systemd/0143-boot-Add-xstrn8_to_16.patch
Jan Macku 79350f79d8 systemd-252-2
Resolves: #2137584,#2138081,#2141979
2022-12-09 09:36:14 +01:00

165 lines
5.1 KiB
Diff

From c48d12fc0abb8f113ed386c47bf02451ea8f853d Mon Sep 17 00:00:00 2001
From: Jan Janssen <medhefgo@web.de>
Date: Tue, 22 Nov 2022 15:55:07 +0100
Subject: [PATCH] boot: Add xstrn8_to_16
(cherry picked from commit 8ad7deffa95d33b5849ad6589dd52ab12e645edc)
Related: #2138081
---
src/boot/efi/efi-string.c | 78 +++++++++++++++++++++++++++++++++-
src/boot/efi/efi-string.h | 5 +++
src/boot/efi/test-efi-string.c | 27 ++++++++++++
3 files changed, 109 insertions(+), 1 deletion(-)
diff --git a/src/boot/efi/efi-string.c b/src/boot/efi/efi-string.c
index b877c6f224..2ba15673c9 100644
--- a/src/boot/efi/efi-string.c
+++ b/src/boot/efi/efi-string.c
@@ -9,7 +9,8 @@
# include "util.h"
#else
# include <stdlib.h>
-# include "macro.h"
+# include "alloc-util.h"
+# define xnew(t, n) ASSERT_SE_PTR(new(t, n))
# define xmalloc(n) ASSERT_SE_PTR(malloc(n))
#endif
@@ -138,6 +139,81 @@ DEFINE_STRCHR(char16_t, strchr16);
DEFINE_STRNDUP(char, xstrndup8, strnlen8);
DEFINE_STRNDUP(char16_t, xstrndup16, strnlen16);
+static unsigned utf8_to_unichar(const char *utf8, size_t n, char32_t *c) {
+ char32_t unichar;
+ unsigned len;
+
+ assert(utf8);
+ assert(c);
+
+ if (!(utf8[0] & 0x80)) {
+ *c = utf8[0];
+ return 1;
+ } else if ((utf8[0] & 0xe0) == 0xc0) {
+ len = 2;
+ unichar = utf8[0] & 0x1f;
+ } else if ((utf8[0] & 0xf0) == 0xe0) {
+ len = 3;
+ unichar = utf8[0] & 0x0f;
+ } else if ((utf8[0] & 0xf8) == 0xf0) {
+ len = 4;
+ unichar = utf8[0] & 0x07;
+ } else if ((utf8[0] & 0xfc) == 0xf8) {
+ len = 5;
+ unichar = utf8[0] & 0x03;
+ } else if ((utf8[0] & 0xfe) == 0xfc) {
+ len = 6;
+ unichar = utf8[0] & 0x01;
+ } else {
+ *c = UINT32_MAX;
+ return 1;
+ }
+
+ if (len > n) {
+ *c = UINT32_MAX;
+ return len;
+ }
+
+ for (unsigned i = 1; i < len; i++) {
+ if ((utf8[i] & 0xc0) != 0x80) {
+ *c = UINT32_MAX;
+ return len;
+ }
+ unichar <<= 6;
+ unichar |= utf8[i] & 0x3f;
+ }
+
+ *c = unichar;
+ return len;
+}
+
+/* Convert UTF-8 to UCS-2, skipping any invalid or short byte sequences. */
+char16_t *xstrn8_to_16(const char *str8, size_t n) {
+ if (!str8 || n == 0)
+ return NULL;
+
+ size_t i = 0;
+ char16_t *str16 = xnew(char16_t, n + 1);
+
+ while (n > 0 && *str8 != '\0') {
+ char32_t unichar;
+
+ size_t utf8len = utf8_to_unichar(str8, n, &unichar);
+ str8 += utf8len;
+ n = LESS_BY(n, utf8len);
+
+ switch (unichar) {
+ case 0 ... 0xd7ffU:
+ case 0xe000U ... 0xffffU:
+ str16[i++] = unichar;
+ break;
+ }
+ }
+
+ str16[i] = '\0';
+ return str16;
+}
+
static bool efi_fnmatch_prefix(const char16_t *p, const char16_t *h, const char16_t **ret_p, const char16_t **ret_h) {
assert(p);
assert(h);
diff --git a/src/boot/efi/efi-string.h b/src/boot/efi/efi-string.h
index 1ebd5fd6b7..9b2a9ad1c5 100644
--- a/src/boot/efi/efi-string.h
+++ b/src/boot/efi/efi-string.h
@@ -99,6 +99,11 @@ static inline char16_t *xstrdup16(const char16_t *s) {
return xstrndup16(s, SIZE_MAX);
}
+char16_t *xstrn8_to_16(const char *str8, size_t n);
+static inline char16_t *xstr8_to_16(const char *str8) {
+ return xstrn8_to_16(str8, strlen8(str8));
+}
+
bool efi_fnmatch(const char16_t *pattern, const char16_t *haystack);
bool parse_number8(const char *s, uint64_t *ret_u, const char **ret_tail);
diff --git a/src/boot/efi/test-efi-string.c b/src/boot/efi/test-efi-string.c
index 2b2359fe5c..7b43e1d629 100644
--- a/src/boot/efi/test-efi-string.c
+++ b/src/boot/efi/test-efi-string.c
@@ -324,6 +324,33 @@ TEST(xstrdup16) {
free(s);
}
+TEST(xstrn8_to_16) {
+ char16_t *s = NULL;
+
+ assert_se(xstrn8_to_16(NULL, 1) == NULL);
+ assert_se(xstrn8_to_16("a", 0) == NULL);
+
+ assert_se(s = xstrn8_to_16("", 1));
+ assert_se(streq16(s, u""));
+ free(s);
+
+ assert_se(s = xstrn8_to_16("1", 1));
+ assert_se(streq16(s, u"1"));
+ free(s);
+
+ assert_se(s = xstr8_to_16("abcxyzABCXYZ09 .,-_#*!\"§$%&/()=?`~"));
+ assert_se(streq16(s, u"abcxyzABCXYZ09 .,-_#*!\"§$%&/()=?`~"));
+ free(s);
+
+ assert_se(s = xstr8_to_16("ÿⱿ𝇉 😺"));
+ assert_se(streq16(s, u"ÿⱿ "));
+ free(s);
+
+ assert_se(s = xstrn8_to_16("¶¶", 3));
+ assert_se(streq16(s, u"¶"));
+ free(s);
+}
+
#define TEST_FNMATCH_ONE(pattern, haystack, expect) \
({ \
assert_se(fnmatch(pattern, haystack, 0) == (expect ? 0 : FNM_NOMATCH)); \