static int lexstring(js_State *J)
{
const char *s;
int q = J->lexchar;
jsY_next(J);
textinit(J);
while (J->lexchar != q) {
if (J->lexchar == 0 || J->lexchar == '\n')
jsY_error(J, "string not terminated");
if (jsY_accept(J, '\\')) {
if (lexescape(J))
jsY_error(J, "malformed escape sequence");
} else {
textpush(J, J->lexchar);
jsY_next(J);
}
}
jsY_expect(J, q);
s = textend(J);
J->text = js_intern(J, s);
return TK_STRING;
}
/*
* Read a regular expression.
*/
static int
lexregexp(wint_t endc)
{
(void) lexescape(endc, 1, 0);
yylval.node = renode(linebuf);
return (URE);
}
/*
* Process a string, converting the escape characters as required by
* 1003.2. The processed string ends up in the global linebuf[]. This
* routine also changes the value of 'progfd' - the program file
* descriptor, so it should be used with some care. It is presently used to
* process -v (awk1.c) and var=str type arguments (awk2.c, nextrecord()).
*/
void
strescape(wchar_t *str)
{
progptr = str;
proglen = wcslen(str) + 1; /* Include \0 */
(void) lexescape('\0', 0, 1);
progptr = NULL;
}
/*
* Read a string for the lexical analyzer.
* `endc' terminates the string.
*/
static int
lexstring(wint_t endc)
{
size_t length = lexescape(endc, 0, 0);
yylval.node = stringnode(linebuf, FALLOC, length);
return (CONSTANT);
}
/*
* next - get next token
*/
static int /* 1 normal, 0 failure */
next(struct vars * v)
{
chr c;
/* errors yield an infinite sequence of failures */
if (ISERR())
return 0; /* the error has set nexttype to EOS */
/* remember flavor of last token */
v->lasttype = v->nexttype;
/* REG_BOSONLY */
if (v->nexttype == EMPTY && (v->cflags & REG_BOSONLY))
{
/* at start of a REG_BOSONLY RE */
RETV(SBEGIN, 0); /* same as \A */
}
/* if we're nested and we've hit end, return to outer level */
if (v->savenow != NULL && ATEOS())
{
v->now = v->savenow;
v->stop = v->savestop;
v->savenow = v->savestop = NULL;
}
/* skip white space etc. if appropriate (not in literal or []) */
if (v->cflags & REG_EXPANDED)
switch (v->lexcon)
{
case L_ERE:
case L_BRE:
case L_EBND:
case L_BBND:
skip(v);
break;
}
/* handle EOS, depending on context */
if (ATEOS())
{
switch (v->lexcon)
{
case L_ERE:
case L_BRE:
case L_Q:
RET(EOS);
break;
case L_EBND:
case L_BBND:
FAILW(REG_EBRACE);
break;
case L_BRACK:
case L_CEL:
case L_ECL:
case L_CCL:
FAILW(REG_EBRACK);
break;
}
assert(NOTREACHED);
}
/* okay, time to actually get a character */
c = *v->now++;
/* deal with the easy contexts, punt EREs to code below */
switch (v->lexcon)
{
case L_BRE: /* punt BREs to separate function */
return brenext(v, c);
break;
case L_ERE: /* see below */
break;
case L_Q: /* literal strings are easy */
RETV(PLAIN, c);
break;
case L_BBND: /* bounds are fairly simple */
case L_EBND:
switch (c)
{
case CHR('0'):
case CHR('1'):
case CHR('2'):
case CHR('3'):
case CHR('4'):
case CHR('5'):
case CHR('6'):
case CHR('7'):
case CHR('8'):
case CHR('9'):
RETV(DIGIT, (chr) DIGITVAL(c));
break;
case CHR(','):
RET(',');
break;
case CHR('}'): /* ERE bound ends with } */
if (INCON(L_EBND))
{
INTOCON(L_ERE);
if ((v->cflags & REG_ADVF) && NEXT1('?'))
{
v->now++;
NOTE(REG_UNONPOSIX);
RETV('}', 0);
}
RETV('}', 1);
}
else
FAILW(REG_BADBR);
break;
case CHR('\\'): /* BRE bound ends with \} */
if (INCON(L_BBND) && NEXT1('}'))
{
v->now++;
INTOCON(L_BRE);
RET('}');
}
else
FAILW(REG_BADBR);
break;
default:
FAILW(REG_BADBR);
break;
}
assert(NOTREACHED);
break;
case L_BRACK: /* brackets are not too hard */
switch (c)
{
case CHR(']'):
if (LASTTYPE('['))
RETV(PLAIN, c);
else
{
INTOCON((v->cflags & REG_EXTENDED) ?
L_ERE : L_BRE);
RET(']');
}
break;
case CHR('\\'):
NOTE(REG_UBBS);
if (!(v->cflags & REG_ADVF))
RETV(PLAIN, c);
NOTE(REG_UNONPOSIX);
if (ATEOS())
FAILW(REG_EESCAPE);
(DISCARD) lexescape(v);
switch (v->nexttype)
{ /* not all escapes okay here */
case PLAIN:
return 1;
break;
case CCLASS:
switch (v->nextvalue)
{
case 'd':
lexnest(v, brbackd, ENDOF(brbackd));
break;
case 's':
lexnest(v, brbacks, ENDOF(brbacks));
break;
case 'w':
lexnest(v, brbackw, ENDOF(brbackw));
break;
default:
FAILW(REG_EESCAPE);
break;
}
/* lexnest done, back up and try again */
v->nexttype = v->lasttype;
return next(v);
break;
}
/* not one of the acceptable escapes */
FAILW(REG_EESCAPE);
break;
case CHR('-'):
if (LASTTYPE('[') || NEXT1(']'))
RETV(PLAIN, c);
else
RETV(RANGE, c);
break;
case CHR('['):
if (ATEOS())
FAILW(REG_EBRACK);
switch (*v->now++)
{
case CHR('.'):
INTOCON(L_CEL);
/* might or might not be locale-specific */
RET(COLLEL);
break;
case CHR('='):
INTOCON(L_ECL);
NOTE(REG_ULOCALE);
RET(ECLASS);
break;
case CHR(':'):
INTOCON(L_CCL);
NOTE(REG_ULOCALE);
RET(CCLASS);
break;
default: /* oops */
v->now--;
RETV(PLAIN, c);
break;
}
assert(NOTREACHED);
break;
default:
RETV(PLAIN, c);
break;
}
assert(NOTREACHED);
break;
case L_CEL: /* collating elements are easy */
if (c == CHR('.') && NEXT1(']'))
{
v->now++;
INTOCON(L_BRACK);
RETV(END, '.');
}
else
RETV(PLAIN, c);
break;
case L_ECL: /* ditto equivalence classes */
if (c == CHR('=') && NEXT1(']'))
{
v->now++;
INTOCON(L_BRACK);
RETV(END, '=');
}
else
RETV(PLAIN, c);
break;
case L_CCL: /* ditto character classes */
if (c == CHR(':') && NEXT1(']'))
{
v->now++;
INTOCON(L_BRACK);
RETV(END, ':');
}
else
RETV(PLAIN, c);
break;
default:
assert(NOTREACHED);
break;
}
/* that got rid of everything except EREs and AREs */
assert(INCON(L_ERE));
/* deal with EREs and AREs, except for backslashes */
switch (c)
{
case CHR('|'):
RET('|');
break;
case CHR('*'):
if ((v->cflags & REG_ADVF) && NEXT1('?'))
{
v->now++;
NOTE(REG_UNONPOSIX);
RETV('*', 0);
}
RETV('*', 1);
break;
case CHR('+'):
if ((v->cflags & REG_ADVF) && NEXT1('?'))
{
v->now++;
NOTE(REG_UNONPOSIX);
RETV('+', 0);
}
RETV('+', 1);
break;
case CHR('?'):
if ((v->cflags & REG_ADVF) && NEXT1('?'))
{
v->now++;
NOTE(REG_UNONPOSIX);
RETV('?', 0);
}
RETV('?', 1);
break;
case CHR('{'): /* bounds start or plain character */
if (v->cflags & REG_EXPANDED)
skip(v);
if (ATEOS() || !iscdigit(*v->now))
{
NOTE(REG_UBRACES);
NOTE(REG_UUNSPEC);
RETV(PLAIN, c);
}
else
{
NOTE(REG_UBOUNDS);
INTOCON(L_EBND);
RET('{');
}
assert(NOTREACHED);
break;
case CHR('('): /* parenthesis, or advanced extension */
if ((v->cflags & REG_ADVF) && NEXT1('?'))
{
NOTE(REG_UNONPOSIX);
v->now++;
if (ATEOS())
FAILW(REG_BADRPT);
switch (*v->now++)
{
case CHR(':'): /* non-capturing paren */
RETV('(', 0);
break;
case CHR('#'): /* comment */
while (!ATEOS() && *v->now != CHR(')'))
v->now++;
if (!ATEOS())
v->now++;
assert(v->nexttype == v->lasttype);
return next(v);
break;
case CHR('='): /* positive lookahead */
NOTE(REG_ULOOKAROUND);
RETV(LACON, LATYPE_AHEAD_POS);
break;
case CHR('!'): /* negative lookahead */
NOTE(REG_ULOOKAROUND);
RETV(LACON, LATYPE_AHEAD_NEG);
break;
case CHR('<'):
if (ATEOS())
FAILW(REG_BADRPT);
switch (*v->now++)
{
case CHR('='): /* positive lookbehind */
NOTE(REG_ULOOKAROUND);
RETV(LACON, LATYPE_BEHIND_POS);
break;
case CHR('!'): /* negative lookbehind */
NOTE(REG_ULOOKAROUND);
RETV(LACON, LATYPE_BEHIND_NEG);
break;
default:
FAILW(REG_BADRPT);
break;
}
assert(NOTREACHED);
break;
default:
FAILW(REG_BADRPT);
break;
}
assert(NOTREACHED);
}
if (v->cflags & REG_NOSUB)
RETV('(', 0); /* all parens non-capturing */
else
RETV('(', 1);
break;
case CHR(')'):
if (LASTTYPE('('))
NOTE(REG_UUNSPEC);
RETV(')', c);
break;
case CHR('['): /* easy except for [[:<:]] and [[:>:]] */
if (HAVE(6) && *(v->now + 0) == CHR('[') &&
*(v->now + 1) == CHR(':') &&
(*(v->now + 2) == CHR('<') ||
*(v->now + 2) == CHR('>')) &&
*(v->now + 3) == CHR(':') &&
*(v->now + 4) == CHR(']') &&
*(v->now + 5) == CHR(']'))
{
c = *(v->now + 2);
v->now += 6;
NOTE(REG_UNONPOSIX);
RET((c == CHR('<')) ? '<' : '>');
}
INTOCON(L_BRACK);
if (NEXT1('^'))
{
v->now++;
RETV('[', 0);
}
RETV('[', 1);
break;
case CHR('.'):
RET('.');
break;
case CHR('^'):
RET('^');
break;
case CHR('$'):
RET('$');
break;
case CHR('\\'): /* mostly punt backslashes to code below */
if (ATEOS())
FAILW(REG_EESCAPE);
break;
default: /* ordinary character */
RETV(PLAIN, c);
break;
}
/* ERE/ARE backslash handling; backslash already eaten */
assert(!ATEOS());
if (!(v->cflags & REG_ADVF))
{ /* only AREs have non-trivial escapes */
if (iscalnum(*v->now))
{
NOTE(REG_UBSALNUM);
NOTE(REG_UUNSPEC);
}
RETV(PLAIN, *v->now++);
}
(DISCARD) lexescape(v);
if (ISERR())
FAILW(REG_EESCAPE);
if (v->nexttype == CCLASS)
{ /* fudge at lexical level */
switch (v->nextvalue)
{
case 'd':
lexnest(v, backd, ENDOF(backd));
break;
case 'D':
lexnest(v, backD, ENDOF(backD));
break;
case 's':
lexnest(v, backs, ENDOF(backs));
break;
case 'S':
lexnest(v, backS, ENDOF(backS));
break;
case 'w':
lexnest(v, backw, ENDOF(backw));
break;
case 'W':
lexnest(v, backW, ENDOF(backW));
break;
default:
assert(NOTREACHED);
FAILW(REG_ASSERT);
break;
}
/* lexnest done, back up and try again */
v->nexttype = v->lasttype;
return next(v);
}
/* otherwise, lexescape has already done the work */
return !ISERR();
}
