summaryrefslogtreecommitdiff
path: root/fs/udf/unicode.c
diff options
context:
space:
mode:
authorJan Kara <jack@suse.cz>2018-04-16 17:30:14 +0200
committerJan Kara <jack@suse.cz>2018-04-19 16:00:48 +0200
commitef2e18f1fa3958f8ee1a38acaebb6991d49dce18 (patch)
tree01f660cf8b584674b178200f89e4eb529306e2a0 /fs/udf/unicode.c
parentd504adc29142755edda4ef0f24ec81b7088564a4 (diff)
downloadlwn-ef2e18f1fa3958f8ee1a38acaebb6991d49dce18.tar.gz
lwn-ef2e18f1fa3958f8ee1a38acaebb6991d49dce18.zip
udf: Add support for encoding UTF-16 characters
Add support to store characters outside of Base Multilingual Plane of UTF-16 in CS0 encoding of UDF. Signed-off-by: Jan Kara <jack@suse.cz>
Diffstat (limited to 'fs/udf/unicode.c')
-rw-r--r--fs/udf/unicode.c79
1 files changed, 43 insertions, 36 deletions
diff --git a/fs/udf/unicode.c b/fs/udf/unicode.c
index 329be783f98a..616ffee441c5 100644
--- a/fs/udf/unicode.c
+++ b/fs/udf/unicode.c
@@ -28,9 +28,13 @@
#include "udf_sb.h"
+#define PLANE_SIZE 0x10000
#define UNICODE_MAX 0x10ffff
#define SURROGATE_MASK 0xfffff800
#define SURROGATE_PAIR 0x0000d800
+#define SURROGATE_LOW 0x00000400
+#define SURROGATE_CHAR_BITS 10
+#define SURROGATE_CHAR_MASK ((1 << SURROGATE_CHAR_BITS) - 1)
static int udf_uni2char_utf8(wchar_t uni,
unsigned char *out,
@@ -51,26 +55,6 @@ static int udf_uni2char_utf8(wchar_t uni,
return u_len;
}
-static int udf_char2uni_utf8(const unsigned char *in,
- int boundlen,
- wchar_t *uni)
-{
- int u_len;
- unicode_t c;
-
- u_len = utf8_to_utf32(in, boundlen, &c);
- if (u_len < 0) {
- *uni = '?';
- return -EINVAL;
- }
-
- if (c > MAX_WCHAR_T)
- *uni = '?';
- else
- *uni = c;
- return u_len;
-}
-
#define ILLEGAL_CHAR_MARK '_'
#define EXT_MARK '.'
#define CRC_MARK '#'
@@ -261,19 +245,17 @@ static int udf_name_to_CS0(struct super_block *sb,
{
int i, len;
unsigned int max_val;
- wchar_t uni_char;
int u_len, u_ch;
+ unicode_t uni_char;
int (*conv_f)(const unsigned char *, int, wchar_t *);
if (ocu_max_len <= 0)
return 0;
- if (UDF_QUERY_FLAG(sb, UDF_FLAG_UTF8)) {
- conv_f = udf_char2uni_utf8;
- } else if (UDF_QUERY_FLAG(sb, UDF_FLAG_NLS_MAP)) {
+ if (UDF_QUERY_FLAG(sb, UDF_FLAG_NLS_MAP))
conv_f = UDF_SB(sb)->s_nls_map->char2uni;
- } else
- BUG();
+ else
+ conv_f = NULL;
memset(ocu, 0, ocu_max_len);
ocu[0] = 8;
@@ -282,30 +264,55 @@ static int udf_name_to_CS0(struct super_block *sb,
try_again:
u_len = 1;
- for (i = 0; i < str_len; i++) {
+ for (i = 0; i < str_len; i += len) {
/* Name didn't fit? */
if (u_len + u_ch > ocu_max_len)
return 0;
- len = conv_f(&str_i[i], str_len - i, &uni_char);
- if (!len)
- continue;
+ if (conv_f) {
+ wchar_t wchar;
+
+ len = conv_f(&str_i[i], str_len - i, &wchar);
+ if (len > 0)
+ uni_char = wchar;
+ } else {
+ len = utf8_to_utf32(&str_i[i], str_len - i,
+ &uni_char);
+ }
/* Invalid character, deal with it */
- if (len < 0) {
+ if (len <= 0 || uni_char > UNICODE_MAX) {
len = 1;
uni_char = '?';
}
if (uni_char > max_val) {
- max_val = 0xffff;
- ocu[0] = 0x10;
- u_ch = 2;
- goto try_again;
+ unicode_t c;
+
+ if (max_val == 0xff) {
+ max_val = 0xffff;
+ ocu[0] = 0x10;
+ u_ch = 2;
+ goto try_again;
+ }
+ /*
+ * Use UTF-16 encoding for chars outside we
+ * cannot encode directly.
+ */
+ if (u_len + 2 * u_ch > ocu_max_len)
+ return 0;
+
+ uni_char -= PLANE_SIZE;
+ c = SURROGATE_PAIR |
+ ((uni_char >> SURROGATE_CHAR_BITS) &
+ SURROGATE_CHAR_MASK);
+ ocu[u_len++] = (uint8_t)(c >> 8);
+ ocu[u_len++] = (uint8_t)(c & 0xff);
+ uni_char = SURROGATE_PAIR | SURROGATE_LOW |
+ (uni_char & SURROGATE_CHAR_MASK);
}
if (max_val == 0xffff)
ocu[u_len++] = (uint8_t)(uni_char >> 8);
ocu[u_len++] = (uint8_t)(uni_char & 0xff);
- i += len - 1;
}
return u_len;