/*
* Convert a multibyte character string argument for the %s format to a wide
* string representation. ``prec'' specifies the maximum number of bytes
* to output. If ``prec'' is greater than or equal to zero, we can't assume
* that the multibyte char. string ends in a null character.
*/
static wchar_t *
__mbsconv(char *mbsarg, int prec)
{
mbstate_t mbs;
wchar_t *convbuf, *wcp;
const char *p;
size_t insize, nchars, nconv;
if (mbsarg == NULL)
return (NULL);
/*
* Supplied argument is a multibyte string; convert it to wide
* characters first.
*/
if (prec >= 0) {
/*
* String is not guaranteed to be NUL-terminated. Find the
* number of characters to print.
*/
p = mbsarg;
insize = nchars = nconv = 0;
mbs = initial_mbs;
while (nchars != (size_t)prec) {
nconv = mbrlen(p, MB_CUR_MAX, &mbs);
if (nconv == 0 || nconv == (size_t)-1 ||
nconv == (size_t)-2)
break;
p += nconv;
nchars++;
insize += nconv;
}
if (nconv == (size_t)-1 || nconv == (size_t)-2)
return (NULL);
} else {
insize = strlen(mbsarg);
nconv = 0;
}
/*
* Allocate buffer for the result and perform the conversion,
* converting at most `size' bytes of the input multibyte string to
* wide characters for printing.
*/
convbuf = malloc((insize + 1) * sizeof(*convbuf));
if (convbuf == NULL)
return (NULL);
wcp = convbuf;
p = mbsarg;
mbs = initial_mbs;
while (insize != 0) {
nconv = mbrtowc(wcp, p, insize, &mbs);
if (nconv == 0 || nconv == (size_t)-1 || nconv == (size_t)-2)
break;
wcp++;
p += nconv;
insize -= nconv;
}
if (nconv == (size_t)-1 || nconv == (size_t)-2) {
free(convbuf);
return (NULL);
}
*wcp = L'\0';
return (convbuf);
}
TEST(wchar, mbrtowc) {
wchar_t out[8];
out[0] = 'x';
ASSERT_EQ(0U, mbrtowc(out, "hello", 0, NULL));
ASSERT_EQ('x', out[0]);
ASSERT_EQ(0U, mbrtowc(out, "hello", 0, NULL));
ASSERT_EQ(0U, mbrtowc(out, "", 0, NULL));
ASSERT_EQ(1U, mbrtowc(out, "hello", 1, NULL));
ASSERT_EQ(L'h', out[0]);
ASSERT_EQ(0U, mbrtowc(NULL, "hello", 0, NULL));
ASSERT_EQ(0U, mbrtowc(NULL, "", 0, NULL));
ASSERT_EQ(1U, mbrtowc(NULL, "hello", 1, NULL));
ASSERT_EQ(0U, mbrtowc(NULL, NULL, 0, NULL));
ASSERT_STREQ("C.UTF-8", setlocale(LC_CTYPE, "C.UTF-8"));
uselocale(LC_GLOBAL_LOCALE);
// 1-byte UTF-8.
ASSERT_EQ(1U, mbrtowc(out, "abcdef", 6, NULL));
ASSERT_EQ(L'a', out[0]);
// 2-byte UTF-8.
ASSERT_EQ(2U, mbrtowc(out, "\xc2\xa2" "cdef", 6, NULL));
ASSERT_EQ(static_cast<wchar_t>(0x00a2), out[0]);
// 3-byte UTF-8.
ASSERT_EQ(3U, mbrtowc(out, "\xe2\x82\xac" "def", 6, NULL));
ASSERT_EQ(static_cast<wchar_t>(0x20ac), out[0]);
// 4-byte UTF-8.
ASSERT_EQ(4U, mbrtowc(out, "\xf0\xa4\xad\xa2" "ef", 6, NULL));
ASSERT_EQ(static_cast<wchar_t>(0x24b62), out[0]);
#if defined(__BIONIC__) // glibc allows this.
// Illegal 5-byte UTF-8.
ASSERT_EQ(static_cast<size_t>(-1), mbrtowc(out, "\xf8\xa1\xa2\xa3\xa4" "f", 6, NULL));
ASSERT_EQ(EILSEQ, errno);
#endif
// Illegal over-long sequence.
ASSERT_EQ(static_cast<size_t>(-1), mbrtowc(out, "\xf0\x82\x82\xac" "ef", 6, NULL));
ASSERT_EQ(EILSEQ, errno);
}
const char *file_getbuffer(RMagic *ms) {
char *pbuf, *op, *np;
size_t psize, len;
if (ms->haderr)
return NULL;
if (ms->flags & R_MAGIC_RAW)
return ms->o.buf;
if (ms->o.buf == NULL) {
eprintf ("ms->o.buf = NULL\n");
return NULL;
}
/* * 4 is for octal representation, + 1 is for NUL */
len = strlen (ms->o.buf);
if (len > (SIZE_MAX - 1) / 4) {
file_oomem (ms, len);
return NULL;
}
psize = len * 4 + 1;
if ((pbuf = realloc (ms->o.pbuf, psize)) == NULL) {
file_oomem (ms, psize);
return NULL;
}
ms->o.pbuf = pbuf;
#if 1
//defined(HAVE_WCHAR_H) && defined(HAVE_MBRTOWC) && defined(HAVE_WCWIDTH)
{
mbstate_t state;
wchar_t nextchar;
int mb_conv = 1;
size_t bytesconsumed;
char *eop;
(void)memset(&state, 0, sizeof(mbstate_t));
np = ms->o.pbuf;
op = ms->o.buf;
eop = op + len;
while (op < eop) {
bytesconsumed = mbrtowc(&nextchar, op,
(size_t)(eop - op), &state);
if (bytesconsumed == (size_t)(-1) ||
bytesconsumed == (size_t)(-2)) {
mb_conv = 0;
break;
}
if (iswprint(nextchar)) {
(void)memcpy(np, op, bytesconsumed);
op += bytesconsumed;
np += bytesconsumed;
} else {
while (bytesconsumed-- > 0)
OCTALIFY(np, op);
}
}
*np = '\0';
/* Parsing succeeded as a multi-byte sequence */
if (mb_conv != 0)
return ms->o.pbuf;
}
#endif
for (np = ms->o.pbuf, op = ms->o.buf; *op; op++) {
if (isprint ((ut8)*op)) {
*np++ = *op;
} else {
OCTALIFY (np, op);
}
}
*np = '\0';
return ms->o.pbuf;
}
int
main(int argc, char *argv[])
{
FILE *fp;
int (*fcn)(FILE *, const char *);
int ch, rval;
size_t n;
setlocale(LC_ALL, "");
fcn = NULL;
dchar = '\t'; /* default delimiter is \t */
strcpy(dcharmb, "\t");
while ((ch = getopt(argc, argv, "b:c:d:f:snw")) != -1)
switch(ch) {
case 'b':
get_list(optarg);
bflag = 1;
break;
case 'c':
get_list(optarg);
cflag = 1;
break;
case 'd':
n = mbrtowc(&dchar, optarg, MB_LEN_MAX, NULL);
if (dchar == '\0' || n != strlen(optarg))
errx(1, "bad delimiter");
strcpy(dcharmb, optarg);
dflag = 1;
break;
case 'f':
get_list(optarg);
fflag = 1;
break;
case 's':
sflag = 1;
break;
case 'n':
nflag = 1;
break;
case 'w':
wflag = 1;
break;
case '?':
default:
usage();
}
argc -= optind;
argv += optind;
if (fflag) {
if (bflag || cflag || nflag || (wflag && dflag))
usage();
} else if (!(bflag || cflag) || dflag || sflag || wflag)
usage();
else if (!bflag && nflag)
usage();
if (fflag)
fcn = f_cut;
else if (cflag)
fcn = MB_CUR_MAX > 1 ? c_cut : b_cut;
else if (bflag)
fcn = nflag && MB_CUR_MAX > 1 ? b_n_cut : b_cut;
rval = 0;
if (*argv)
for (; *argv; ++argv) {
if (strcmp(*argv, "-") == 0)
rval |= fcn(stdin, "stdin");
else {
if (!(fp = fopen(*argv, "r"))) {
warn("%s", *argv);
rval = 1;
continue;
}
fcn(fp, *argv);
(void)fclose(fp);
}
}
else
rval = fcn(stdin, "stdin");
exit(rval);
}
void fold_text(const char *text,
int linelen,
int line_cb(void *arg, const char *start, int len),
void *arg)
{
const char *start, *end, *sep;
size_t sep_bytes, len;
int col, rc = 0;
mbstate_t ps;
/* start, end and sep are byte-positions in the string, and should always
* lie on the start of a multibyte sequence */
start = end = sep = text;
sep_bytes = 0;
col = 0;
len = strlen(text);
memset(&ps, 0, sizeof(ps));
while (!rc) {
size_t bytes;
wchar_t wc;
int width;
bytes = mbrtowc(&wc, end, len - (end - text), &ps);
assert(bytes != (size_t)-1);
/* we'll get a zero size for the nul terminator, (size_t) -2
* if we've reached the end of the buffer, or (size_t) -1 on
* error */
if (!bytes || bytes == (size_t) -2 || bytes == (size_t) -1) {
line_cb(arg, start, end - start);
break;
}
if (wc == L'\n') {
rc = line_cb(arg, start, end - start);
start = sep = end += bytes;
sep_bytes = 0;
col = 0;
continue;
}
width = wcwidth(wc);
/* we should have caught this in the !bytes check... */
if (width == 0) {
line_cb(arg, start, end - start);
break;
}
/* unprintable character? just add it to the current line */
if (width < 0) {
end += bytes;
continue;
}
col += width;
if (col > linelen) {
if (sep != start) {
/* split on a previous word boundary, if
* possible */
rc = line_cb(arg, start, sep - start);
end = sep + sep_bytes;
} else {
/* otherwise, break the word */
rc = line_cb(arg, start, end - start);
}
sep_bytes = 0;
start = sep = end;
col = 0;
} else {
/* record our last separator */
if (wc == L' ') {
sep = end;
sep_bytes = bytes;
}
end += bytes;
}
}
}
wchar_t input_common_readch(int timed)
{
if (! has_lookahead())
{
if (timed)
{
int count;
fd_set fds;
struct timeval tm=
{
0,
1000 * WAIT_ON_ESCAPE
}
;
FD_ZERO(&fds);
FD_SET(0, &fds);
count = select(1, &fds, 0, 0, &tm);
switch (count)
{
case 0:
return WEOF;
case -1:
return WEOF;
break;
default:
break;
}
}
wchar_t res;
mbstate_t state = {};
while (1)
{
wint_t b = readb();
char bb;
size_t sz;
if ((b >= R_NULL) && (b < R_NULL + 1000))
return b;
bb=b;
sz = mbrtowc(&res, &bb, 1, &state);
switch (sz)
{
case (size_t)(-1):
memset(&state, '\0', sizeof(state));
debug(2, L"Illegal input");
return R_NULL;
case (size_t)(-2):
break;
case 0:
return 0;
default:
return res;
}
}
}
else
{
if (!timed)
{
while (has_lookahead() && lookahead_top() == WEOF)
lookahead_pop();
if (! has_lookahead())
return input_common_readch(0);
}
return lookahead_pop();
}
}
/* Get a logical line */
static int get_line(register FILE *f, int *length)
{
int c;
char *p;
int column;
static int colflg;
#ifdef HAVE_WIDECHAR
size_t i;
wchar_t wc;
int wc_width;
mbstate_t state, state_bak; /* Current status of the stream. */
char mbc[MB_LEN_MAX]; /* Buffer for one multibyte char. */
size_t mblength; /* Byte length of multibyte char. */
size_t mbc_pos = 0; /* Position of the MBC. */
int use_mbc_buffer_flag = 0; /* If 1, mbc has data. */
int break_flag = 0; /* If 1, exit while(). */
long file_pos_bak = Ftell(f);
memset(&state, '\0', sizeof(mbstate_t));
#endif
prepare_line_buffer();
p = Line;
column = 0;
c = Getc(f);
if (colflg && c == '\n') {
Currline++;
c = Getc(f);
}
while (p < &Line[LineLen - 1]) {
#ifdef HAVE_WIDECHAR
if (fold_opt && use_mbc_buffer_flag && MB_CUR_MAX > 1) {
use_mbc_buffer_flag = 0;
state_bak = state;
mbc[mbc_pos++] = c;
process_mbc:
mblength = mbrtowc(&wc, mbc, mbc_pos, &state);
switch (mblength) {
case (size_t)-2: /* Incomplete multibyte character. */
use_mbc_buffer_flag = 1;
state = state_bak;
break;
case (size_t)-1: /* Invalid as a multibyte character. */
*p++ = mbc[0];
state = state_bak;
column++;
file_pos_bak++;
if (column >= Mcol) {
Fseek(f, file_pos_bak);
} else {
memmove(mbc, mbc + 1, --mbc_pos);
if (mbc_pos > 0) {
mbc[mbc_pos] = '\0';
goto process_mbc;
}
}
break;
default:
wc_width = wcwidth(wc);
if (column + wc_width > Mcol) {
Fseek(f, file_pos_bak);
break_flag = 1;
} else {
for (i = 0; p < &Line[LineLen - 1] &&
i < mbc_pos; i++)
*p++ = mbc[i];
if (wc_width > 0)
column += wc_width;
}
}
if (break_flag || column >= Mcol)
break;
c = Getc(f);
continue;
}
#endif /* HAVE_WIDECHAR */
if (c == EOF) {
if (p > Line) {
*p = '\0';
*length = p - Line;
return (column);
}
*length = p - Line;
return (EOF);
}
if (c == '\n') {
Currline++;
break;
}
*p++ = c;
#if 0
if (c == '\033') { /* ESC */
c = Getc(f);
while (c > ' ' && c < '0' && p < &Line[LineLen - 1]) {
*p++ = c;
c = Getc(f);
}
if (c >= '0' && c < '\177' && p < &Line[LineLen - 1]) {
*p++ = c;
c = Getc(f);
continue;
}
}
#endif /* 0 */
if (c == '\t') {
if (!hardtabs || (column < promptlen && !hard)) {
if (hardtabs && eraseln && !dumb) {
column = 1 + (column | 7);
putstring(eraseln);
promptlen = 0;
} else {
for (--p; p < &Line[LineLen - 1];) {
*p++ = ' ';
if ((++column & 7) == 0)
break;
}
if (column >= promptlen)
promptlen = 0;
}
} else
column = 1 + (column | 7);
} else if (c == '\b' && column > 0) {
column--;
} else if (c == '\r') {
int next = Getc(f);
if (next == '\n') {
p--;
Currline++;
break;
}
Ungetc(next, f);
column = 0;
} else if (c == '\f' && stop_opt) {
p[-1] = '^';
*p++ = 'L';
column += 2;
Pause++;
} else if (c == EOF) {
*length = p - Line;
return (column);
} else {
#ifdef HAVE_WIDECHAR
if (fold_opt && MB_CUR_MAX > 1) {
memset(mbc, '\0', MB_LEN_MAX);
mbc_pos = 0;
mbc[mbc_pos++] = c;
state_bak = state;
mblength = mbrtowc(&wc, mbc, mbc_pos, &state);
/* The value of mblength is always less than 2 here. */
switch (mblength) {
case (size_t)-2:
p--;
file_pos_bak = Ftell(f) - 1;
state = state_bak;
use_mbc_buffer_flag = 1;
break;
case (size_t)-1:
state = state_bak;
column++;
break;
default:
wc_width = wcwidth(wc);
if (wc_width > 0)
column += wc_width;
}
} else
#endif /* HAVE_WIDECHAR */
{
if (isprint(c))
column++;
}
}
if (column >= Mcol && fold_opt)
break;
#ifdef HAVE_WIDECHAR
if (use_mbc_buffer_flag == 0 && p >= &Line[LineLen - 1 - 4])
/* don't read another char if there is no space for
* whole multibyte sequence */
break;
#endif
c = Getc(f);
}
if (column >= Mcol && Mcol > 0) {
if (!Wrap) {
*p++ = '\n';
}
}
colflg = column == Mcol && fold_opt;
if (colflg && eatnl && Wrap) {
*p++ = '\n'; /* simulate normal wrap */
}
*length = p - Line;
*p = 0;
return (column);
}
long ILAnsiGetChars(const unsigned char *bytes, unsigned long byteCount,
unsigned short *chars, unsigned long charCount)
{
#ifdef IL_CONFIG_LATIN1
unsigned long len;
/* Check for enough space in the output buffer */
if(byteCount > charCount)
{
return -1;
}
/* Convert the bytes */
len = byteCount;
while(len > 0)
{
*chars++ = (unsigned short)(*bytes++);
--len;
}
return (long)byteCount;
#else
#if HAVE_MBRTOWC
/* Use the re-entrant function to perform the conversion */
mbstate_t state;
size_t chlen;
unsigned long len = 0;
wchar_t ch;
int wrlen;
ILMemZero(&state, sizeof(state));
mbrtowc((wchar_t *)0, (char *)0, 0, &state);
while(byteCount > 0)
{
chlen = mbrtowc(&ch, (char *)bytes, (size_t)byteCount, &state);
if(chlen == (size_t)(-1) || chlen == (size_t)(-2))
{
/* Invalid character */
++bytes;
--byteCount;
}
else if(chlen != 0)
{
/* Ordinary character */
wrlen = ILUTF16WriteChar((unsigned short *)0, (unsigned long)ch);
if(charCount < (unsigned long)wrlen)
{
return -1;
}
ILUTF16WriteChar(chars, (unsigned long)ch);
chars += wrlen;
len += wrlen;
bytes += chlen;
byteCount -= (unsigned long)chlen;
}
else
{
/* Embedded NUL character */
if(charCount <= 0)
{
return -1;
}
*chars++ = '\0';
++len;
++bytes;
--byteCount;
}
}
return (long)len;
#else
/* Use the non re-entrant function to perform the conversion
and just hope that the underlying libc takes care of the
thread-safety issues for us */
int chlen;
unsigned long len = 0;
wchar_t ch;
int wrlen;
mbtowc((wchar_t *)0, (char *)0, 0);
while(byteCount > 0)
{
chlen = mbtowc(&ch, (char *)bytes, (size_t)byteCount);
if(chlen > 0)
{
/* Ordinary character */
wrlen = ILUTF16WriteChar((unsigned short *)0, (unsigned long)ch);
if(charCount < (unsigned long)wrlen)
{
return -1;
}
ILUTF16WriteChar(chars, (unsigned long)ch);
chars += wrlen;
len += wrlen;
bytes += chlen;
byteCount -= (unsigned long)chlen;
}
else if(!chlen)
{
/* Embedded NUL character */
if(charCount <= 0)
{
return -1;
}
*chars++ = '\0';
++len;
++bytes;
--byteCount;
}
else
{
/* Invalid character */
++bytes;
--byteCount;
}
}
return (long)len;
#endif
#endif
}
int
main(int argc, char *argv[])
{
INPUT *F1, *F2;
int aflag, ch, cval, vflag;
char *end;
setlocale(LC_ALL, "");
F1 = &input1;
F2 = &input2;
aflag = vflag = 0;
obsolete(argv);
while ((ch = getopt(argc, argv, "\01a:e:j:1:2:o:t:v:")) != -1) {
switch (ch) {
case '\01': /* See comment in obsolete(). */
aflag = 1;
F1->unpair = F2->unpair = 1;
break;
case '1':
if ((F1->joinf = strtol(optarg, &end, 10)) < 1)
errx(1, "-1 option field number less than 1");
if (*end)
errx(1, "illegal field number -- %s", optarg);
--F1->joinf;
break;
case '2':
if ((F2->joinf = strtol(optarg, &end, 10)) < 1)
errx(1, "-2 option field number less than 1");
if (*end)
errx(1, "illegal field number -- %s", optarg);
--F2->joinf;
break;
case 'a':
aflag = 1;
switch(strtol(optarg, &end, 10)) {
case 1:
F1->unpair = 1;
break;
case 2:
F2->unpair = 1;
break;
default:
errx(1, "-a option file number not 1 or 2");
break;
}
if (*end)
errx(1, "illegal file number -- %s", optarg);
break;
case 'e':
empty = optarg;
break;
case 'j':
if ((F1->joinf = F2->joinf =
strtol(optarg, &end, 10)) < 1)
errx(1, "-j option field number less than 1");
if (*end)
errx(1, "illegal field number -- %s", optarg);
--F1->joinf;
--F2->joinf;
break;
case 'o':
fieldarg(optarg);
break;
case 't':
spans = 0;
if (mbrtowc(&tabchar[0], optarg, MB_LEN_MAX, NULL) !=
strlen(optarg))
errx(1, "illegal tab character specification");
tabchar[1] = L'\0';
break;
case 'v':
vflag = 1;
joinout = 0;
switch (strtol(optarg, &end, 10)) {
case 1:
F1->unpair = 1;
break;
case 2:
F2->unpair = 1;
break;
default:
errx(1, "-v option file number not 1 or 2");
break;
}
if (*end)
errx(1, "illegal file number -- %s", optarg);
break;
case '?':
default:
usage();
}
}
argc -= optind;
argv += optind;
if (aflag && vflag)
errx(1, "the -a and -v options are mutually exclusive");
if (argc != 2)
usage();
/* Open the files; "-" means stdin. */
if (!strcmp(*argv, "-"))
F1->fp = stdin;
else if ((F1->fp = fopen(*argv, "r")) == NULL)
err(1, "%s", *argv);
++argv;
if (!strcmp(*argv, "-"))
F2->fp = stdin;
else if ((F2->fp = fopen(*argv, "r")) == NULL)
err(1, "%s", *argv);
if (F1->fp == stdin && F2->fp == stdin)
errx(1, "only one input file may be stdin");
slurp(F1);
slurp(F2);
while (F1->setcnt && F2->setcnt) {
cval = cmp(F1->set, F1->joinf, F2->set, F2->joinf);
if (cval == 0) {
/* Oh joy, oh rapture, oh beauty divine! */
if (joinout)
joinlines(F1, F2);
slurp(F1);
slurp(F2);
} else if (cval < 0) {
/* File 1 takes the lead... */
if (F1->unpair)
joinlines(F1, NULL);
slurp(F1);
} else {
/* File 2 takes the lead... */
if (F2->unpair)
joinlines(F2, NULL);
slurp(F2);
}
}
/*
* Now that one of the files is used up, optionally output any
* remaining lines from the other file.
*/
if (F1->unpair)
while (F1->setcnt) {
joinlines(F1, NULL);
slurp(F1);
}
if (F2->unpair)
while (F2->setcnt) {
joinlines(F2, NULL);
slurp(F2);
}
exit(0);
}
/*
* __svfscanf_unlocked - non-MT-safe version of __svfscanf
*/
int
__svfscanf_unlocked(FILE *fp, const char *fmt0, va_list ap)
{
const u_char *fmt = (const u_char *)fmt0;
int c; /* character from format, or conversion */
size_t width; /* field width, or 0 */
char *p; /* points into all kinds of strings */
size_t n; /* handy size_t */
int flags; /* flags as defined above */
char *p0; /* saves original value of p when necessary */
int nassigned; /* number of fields assigned */
int nconversions; /* number of conversions */
int nread; /* number of characters consumed from fp */
int base; /* base argument to conversion function */
char ccltab[256]; /* character class table for %[...] */
char buf[BUF]; /* buffer for numeric and mb conversions */
wchar_t *wcp; /* handy wide-character pointer */
size_t nconv; /* length of multibyte sequence converted */
static const mbstate_t initial;
mbstate_t mbs;
/* `basefix' is used to avoid `if' tests in the integer scanner */
static const short basefix[17] =
{ 10, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 };
_DIAGASSERT(fp != NULL);
_DIAGASSERT(fmt0 != NULL);
_SET_ORIENTATION(fp, -1);
nassigned = 0;
nconversions = 0;
nread = 0;
base = 0;
for (;;) {
c = (unsigned char)*fmt++;
if (c == 0)
return (nassigned);
if (isspace(c)) {
while ((fp->_r > 0 || __srefill(fp) == 0) &&
isspace(*fp->_p))
nread++, fp->_r--, fp->_p++;
continue;
}
if (c != '%')
goto literal;
width = 0;
flags = 0;
/*
* switch on the format. continue if done;
* break once format type is derived.
*/
again: c = *fmt++;
switch (c) {
case '%':
SCANF_SKIP_SPACE();
literal:
if (fp->_r <= 0 && __srefill(fp))
goto input_failure;
if (*fp->_p != c)
goto match_failure;
fp->_r--, fp->_p++;
nread++;
continue;
case '*':
flags |= SUPPRESS;
goto again;
case 'j':
flags |= INTMAXT;
goto again;
case 'l':
if (flags & LONG) {
flags &= ~LONG;
flags |= LONGLONG;
} else
flags |= LONG;
goto again;
case 'q':
flags |= LONGLONG; /* not quite */
goto again;
case 't':
flags |= PTRDIFFT;
goto again;
case 'z':
flags |= SIZET;
goto again;
case 'L':
flags |= LONGDBL;
goto again;
case 'h':
if (flags & SHORT) {
flags &= ~SHORT;
flags |= SHORTSHORT;
} else
flags |= SHORT;
goto again;
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
width = width * 10 + c - '0';
goto again;
/*
* Conversions.
*/
case 'd':
c = CT_INT;
base = 10;
break;
case 'i':
c = CT_INT;
base = 0;
break;
case 'o':
c = CT_INT;
flags |= UNSIGNED;
base = 8;
break;
case 'u':
c = CT_INT;
flags |= UNSIGNED;
base = 10;
break;
case 'X':
case 'x':
flags |= PFXOK; /* enable 0x prefixing */
c = CT_INT;
flags |= UNSIGNED;
base = 16;
break;
#ifndef NO_FLOATING_POINT
case 'A': case 'E': case 'F': case 'G':
case 'a': case 'e': case 'f': case 'g':
c = CT_FLOAT;
break;
#endif
case 'S':
flags |= LONG;
/* FALLTHROUGH */
case 's':
c = CT_STRING;
break;
case '[':
fmt = __sccl(ccltab, fmt);
flags |= NOSKIP;
c = CT_CCL;
break;
case 'C':
flags |= LONG;
/* FALLTHROUGH */
case 'c':
flags |= NOSKIP;
c = CT_CHAR;
break;
case 'p': /* pointer format is like hex */
flags |= POINTER | PFXOK;
c = CT_INT; /* assumes sizeof(uintmax_t) */
flags |= UNSIGNED; /* >= sizeof(uintptr_t) */
base = 16;
break;
case 'n':
nconversions++;
if (flags & SUPPRESS) /* ??? */
continue;
if (flags & SHORTSHORT)
*va_arg(ap, char *) = nread;
else if (flags & SHORT)
*va_arg(ap, short *) = nread;
else if (flags & LONG)
*va_arg(ap, long *) = nread;
else if (flags & LONGLONG)
*va_arg(ap, long long *) = nread;
else if (flags & INTMAXT)
*va_arg(ap, intmax_t *) = nread;
else if (flags & SIZET)
*va_arg(ap, size_t *) = nread;
else if (flags & PTRDIFFT)
*va_arg(ap, ptrdiff_t *) = nread;
else
*va_arg(ap, int *) = nread;
continue;
default:
goto match_failure;
/*
* Disgusting backwards compatibility hack. XXX
*/
case '\0': /* compat */
return (EOF);
}
/*
* We have a conversion that requires input.
*/
if (fp->_r <= 0 && __srefill(fp))
goto input_failure;
/*
* Consume leading white space, except for formats
* that suppress this.
*/
if ((flags & NOSKIP) == 0) {
while (isspace(*fp->_p)) {
nread++;
if (--fp->_r > 0)
fp->_p++;
else if (__srefill(fp))
goto input_failure;
}
/*
* Note that there is at least one character in
* the buffer, so conversions that do not set NOSKIP
* ca no longer result in an input failure.
*/
}
/*
* Do the conversion.
*/
switch (c) {
case CT_CHAR:
/* scan arbitrary characters (sets NOSKIP) */
if (width == 0)
width = 1;
if (flags & LONG) {
if ((flags & SUPPRESS) == 0)
wcp = va_arg(ap, wchar_t *);
else
wcp = NULL;
n = 0;
while (width != 0) {
if (n == MB_CUR_MAX) {
fp->_flags |= __SERR;
goto input_failure;
}
buf[n++] = *fp->_p;
fp->_p++;
fp->_r--;
mbs = initial;
nconv = mbrtowc(wcp, buf, n, &mbs);
if (nconv == (size_t)-1) {
fp->_flags |= __SERR;
goto input_failure;
}
if (nconv == 0 && !(flags & SUPPRESS))
*wcp = L'\0';
if (nconv != (size_t)-2) {
nread += n;
width--;
if (!(flags & SUPPRESS))
wcp++;
n = 0;
}
if (fp->_r <= 0 && __srefill(fp)) {
if (n != 0) {
fp->_flags |= __SERR;
goto input_failure;
}
break;
}
}
if (!(flags & SUPPRESS))
nassigned++;
} else if (flags & SUPPRESS) {
size_t sum = 0;
for (;;) {
if ((n = fp->_r) < width) {
sum += n;
width -= n;
fp->_p += n;
if (__srefill(fp)) {
if (sum == 0)
goto input_failure;
break;
}
} else {
sum += width;
fp->_r -= width;
fp->_p += width;
break;
}
}
nread += sum;
} else {
size_t r = fread(va_arg(ap, char *), 1,
width, fp);
if (r == 0)
goto input_failure;
nread += r;
nassigned++;
}
nconversions++;
break;
case CT_CCL:
/* scan a (nonempty) character class (sets NOSKIP) */
if (width == 0)
width = (size_t)~0; /* `infinity' */
/* take only those things in the class */
if (flags & LONG) {
wchar_t twc;
int nchars;
if ((flags & SUPPRESS) == 0)
wcp = va_arg(ap, wchar_t *);
else
unsigned long ILAnsiGetCharCount(const unsigned char *bytes,
unsigned long count)
{
#ifdef IL_CONFIG_LATIN1
return count;
#else
#if HAVE_MBRTOWC
/* Use the re-entrant function to perform the conversion */
mbstate_t state;
size_t chlen;
unsigned long len = 0;
wchar_t ch;
ILMemZero(&state, sizeof(state));
mbrtowc((wchar_t *)0, (char *)0, 0, &state);
while(count > 0)
{
chlen = mbrtowc(&ch, (char *)bytes, (size_t)count, &state);
if(chlen == (size_t)(-1) || chlen == (size_t)(-2))
{
/* Invalid character */
++bytes;
--count;
}
else if(chlen != 0)
{
/* Ordinary character */
len += ILUTF16WriteChar((unsigned short *)0, (unsigned long)ch);
bytes += chlen;
count -= (unsigned long)chlen;
}
else
{
/* Embedded NUL character */
++len;
++bytes;
--count;
}
}
return len;
#else
/* Use the non re-entrant function to perform the conversion
and just hope that the underlying libc takes care of the
thread-safety issues for us */
int chlen;
unsigned long len = 0;
wchar_t ch;
mbtowc((wchar_t *)0, (char *)0, 0);
while(count > 0)
{
chlen = mbtowc(&ch, (char *)bytes, (size_t)count);
if(chlen > 0)
{
/* Ordinary character */
len += ILUTF16WriteChar((unsigned short *)0, (unsigned long)ch);
bytes += chlen;
count -= (unsigned long)chlen;
}
else if(!chlen)
{
/* Embedded NUL character */
++len;
++bytes;
--count;
}
else
{
/* Invalid character */
++bytes;
--count;
}
}
return len;
#endif
#endif
}
size_t
rpl_mbrtowc (wchar_t *pwc, const char *s, size_t n, mbstate_t *ps)
{
# if MBRTOWC_NULL_ARG_BUG || MBRTOWC_RETVAL_BUG
if (s == NULL)
{
pwc = NULL;
s = "";
n = 1;
}
# endif
# if MBRTOWC_RETVAL_BUG
{
static mbstate_t internal_state;
/* Override mbrtowc's internal state. We can not call mbsinit() on the
hidden internal state, but we can call it on our variable. */
if (ps == NULL)
ps = &internal_state;
if (!mbsinit (ps))
{
/* Parse the rest of the multibyte character byte for byte. */
size_t count = 0;
for (; n > 0; s++, n--)
{
wchar_t wc;
size_t ret = mbrtowc (&wc, s, 1, ps);
if (ret == (size_t)(-1))
return (size_t)(-1);
count++;
if (ret != (size_t)(-2))
{
/* The multibyte character has been completed. */
if (pwc != NULL)
*pwc = wc;
return (wc == 0 ? 0 : count);
}
}
return (size_t)(-2);
}
}
# endif
# if MBRTOWC_NUL_RETVAL_BUG
{
wchar_t wc;
size_t ret = mbrtowc (&wc, s, n, ps);
if (ret != (size_t)(-1) && ret != (size_t)(-2))
{
if (pwc != NULL)
*pwc = wc;
if (wc == 0)
ret = 0;
}
return ret;
}
# else
return mbrtowc (pwc, s, n, ps);
# endif
}
int main()
{
// mbstate_t comes from the underlying C library; it is defined (in C99) as:
// a complete object type other than an array type that can hold the conversion
// state information necessary to convert between sequences of multibyte
// characters and wide characters
#if defined(__clang__)
#pragma clang diagnostic push
#pragma clang diagnostic ignored "-Wmissing-braces"
#endif
mbstate_t mb = {0};
#if defined(__clang__)
#pragma clang diagnostic pop
#endif
size_t s = 0;
tm *tm = 0;
wint_t w = 0;
::FILE* fp = 0;
#ifdef __APPLE__
__darwin_va_list va;
#else
__builtin_va_list va;
#endif
char* ns = 0;
wchar_t* ws = 0;
static_assert((std::is_same<decltype(fwprintf(fp, L"")), int>::value), "");
static_assert((std::is_same<decltype(fwscanf(fp, L"")), int>::value), "");
static_assert((std::is_same<decltype(swprintf(ws, s, L"")), int>::value), "");
static_assert((std::is_same<decltype(swscanf(L"", L"")), int>::value), "");
static_assert((std::is_same<decltype(vfwprintf(fp, L"", va)), int>::value), "");
static_assert((std::is_same<decltype(vfwscanf(fp, L"", va)), int>::value), "");
static_assert((std::is_same<decltype(vswprintf(ws, s, L"", va)), int>::value), "");
static_assert((std::is_same<decltype(vswscanf(L"", L"", va)), int>::value), "");
static_assert((std::is_same<decltype(fgetwc(fp)), wint_t>::value), "");
static_assert((std::is_same<decltype(fgetws(ws, 0, fp)), wchar_t*>::value), "");
static_assert((std::is_same<decltype(fputwc(L' ', fp)), wint_t>::value), "");
static_assert((std::is_same<decltype(fputws(L"", fp)), int>::value), "");
static_assert((std::is_same<decltype(fwide(fp, 0)), int>::value), "");
static_assert((std::is_same<decltype(getwc(fp)), wint_t>::value), "");
static_assert((std::is_same<decltype(putwc(L' ', fp)), wint_t>::value), "");
static_assert((std::is_same<decltype(ungetwc(L' ', fp)), wint_t>::value), "");
static_assert((std::is_same<decltype(wcstod(L"", (wchar_t**)0)), double>::value), "");
static_assert((std::is_same<decltype(wcstof(L"", (wchar_t**)0)), float>::value), "");
static_assert((std::is_same<decltype(wcstold(L"", (wchar_t**)0)), long double>::value), "");
static_assert((std::is_same<decltype(wcstol(L"", (wchar_t**)0, 0)), long>::value), "");
static_assert((std::is_same<decltype(wcstoll(L"", (wchar_t**)0, 0)), long long>::value), "");
static_assert((std::is_same<decltype(wcstoul(L"", (wchar_t**)0, 0)), unsigned long>::value), "");
static_assert((std::is_same<decltype(wcstoull(L"", (wchar_t**)0, 0)), unsigned long long>::value), "");
static_assert((std::is_same<decltype(wcscpy(ws, L"")), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcsncpy(ws, L"", s)), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcscat(ws, L"")), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcsncat(ws, L"", s)), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcscmp(L"", L"")), int>::value), "");
static_assert((std::is_same<decltype(wcscoll(L"", L"")), int>::value), "");
static_assert((std::is_same<decltype(wcsncmp(L"", L"", s)), int>::value), "");
static_assert((std::is_same<decltype(wcsxfrm(ws, L"", s)), size_t>::value), "");
static_assert((std::is_same<decltype(wcschr((wchar_t*)0, L' ')), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcscspn(L"", L"")), size_t>::value), "");
static_assert((std::is_same<decltype(wcslen(L"")), size_t>::value), "");
static_assert((std::is_same<decltype(wcspbrk((wchar_t*)0, L"")), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcsrchr((wchar_t*)0, L' ')), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcsspn(L"", L"")), size_t>::value), "");
static_assert((std::is_same<decltype(wcsstr((wchar_t*)0, L"")), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcstok(ws, L"", (wchar_t**)0)), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wmemchr((wchar_t*)0, L' ', s)), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wmemcmp(L"", L"", s)), int>::value), "");
static_assert((std::is_same<decltype(wmemcpy(ws, L"", s)), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wmemmove(ws, L"", s)), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wmemset(ws, L' ', s)), wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcsftime(ws, s, L"", tm)), size_t>::value), "");
static_assert((std::is_same<decltype(btowc(0)), wint_t>::value), "");
static_assert((std::is_same<decltype(wctob(w)), int>::value), "");
static_assert((std::is_same<decltype(mbsinit(&mb)), int>::value), "");
static_assert((std::is_same<decltype(mbrlen("", s, &mb)), size_t>::value), "");
static_assert((std::is_same<decltype(mbrtowc(ws, "", s, &mb)), size_t>::value), "");
static_assert((std::is_same<decltype(wcrtomb(ns, L' ', &mb)), size_t>::value), "");
static_assert((std::is_same<decltype(mbsrtowcs(ws, (const char**)0, s, &mb)), size_t>::value), "");
static_assert((std::is_same<decltype(wcsrtombs(ns, (const wchar_t**)0, s, &mb)), size_t>::value), "");
// These tests fail on systems whose C library doesn't provide a correct overload
// set for wcschr, wcspbrk, wcsrchr, wcsstr, and wmemchr, unless the compiler is
// a suitably recent version of Clang.
#if !defined(__APPLE__) || defined(_LIBCPP_PREFERRED_OVERLOAD)
static_assert((std::is_same<decltype(wcschr((const wchar_t*)0, L' ')), const wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcspbrk((const wchar_t*)0, L"")), const wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcsrchr((const wchar_t*)0, L' ')), const wchar_t*>::value), "");
static_assert((std::is_same<decltype(wcsstr((const wchar_t*)0, L"")), const wchar_t*>::value), "");
static_assert((std::is_same<decltype(wmemchr((const wchar_t*)0, L' ', s)), const wchar_t*>::value), "");
#endif
#ifndef _LIBCPP_HAS_NO_STDIN
static_assert((std::is_same<decltype(getwchar()), wint_t>::value), "");
static_assert((std::is_same<decltype(vwscanf(L"", va)), int>::value), "");
static_assert((std::is_same<decltype(wscanf(L"")), int>::value), "");
#endif
#ifndef _LIBCPP_HAS_NO_STDOUT
static_assert((std::is_same<decltype(putwchar(L' ')), wint_t>::value), "");
static_assert((std::is_same<decltype(vwprintf(L"", va)), int>::value), "");
static_assert((std::is_same<decltype(wprintf(L"")), int>::value), "");
#endif
}
int
main(int argc, char *argv[])
{
struct stat sb;
int ch, fd, match;
wchar_t termchar;
unsigned char *back, *front;
unsigned const char *file;
wchar_t *key;
(void) setlocale(LC_CTYPE, "");
file = _path_words;
termchar = L'\0';
while ((ch = getopt(argc, argv, "dft:")) != -1)
switch(ch) {
case 'd':
dflag = 1;
break;
case 'f':
fflag = 1;
break;
case 't':
if (mbrtowc(&termchar, optarg, MB_LEN_MAX, NULL) !=
strlen(optarg))
errx(2, "invalid termination character");
break;
case '?':
default:
usage();
}
argc -= optind;
argv += optind;
if (argc == 0)
usage();
if (argc == 1) /* But set -df by default. */
dflag = fflag = 1;
key = prepkey(*argv++, termchar);
if (argc >= 2)
file = *argv++;
match = 1;
do {
if ((fd = open(file, O_RDONLY, 0)) < 0 || fstat(fd, &sb))
err(2, "%s", file);
if ((uintmax_t)sb.st_size > (uintmax_t)SIZE_T_MAX)
errx(2, "%s: %s", file, strerror(EFBIG));
if (sb.st_size == 0) {
close(fd);
continue;
}
if ((front = mmap(NULL, (size_t)sb.st_size, PROT_READ, MAP_SHARED, fd, (off_t)0)) == MAP_FAILED)
err(2, "%s", file);
back = front + sb.st_size;
match *= (look(key, front, back));
close(fd);
} while (argc-- > 2 && (file = *argv++));
exit(match);
}
static void
h_ctype2(const struct test *t, bool use_mbstate)
{
mbstate_t *stp;
mbstate_t st;
char buf[SIZE];
char *str;
size_t n;
ATF_REQUIRE_STREQ(setlocale(LC_ALL, "C"), "C");
#if defined(__NetBSD__)
ATF_REQUIRE(setlocale(LC_CTYPE, t->locale) != NULL);
#else
if (setlocale(LC_CTYPE, t->locale) == NULL) {
fprintf(stderr, "Locale %s not found.\n", t->locale);
return;
}
#endif
(void)strvis(buf, t->data, VIS_WHITE | VIS_OCTAL);
(void)printf("Checking string: \"%s\"\n", buf);
ATF_REQUIRE((str = setlocale(LC_ALL, NULL)) != NULL);
(void)printf("Using locale: %s\n", str);
(void)printf("Using mbstate: %s\n", use_mbstate ? "yes" : "no");
(void)memset(&st, 0, sizeof(st));
// mbrtowc(0, 0, 0, &st); /* XXX for ISO2022-JP */
stp = use_mbstate ? &st : 0;
for (n = 9; n > 0; n--) {
const char *src = t->data;
wchar_t dst;
size_t nchar = 0;
int width = 0;
ATF_REQUIRE(mbsinit(stp) != 0);
for (;;) {
size_t rv = mbrtowc(&dst, src, n, stp);
if (rv == 0)
break;
if (rv == (size_t)-2) {
src += n;
width += n;
continue;
}
if (rv == (size_t)-1) {
ATF_REQUIRE_EQ(errno, EILSEQ);
atf_tc_fail("Invalid sequence");
/* NOTREACHED */
}
width += rv;
src += rv;
if (dst != t->wchars[nchar] ||
width != t->widths[nchar]) {
(void)printf("At position %zd:\n", nchar);
(void)printf(" expected: 0x%04X (%u)\n",
t->wchars[nchar], t->widths[nchar]);
(void)printf(" got : 0x%04X (%u)\n",
dst, width);
atf_tc_fail("Test failed");
}
nchar++;
width = 0;
}
ATF_REQUIRE_EQ_MSG(dst, 0, "Incorrect terminating character: "
"0x%04X (expected: 0x00)", dst);
ATF_REQUIRE_EQ_MSG(nchar, t->length, "Incorrect length: "
"%zd (expected: %zd)", nchar, t->length);
}
{
wchar_t wbuf[SIZE];
size_t rv;
char const *src = t->data;
int i;
(void)memset(wbuf, 0xFF, sizeof(wbuf));
rv = mbsrtowcs(wbuf, &src, SIZE, stp);
ATF_REQUIRE_EQ_MSG(rv, t->length, "Incorrect length: %zd "
"(expected: %zd)", rv, t->length);
ATF_REQUIRE_EQ(src, NULL);
for (i = 0; wbuf[i] != 0; ++i) {
if (wbuf[i] == t->wchars[i])
continue;
(void)printf("At position %d:\n", i);
(void)printf(" expected: 0x%04X\n", t->wchars[i]);
(void)printf(" got : 0x%04X\n", wbuf[i]);
atf_tc_fail("Test failed");
}
ATF_REQUIRE_EQ_MSG((size_t)i, t->length, "Incorrect length: "
"%d (expected: %zd)", i, t->length);
}
(void)printf("Ok.\n");
}
/* Decode an item via the fish 1.x format. Adapted from fish 1.x's item_get(). */
history_item_t history_t::decode_item_fish_1_x(const char *begin, size_t length) {
const char *end = begin + length;
const char *pos=begin;
bool was_backslash = 0;
wcstring out;
bool first_char = true;
bool timestamp_mode = false;
time_t timestamp = 0;
while( 1 )
{
wchar_t c;
mbstate_t state;
size_t res;
memset( &state, 0, sizeof(state) );
res = mbrtowc( &c, pos, end-pos, &state );
if( res == (size_t)-1 )
{
pos++;
continue;
}
else if( res == (size_t)-2 )
{
break;
}
else if( res == (size_t)0 )
{
pos++;
continue;
}
pos += res;
if( c == L'\n' )
{
if( timestamp_mode )
{
const wchar_t *time_string = out.c_str();
while( *time_string && !iswdigit(*time_string))
time_string++;
errno=0;
if( *time_string )
{
time_t tm;
wchar_t *end;
errno = 0;
tm = (time_t)wcstol( time_string, &end, 10 );
if( tm && !errno && !*end )
{
timestamp = tm;
}
}
out.clear();
timestamp_mode = false;
continue;
}
if( !was_backslash )
break;
}
if( first_char )
{
if( c == L'#' )
timestamp_mode = true;
}
first_char = false;
out.push_back(c);
was_backslash = ( (c == L'\\') && !was_backslash);
}
out = history_unescape_newlines_fish_1_x(out);
return history_item_t(out, timestamp);
}
wchar_t *str2wcs_internal( const char *in, wchar_t *out )
{
size_t res=0;
int in_pos=0;
int out_pos = 0;
mbstate_t state;
size_t len;
CHECK( in, 0 );
CHECK( out, 0 );
len = strlen(in);
memset( &state, 0, sizeof(state) );
while( in[in_pos] )
{
res = mbrtowc( &out[out_pos], &in[in_pos], len-in_pos, &state );
if( ( ( out[out_pos] >= ENCODE_DIRECT_BASE) &&
( out[out_pos] < ENCODE_DIRECT_BASE+256)) ||
( out[out_pos] == INTERNAL_SEPARATOR ) )
{
out[out_pos] = ENCODE_DIRECT_BASE + (unsigned char)in[in_pos];
in_pos++;
memset( &state, 0, sizeof(state) );
out_pos++;
}
else
{
switch( res )
{
case (size_t)(-2):
case (size_t)(-1):
{
out[out_pos] = ENCODE_DIRECT_BASE + (unsigned char)in[in_pos];
in_pos++;
memset( &state, 0, sizeof(state) );
break;
}
case 0:
{
return out;
}
default:
{
in_pos += res;
break;
}
}
out_pos++;
}
}
out[out_pos] = 0;
return out;
}
mod_export ZLE_INT_T
getrestchar(int inchar)
{
char c = inchar;
wchar_t outchar;
int timeout;
static mbstate_t mbs;
/*
* We are guaranteed to set a valid wide last character,
* although it may be WEOF (which is technically not
* a wide character at all...)
*/
lastchar_wide_valid = 1;
if (inchar == EOF) {
/* End of input, so reset the shift state. */
memset(&mbs, 0, sizeof mbs);
return lastchar_wide = WEOF;
}
/*
* Return may be zero if we have a NULL; handle this like
* any other character.
*/
while (1) {
size_t cnt = mbrtowc(&outchar, &c, 1, &mbs);
if (cnt == MB_INVALID) {
/*
* Invalid input. Hmm, what's the right thing to do here?
*/
memset(&mbs, 0, sizeof mbs);
return lastchar_wide = WEOF;
}
if (cnt != MB_INCOMPLETE)
break;
/*
* Always apply KEYTIMEOUT to the remains of the input
* character. The parts of a multibyte character should
* arrive together. If we don't do this the input can
* get stuck if an invalid byte sequence arrives.
*/
inchar = getbyte(1L, &timeout);
/* getbyte deliberately resets lastchar_wide_valid */
lastchar_wide_valid = 1;
if (inchar == EOF) {
memset(&mbs, 0, sizeof mbs);
if (timeout)
{
/*
* This case means that we got a valid initial byte
* (since we tested for EOF above), but the followup
* timed out. This probably indicates a duff character.
* Return a '?'.
*/
lastchar = '?';
return lastchar_wide = L'?';
}
else
return lastchar_wide = WEOF;
}
c = inchar;
}
return lastchar_wide = (ZLE_INT_T)outchar;
}
static int
default_char2int(SCR *sp, const char * str, ssize_t len, CONVWIN *cw,
size_t *tolen, const CHAR_T **dst, const char *enc)
{
int j;
size_t i = 0;
CHAR_T **tostr = (CHAR_T **)(void *)&cw->bp1;
size_t *blen = &cw->blen1;
mbstate_t mbs;
size_t n;
ssize_t nlen = len;
const char *src = (const char *)str;
iconv_t id = (iconv_t)-1;
char buffer[CONV_BUFFER_SIZE];
size_t left = len;
int error = 1;
MEMSET(&mbs, 0, 1);
BINC_RETW(NULL, *tostr, *blen, nlen);
#ifdef USE_ICONV
if (strcmp(nl_langinfo(CODESET), enc)) {
id = iconv_open(nl_langinfo(CODESET), enc);
if (id == (iconv_t)-1)
goto err;
CONVERT(str, left, src, len);
}
#endif
for (i = 0, j = 0; j < len; ) {
n = mbrtowc((*tostr)+i, src+j, len-j, &mbs);
/* NULL character converted */
if (n == (size_t)-2) error = -(len-j);
if (n == (size_t)-1 || n == (size_t)-2)
HANDLE_MBR_ERROR(n, mbs, (*tostr)[i], src[j]);
if (n == 0) n = 1;
j += n;
if (++i >= *blen) {
nlen += 256;
BINC_RETW(NULL, *tostr, *blen, nlen);
}
if (id != (iconv_t)-1 && j == len && left) {
CONVERT(str, left, src, len);
j = 0;
}
}
*tolen = i;
if (id != (iconv_t)-1)
iconv_close(id);
*dst = cw->bp1;
return 0;
err:
*tolen = i;
if (id != (iconv_t)-1)
iconv_close(id);
*dst = cw->bp1;
return error;
}
int
main (int argc, char *argv[])
{
mbstate_t state;
wchar_t wc;
size_t ret;
/* configure should already have checked that the locale is supported. */
if (setlocale (LC_ALL, "") == NULL)
return 1;
/* Test NUL byte input. */
{
const char *src;
memset (&state, '\0', sizeof (mbstate_t));
src = "";
ret = mbsnrtowcs (NULL, &src, 1, 0, &state);
ASSERT (ret == 0);
ASSERT (mbsinit (&state));
src = "";
ret = mbsnrtowcs (NULL, &src, 1, 1, &state);
ASSERT (ret == 0);
ASSERT (mbsinit (&state));
wc = (wchar_t) 0xBADFACE;
src = "";
ret = mbsnrtowcs (&wc, &src, 1, 0, &state);
ASSERT (ret == 0);
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (mbsinit (&state));
wc = (wchar_t) 0xBADFACE;
src = "";
ret = mbsnrtowcs (&wc, &src, 1, 1, &state);
ASSERT (ret == 0);
ASSERT (wc == 0);
ASSERT (mbsinit (&state));
}
if (argc > 1)
{
int unlimited;
for (unlimited = 0; unlimited < 2; unlimited++)
{
#define BUFSIZE 10
wchar_t buf[BUFSIZE];
const char *src;
mbstate_t temp_state;
{
size_t i;
for (i = 0; i < BUFSIZE; i++)
buf[i] = (wchar_t) 0xBADFACE;
}
switch (argv[1][0])
{
case '1':
/* Locale encoding is ISO-8859-1 or ISO-8859-15. */
{
char input[] = "B\374\337er"; /* "Büßer" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'B');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 1, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == (unsigned char) '\374');
ASSERT (mbsinit (&state));
input[1] = '\0';
src = input + 2;
temp_state = state;
ret = mbsnrtowcs (NULL, &src, 4, unlimited ? BUFSIZE : 1, &temp_state);
ASSERT (ret == 3);
ASSERT (src == input + 2);
ASSERT (mbsinit (&state));
src = input + 2;
ret = mbsnrtowcs (buf, &src, 4, unlimited ? BUFSIZE : 1, &state);
ASSERT (ret == (unlimited ? 3 : 1));
ASSERT (src == (unlimited ? NULL : input + 3));
ASSERT (wctob (buf[0]) == (unsigned char) '\337');
if (unlimited)
{
ASSERT (buf[1] == 'e');
ASSERT (buf[2] == 'r');
ASSERT (buf[3] == 0);
ASSERT (buf[4] == (wchar_t) 0xBADFACE);
}
else
ASSERT (buf[1] == (wchar_t) 0xBADFACE);
ASSERT (mbsinit (&state));
}
break;
case '2':
/* Locale encoding is UTF-8. */
{
char input[] = "B\303\274\303\237er"; /* "Büßer" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'B');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 1, &state);
ASSERT (ret == (size_t)(-2));
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (!mbsinit (&state));
input[1] = '\0';
src = input + 2;
temp_state = state;
ret = mbsnrtowcs (NULL, &src, 6, unlimited ? BUFSIZE : 2, &temp_state);
ASSERT (ret == 4);
ASSERT (src == input + 2);
ASSERT (!mbsinit (&state));
src = input + 2;
ret = mbsnrtowcs (buf, &src, 6, unlimited ? BUFSIZE : 2, &state);
ASSERT (ret == (unlimited ? 4 : 2));
ASSERT (src == (unlimited ? NULL : input + 5));
ASSERT (wctob (buf[0]) == EOF);
ASSERT (wctob (buf[1]) == EOF);
if (unlimited)
{
ASSERT (buf[2] == 'e');
ASSERT (buf[3] == 'r');
ASSERT (buf[4] == 0);
ASSERT (buf[5] == (wchar_t) 0xBADFACE);
}
else
ASSERT (buf[2] == (wchar_t) 0xBADFACE);
ASSERT (mbsinit (&state));
}
break;
case '3':
/* Locale encoding is EUC-JP. */
{
char input[] = "<\306\374\313\334\270\354>"; /* "<日本語>" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == '<');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 2, &state);
ASSERT (ret == 2);
ASSERT (wctob (wc) == EOF);
ASSERT (mbsinit (&state));
input[1] = '\0';
input[2] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 3, 1, &state);
ASSERT (ret == (size_t)(-2));
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (!mbsinit (&state));
input[3] = '\0';
src = input + 4;
temp_state = state;
ret = mbsnrtowcs (NULL, &src, 5, unlimited ? BUFSIZE : 2, &temp_state);
ASSERT (ret == 3);
ASSERT (src == input + 4);
ASSERT (!mbsinit (&state));
src = input + 4;
ret = mbsnrtowcs (buf, &src, 5, unlimited ? BUFSIZE : 2, &state);
ASSERT (ret == (unlimited ? 3 : 2));
ASSERT (src == (unlimited ? NULL : input + 7));
ASSERT (wctob (buf[0]) == EOF);
ASSERT (wctob (buf[1]) == EOF);
if (unlimited)
{
ASSERT (buf[2] == '>');
ASSERT (buf[3] == 0);
ASSERT (buf[4] == (wchar_t) 0xBADFACE);
}
else
ASSERT (buf[2] == (wchar_t) 0xBADFACE);
ASSERT (mbsinit (&state));
}
break;
case '4':
/* Locale encoding is GB18030. */
{
char input[] = "B\250\271\201\060\211\070er"; /* "Büßer" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'B');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 1, &state);
ASSERT (ret == (size_t)(-2));
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (!mbsinit (&state));
input[1] = '\0';
src = input + 2;
temp_state = state;
ret = mbsnrtowcs (NULL, &src, 8, unlimited ? BUFSIZE : 2, &temp_state);
ASSERT (ret == 4);
ASSERT (src == input + 2);
ASSERT (!mbsinit (&state));
src = input + 2;
ret = mbsnrtowcs (buf, &src, 8, unlimited ? BUFSIZE : 2, &state);
ASSERT (ret == (unlimited ? 4 : 2));
ASSERT (src == (unlimited ? NULL : input + 7));
ASSERT (wctob (buf[0]) == EOF);
ASSERT (wctob (buf[1]) == EOF);
if (unlimited)
{
ASSERT (buf[2] == 'e');
ASSERT (buf[3] == 'r');
ASSERT (buf[4] == 0);
ASSERT (buf[5] == (wchar_t) 0xBADFACE);
}
else
ASSERT (buf[2] == (wchar_t) 0xBADFACE);
ASSERT (mbsinit (&state));
}
break;
default:
return 1;
}
}
return 0;
}
return 1;
}
static void ttyin(char buf[], register int nmax, char pchar)
{
char *sp;
int c;
int slash = 0;
int maxlen;
sp = buf;
maxlen = 0;
while (sp - buf < nmax) {
if (promptlen > maxlen)
maxlen = promptlen;
c = readch();
if (c == '\\') {
slash++;
} else if (((cc_t) c == otty.c_cc[VERASE]) && !slash) {
if (sp > buf) {
#ifdef HAVE_WIDECHAR
if (MB_CUR_MAX > 1) {
wchar_t wc;
size_t pos = 0, mblength;
mbstate_t state, state_bak;
memset(&state, '\0', sizeof(mbstate_t));
while (1) {
state_bak = state;
mblength =
mbrtowc(&wc, buf + pos,
sp - buf, &state);
state = (mblength == (size_t)-2
|| mblength ==
(size_t)-1) ? state_bak
: state;
mblength =
(mblength == (size_t)-2
|| mblength == (size_t)-1
|| mblength ==
0) ? 1 : mblength;
if (buf + pos + mblength >= sp)
break;
pos += mblength;
}
if (mblength == 1) {
ERASEONECOLUMN(docrterase);
} else {
int wc_width;
wc_width = wcwidth(wc);
wc_width =
(wc_width <
1) ? 1 : wc_width;
while (wc_width--) {
ERASEONECOLUMN(docrterase);
}
}
while (mblength--) {
--promptlen;
--sp;
}
} else
#endif /* HAVE_WIDECHAR */
{
--promptlen;
ERASEONECOLUMN(docrterase);
--sp;
}
if ((*sp < ' ' && *sp != '\n') || *sp == RUBOUT) {
--promptlen;
ERASEONECOLUMN(docrterase);
}
continue;
} else {
if (!eraseln)
promptlen = maxlen;
siglongjmp(restore, 1);
}
} else if (((cc_t) c == otty.c_cc[VKILL]) && !slash) {
if (hard) {
show(c);
putchar('\n');
putchar(pchar);
} else {
putchar('\r');
putchar(pchar);
if (eraseln)
erasep(1);
else if (docrtkill)
while (promptlen-- > 1)
putserr(BSB);
promptlen = 1;
}
sp = buf;
fflush(stdout);
continue;
}
if (slash && ((cc_t) c == otty.c_cc[VKILL]
|| (cc_t) c == otty.c_cc[VERASE])) {
ERASEONECOLUMN(docrterase);
--sp;
}
if (c != '\\')
slash = 0;
*sp++ = c;
if ((c < ' ' && c != '\n' && c != ESC) || c == RUBOUT) {
c += (c == RUBOUT) ? -0100 : 0100;
putserr(CARAT);
promptlen++;
}
if (c != '\n' && c != ESC) {
putcerr(c);
promptlen++;
} else
break;
}
*--sp = '\0';
if (!eraseln)
promptlen = maxlen;
if (sp - buf >= nmax - 1)
more_error(_("Line too long"));
}
int
glob(const char *pattern, int flags, int (*errfunc)(const char *, int), glob_t *pglob)
{
const char *patnext;
size_t limit;
Char *bufnext, *bufend, patbuf[MAXPATHLEN], prot;
mbstate_t mbs;
wchar_t wc;
size_t clen;
patnext = pattern;
if (!(flags & GLOB_APPEND)) {
pglob->gl_pathc = 0;
pglob->gl_pathv = NULL;
if (!(flags & GLOB_DOOFFS))
pglob->gl_offs = 0;
}
limit = 0;
pglob->gl_flags = flags & ~GLOB_MAGCHAR;
pglob->gl_errfunc = errfunc;
pglob->gl_matchc = 0;
bufnext = patbuf;
bufend = bufnext + MAXPATHLEN - 1;
if (flags & GLOB_NOESCAPE) {
memset(&mbs, 0, sizeof(mbs));
while (bufend - bufnext >= MB_CUR_MAX) {
clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs);
if (clen == (size_t)-1 || clen == (size_t)-2)
return (GLOB_NOMATCH);
else if (clen == 0)
break;
*bufnext++ = wc;
patnext += clen;
}
} else {
/* Protect the quoted characters. */
memset(&mbs, 0, sizeof(mbs));
while (bufend - bufnext >= MB_CUR_MAX) {
if (*patnext == QUOTE) {
if (*++patnext == EOS) {
*bufnext++ = QUOTE | M_PROTECT;
continue;
}
prot = M_PROTECT;
} else
prot = 0;
clen = mbrtowc(&wc, patnext, MB_LEN_MAX, &mbs);
if (clen == (size_t)-1 || clen == (size_t)-2)
return (GLOB_NOMATCH);
else if (clen == 0)
break;
*bufnext++ = wc | prot;
patnext += clen;
}
}
*bufnext = EOS;
if (flags & GLOB_BRACE)
return globexp1(patbuf, pglob, &limit);
else
return glob0(patbuf, pglob, &limit);
}
static int
decode_current_locale(const char* arg, wchar_t **wstr, size_t *wlen,
const char **reason, int surrogateescape)
{
wchar_t *res;
size_t argsize;
size_t count;
#ifdef HAVE_MBRTOWC
unsigned char *in;
wchar_t *out;
mbstate_t mbs;
#endif
#ifdef HAVE_BROKEN_MBSTOWCS
/* Some platforms have a broken implementation of
* mbstowcs which does not count the characters that
* would result from conversion. Use an upper bound.
*/
argsize = strlen(arg);
#else
argsize = mbstowcs(NULL, arg, 0);
#endif
if (argsize != (size_t)-1) {
if (argsize > PY_SSIZE_T_MAX / sizeof(wchar_t) - 1) {
return -1;
}
res = (wchar_t *)PyMem_RawMalloc((argsize + 1) * sizeof(wchar_t));
if (!res) {
return -1;
}
count = mbstowcs(res, arg, argsize + 1);
if (count != (size_t)-1) {
wchar_t *tmp;
/* Only use the result if it contains no
surrogate characters. */
for (tmp = res; *tmp != 0 &&
!Py_UNICODE_IS_SURROGATE(*tmp); tmp++)
;
if (*tmp == 0) {
if (wlen != NULL) {
*wlen = count;
}
*wstr = res;
return 0;
}
}
PyMem_RawFree(res);
}
/* Conversion failed. Fall back to escaping with surrogateescape. */
#ifdef HAVE_MBRTOWC
/* Try conversion with mbrtwoc (C99), and escape non-decodable bytes. */
/* Overallocate; as multi-byte characters are in the argument, the
actual output could use less memory. */
argsize = strlen(arg) + 1;
if (argsize > PY_SSIZE_T_MAX / sizeof(wchar_t)) {
return -1;
}
res = (wchar_t*)PyMem_RawMalloc(argsize * sizeof(wchar_t));
if (!res) {
return -1;
}
in = (unsigned char*)arg;
out = res;
memset(&mbs, 0, sizeof mbs);
while (argsize) {
size_t converted = mbrtowc(out, (char*)in, argsize, &mbs);
if (converted == 0) {
/* Reached end of string; null char stored. */
break;
}
if (converted == (size_t)-2) {
/* Incomplete character. This should never happen,
since we provide everything that we have -
unless there is a bug in the C library, or I
misunderstood how mbrtowc works. */
goto decode_error;
}
if (converted == (size_t)-1) {
if (!surrogateescape) {
goto decode_error;
}
/* Conversion error. Escape as UTF-8b, and start over
in the initial shift state. */
*out++ = 0xdc00 + *in++;
argsize--;
memset(&mbs, 0, sizeof mbs);
continue;
}
if (Py_UNICODE_IS_SURROGATE(*out)) {
if (!surrogateescape) {
goto decode_error;
}
/* Surrogate character. Escape the original
byte sequence with surrogateescape. */
argsize -= converted;
while (converted--) {
*out++ = 0xdc00 + *in++;
}
continue;
}
/* successfully converted some bytes */
in += converted;
argsize -= converted;
out++;
}
if (wlen != NULL) {
*wlen = out - res;
}
*wstr = res;
return 0;
decode_error:
PyMem_RawFree(res);
if (wlen) {
*wlen = in - (unsigned char*)arg;
}
if (reason) {
*reason = "decoding error";
}
return -2;
#else /* HAVE_MBRTOWC */
/* Cannot use C locale for escaping; manually escape as if charset
is ASCII (i.e. escape all bytes > 128. This will still roundtrip
correctly in the locale's charset, which must be an ASCII superset. */
return decode_ascii(arg, wstr, wlen, reason, surrogateescape);
#endif /* HAVE_MBRTOWC */
}
static int
glob3(Char *pathbuf, Char *pathend, Char *pathend_last,
Char *pattern, Char *restpattern,
glob_t *pglob, size_t *limit)
{
struct dirent *dp;
DIR *dirp;
int err;
char buf[MAXPATHLEN];
/*
* The readdirfunc declaration can't be prototyped, because it is
* assigned, below, to two functions which are prototyped in glob.h
* and dirent.h as taking pointers to differently typed opaque
* structures.
*/
struct dirent *(*readdirfunc)();
if (pathend > pathend_last)
return (GLOB_ABORTED);
*pathend = EOS;
errno = 0;
if ((dirp = g_opendir(pathbuf, pglob)) == NULL) {
/* TODO: don't call for ENOENT or ENOTDIR? */
if (pglob->gl_errfunc) {
if (g_Ctoc(pathbuf, buf, sizeof(buf)))
return (GLOB_ABORTED);
if (pglob->gl_errfunc(buf, errno) ||
pglob->gl_flags & GLOB_ERR)
return (GLOB_ABORTED);
}
return(0);
}
err = 0;
/* Search directory for matching names. */
if (pglob->gl_flags & GLOB_ALTDIRFUNC)
readdirfunc = pglob->gl_readdir;
else
readdirfunc = readdir;
while ((dp = (*readdirfunc)(dirp))) {
char *sc;
Char *dc;
wchar_t wc;
size_t clen;
mbstate_t mbs;
/* Initial DOT must be matched literally. */
if (dp->d_name[0] == DOT && *pattern != DOT)
continue;
memset(&mbs, 0, sizeof(mbs));
dc = pathend;
sc = dp->d_name;
while (dc < pathend_last) {
clen = mbrtowc(&wc, sc, MB_LEN_MAX, &mbs);
if (clen == (size_t)-1 || clen == (size_t)-2) {
wc = *sc;
clen = 1;
memset(&mbs, 0, sizeof(mbs));
}
if ((*dc++ = wc) == EOS)
break;
sc += clen;
}
if (!match(pathend, pattern, restpattern)) {
*pathend = EOS;
continue;
}
err = glob2(pathbuf, --dc, pathend_last, restpattern,
pglob, limit);
if (err)
break;
}
if (pglob->gl_flags & GLOB_ALTDIRFUNC)
(*pglob->gl_closedir)(dirp);
else
closedir(dirp);
return(err);
}
TEST(wchar, wcstombs_wcrtombs) {
const wchar_t chars[] = { L'h', L'e', L'l', L'l', L'o', 0 };
const wchar_t bad_chars[] = { L'h', L'i', static_cast<wchar_t>(0xffffffff), 0 };
const wchar_t* src;
char bytes[BUFSIZ];
// Given a NULL destination, these functions count valid characters.
EXPECT_EQ(5U, wcstombs(NULL, chars, 0));
EXPECT_EQ(5U, wcstombs(NULL, chars, 4));
EXPECT_EQ(5U, wcstombs(NULL, chars, 256));
src = chars;
EXPECT_EQ(5U, wcsrtombs(NULL, &src, 0, NULL));
EXPECT_EQ(&chars[0], src);
src = chars;
EXPECT_EQ(5U, wcsrtombs(NULL, &src, 4, NULL));
EXPECT_EQ(&chars[0], src);
src = chars;
EXPECT_EQ(5U, wcsrtombs(NULL, &src, 256, NULL));
EXPECT_EQ(&chars[0], src);
// An unrepresentable char just returns an error from wcstombs...
errno = 0;
EXPECT_EQ(static_cast<size_t>(-1), wcstombs(NULL, bad_chars, 0));
EXPECT_EQ(EILSEQ, errno);
errno = 0;
EXPECT_EQ(static_cast<size_t>(-1), wcstombs(NULL, bad_chars, 256));
EXPECT_EQ(EILSEQ, errno);
// And wcsrtombs doesn't tell us where it got stuck because we didn't ask it
// to actually convert anything...
errno = 0;
src = bad_chars;
EXPECT_EQ(static_cast<size_t>(-1), wcsrtombs(NULL, &src, 0, NULL));
EXPECT_EQ(&bad_chars[0], src);
EXPECT_EQ(EILSEQ, errno);
errno = 0;
src = bad_chars;
EXPECT_EQ(static_cast<size_t>(-1), wcsrtombs(NULL, &src, 256, NULL));
EXPECT_EQ(&bad_chars[0], src);
EXPECT_EQ(EILSEQ, errno);
// Okay, now let's test actually converting something...
memset(bytes, 'x', sizeof(bytes));
EXPECT_EQ(0U, wcstombs(bytes, chars, 0));
memset(bytes, 'x', sizeof(bytes));
EXPECT_EQ(4U, wcstombs(bytes, chars, 4));
bytes[5] = 0;
EXPECT_STREQ("hellx", bytes);
memset(bytes, 'x', sizeof(bytes));
EXPECT_EQ(5U, wcstombs(bytes, chars, 256));
EXPECT_STREQ("hello", bytes);
memset(bytes, 'x', sizeof(bytes));
EXPECT_EQ(5U, wcstombs(bytes, chars, 6));
EXPECT_STREQ("hello", bytes);
errno = 0;
memset(bytes, 'x', sizeof(bytes));
EXPECT_EQ(static_cast<size_t>(-1), wcstombs(bytes, bad_chars, 256));
EXPECT_EQ(EILSEQ, errno);
bytes[3] = 0;
EXPECT_STREQ("hix", bytes);
// wcsrtombs is a bit more informative...
memset(bytes, 'x', sizeof(bytes));
src = chars;
EXPECT_EQ(0U, wcsrtombs(bytes, &src, 0, NULL));
EXPECT_EQ(&chars[0], src); // No input consumed.
EXPECT_EQ(EILSEQ, errno);
memset(bytes, 'x', sizeof(bytes));
src = chars;
EXPECT_EQ(4U, wcsrtombs(bytes, &src, 4, NULL));
EXPECT_EQ(&chars[4], src); // Some input consumed.
EXPECT_EQ(EILSEQ, errno);
bytes[5] = 0;
EXPECT_STREQ("hellx", bytes);
memset(bytes, 'x', sizeof(bytes));
src = chars;
EXPECT_EQ(5U, wcsrtombs(bytes, &src, 256, NULL));
EXPECT_EQ(NULL, src); // All input consumed!
EXPECT_EQ(EILSEQ, errno);
EXPECT_STREQ("hello", bytes);
memset(bytes, 'x', sizeof(bytes));
src = chars;
EXPECT_EQ(5U, wcsrtombs(bytes, &src, 6, NULL));
EXPECT_EQ(NULL, src); // All input consumed.
EXPECT_EQ(EILSEQ, errno);
EXPECT_STREQ("hello", bytes);
memset(bytes, 'x', sizeof(bytes));
src = bad_chars;
EXPECT_EQ(static_cast<size_t>(-1), wcsrtombs(bytes, &src, 256, NULL));
EXPECT_EQ(&bad_chars[2], src);
EXPECT_EQ(EILSEQ, errno);
bytes[3] = 0;
EXPECT_STREQ("hix", bytes);
// Any non-initial state is invalid when calling wcsrtombs.
mbstate_t ps;
src = chars;
memset(&ps, 0, sizeof(ps));
ASSERT_EQ(static_cast<size_t>(-2), mbrtowc(NULL, "\xc2", 1, &ps));
EXPECT_EQ(static_cast<size_t>(-1), wcsrtombs(NULL, &src, 0, &ps));
EXPECT_EQ(EILSEQ, errno);
}
static inline void do_layout(VisualRow *vrow, VTLayout *vtlayout)
{
#define SIZE 4096
size_t logical_char2cell_index[SIZE];
size_t in2out[SIZE];
size_t out2in[SIZE];
char logical[SIZE]; /* is this enough? */
int logical_width, logical_length;
char visual[SIZE]; /* is this enough? */
size_t visual_length = SIZE;
wchar_t wc;
mbstate_t ps;
size_t mb_len;
int consumed = 0;
int cursor_pos = 0;
int current_char = 0;
int width;
int logical_cell_index;
Char mbchar;
memset(logical_char2cell_index,0,SIZE);
memset(logical,0,SIZE);
memset(visual,0,SIZE);
memset(in2out,0,SIZE);
memset(out2in,0,SIZE);
memset(&ps,0,sizeof(ps));
construct_char2cell_index(vrow->row,logical_char2cell_index);
Row_compose_bytes(vrow->row,
0,vrow->row->num_columns,
logical,SIZE,
&logical_length,&logical_width,1);
#undef SIZE
#if debug
printf("logical: [%s]\n",logical);
printf("logical_length = %u\n",logical_length);
#endif
vtlayout->transform(vtlayout,
logical, logical_length,
visual, &visual_length,
in2out,out2in,NULL);
#if debug
printf("visual : [%s]\n",visual);
printf("visual_length = %u\n\n",visual_length);
#endif
/* Well off course, I need to refine this code ;-) */
while(consumed < visual_length)
{
mb_len = mbrtowc(&wc,visual+consumed,visual_length-consumed,&ps);
if(mb_len == 0 || mb_len == (size_t)-1 || mb_len == (size_t)-2 )
{
#if debug
printf("found invalid or NULL character(%u)\n",mb_len);
#endif
break;
}
else
{
logical_cell_index = logical_char2cell_index[out2in[current_char]];
#if debug
printf("logical char index[%d] <-> visual char index[%d]\n",
out2in[current_char],current_char);
#endif
if(Row_is_char_drawn(vrow->row,logical_cell_index))
{
width = wcwidth(wc);
Char_init(mbchar,visual+consumed,mb_len,width,(width>0));
Row_add_char(vrow->vrow,
cursor_pos,
&mbchar,
Row_is_bold(vrow->row,logical_cell_index),
Row_is_blink(vrow->row,logical_cell_index),
Row_is_inverse(vrow->row,logical_cell_index),
Row_is_underline(vrow->row,logical_cell_index),
Row_get_foreground(vrow->row,logical_cell_index),
Row_get_background(vrow->row,logical_cell_index),
Row_get_charset(vrow->row,logical_cell_index));
}
else
{
Row_set_char_drawn(vrow->vrow,cursor_pos,0);
width = 1;
}
vrow->visual2logical_index[cursor_pos] = logical_cell_index;
vrow->logical2visual_index[logical_cell_index] = cursor_pos;
#if debug
printf("logical cell index[%d] <-> visual cell index[%d]\n",
logical_cell_index,cursor_pos);
#endif
}
/* process for multi column character */
if(width > 1 && Row_is_char_drawn(vrow->row,logical_cell_index) &&
width == Row_get_cell_width(vrow->row,logical_cell_index) )
{
int i;
for(i=1;i<width;i++)
{
vrow->visual2logical_index[cursor_pos+i] = logical_cell_index+i;
vrow->logical2visual_index[logical_cell_index+i] = cursor_pos+i;
}
}
cursor_pos+=width;
current_char++;
consumed+=mb_len;
}
/* The visual row possibly has ligatures */
if(cursor_pos < vrow->row->num_columns &&
!vtlayout->is_direction_LTR(vtlayout))
Row_insert_cells(vrow->vrow,0,vrow->row->num_columns-cursor_pos);
}
size_t
mbsrtowcs (wchar_t *dest, const char **srcp, size_t len, mbstate_t *ps)
{
if (ps == NULL)
ps = &_gl_mbsrtowcs_state;
{
const char *src = *srcp;
if (dest != NULL)
{
wchar_t *destptr = dest;
for (; len > 0; destptr++, len--)
{
size_t src_avail;
size_t ret;
/* An optimized variant of
src_avail = strnlen1 (src, MB_LEN_MAX); */
if (src[0] == '\0')
src_avail = 1;
else if (src[1] == '\0')
src_avail = 2;
else if (src[2] == '\0')
src_avail = 3;
else if (MB_LEN_MAX <= 4 || src[3] == '\0')
src_avail = 4;
else
src_avail = 4 + strnlen1 (src + 4, MB_LEN_MAX - 4);
/* Parse the next multibyte character. */
ret = mbrtowc (destptr, src, src_avail, ps);
if (ret == (size_t)(-2))
/* Encountered a multibyte character that extends past a '\0' byte
or that is longer than MB_LEN_MAX bytes. Cannot happen. */
abort ();
if (ret == (size_t)(-1))
goto bad_input;
if (ret == 0)
{
src = NULL;
/* Here mbsinit (ps). */
break;
}
src += ret;
}
*srcp = src;
return destptr - dest;
}
else
{
/* Ignore dest and len, don't store *srcp at the end, and
don't clobber *ps. */
mbstate_t state = *ps;
size_t totalcount = 0;
for (;; totalcount++)
{
size_t src_avail;
size_t ret;
/* An optimized variant of
src_avail = strnlen1 (src, MB_LEN_MAX); */
if (src[0] == '\0')
src_avail = 1;
else if (src[1] == '\0')
src_avail = 2;
else if (src[2] == '\0')
src_avail = 3;
else if (MB_LEN_MAX <= 4 || src[3] == '\0')
src_avail = 4;
else
src_avail = 4 + strnlen1 (src + 4, MB_LEN_MAX - 4);
/* Parse the next multibyte character. */
ret = mbrtowc (NULL, src, src_avail, &state);
if (ret == (size_t)(-2))
/* Encountered a multibyte character that extends past a '\0' byte
or that is longer than MB_LEN_MAX bytes. Cannot happen. */
abort ();
if (ret == (size_t)(-1))
goto bad_input2;
if (ret == 0)
{
/* Here mbsinit (&state). */
break;
}
src += ret;
}
return totalcount;
}
bad_input:
*srcp = src;
bad_input2:
errno = EILSEQ;
return (size_t)(-1);
}
}
void
conv_c(PR *pr, u_char *p, size_t bufsize)
{
char buf[10];
char const *str;
wchar_t wc;
size_t clen, oclen;
int converr, pad, width;
u_char peekbuf[MB_LEN_MAX];
if (pr->mbleft > 0) {
str = "**";
pr->mbleft--;
goto strpr;
}
switch(*p) {
case '\0':
str = "\\0";
goto strpr;
/* case '\a': */
case '\007':
str = "\\a";
goto strpr;
case '\b':
str = "\\b";
goto strpr;
case '\f':
str = "\\f";
goto strpr;
case '\n':
str = "\\n";
goto strpr;
case '\r':
str = "\\r";
goto strpr;
case '\t':
str = "\\t";
goto strpr;
case '\v':
str = "\\v";
goto strpr;
default:
break;
}
/*
* Multibyte characters are disabled for hexdump(1) for backwards
* compatibility and consistency (none of its other output formats
* recognize them correctly).
*/
converr = 0;
if (odmode && MB_CUR_MAX > 1) {
oclen = 0;
retry:
clen = mbrtowc(&wc, (const char *)p, bufsize, &pr->mbstate);
if (clen == 0)
clen = 1;
else if (clen == (size_t)-1 || (clen == (size_t)-2 &&
p == peekbuf)) {
memset(&pr->mbstate, 0, sizeof(pr->mbstate));
wc = *p;
clen = 1;
converr = 1;
} else if (clen == (size_t)-2) {
/*
* Incomplete character; peek ahead and see if we
* can complete it.
*/
oclen = bufsize;
bufsize = peek(p = peekbuf, MB_CUR_MAX);
goto retry;
}
clen += oclen;
} else {
wc = *p;
clen = 1;
}
if (!converr && iswprint(wc)) {
if (!odmode) {
*pr->cchar = 'c';
(void)printf(pr->fmt, (int)wc);
} else {
*pr->cchar = 'C';
assert(strcmp(pr->fmt, "%3C") == 0);
width = wcwidth(wc);
assert(width >= 0);
pad = 3 - width;
if (pad < 0)
pad = 0;
(void)printf("%*s%C", pad, "", wc);
pr->mbleft = clen - 1;
}
} else {
(void)sprintf(buf, "%03o", (int)*p);
str = buf;
strpr: *pr->cchar = 's';
(void)printf(pr->fmt, str);
}
}
static int
test_one_locale (const char *name, int codepage)
{
mbstate_t state;
wchar_t wc;
size_t ret;
# if 1
/* Portable code to set the locale. */
{
char name_with_codepage[1024];
sprintf (name_with_codepage, "%s.%d", name, codepage);
/* Set the locale. */
if (setlocale (LC_ALL, name_with_codepage) == NULL)
return 77;
}
# else
/* Hacky way to set a locale.codepage combination that setlocale() refuses
to set. */
{
/* Codepage of the current locale, set with setlocale().
Not necessarily the same as GetACP(). */
extern __declspec(dllimport) unsigned int __lc_codepage;
/* Set the locale. */
if (setlocale (LC_ALL, name) == NULL)
return 77;
/* Clobber the codepage and MB_CUR_MAX, both set by setlocale(). */
__lc_codepage = codepage;
switch (codepage)
{
case 1252:
case 1256:
MB_CUR_MAX = 1;
break;
case 932:
case 950:
case 936:
MB_CUR_MAX = 2;
break;
case 54936:
case 65001:
MB_CUR_MAX = 4;
break;
}
/* Test whether the codepage is really available. */
memset (&state, '\0', sizeof (mbstate_t));
if (mbrtowc (&wc, " ", 1, &state) == (size_t)(-1))
return 77;
}
# endif
/* Test zero-length input. */
{
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "x", 0, &state);
/* gnulib's implementation returns (size_t)(-2).
The AIX 5.1 implementation returns (size_t)(-1).
glibc's implementation returns 0. */
ASSERT (ret == (size_t)(-2) || ret == (size_t)(-1) || ret == 0);
ASSERT (mbsinit (&state));
}
/* Test NUL byte input. */
{
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "", 1, &state);
ASSERT (ret == 0);
ASSERT (wc == 0);
ASSERT (mbsinit (&state));
ret = mbrtowc (NULL, "", 1, &state);
ASSERT (ret == 0);
ASSERT (mbsinit (&state));
}
/* Test single-byte input. */
{
int c;
char buf[1];
memset (&state, '\0', sizeof (mbstate_t));
for (c = 0; c < 0x100; c++)
switch (c)
{
case '\t': case '\v': case '\f':
case ' ': case '!': case '"': case '#': case '%':
case '&': case '\'': case '(': case ')': case '*':
case '+': case ',': case '-': case '.': case '/':
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
case ':': case ';': case '<': case '=': case '>':
case '?':
case 'A': case 'B': case 'C': case 'D': case 'E':
case 'F': case 'G': case 'H': case 'I': case 'J':
case 'K': case 'L': case 'M': case 'N': case 'O':
case 'P': case 'Q': case 'R': case 'S': case 'T':
case 'U': case 'V': case 'W': case 'X': case 'Y':
case 'Z':
case '[': case '\\': case ']': case '^': case '_':
case 'a': case 'b': case 'c': case 'd': case 'e':
case 'f': case 'g': case 'h': case 'i': case 'j':
case 'k': case 'l': case 'm': case 'n': case 'o':
case 'p': case 'q': case 'r': case 's': case 't':
case 'u': case 'v': case 'w': case 'x': case 'y':
case 'z': case '{': case '|': case '}': case '~':
/* c is in the ISO C "basic character set". */
buf[0] = c;
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, buf, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == c);
ASSERT (mbsinit (&state));
ret = mbrtowc (NULL, buf, 1, &state);
ASSERT (ret == 1);
ASSERT (mbsinit (&state));
break;
}
}
/* Test special calling convention, passing a NULL pointer. */
{
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, NULL, 5, &state);
ASSERT (ret == 0);
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (mbsinit (&state));
}
switch (codepage)
{
case 1252:
/* Locale encoding is CP1252, an extension of ISO-8859-1. */
{
char input[] = "B\374\337er"; /* "Büßer" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'B');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 1, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == (unsigned char) '\374');
ASSERT (wc == 0x00FC);
ASSERT (mbsinit (&state));
input[1] = '\0';
/* Test support of NULL first argument. */
ret = mbrtowc (NULL, input + 2, 3, &state);
ASSERT (ret == 1);
ASSERT (mbsinit (&state));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 2, 3, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == (unsigned char) '\337');
ASSERT (wc == 0x00DF);
ASSERT (mbsinit (&state));
input[2] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 3, 2, &state);
ASSERT (ret == 1);
ASSERT (wc == 'e');
ASSERT (mbsinit (&state));
input[3] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 4, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'r');
ASSERT (mbsinit (&state));
}
return 0;
case 1256:
/* Locale encoding is CP1256, not the same as ISO-8859-6. */
{
char input[] = "x\302\341\346y"; /* "xآلوy" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'x');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 1, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == (unsigned char) '\302');
ASSERT (wc == 0x0622);
ASSERT (mbsinit (&state));
input[1] = '\0';
/* Test support of NULL first argument. */
ret = mbrtowc (NULL, input + 2, 3, &state);
ASSERT (ret == 1);
ASSERT (mbsinit (&state));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 2, 3, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == (unsigned char) '\341');
ASSERT (wc == 0x0644);
ASSERT (mbsinit (&state));
input[2] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 3, 2, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == (unsigned char) '\346');
ASSERT (wc == 0x0648);
ASSERT (mbsinit (&state));
input[3] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 4, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'y');
ASSERT (mbsinit (&state));
}
return 0;
case 932:
/* Locale encoding is CP932, similar to Shift_JIS. */
{
char input[] = "<\223\372\226\173\214\352>"; /* "<日本語>" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == '<');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 2, &state);
ASSERT (ret == 2);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x65E5);
ASSERT (mbsinit (&state));
input[1] = '\0';
input[2] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 3, 1, &state);
ASSERT (ret == (size_t)(-2));
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (!mbsinit (&state));
input[3] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 4, 4, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x672C);
ASSERT (mbsinit (&state));
input[4] = '\0';
/* Test support of NULL first argument. */
ret = mbrtowc (NULL, input + 5, 3, &state);
ASSERT (ret == 2);
ASSERT (mbsinit (&state));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 5, 3, &state);
ASSERT (ret == 2);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x8A9E);
ASSERT (mbsinit (&state));
input[5] = '\0';
input[6] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 7, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == '>');
ASSERT (mbsinit (&state));
/* Test some invalid input. */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\377", 1, &state); /* 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\225\377", 2, &state); /* 0x95 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
}
return 0;
case 950:
/* Locale encoding is CP950, similar to Big5. */
{
char input[] = "<\244\351\245\273\273\171>"; /* "<日本語>" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == '<');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 2, &state);
ASSERT (ret == 2);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x65E5);
ASSERT (mbsinit (&state));
input[1] = '\0';
input[2] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 3, 1, &state);
ASSERT (ret == (size_t)(-2));
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (!mbsinit (&state));
input[3] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 4, 4, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x672C);
ASSERT (mbsinit (&state));
input[4] = '\0';
/* Test support of NULL first argument. */
ret = mbrtowc (NULL, input + 5, 3, &state);
ASSERT (ret == 2);
ASSERT (mbsinit (&state));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 5, 3, &state);
ASSERT (ret == 2);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x8A9E);
ASSERT (mbsinit (&state));
input[5] = '\0';
input[6] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 7, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == '>');
ASSERT (mbsinit (&state));
/* Test some invalid input. */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\377", 1, &state); /* 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\225\377", 2, &state); /* 0x95 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
}
return 0;
case 936:
/* Locale encoding is CP936 = GBK, an extension of GB2312. */
{
char input[] = "<\310\325\261\276\325\132>"; /* "<日本語>" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == '<');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 2, &state);
ASSERT (ret == 2);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x65E5);
ASSERT (mbsinit (&state));
input[1] = '\0';
input[2] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 3, 1, &state);
ASSERT (ret == (size_t)(-2));
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (!mbsinit (&state));
input[3] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 4, 4, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x672C);
ASSERT (mbsinit (&state));
input[4] = '\0';
/* Test support of NULL first argument. */
ret = mbrtowc (NULL, input + 5, 3, &state);
ASSERT (ret == 2);
ASSERT (mbsinit (&state));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 5, 3, &state);
ASSERT (ret == 2);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x8A9E);
ASSERT (mbsinit (&state));
input[5] = '\0';
input[6] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 7, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == '>');
ASSERT (mbsinit (&state));
/* Test some invalid input. */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\377", 1, &state); /* 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\225\377", 2, &state); /* 0x95 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
}
return 0;
case 54936:
/* Locale encoding is CP54936 = GB18030. */
{
char input[] = "B\250\271\201\060\211\070er"; /* "Büßer" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'B');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 1, &state);
ASSERT (ret == (size_t)(-2));
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (!mbsinit (&state));
input[1] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 2, 7, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x00FC);
ASSERT (mbsinit (&state));
input[2] = '\0';
/* Test support of NULL first argument. */
ret = mbrtowc (NULL, input + 3, 6, &state);
ASSERT (ret == 4);
ASSERT (mbsinit (&state));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 3, 6, &state);
ASSERT (ret == 4);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x00DF);
ASSERT (mbsinit (&state));
input[3] = '\0';
input[4] = '\0';
input[5] = '\0';
input[6] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 7, 2, &state);
ASSERT (ret == 1);
ASSERT (wc == 'e');
ASSERT (mbsinit (&state));
input[5] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 8, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'r');
ASSERT (mbsinit (&state));
/* Test some invalid input. */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\377", 1, &state); /* 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\225\377", 2, &state); /* 0x95 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\201\045", 2, &state); /* 0x81 0x25 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\201\060\377", 3, &state); /* 0x81 0x30 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\201\060\377\064", 4, &state); /* 0x81 0x30 0xFF 0x34 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\201\060\211\072", 4, &state); /* 0x81 0x30 0x89 0x3A */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
}
return 0;
case 65001:
/* Locale encoding is CP65001 = UTF-8. */
{
char input[] = "B\303\274\303\237er"; /* "Büßer" */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'B');
ASSERT (mbsinit (&state));
input[0] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 1, 1, &state);
ASSERT (ret == (size_t)(-2));
ASSERT (wc == (wchar_t) 0xBADFACE);
ASSERT (!mbsinit (&state));
input[1] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 2, 5, &state);
ASSERT (ret == 1);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x00FC);
ASSERT (mbsinit (&state));
input[2] = '\0';
/* Test support of NULL first argument. */
ret = mbrtowc (NULL, input + 3, 4, &state);
ASSERT (ret == 2);
ASSERT (mbsinit (&state));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 3, 4, &state);
ASSERT (ret == 2);
ASSERT (wctob (wc) == EOF);
ASSERT (wc == 0x00DF);
ASSERT (mbsinit (&state));
input[3] = '\0';
input[4] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 5, 2, &state);
ASSERT (ret == 1);
ASSERT (wc == 'e');
ASSERT (mbsinit (&state));
input[5] = '\0';
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, input + 6, 1, &state);
ASSERT (ret == 1);
ASSERT (wc == 'r');
ASSERT (mbsinit (&state));
/* Test some invalid input. */
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\377", 1, &state); /* 0xFF */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\303\300", 2, &state); /* 0xC3 0xC0 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\343\300", 2, &state); /* 0xE3 0xC0 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\343\300\200", 3, &state); /* 0xE3 0xC0 0x80 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\343\200\300", 3, &state); /* 0xE3 0x80 0xC0 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\363\300", 2, &state); /* 0xF3 0xC0 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\363\300\200\200", 4, &state); /* 0xF3 0xC0 0x80 0x80 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\363\200\300", 3, &state); /* 0xF3 0x80 0xC0 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\363\200\300\200", 4, &state); /* 0xF3 0x80 0xC0 0x80 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
memset (&state, '\0', sizeof (mbstate_t));
wc = (wchar_t) 0xBADFACE;
ret = mbrtowc (&wc, "\363\200\200\300", 4, &state); /* 0xF3 0x80 0x80 0xC0 */
ASSERT (ret == (size_t)-1);
ASSERT (errno == EILSEQ);
}
return 0;
default:
return 1;
}
}
static int
cnt(const char *file)
{
struct stat sb;
uintmax_t linect, wordct, charct, llct, tmpll;
int fd, len, warned;
size_t clen;
short gotsp;
u_char *p;
u_char buf[MAXBSIZE];
wchar_t wch;
mbstate_t mbs;
linect = wordct = charct = llct = tmpll = 0;
if (file == NULL)
fd = STDIN_FILENO;
else {
if ((fd = open(file, O_RDONLY, 0)) < 0) {
warn("%s: open", file);
return (1);
}
if (doword || (domulti && MB_CUR_MAX != 1))
goto word;
/*
* Line counting is split out because it's a lot faster to get
* lines than to get words, since the word count requires some
* logic.
*/
if (doline) {
while ((len = read(fd, buf, MAXBSIZE))) {
if (len == -1) {
warn("%s: read", file);
(void)close(fd);
return (1);
}
if (siginfo) {
show_cnt(file, linect, wordct, charct,
llct);
}
charct += len;
for (p = buf; len--; ++p)
if (*p == '\n') {
if (tmpll > llct)
llct = tmpll;
tmpll = 0;
++linect;
} else
tmpll++;
}
reset_siginfo();
tlinect += linect;
if (dochar)
tcharct += charct;
if (dolongline) {
if (llct > tlongline)
tlongline = llct;
}
show_cnt(file, linect, wordct, charct, llct);
(void)close(fd);
return (0);
}
/*
* If all we need is the number of characters and it's a
* regular file, just stat the puppy.
*/
if (dochar || domulti) {
if (fstat(fd, &sb)) {
warn("%s: fstat", file);
(void)close(fd);
return (1);
}
if (S_ISREG(sb.st_mode)) {
reset_siginfo();
charct = sb.st_size;
show_cnt(file, linect, wordct, charct, llct);
tcharct += charct;
(void)close(fd);
return (0);
}
}
}
/* Do it the hard way... */
word: gotsp = 1;
warned = 0;
memset(&mbs, 0, sizeof(mbs));
while ((len = read(fd, buf, MAXBSIZE)) != 0) {
if (len == -1) {
warn("%s: read", file != NULL ? file : "stdin");
(void)close(fd);
return (1);
}
p = buf;
while (len > 0) {
if (siginfo)
show_cnt(file, linect, wordct, charct, llct);
if (!domulti || MB_CUR_MAX == 1) {
clen = 1;
wch = (unsigned char)*p;
} else if ((clen = mbrtowc(&wch, p, len, &mbs)) ==
(size_t)-1) {
if (!warned) {
errno = EILSEQ;
warn("%s",
file != NULL ? file : "stdin");
warned = 1;
}
memset(&mbs, 0, sizeof(mbs));
clen = 1;
wch = (unsigned char)*p;
} else if (clen == (size_t)-2)
break;
else if (clen == 0)
clen = 1;
charct++;
if (wch != L'\n')
tmpll++;
len -= clen;
p += clen;
if (wch == L'\n') {
if (tmpll > llct)
llct = tmpll;
tmpll = 0;
++linect;
}
if (iswspace(wch))
gotsp = 1;
else if (gotsp) {
gotsp = 0;
++wordct;
}
}
}
reset_siginfo();
if (domulti && MB_CUR_MAX > 1)
if (mbrtowc(NULL, NULL, 0, &mbs) == (size_t)-1 && !warned)
warn("%s", file != NULL ? file : "stdin");
if (doline)
tlinect += linect;
if (doword)
twordct += wordct;
if (dochar || domulti)
tcharct += charct;
if (dolongline) {
if (llct > tlongline)
tlongline = llct;
}
show_cnt(file, linect, wordct, charct, llct);
(void)close(fd);
return (0);
}
