Merge pull request #10173 from poettering/sd-boot-utf
the utf8.c changes from PR #9437 (i.e. the boot counting PR)
This commit is contained in:
Zbigniew Jędrzejewski-Szmek
GitHub
commit
b6de93096e
123
src/basic/utf8.c
123
src/basic/utf8.c
|
|
@ -61,12 +61,12 @@ static bool unichar_is_control(char32_t ch) {
|
|||
}
|
||||
|
||||
/* count of characters used to encode one unicode char */
|
||||
static int utf8_encoded_expected_len(const char *str) {
|
||||
unsigned char c;
|
||||
static size_t utf8_encoded_expected_len(const char *str) {
|
||||
uint8_t c;
|
||||
|
||||
assert(str);
|
||||
|
||||
c = (unsigned char) str[0];
|
||||
c = (uint8_t) str[0];
|
||||
if (c < 0x80)
|
||||
return 1;
|
||||
if ((c & 0xe0) == 0xc0)
|
||||
|
|
@ -86,7 +86,7 @@ static int utf8_encoded_expected_len(const char *str) {
|
|||
/* decode one unicode char */
|
||||
int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar) {
|
||||
char32_t unichar;
|
||||
int len, i;
|
||||
size_t len, i;
|
||||
|
||||
assert(str);
|
||||
|
||||
|
|
@ -118,6 +118,7 @@ int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar) {
|
|||
for (i = 1; i < len; i++) {
|
||||
if (((char32_t)str[i] & 0xc0) != 0x80)
|
||||
return -EINVAL;
|
||||
|
||||
unichar <<= 6;
|
||||
unichar |= (char32_t)str[i] & 0x3f;
|
||||
}
|
||||
|
|
@ -154,22 +155,23 @@ bool utf8_is_printable_newline(const char* str, size_t length, bool newline) {
|
|||
return true;
|
||||
}
|
||||
|
||||
const char *utf8_is_valid(const char *str) {
|
||||
const uint8_t *p;
|
||||
char *utf8_is_valid(const char *str) {
|
||||
const char *p;
|
||||
|
||||
assert(str);
|
||||
|
||||
for (p = (const uint8_t*) str; *p; ) {
|
||||
p = str;
|
||||
while (*p) {
|
||||
int len;
|
||||
|
||||
len = utf8_encoded_valid_unichar((const char *)p);
|
||||
len = utf8_encoded_valid_unichar(p);
|
||||
if (len < 0)
|
||||
return NULL;
|
||||
|
||||
p += len;
|
||||
}
|
||||
|
||||
return str;
|
||||
return (char*) str;
|
||||
}
|
||||
|
||||
char *utf8_escape_invalid(const char *str) {
|
||||
|
|
@ -312,18 +314,25 @@ size_t utf8_encode_unichar(char *out_utf8, char32_t g) {
|
|||
return 0;
|
||||
}
|
||||
|
||||
char *utf16_to_utf8(const void *s, size_t length) {
|
||||
char *utf16_to_utf8(const char16_t *s, size_t length /* bytes! */) {
|
||||
const uint8_t *f;
|
||||
char *r, *t;
|
||||
|
||||
r = new(char, (length * 4 + 1) / 2 + 1);
|
||||
assert(s);
|
||||
|
||||
/* Input length is in bytes, i.e. the shortest possible character takes 2 bytes. Each unicode character may
|
||||
* take up to 4 bytes in UTF-8. Let's also account for a trailing NUL byte. */
|
||||
if (length * 2 < length)
|
||||
return NULL; /* overflow */
|
||||
|
||||
r = new(char, length * 2 + 1);
|
||||
if (!r)
|
||||
return NULL;
|
||||
|
||||
f = s;
|
||||
f = (const uint8_t*) s;
|
||||
t = r;
|
||||
|
||||
while (f < (const uint8_t*) s + length) {
|
||||
while (f + 1 < (const uint8_t*) s + length) {
|
||||
char16_t w1, w2;
|
||||
|
||||
/* see RFC 2781 section 2.2 */
|
||||
|
|
@ -333,13 +342,13 @@ char *utf16_to_utf8(const void *s, size_t length) {
|
|||
|
||||
if (!utf16_is_surrogate(w1)) {
|
||||
t += utf8_encode_unichar(t, w1);
|
||||
|
||||
continue;
|
||||
}
|
||||
|
||||
if (utf16_is_trailing_surrogate(w1))
|
||||
continue;
|
||||
else if (f >= (const uint8_t*) s + length)
|
||||
continue; /* spurious trailing surrogate, ignore */
|
||||
|
||||
if (f + 1 >= (const uint8_t*) s + length)
|
||||
break;
|
||||
|
||||
w2 = f[1] << 8 | f[0];
|
||||
|
|
@ -347,7 +356,7 @@ char *utf16_to_utf8(const void *s, size_t length) {
|
|||
|
||||
if (!utf16_is_trailing_surrogate(w2)) {
|
||||
f -= 2;
|
||||
continue;
|
||||
continue; /* surrogate missing its trailing surrogate, ignore */
|
||||
}
|
||||
|
||||
t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2));
|
||||
|
|
@ -357,6 +366,79 @@ char *utf16_to_utf8(const void *s, size_t length) {
|
|||
return r;
|
||||
}
|
||||
|
||||
size_t utf16_encode_unichar(char16_t *out, char32_t c) {
|
||||
|
||||
/* Note that this encodes as little-endian. */
|
||||
|
||||
switch (c) {
|
||||
|
||||
case 0 ... 0xd7ffU:
|
||||
case 0xe000U ... 0xffffU:
|
||||
out[0] = htole16(c);
|
||||
return 1;
|
||||
|
||||
case 0x10000U ... 0x10ffffU:
|
||||
c -= 0x10000U;
|
||||
out[0] = htole16((c >> 10) + 0xd800U);
|
||||
out[1] = htole16((c & 0x3ffU) + 0xdc00U);
|
||||
return 2;
|
||||
|
||||
default: /* A surrogate (invalid) */
|
||||
return 0;
|
||||
}
|
||||
}
|
||||
|
||||
char16_t *utf8_to_utf16(const char *s, size_t length) {
|
||||
char16_t *n, *p;
|
||||
size_t i;
|
||||
int r;
|
||||
|
||||
assert(s);
|
||||
|
||||
n = new(char16_t, length + 1);
|
||||
if (!n)
|
||||
return NULL;
|
||||
|
||||
p = n;
|
||||
|
||||
for (i = 0; i < length;) {
|
||||
char32_t unichar;
|
||||
size_t e;
|
||||
|
||||
e = utf8_encoded_expected_len(s + i);
|
||||
if (e <= 1) /* Invalid and single byte characters are copied as they are */
|
||||
goto copy;
|
||||
|
||||
if (i + e > length) /* sequence longer than input buffer, then copy as-is */
|
||||
goto copy;
|
||||
|
||||
r = utf8_encoded_to_unichar(s + i, &unichar);
|
||||
if (r < 0) /* sequence invalid, then copy as-is */
|
||||
goto copy;
|
||||
|
||||
p += utf16_encode_unichar(p, unichar);
|
||||
i += e;
|
||||
continue;
|
||||
|
||||
copy:
|
||||
*(p++) = htole16(s[i++]);
|
||||
}
|
||||
|
||||
*p = 0;
|
||||
return n;
|
||||
}
|
||||
|
||||
size_t char16_strlen(const char16_t *s) {
|
||||
size_t n = 0;
|
||||
|
||||
assert(s);
|
||||
|
||||
while (*s != 0)
|
||||
n++, s++;
|
||||
|
||||
return n;
|
||||
}
|
||||
|
||||
/* expected size used to encode one unicode char */
|
||||
static int utf8_unichar_to_encoded_len(char32_t unichar) {
|
||||
|
||||
|
|
@ -376,8 +458,9 @@ static int utf8_unichar_to_encoded_len(char32_t unichar) {
|
|||
|
||||
/* validate one encoded unicode char and return its length */
|
||||
int utf8_encoded_valid_unichar(const char *str) {
|
||||
int len, i, r;
|
||||
char32_t unichar;
|
||||
size_t len, i;
|
||||
int r;
|
||||
|
||||
assert(str);
|
||||
|
||||
|
|
@ -399,14 +482,14 @@ int utf8_encoded_valid_unichar(const char *str) {
|
|||
return r;
|
||||
|
||||
/* check if encoded length matches encoded value */
|
||||
if (utf8_unichar_to_encoded_len(unichar) != len)
|
||||
if (utf8_unichar_to_encoded_len(unichar) != (int) len)
|
||||
return -EINVAL;
|
||||
|
||||
/* check if value has valid range */
|
||||
if (!unichar_is_valid(unichar))
|
||||
return -EINVAL;
|
||||
|
||||
return len;
|
||||
return (int) len;
|
||||
}
|
||||
|
||||
size_t utf8_n_codepoints(const char *str) {
|
||||
|
|
|
|||
|
|
@ -14,7 +14,7 @@
|
|||
|
||||
bool unichar_is_valid(char32_t c);
|
||||
|
||||
const char *utf8_is_valid(const char *s) _pure_;
|
||||
char *utf8_is_valid(const char *s) _pure_;
|
||||
char *ascii_is_valid(const char *s) _pure_;
|
||||
char *ascii_is_valid_n(const char *str, size_t len);
|
||||
|
||||
|
|
@ -25,21 +25,26 @@ char *utf8_escape_invalid(const char *s);
|
|||
char *utf8_escape_non_printable(const char *str);
|
||||
|
||||
size_t utf8_encode_unichar(char *out_utf8, char32_t g);
|
||||
char *utf16_to_utf8(const void *s, size_t length);
|
||||
size_t utf16_encode_unichar(char16_t *out, char32_t c);
|
||||
|
||||
char *utf16_to_utf8(const char16_t *s, size_t length /* bytes! */);
|
||||
char16_t *utf8_to_utf16(const char *s, size_t length);
|
||||
|
||||
size_t char16_strlen(const char16_t *s);
|
||||
|
||||
int utf8_encoded_valid_unichar(const char *str);
|
||||
int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar);
|
||||
|
||||
static inline bool utf16_is_surrogate(char16_t c) {
|
||||
return (0xd800 <= c && c <= 0xdfff);
|
||||
return c >= 0xd800U && c <= 0xdfffU;
|
||||
}
|
||||
|
||||
static inline bool utf16_is_trailing_surrogate(char16_t c) {
|
||||
return (0xdc00 <= c && c <= 0xdfff);
|
||||
return c >= 0xdc00U && c <= 0xdfffU;
|
||||
}
|
||||
|
||||
static inline char32_t utf16_surrogate_pair_to_unichar(char16_t lead, char16_t trail) {
|
||||
return ((lead - 0xd800) << 10) + (trail - 0xdc00) + 0x10000;
|
||||
return ((((char32_t) lead - 0xd800U) << 10) + ((char32_t) trail - 0xdc00U) + 0x10000U);
|
||||
}
|
||||
|
||||
size_t utf8_n_codepoints(const char *str);
|
||||
|
|
|
|||
|
|
@ -2,6 +2,7 @@
|
|||
|
||||
#include "alloc-util.h"
|
||||
#include "string-util.h"
|
||||
#include "strv.h"
|
||||
#include "utf8.h"
|
||||
#include "util.h"
|
||||
|
||||
|
|
@ -87,15 +88,25 @@ static void test_utf8_escaping_printable(void) {
|
|||
}
|
||||
|
||||
static void test_utf16_to_utf8(void) {
|
||||
char *a = NULL;
|
||||
const uint16_t utf16[] = { htole16('a'), htole16(0xd800), htole16('b'), htole16(0xdc00), htole16('c'), htole16(0xd801), htole16(0xdc37) };
|
||||
const char utf8[] = { 'a', 'b', 'c', 0xf0, 0x90, 0x90, 0xb7, 0 };
|
||||
const char16_t utf16[] = { htole16('a'), htole16(0xd800), htole16('b'), htole16(0xdc00), htole16('c'), htole16(0xd801), htole16(0xdc37) };
|
||||
static const char utf8[] = { 'a', 'b', 'c', 0xf0, 0x90, 0x90, 0xb7 };
|
||||
_cleanup_free_ char16_t *b = NULL;
|
||||
_cleanup_free_ char *a = NULL;
|
||||
|
||||
a = utf16_to_utf8(utf16, 14);
|
||||
/* Convert UTF-16 to UTF-8, filtering embedded bad chars */
|
||||
a = utf16_to_utf8(utf16, sizeof(utf16));
|
||||
assert_se(a);
|
||||
assert_se(streq(a, utf8));
|
||||
assert_se(memcmp(a, utf8, sizeof(utf8)) == 0);
|
||||
|
||||
/* Convert UTF-8 to UTF-16, and back */
|
||||
b = utf8_to_utf16(utf8, sizeof(utf8));
|
||||
assert_se(b);
|
||||
|
||||
free(a);
|
||||
a = utf16_to_utf8(b, char16_strlen(b) * 2);
|
||||
assert_se(a);
|
||||
assert_se(strlen(a) == sizeof(utf8));
|
||||
assert_se(memcmp(a, utf8, sizeof(utf8)) == 0);
|
||||
}
|
||||
|
||||
static void test_utf8_n_codepoints(void) {
|
||||
|
|
@ -116,6 +127,28 @@ static void test_utf8_console_width(void) {
|
|||
assert_se(utf8_console_width("\xF1") == (size_t) -1);
|
||||
}
|
||||
|
||||
static void test_utf8_to_utf16(void) {
|
||||
const char *p;
|
||||
|
||||
FOREACH_STRING(p,
|
||||
"abc",
|
||||
"zażółcić gęślą jaźń",
|
||||
"串",
|
||||
"",
|
||||
"…👊🔪💐…") {
|
||||
|
||||
_cleanup_free_ char16_t *a = NULL;
|
||||
_cleanup_free_ char *b = NULL;
|
||||
|
||||
a = utf8_to_utf16(p, strlen(p));
|
||||
assert_se(a);
|
||||
|
||||
b = utf16_to_utf8(a, char16_strlen(a) * 2);
|
||||
assert_se(b);
|
||||
assert_se(streq(p, b));
|
||||
}
|
||||
}
|
||||
|
||||
int main(int argc, char *argv[]) {
|
||||
test_utf8_is_valid();
|
||||
test_utf8_is_printable();
|
||||
|
|
@ -127,6 +160,7 @@ int main(int argc, char *argv[]) {
|
|||
test_utf16_to_utf8();
|
||||
test_utf8_n_codepoints();
|
||||
test_utf8_console_width();
|
||||
test_utf8_to_utf16();
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
|
|
|||
Loading…
Reference in New Issue