vt: rearrange vc_translate_unicode

The code was too overcomplicated. Extract vc_sanitize_unicode to a
separate function and flatten the code. I believe the code is
straightforward now.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Link: https://lore.kernel.org/r/20200615074910.19267-13-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This commit is contained in:
Jiri Slaby 2020-06-15 09:48:45 +02:00 committed by Greg Kroah-Hartman
parent 694d8a487c
commit ede98d12b6
1 changed files with 67 additions and 60 deletions
drivers/tty/vt

View File

@ -2572,6 +2572,19 @@ static inline int vc_translate_ascii(const struct vc_data *vc, int c)
return c;
}
/**
* vc_sanitize_unicode -- Replace invalid Unicode code points with U+FFFD
* @c: the received character, or U+FFFD for invalid sequences.
*/
static inline int vc_sanitize_unicode(const int c)
{
if ((c >= 0xd800 && c <= 0xdfff) || c == 0xfffe || c == 0xffff)
return 0xfffd;
return c;
}
/**
* vc_translate_unicode -- Combine UTF-8 into Unicode in @vc_utf_char
*
@ -2586,69 +2599,63 @@ static int vc_translate_unicode(struct vc_data *vc, int c, bool *rescan)
0x001fffff, 0x03ffffff, 0x7fffffff
};
/* Continuation byte received */
if ((c & 0xc0) == 0x80) {
/* Continuation byte received */
if (vc->vc_utf_count) {
vc->vc_utf_char = (vc->vc_utf_char << 6) | (c & 0x3f);
vc->vc_npar++;
if (--vc->vc_utf_count) {
/* Still need some bytes */
return -1;
}
/* Got a whole character */
c = vc->vc_utf_char;
/* Reject overlong sequences */
if (c <= utf8_length_changes[vc->vc_npar - 1] ||
c > utf8_length_changes[vc->vc_npar])
return 0xfffd;
} else {
/* Unexpected continuation byte */
vc->vc_utf_count = 0;
/* Unexpected continuation byte? */
if (!vc->vc_utf_count)
return 0xfffd;
}
} else {
/* Single ASCII byte or first byte of a sequence received */
if (vc->vc_utf_count) {
/* Continuation byte expected */
*rescan = true;
vc->vc_utf_count = 0;
return 0xfffd;
} else if (c > 0x7f) {
/* First byte of a multibyte sequence received */
vc->vc_npar = 0;
if ((c & 0xe0) == 0xc0) {
vc->vc_utf_count = 1;
vc->vc_utf_char = (c & 0x1f);
} else if ((c & 0xf0) == 0xe0) {
vc->vc_utf_count = 2;
vc->vc_utf_char = (c & 0x0f);
} else if ((c & 0xf8) == 0xf0) {
vc->vc_utf_count = 3;
vc->vc_utf_char = (c & 0x07);
} else if ((c & 0xfc) == 0xf8) {
vc->vc_utf_count = 4;
vc->vc_utf_char = (c & 0x03);
} else if ((c & 0xfe) == 0xfc) {
vc->vc_utf_count = 5;
vc->vc_utf_char = (c & 0x01);
} else {
/* 254 and 255 are invalid */
return 0xfffd;
}
if (vc->vc_utf_count) {
/* Still need some bytes */
return -1;
}
}
/* Nothing to do if an ASCII byte was received */
}
/* End of UTF-8 decoding. */
/* c is the received character, or U+FFFD for invalid sequences. */
/* Replace invalid Unicode code points with U+FFFD too */
if ((c >= 0xd800 && c <= 0xdfff) || c == 0xfffe || c == 0xffff)
return 0xfffd;
return c;
vc->vc_utf_char = (vc->vc_utf_char << 6) | (c & 0x3f);
vc->vc_npar++;
if (--vc->vc_utf_count)
goto need_more_bytes;
/* Got a whole character */
c = vc->vc_utf_char;
/* Reject overlong sequences */
if (c <= utf8_length_changes[vc->vc_npar - 1] ||
c > utf8_length_changes[vc->vc_npar])
return 0xfffd;
return vc_sanitize_unicode(c);
}
/* Single ASCII byte or first byte of a sequence received */
if (vc->vc_utf_count) {
/* Continuation byte expected */
*rescan = true;
vc->vc_utf_count = 0;
return 0xfffd;
}
/* Nothing to do if an ASCII byte was received */
if (c <= 0x7f)
return c;
/* First byte of a multibyte sequence received */
vc->vc_npar = 0;
if ((c & 0xe0) == 0xc0) {
vc->vc_utf_count = 1;
vc->vc_utf_char = (c & 0x1f);
} else if ((c & 0xf0) == 0xe0) {
vc->vc_utf_count = 2;
vc->vc_utf_char = (c & 0x0f);
} else if ((c & 0xf8) == 0xf0) {
vc->vc_utf_count = 3;
vc->vc_utf_char = (c & 0x07);
} else if ((c & 0xfc) == 0xf8) {
vc->vc_utf_count = 4;
vc->vc_utf_char = (c & 0x03);
} else if ((c & 0xfe) == 0xfc) {
vc->vc_utf_count = 5;
vc->vc_utf_char = (c & 0x01);
} else {
/* 254 and 255 are invalid */
return 0xfffd;
}
need_more_bytes:
return -1;
}
/* acquires console_lock */