mirror of
https://github.com/tmux/tmux.git
synced 2024-12-12 17:38:48 +00:00
Be more strict about invalid UTF-8.
This commit is contained in:
parent
c56b81a2ce
commit
64333e3ef8
13
input.c
13
input.c
@ -446,11 +446,11 @@ const struct input_transition input_state_ground_table[] = {
|
||||
{ 0x1c, 0x1f, input_c0_dispatch, NULL },
|
||||
{ 0x20, 0x7e, input_print, NULL },
|
||||
{ 0x7f, 0x7f, NULL, NULL },
|
||||
{ 0x80, 0xc1, input_print, NULL },
|
||||
{ 0x80, 0xc1, NULL, NULL },
|
||||
{ 0xc2, 0xdf, input_utf8_open, &input_state_utf8_one },
|
||||
{ 0xe0, 0xef, input_utf8_open, &input_state_utf8_two },
|
||||
{ 0xf0, 0xf4, input_utf8_open, &input_state_utf8_three },
|
||||
{ 0xf5, 0xff, input_print, NULL },
|
||||
{ 0xf5, 0xff, NULL, NULL },
|
||||
|
||||
{ -1, -1, NULL, NULL }
|
||||
};
|
||||
@ -1923,7 +1923,8 @@ input_utf8_open(struct input_ctx *ictx)
|
||||
{
|
||||
struct utf8_data *ud = &ictx->utf8data;
|
||||
|
||||
utf8_open(ud, ictx->ch);
|
||||
if (!utf8_open(ud, ictx->ch))
|
||||
log_fatalx("UTF-8 open invalid %#hhx", ictx->ch);
|
||||
|
||||
log_debug("%s %hhu", __func__, ud->size);
|
||||
|
||||
@ -1936,7 +1937,8 @@ input_utf8_add(struct input_ctx *ictx)
|
||||
{
|
||||
struct utf8_data *ud = &ictx->utf8data;
|
||||
|
||||
utf8_append(ud, ictx->ch);
|
||||
if (utf8_append(ud, ictx->ch) != 1)
|
||||
log_fatalx("UTF-8 add invalid %#hhx", ictx->ch);
|
||||
|
||||
log_debug("%s", __func__);
|
||||
|
||||
@ -1949,7 +1951,8 @@ input_utf8_close(struct input_ctx *ictx)
|
||||
{
|
||||
struct utf8_data *ud = &ictx->utf8data;
|
||||
|
||||
utf8_append(ud, ictx->ch);
|
||||
if (utf8_append(ud, ictx->ch) != 0)
|
||||
log_fatalx("UTF-8 close invalid %#hhx", ictx->ch);
|
||||
|
||||
log_debug("%s %hhu '%*s' (width %hhu)", __func__, ud->size,
|
||||
(int)ud->size, ud->data, ud->width);
|
||||
|
@ -144,7 +144,7 @@ key_string_lookup_string(const char *string)
|
||||
static const char *other = "!#()+,-.0123456789:;<=>?'\r\t";
|
||||
key_code key;
|
||||
u_short u;
|
||||
int size;
|
||||
int size, more;
|
||||
key_code modifiers;
|
||||
struct utf8_data ud;
|
||||
u_int i;
|
||||
@ -177,7 +177,9 @@ key_string_lookup_string(const char *string)
|
||||
if (strlen(string) != ud.size)
|
||||
return (KEYC_NONE);
|
||||
for (i = 1; i < ud.size; i++)
|
||||
utf8_append(&ud, (u_char)string[i]);
|
||||
more = utf8_append(&ud, (u_char)string[i]);
|
||||
if (more != 0)
|
||||
return (KEYC_NONE);
|
||||
key = utf8_combine(&ud);
|
||||
return (key | modifiers);
|
||||
}
|
||||
|
@ -115,6 +115,7 @@ screen_write_strlen(const char *fmt, ...)
|
||||
struct utf8_data ud;
|
||||
u_char *ptr;
|
||||
size_t left, size = 0;
|
||||
int more;
|
||||
|
||||
va_start(ap, fmt);
|
||||
xvasprintf(&msg, fmt, ap);
|
||||
@ -128,11 +129,12 @@ screen_write_strlen(const char *fmt, ...)
|
||||
left = strlen(ptr);
|
||||
if (left < (size_t)ud.size - 1)
|
||||
break;
|
||||
while (utf8_append(&ud, *ptr))
|
||||
while ((more = utf8_append(&ud, *ptr)) == 1)
|
||||
ptr++;
|
||||
ptr++;
|
||||
|
||||
size += ud.width;
|
||||
if (more == 0)
|
||||
size += ud.width;
|
||||
} else {
|
||||
if (*ptr > 0x1f && *ptr < 0x7f)
|
||||
size++;
|
||||
@ -176,6 +178,7 @@ screen_write_vnputs(struct screen_write_ctx *ctx, ssize_t maxlen,
|
||||
struct utf8_data ud;
|
||||
u_char *ptr;
|
||||
size_t left, size = 0;
|
||||
int more;
|
||||
|
||||
xvasprintf(&msg, fmt, ap);
|
||||
|
||||
@ -187,22 +190,24 @@ screen_write_vnputs(struct screen_write_ctx *ctx, ssize_t maxlen,
|
||||
left = strlen(ptr);
|
||||
if (left < (size_t)ud.size - 1)
|
||||
break;
|
||||
while (utf8_append(&ud, *ptr))
|
||||
while ((more = utf8_append(&ud, *ptr)) == 1)
|
||||
ptr++;
|
||||
ptr++;
|
||||
|
||||
if (maxlen > 0 &&
|
||||
size + ud.width > (size_t) maxlen) {
|
||||
while (size < (size_t) maxlen) {
|
||||
screen_write_putc(ctx, gc, ' ');
|
||||
size++;
|
||||
if (more == 0) {
|
||||
if (maxlen > 0 &&
|
||||
size + ud.width > (size_t) maxlen) {
|
||||
while (size < (size_t) maxlen) {
|
||||
screen_write_putc(ctx, gc, ' ');
|
||||
size++;
|
||||
}
|
||||
break;
|
||||
}
|
||||
break;
|
||||
}
|
||||
size += ud.width;
|
||||
size += ud.width;
|
||||
|
||||
utf8_copy(&gc->data, &ud);
|
||||
screen_write_cell(ctx, gc);
|
||||
utf8_copy(&gc->data, &ud);
|
||||
screen_write_cell(ctx, gc);
|
||||
}
|
||||
} else {
|
||||
if (maxlen > 0 && size + 1 > (size_t) maxlen)
|
||||
break;
|
||||
@ -231,6 +236,7 @@ screen_write_cnputs(struct screen_write_ctx *ctx, ssize_t maxlen,
|
||||
char *msg;
|
||||
u_char *ptr, *last;
|
||||
size_t left, size = 0;
|
||||
int more;
|
||||
|
||||
va_start(ap, fmt);
|
||||
xvasprintf(&msg, fmt, ap);
|
||||
@ -260,22 +266,24 @@ screen_write_cnputs(struct screen_write_ctx *ctx, ssize_t maxlen,
|
||||
left = strlen(ptr);
|
||||
if (left < (size_t)ud.size - 1)
|
||||
break;
|
||||
while (utf8_append(&ud, *ptr))
|
||||
while ((more = utf8_append(&ud, *ptr)) == 1)
|
||||
ptr++;
|
||||
ptr++;
|
||||
|
||||
if (maxlen > 0 &&
|
||||
size + ud.width > (size_t) maxlen) {
|
||||
while (size < (size_t) maxlen) {
|
||||
screen_write_putc(ctx, gc, ' ');
|
||||
size++;
|
||||
if (more == 0) {
|
||||
if (maxlen > 0 &&
|
||||
size + ud.width > (size_t) maxlen) {
|
||||
while (size < (size_t) maxlen) {
|
||||
screen_write_putc(ctx, gc, ' ');
|
||||
size++;
|
||||
}
|
||||
break;
|
||||
}
|
||||
break;
|
||||
}
|
||||
size += ud.width;
|
||||
size += ud.width;
|
||||
|
||||
utf8_copy(&lgc.data, &ud);
|
||||
screen_write_cell(ctx, &lgc);
|
||||
utf8_copy(&lgc.data, &ud);
|
||||
screen_write_cell(ctx, &lgc);
|
||||
}
|
||||
} else {
|
||||
if (maxlen > 0 && size + 1 > (size_t) maxlen)
|
||||
break;
|
||||
|
2
tmux.h
2
tmux.h
@ -624,7 +624,7 @@ struct utf8_data {
|
||||
u_char have;
|
||||
u_char size;
|
||||
|
||||
u_char width;
|
||||
u_char width; /* 0xff if invalid */
|
||||
} __packed;
|
||||
|
||||
/* Grid attributes. */
|
||||
|
11
tty-keys.c
11
tty-keys.c
@ -472,7 +472,7 @@ tty_keys_next(struct tty *tty)
|
||||
const char *buf;
|
||||
size_t len, size;
|
||||
cc_t bspace;
|
||||
int delay, expired = 0;
|
||||
int delay, expired = 0, more;
|
||||
key_code key;
|
||||
struct utf8_data ud;
|
||||
u_int i;
|
||||
@ -547,7 +547,9 @@ first_key:
|
||||
goto partial_key;
|
||||
}
|
||||
for (i = 1; i < size; i++)
|
||||
utf8_append(&ud, (u_char)buf[i]);
|
||||
more = utf8_append(&ud, (u_char)buf[i]);
|
||||
if (more != 0)
|
||||
goto discard_key;
|
||||
key = utf8_combine(&ud);
|
||||
log_debug("UTF-8 key %.*s %#llx", (int)size, buf, key);
|
||||
goto complete_key;
|
||||
@ -653,6 +655,7 @@ tty_keys_mouse(struct tty *tty, const char *buf, size_t len, size_t *size)
|
||||
struct utf8_data ud;
|
||||
u_int i, value, x, y, b, sgr_b;
|
||||
u_char sgr_type, c;
|
||||
int more;
|
||||
|
||||
/*
|
||||
* Standard mouse sequences are \033[M followed by three characters
|
||||
@ -699,7 +702,9 @@ tty_keys_mouse(struct tty *tty, const char *buf, size_t len, size_t *size)
|
||||
(*size)++;
|
||||
if (len <= *size)
|
||||
return (1);
|
||||
utf8_append(&ud, buf[*size]);
|
||||
more = utf8_append(&ud, buf[*size]);
|
||||
if (more != 0)
|
||||
return (-1);
|
||||
value = utf8_combine(&ud);
|
||||
} else
|
||||
value = (u_char)buf[*size];
|
||||
|
42
utf8.c
42
utf8.c
@ -403,22 +403,26 @@ utf8_open(struct utf8_data *ud, u_char ch)
|
||||
/*
|
||||
* Append character to UTF-8, closing if finished.
|
||||
*
|
||||
* Returns 1 if more UTF-8 data to come, 0 if finished.
|
||||
* Returns 1 if more UTF-8 data to come, 0 if finished and valid, -1 if
|
||||
* finished and invalid.
|
||||
*/
|
||||
int
|
||||
utf8_append(struct utf8_data *ud, u_char ch)
|
||||
{
|
||||
/* XXX this should do validity checks too! */
|
||||
|
||||
if (ud->have >= ud->size)
|
||||
fatalx("UTF-8 character overflow");
|
||||
if (ud->size > sizeof ud->data)
|
||||
fatalx("UTF-8 character size too large");
|
||||
|
||||
if (ud->have != 0 && (ch & 0xc0) != 0x80)
|
||||
ud->width = 0xff;
|
||||
|
||||
ud->data[ud->have++] = ch;
|
||||
if (ud->have != ud->size)
|
||||
return (1);
|
||||
|
||||
if (ud->width == 0xff)
|
||||
return (-1);
|
||||
ud->width = utf8_width(utf8_combine(ud));
|
||||
return (0);
|
||||
}
|
||||
@ -556,15 +560,15 @@ utf8_strvis(char *dst, const char *src, size_t len, int flag)
|
||||
while (src < end) {
|
||||
if (utf8_open(&ud, *src)) {
|
||||
more = 1;
|
||||
while (++src < end && more)
|
||||
while (++src < end && more == 1)
|
||||
more = utf8_append(&ud, *src);
|
||||
if (!more) {
|
||||
if (more == 0) {
|
||||
/* UTF-8 character finished. */
|
||||
for (i = 0; i < ud.size; i++)
|
||||
*dst++ = ud.data[i];
|
||||
continue;
|
||||
} else if (ud.have > 0) {
|
||||
/* Not a complete UTF-8 character. */
|
||||
/* Not a complete, valid UTF-8 character. */
|
||||
src -= ud.have;
|
||||
}
|
||||
}
|
||||
@ -600,9 +604,9 @@ utf8_sanitize(const char *src)
|
||||
dst = xreallocarray(dst, n + 1, sizeof *dst);
|
||||
if (utf8_open(&ud, *src)) {
|
||||
more = 1;
|
||||
while (*++src != '\0' && more)
|
||||
while (*++src != '\0' && more == 1)
|
||||
more = utf8_append(&ud, *src);
|
||||
if (!more) {
|
||||
if (more != 1) {
|
||||
dst = xreallocarray(dst, n + ud.width,
|
||||
sizeof *dst);
|
||||
for (i = 0; i < ud.width; i++)
|
||||
@ -612,10 +616,8 @@ utf8_sanitize(const char *src)
|
||||
src -= ud.have;
|
||||
}
|
||||
if (*src > 0x1f && *src < 0x7f)
|
||||
dst[n] = *src;
|
||||
dst[n++] = *src;
|
||||
src++;
|
||||
|
||||
n++;
|
||||
}
|
||||
|
||||
dst = xreallocarray(dst, n + 1, sizeof *dst);
|
||||
@ -641,18 +643,19 @@ utf8_fromcstr(const char *src)
|
||||
dst = xreallocarray(dst, n + 1, sizeof *dst);
|
||||
if (utf8_open(&dst[n], *src)) {
|
||||
more = 1;
|
||||
while (*++src != '\0' && more)
|
||||
while (*++src != '\0' && more == 1)
|
||||
more = utf8_append(&dst[n], *src);
|
||||
if (!more) {
|
||||
if (more != 1) {
|
||||
n++;
|
||||
continue;
|
||||
}
|
||||
src -= dst[n].have;
|
||||
}
|
||||
utf8_set(&dst[n], *src);
|
||||
if (*src > 0x1f && *src < 0x7f) {
|
||||
utf8_set(&dst[n], *src);
|
||||
n++;
|
||||
}
|
||||
src++;
|
||||
|
||||
n++;
|
||||
}
|
||||
|
||||
dst = xreallocarray(dst, n + 1, sizeof *dst);
|
||||
@ -693,15 +696,16 @@ utf8_cstrwidth(const char *s)
|
||||
while (*s != '\0') {
|
||||
if (utf8_open(&tmp, *s)) {
|
||||
more = 1;
|
||||
while (*++s != '\0' && more)
|
||||
while (*++s != '\0' && more == 1)
|
||||
more = utf8_append(&tmp, *s);
|
||||
if (!more) {
|
||||
if (more != 1) {
|
||||
width += tmp.width;
|
||||
continue;
|
||||
}
|
||||
s -= tmp.have;
|
||||
}
|
||||
width++;
|
||||
if (*s > 0x1f && *s < 0x7f)
|
||||
width++;
|
||||
s++;
|
||||
}
|
||||
return (width);
|
||||
|
Loading…
Reference in New Issue
Block a user