X-Git-Url: http://www.chiark.greenend.org.uk/ucgi/~ianmdlvl/git?p=elogind.git;a=blobdiff_plain;f=src%2Fbasic%2Fescape.c;h=2e483880c80c53996a73c9d9816c8cbcac524b0c;hp=42a84c93174e619736e33b7a395179c947deca9b;hb=66ecc207e203db5434610395cd04c40ae8727b58;hpb=538a7c37c543255dbcb21c3854191c855150fa1d

diff --git a/src/basic/escape.c b/src/basic/escape.c
index 42a84c931..2e483880c 100644
--- a/src/basic/escape.c
+++ b/src/basic/escape.c
@@ -1,5 +1,3 @@
-/*-*- Mode: C; c-basic-offset: 8; indent-tabs-mode: nil -*-*/
-
 /***
   This file is part of systemd.
 
@@ -19,12 +17,15 @@
   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 "string-util.h"
+#include "macro.h"
 #include "utf8.h"
-#include "util.h"
 
 size_t cescape_char(char c, char *buf) {
         char * buf_old = buf;
@@ -116,16 +117,18 @@ char *cescape(const char *s) {
         return cescape_length(s, strlen(s));
 }
 
-int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode) {
+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,
-         * unless we encountered a \u sequence in which case the full
-         * unicode character is returned in ret_unicode, instead. */
+        /* 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;
@@ -187,7 +190,8 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode
                 if (a == 0 && b == 0)
                         return -EINVAL;
 
-                *ret = (char) ((a << 4U) | b);
+                *ret = (a << 4U) | b;
+                *eight_bit = true;
                 r = 3;
                 break;
         }
@@ -214,16 +218,7 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode
                 if (c == 0)
                         return -EINVAL;
 
-                if (c < 128)
-                        *ret = c;
-                else {
-                        if (!ret_unicode)
-                                return -EINVAL;
-
-                        *ret = 0;
-                        *ret_unicode = c;
-                }
-
+                *ret = c;
                 r = 5;
                 break;
         }
@@ -233,7 +228,7 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode
 
                 int a[8];
                 unsigned i;
-                uint32_t c;
+                char32_t c;
 
                 if (length != (size_t) -1 && length < 9)
                         return -EINVAL;
@@ -255,16 +250,7 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode
                 if (!unichar_is_valid(c))
                         return -EINVAL;
 
-                if (c < 128)
-                        *ret = c;
-                else {
-                        if (!ret_unicode)
-                                return -EINVAL;
-
-                        *ret = 0;
-                        *ret_unicode = c;
-                }
-
+                *ret = c;
                 r = 9;
                 break;
         }
@@ -279,7 +265,7 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode
         case '7': {
                 /* octal encoding */
                 int a, b, c;
-                uint32_t m;
+                char32_t m;
 
                 if (length != (size_t) -1 && length < 3)
                         return -EINVAL;
@@ -306,6 +292,7 @@ int cunescape_one(const char *p, size_t length, char *ret, uint32_t *ret_unicode
                         return -EINVAL;
 
                 *ret = m;
+                *eight_bit = true;
                 r = 3;
                 break;
         }
@@ -338,8 +325,8 @@ int cunescape_length_with_prefix(const char *s, size_t length, const char *prefi
 
         for (f = s, t = r + pl; f < s + length; f++) {
                 size_t remaining;
-                uint32_t u;
-                char c;
+                bool eight_bit = false;
+                char32_t u;
                 int k;
 
                 remaining = s + length - f;
@@ -362,7 +349,7 @@ int cunescape_length_with_prefix(const char *s, size_t length, const char *prefi
                         return -EINVAL;
                 }
 
-                k = cunescape_one(f + 1, remaining - 1, &c, &u);
+                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 */
@@ -374,14 +361,13 @@ int cunescape_length_with_prefix(const char *s, size_t length, const char *prefi
                         return k;
                 }
 
-                if (c != 0)
-                        /* Non-Unicode? Let's encode this directly */
-                        *(t++) = c;
+                f += k;
+                if (eight_bit)
+                        /* One byte? Set directly as specified */
+                        *(t++) = u;
                 else
-                        /* Unicode? Then let's encode this in UTF-8 */
+                        /* Otherwise encode as multi-byte UTF-8 */
                         t += utf8_encode_unichar(t, u);
-
-                f += k;
         }
 
         *t = 0;