAPE_RELAX mode, keep it verbatim in the
- * output.
- *
- * Unbalanced quotes will only be allowed in EXTRACT_RELAX
- * mode, EXTRACT_CUNESCAPE_RELAX mode does not allow them.
- */
- s[sz++] = '\\';
- goto finish_force_terminate;
- }
- if (flags & EXTRACT_RELAX)
- goto finish_force_terminate;
- return -EINVAL;
- }
-
- if (flags & EXTRACT_CUNESCAPE) {
- uint32_t u;
-
- r = cunescape_one(*p, (size_t) -1, &c, &u);
- if (r < 0) {
- if (flags & EXTRACT_CUNESCAPE_RELAX) {
- s[sz++] = '\\';
- s[sz++] = c;
- goto end_escape;
- }
- return -EINVAL;
- }
-
- (*p) += r - 1;
-
- if (c != 0)
- s[sz++] = c; /* normal explicit char */
- else
- sz += utf8_encode_unichar(s + sz, u); /* unicode chars we'll encode as utf8 */
- } else
- s[sz++] = c;
-
-end_escape:
- backslash = false;
-
- } else if (quote) { /* inside either single or double quotes */
- if (c == 0) {
- if (flags & EXTRACT_RELAX)
- goto finish_force_terminate;
- return -EINVAL;
- } else if (c == quote) /* found the end quote */
- quote = 0;
- else if (c == '\\')
- backslash = true;
- else {
- if (!GREEDY_REALLOC(s, allocated, sz+2))
- return -ENOMEM;
-
- s[sz++] = c;
- }
-
- } else if (separator) {
- if (c == 0)
- goto finish_force_terminate;
- if (!strchr(separators, c))
- goto finish;
-
- } else {
- if (c == 0)
- goto finish_force_terminate;
- else if ((c == '\'' || c == '"') && (flags & EXTRACT_QUOTES))
- quote = c;
- else if (c == '\\')
- backslash = true;
- else if (strchr(separators, c)) {
- if (flags & EXTRACT_DONT_COALESCE_SEPARATORS) {
- (*p) ++;
- goto finish_force_next;
- }
- separator = true;
- } else {
- if (!GREEDY_REALLOC(s, allocated, sz+2))
- return -ENOMEM;
-
- s[sz++] = c;
- }
- }
-
- (*p) ++;
- }
-
-finish_force_terminate:
- *p = NULL;
-finish:
- if (!s) {
- *p = NULL;
- *ret = NULL;
- return 0;
- }
-
-finish_force_next:
- s[sz] = 0;
- *ret = s;
- s = NULL;
-
- return 1;
-}
-
-int extract_first_word_and_warn(
- const char **p,
- char **ret,
- const char *separators,
- ExtractFlags flags,
- const char *unit,
- const char *filename,
- unsigned line,
- const char *rvalue) {
-
- /* Try to unquote it, if it fails, warn about it and try again but this
- * time using EXTRACT_CUNESCAPE_RELAX to keep the backslashes verbatim
- * in invalid escape sequences. */
- const char *save;
- int r;
-
- save = *p;
- r = extract_first_word(p, ret, separators, flags);
- if (r < 0 && !(flags & EXTRACT_CUNESCAPE_RELAX)) {
-
- /* Retry it with EXTRACT_CUNESCAPE_RELAX. */
- *p = save;
- r = extract_first_word(p, ret, separators, flags|EXTRACT_CUNESCAPE_RELAX);
- if (r < 0)
- log_syntax(unit, LOG_ERR, filename, line, r, "Unbalanced quoting in command line, ignoring: \"%s\"", rvalue);
- else
- log_syntax(unit, LOG_WARNING, filename, line, 0, "Invalid escape sequences in command line: \"%s\"", rvalue);
- }
-
- return r;
-}
-
-int extract_many_words(const char **p, const char *separators, ExtractFlags flags, ...) {
- va_list ap;
- char **l;
- int n = 0, i, c, r;
-
- /* Parses a number of words from a string, stripping any
- * quotes if necessary. */
-
- assert(p);
-
- /* Count how many words are expected */
- va_start(ap, flags);
- for (;;) {
- if (!va_arg(ap, char **))
- break;
- n++;
- }
- va_end(ap);
-
- if (n <= 0)
- return 0;
-
- /* Read all words into a temporary array */
- l = newa0(char*, n);
- for (c = 0; c < n; c++) {
-
- r = extract_first_word(p, &l[c], separators, flags);
- if (r < 0) {
- int j;
-
- for (j = 0; j < c; j++)
- free(l[j]);
-
- return r;
- }
-
- if (r == 0)
- break;
- }
-
- /* If we managed to parse all words, return them in the passed
- * in parameters */
- va_start(ap, flags);
- for (i = 0; i < n; i++) {
- char **v;
-
- v = va_arg(ap, char **);
- assert(v);
-
- *v = l[i];
- }
- va_end(ap);
-
- return c;
-}
-
int free_and_strdup(char **p, const char *s) {
char *t;
diff --git a/src/basic/util.h b/src/basic/util.h
index a3ebb987e4..132e6f862b 100644
--- a/src/basic/util.h
+++ b/src/basic/util.h
@@ -270,6 +270,7 @@ typedef enum UnescapeFlags {
int cunescape(const char *s, UnescapeFlags flags, char **ret);
int cunescape_length(const char *s, size_t length, UnescapeFlags flags, char **ret);
int cunescape_length_with_prefix(const char *s, size_t length, const char *prefix, UnescapeFlags flags, char **ret);
+int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode);
char *xescape(const char *s, const char *bad);
@@ -879,18 +880,6 @@ int is_symlink(const char *path);
int is_dir(const char *path, bool follow);
int is_device_node(const char *path);
-typedef enum ExtractFlags {
- EXTRACT_RELAX = 1,
- EXTRACT_CUNESCAPE = 2,
- EXTRACT_CUNESCAPE_RELAX = 4,
- EXTRACT_QUOTES = 8,
- EXTRACT_DONT_COALESCE_SEPARATORS = 16,
-} ExtractFlags;
-
-int extract_first_word(const char **p, char **ret, const char *separators, ExtractFlags flags);
-int extract_first_word_and_warn(const char **p, char **ret, const char *separators, ExtractFlags flags, const char *unit, const char *filename, unsigned line, const char *rvalue);
-int extract_many_words(const char **p, const char *separators, ExtractFlags flags, ...) _sentinel_;
-
int free_and_strdup(char **p, const char *s);
#define INOTIFY_EVENT_MAX (sizeof(struct inotify_event) + NAME_MAX + 1)