diff options
author | Jaegeuk Kim <jaegeuk@kernel.org> | 2015-12-14 14:29:41 -0800 |
---|---|---|
committer | Greg Wallace <greg@gregtwallace.com> | 2016-01-19 22:02:20 -0500 |
commit | cd610cd19355e080571691b06bdf79082af92299 (patch) | |
tree | 08ef814926b7ad31eddd7e3b715d66e689875794 /lib | |
parent | d17426964639cb440256dd70a1e3cdf36eebf019 (diff) | |
download | android_external_f2fs-tools-cd610cd19355e080571691b06bdf79082af92299.tar.gz android_external_f2fs-tools-cd610cd19355e080571691b06bdf79082af92299.tar.bz2 android_external_f2fs-tools-cd610cd19355e080571691b06bdf79082af92299.zip |
mkfs.f2fs: fix storing volume label correctly in utf16
This patch fixes to store volume label as utf16 correctly.
Many conversion codes are copied from exfat-tools.
Change-Id: Iee96b27a5bfc4938f285b4de49d32e01a181d17f
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Diffstat (limited to 'lib')
-rw-r--r-- | lib/libf2fs.c | 177 |
1 files changed, 168 insertions, 9 deletions
diff --git a/lib/libf2fs.c b/lib/libf2fs.c index e80cae0..cd496a1 100644 --- a/lib/libf2fs.c +++ b/lib/libf2fs.c @@ -25,18 +25,177 @@ #include <f2fs_fs.h> -void ASCIIToUNICODE(u_int16_t *out_buf, u_int8_t *in_buf) +/* + * UTF conversion codes are Copied from exfat tools. + */ +static const char *utf8_to_wchar(const char *input, wchar_t *wc, + size_t insize) +{ + if ((input[0] & 0x80) == 0 && insize >= 1) { + *wc = (wchar_t) input[0]; + return input + 1; + } + if ((input[0] & 0xe0) == 0xc0 && insize >= 2) { + *wc = (((wchar_t) input[0] & 0x1f) << 6) | + ((wchar_t) input[1] & 0x3f); + return input + 2; + } + if ((input[0] & 0xf0) == 0xe0 && insize >= 3) { + *wc = (((wchar_t) input[0] & 0x0f) << 12) | + (((wchar_t) input[1] & 0x3f) << 6) | + ((wchar_t) input[2] & 0x3f); + return input + 3; + } + if ((input[0] & 0xf8) == 0xf0 && insize >= 4) { + *wc = (((wchar_t) input[0] & 0x07) << 18) | + (((wchar_t) input[1] & 0x3f) << 12) | + (((wchar_t) input[2] & 0x3f) << 6) | + ((wchar_t) input[3] & 0x3f); + return input + 4; + } + if ((input[0] & 0xfc) == 0xf8 && insize >= 5) { + *wc = (((wchar_t) input[0] & 0x03) << 24) | + (((wchar_t) input[1] & 0x3f) << 18) | + (((wchar_t) input[2] & 0x3f) << 12) | + (((wchar_t) input[3] & 0x3f) << 6) | + ((wchar_t) input[4] & 0x3f); + return input + 5; + } + if ((input[0] & 0xfe) == 0xfc && insize >= 6) { + *wc = (((wchar_t) input[0] & 0x01) << 30) | + (((wchar_t) input[1] & 0x3f) << 24) | + (((wchar_t) input[2] & 0x3f) << 18) | + (((wchar_t) input[3] & 0x3f) << 12) | + (((wchar_t) input[4] & 0x3f) << 6) | + ((wchar_t) input[5] & 0x3f); + return input + 6; + } + return NULL; +} + +static u_int16_t *wchar_to_utf16(u_int16_t *output, wchar_t wc, size_t outsize) +{ + if (wc <= 0xffff) { + if (outsize == 0) + return NULL; + output[0] = cpu_to_le16(wc); + return output + 1; + } + if (outsize < 2) + return NULL; + wc -= 0x10000; + output[0] = cpu_to_le16(0xd800 | ((wc >> 10) & 0x3ff)); + output[1] = cpu_to_le16(0xdc00 | (wc & 0x3ff)); + return output + 2; +} + +int utf8_to_utf16(u_int16_t *output, const char *input, size_t outsize, + size_t insize) +{ + const char *inp = input; + u_int16_t *outp = output; + wchar_t wc; + + while (inp - input < insize && *inp) { + inp = utf8_to_wchar(inp, &wc, insize - (inp - input)); + if (inp == NULL) { + DBG(0, "illegal UTF-8 sequence\n"); + return -EILSEQ; + } + outp = wchar_to_utf16(outp, wc, outsize - (outp - output)); + if (outp == NULL) { + DBG(0, "name is too long\n"); + return -ENAMETOOLONG; + } + } + *outp = cpu_to_le16(0); + return 0; +} + +static const u_int16_t *utf16_to_wchar(const u_int16_t *input, wchar_t *wc, + size_t insize) +{ + if ((le16_to_cpu(input[0]) & 0xfc00) == 0xd800) { + if (insize < 2 || (le16_to_cpu(input[1]) & 0xfc00) != 0xdc00) + return NULL; + *wc = ((wchar_t) (le16_to_cpu(input[0]) & 0x3ff) << 10); + *wc |= (le16_to_cpu(input[1]) & 0x3ff); + *wc += 0x10000; + return input + 2; + } else { + *wc = le16_to_cpu(*input); + return input + 1; + } +} + +static char *wchar_to_utf8(char *output, wchar_t wc, size_t outsize) { - u_int8_t *pchTempPtr = in_buf; - u_int16_t *pwTempPtr = out_buf; + if (wc <= 0x7f) { + if (outsize < 1) + return NULL; + *output++ = (char) wc; + } else if (wc <= 0x7ff) { + if (outsize < 2) + return NULL; + *output++ = 0xc0 | (wc >> 6); + *output++ = 0x80 | (wc & 0x3f); + } else if (wc <= 0xffff) { + if (outsize < 3) + return NULL; + *output++ = 0xe0 | (wc >> 12); + *output++ = 0x80 | ((wc >> 6) & 0x3f); + *output++ = 0x80 | (wc & 0x3f); + } else if (wc <= 0x1fffff) { + if (outsize < 4) + return NULL; + *output++ = 0xf0 | (wc >> 18); + *output++ = 0x80 | ((wc >> 12) & 0x3f); + *output++ = 0x80 | ((wc >> 6) & 0x3f); + *output++ = 0x80 | (wc & 0x3f); + } else if (wc <= 0x3ffffff) { + if (outsize < 5) + return NULL; + *output++ = 0xf8 | (wc >> 24); + *output++ = 0x80 | ((wc >> 18) & 0x3f); + *output++ = 0x80 | ((wc >> 12) & 0x3f); + *output++ = 0x80 | ((wc >> 6) & 0x3f); + *output++ = 0x80 | (wc & 0x3f); + } else if (wc <= 0x7fffffff) { + if (outsize < 6) + return NULL; + *output++ = 0xfc | (wc >> 30); + *output++ = 0x80 | ((wc >> 24) & 0x3f); + *output++ = 0x80 | ((wc >> 18) & 0x3f); + *output++ = 0x80 | ((wc >> 12) & 0x3f); + *output++ = 0x80 | ((wc >> 6) & 0x3f); + *output++ = 0x80 | (wc & 0x3f); + } else + return NULL; + + return output; +} - while (*pchTempPtr != '\0') { - *pwTempPtr = (u_int16_t)*pchTempPtr; - pchTempPtr++; - pwTempPtr++; +int utf16_to_utf8(char *output, const u_int16_t *input, size_t outsize, + size_t insize) +{ + const u_int16_t *inp = input; + char *outp = output; + wchar_t wc; + + while (inp - input < insize && le16_to_cpu(*inp)) { + inp = utf16_to_wchar(inp, &wc, insize - (inp - input)); + if (inp == NULL) { + DBG(0, "illegal UTF-16 sequence\n"); + return -EILSEQ; + } + outp = wchar_to_utf8(outp, wc, outsize - (outp - output)); + if (outp == NULL) { + DBG(0, "name is too long\n"); + return -ENAMETOOLONG; + } } - *pwTempPtr = '\0'; - return; + *outp = '\0'; + return 0; } int log_base_2(u_int32_t num) |