static void
map_funny(char *str)
{
register char *from = str;
register char *to = str;
/*
** Fix 9005: using wrong test, wasn't mapping Kanji chars, so VMS
** filenames were bad.
** WARNING: this means we won't support Kanji filenames!
** Also, this doesn't solve the problem of 8-bit chars in the
** European character sets. We need a CMlegal_filename_char
** routine.
*/
/* map any funny bytes to underscores. */
while (*from != EOS)
{
i4 cnt = CMbytecnt(from);
if ( (!CMalpha(from) && !CMdigit(from)) || CMdbl1st(from) )
*to++ = '_';
else
*to++ = *from;
from += cnt;
}
*to = EOS;
}
/*
* regatom - the lowest level
*
* Optimization: gobbles an entire sequence of ordinary characters so that
* it can turn them into a single node, which is smaller to store and
* faster to run. Backslashed characters are exceptions, each becoming a
* separate node; the code is simpler that way and it's not worth fixing.
*/
static char *
regatom(i4 *flagp)
{
register char *ret;
i4 flags;
char null_byte = '\0';
*flagp = WORST; /* Tentatively. */
switch (*regparse) {
case '^':
CMnext( regparse );
ret = regnode(BOL);
break;
case '$':
CMnext( regparse );
ret = regnode(EOL);
break;
case '.':
CMnext( regparse );
ret = regnode(ANY);
*flagp |= HASWIDTH|SIMPLE;
break;
case '[': {
char *range_start = NULL;
bool double_start;
u_i2 first_u2, last_u2;
u_char first_u1, last_u1;
CMnext( regparse );
if (*regparse == '^') { /* Complement of range. */
ret = regnode(ANYBUT);
CMnext( regparse );
} else
ret = regnode(ANYOF);
if (*regparse == ']' || *regparse == '-') {
regc( regparse );
CMnext( regparse );
}
while (*regparse != '\0' && *regparse != ']') {
if (*regparse == '-') {
char range_op = '-';
CMnext( regparse );
if( *regparse == ']' ||
*regparse == '\0'
)
regc( &range_op );
else {
char *tmp;
bool invalid = FALSE;
bool double_end;
if( range_start == NULL )
invalid = TRUE;
double_end =
CMdbl1st( regparse );
if( !invalid &&
double_end
&& !double_start
)
invalid = TRUE;
if( !invalid &&
double_start
&& !double_start
)
invalid = TRUE;
if( !invalid &&
CMcmpcase( range_start,
regparse ) > 0
)
invalid = TRUE;
if( double_start )
_FAIL("don't know how to support character classes containing double-byte ranges");
if( invalid )
_FAIL("invalid [] range");
/* no double-byte ranges! */
/*
** Initialize the value for the end of the range.
*/
last_u1 = UCHARAT(regparse);
for (; first_u1 <= last_u1;
first_u1++
)
regc( (char *)
&first_u1 );
CMnext( regparse );
}
} else {
range_start = regparse;
if( CMdbl1st( range_start ) )
{
double_start = TRUE;
first_u2 = *(u_i2 *) range_start;
}
else
{
double_start = FALSE;
first_u1 = UCHARAT(range_start);
}
regc( regparse );
CMnext( regparse );
}
}
regc( &null_byte );
if (*regparse != ']')
_FAIL("unmatched []");
CMnext( regparse );
*flagp |= HASWIDTH|SIMPLE;
}
break;
case '(':
CMnext( regparse );
ret = reg(1, &flags);
if (ret == NULL)
return(NULL);
*flagp |= flags&(HASWIDTH|SPSTART);
break;
case '\0':
case '|':
case ')':
CMnext( regparse );
_FAIL("internal urp"); /* Supposed to be caught earlier. */
break;
case '?':
case '+':
case '*':
CMnext( regparse );
_FAIL("?+* follows nothing");
break;
case '\\':
CMnext( regparse );
if (*regparse == '\0')
_FAIL("trailing \\");
ret = regnode(EXACTLY);
regc( regparse );
CMnext( regparse );
regc( &null_byte );
*flagp |= HASWIDTH|SIMPLE;
break;
default: {
register i4 len;
register char ender;
len = my_strcspn(regparse, META);
if (len <= 0)
_FAIL("internal disaster");
ender = *(regparse+len);
if (len > 1 && ISMULT(ender))
len--; /* Back off clear of ?+* operand. */
*flagp |= HASWIDTH;
if (len == 1)
*flagp |= SIMPLE;
ret = regnode(EXACTLY);
while (len > 0) {
regc( regparse );
CMbytedec( len, regparse );
CMnext( regparse );
}
regc( &null_byte );
}
break;
}
return(ret);
}
static
i4
do_readln(FILE *fp,char *buf_ptr,i2 maxchar)
{
char *ptr; /* Working pointer into line buffer */
i4 count; /* Number of chars read */
STATUS status;
ptr = buf_ptr;
while (maxchar > 0)
{
status = SIread(fp,sizeof(char),&count,ptr);
CMbytedec(maxchar,ptr);
if ((CMdbl1st(ptr)) && (maxchar > 0))
{
status = SIread(fp,sizeof(char),&count,(ptr + 1));
CMbytedec(maxchar,ptr);
}
switch(status)
{
case(ENDFILE):
*ptr = EOS;
count = STlength(buf_ptr);
if (count > 0)
{
/*
** We hit EOF on a file that ended abruptly
** without a final end-of-line character.
** Return the final line's count this time -
** we'll return EOF on the next call.
*/
return(count);
}
return(-1);
break;
case(OK):
if ((*ptr == CR) || (*ptr == LF) || (*ptr == FF))
{
/*
** Do NOT include the end-of-line character!
*/
*ptr = EOS;
return(STlength(buf_ptr));
}
CMnext(ptr);
break;
default:
/*
** According to the documentation, we shouldn't
** be able to get here.
*/
IIUGerr(E_DE000E_Dobj_read_fail,UG_ERR_ERROR,0);
return(-1);
}
}
/*
** If we wind up here, then
** we've exceeded maxchars!
*/
*ptr = EOS;
return(STlength(buf_ptr));
}