/* toke.c
*
- * Copyright (c) 1991-1997, Larry Wall
+ * Copyright (c) 1991-1999, Larry Wall
*
* You may distribute under the terms of either the GNU General Public
* License or the Artistic License, as specified in the README file.
* "It all comes from here, the stench and the peril." --Frodo
*/
+/* toke.c
+ *
+ * This file is the tokenizer for Perl. It's closely linked to the
+ * parser, perly.y.
+ *
+ * The main routine is yylex(), which returns the next token.
+ */
+
#include "EXTERN.h"
+#define PERL_IN_TOKE_C
#include "perl.h"
#define yychar PL_yychar
#define yylval PL_yylval
-#ifndef PERL_OBJECT
-static void check_uni _((void));
-static void force_next _((I32 type));
-static char *force_version _((char *start));
-static char *force_word _((char *start, int token, int check_keyword, int allow_pack, int allow_tick));
-static SV *tokeq _((SV *sv));
-static char *scan_const _((char *start));
-static char *scan_formline _((char *s));
-static char *scan_heredoc _((char *s));
-static char *scan_ident _((char *s, char *send, char *dest, STRLEN destlen,
- I32 ck_uni));
-static char *scan_inputsymbol _((char *start));
-static char *scan_pat _((char *start, I32 type));
-static char *scan_str _((char *start));
-static char *scan_subst _((char *start));
-static char *scan_trans _((char *start));
-static char *scan_word _((char *s, char *dest, STRLEN destlen,
- int allow_package, STRLEN *slp));
-static char *skipspace _((char *s));
-static void checkcomma _((char *s, char *name, char *what));
-static void force_ident _((char *s, int kind));
-static void incline _((char *s));
-static int intuit_method _((char *s, GV *gv));
-static int intuit_more _((char *s));
-static I32 lop _((I32 f, expectation x, char *s));
-static void missingterm _((char *s));
-static void no_op _((char *what, char *s));
-static void set_csh _((void));
-static I32 sublex_done _((void));
-static I32 sublex_push _((void));
-static I32 sublex_start _((void));
-#ifdef CRIPPLED_CC
-static int uni _((I32 f, char *s));
-#endif
-static char * filter_gets _((SV *sv, PerlIO *fp, STRLEN append));
-static void restore_rsfp _((void *f));
-static SV *new_constant _((char *s, STRLEN len, char *key, SV *sv, SV *pv, char *type));
-static void restore_expect _((void *e));
-static void restore_lex_expect _((void *e));
-#endif /* PERL_OBJECT */
-
static char ident_too_long[] = "Identifier too long";
-#define UTF (PL_hints & HINT_UTF8)
+static void restore_rsfp(pTHXo_ void *f);
+static void restore_expect(pTHXo_ void *e);
+static void restore_lex_expect(pTHXo_ void *e);
-/* The following are arranged oddly so that the guard on the switch statement
+#define UTF (PL_hints & HINT_UTF8)
+/*
+ * Note: we try to be careful never to call the isXXX_utf8() functions
+ * unless we're pretty sure we've seen the beginning of a UTF-8 character
+ * (that is, the two high bits are set). Otherwise we risk loading in the
+ * heavy-duty SWASHINIT and SWASHGET routines unnecessarily.
+ */
+#define isIDFIRST_lazy(p) ((!UTF || (*((U8*)p) < 0xc0)) \
+ ? isIDFIRST(*(p)) \
+ : isIDFIRST_utf8((U8*)p))
+#define isALNUM_lazy(p) ((!UTF || (*((U8*)p) < 0xc0)) \
+ ? isALNUM(*(p)) \
+ : isALNUM_utf8((U8*)p))
+
+/* In variables name $^X, these are the legal values for X.
+ * 1999-02-27 mjd-perl-patch@plover.com */
+#define isCONTROLVAR(x) (isUPPER(x) || strchr("[\\]^_?", (x)))
+
+/* LEX_* are values for PL_lex_state, the state of the lexer.
+ * They are arranged oddly so that the guard on the switch statement
* can get by with a single comparison (if the compiler is smart enough).
*/
#undef ff_next
#endif
+#ifdef USE_PURE_BISON
+YYSTYPE* yylval_pointer = NULL;
+int* yychar_pointer = NULL;
+# undef yylval
+# undef yychar
+# define yylval (*yylval_pointer)
+# define yychar (*yychar_pointer)
+# define PERL_YYLEX_PARAM yylval_pointer,yychar_pointer
+# undef yylex
+# define yylex() Perl_yylex(aTHX_ yylval_pointer, yychar_pointer)
+#endif
+
#include "keywords.h"
+/* CLINE is a macro that ensures PL_copline has a sane value */
+
#ifdef CLINE
#undef CLINE
#endif
#define CLINE (PL_copline = (PL_curcop->cop_line < PL_copline ? PL_curcop->cop_line : PL_copline))
+/*
+ * Convenience functions to return different tokens and prime the
+ * tokenizer for the next token. They all take an argument.
+ *
+ * TOKEN : generic token (used for '(', DOLSHARP, etc)
+ * OPERATOR : generic operator
+ * AOPERATOR : assignment operator
+ * PREBLOCK : beginning the block after an if, while, foreach, ...
+ * PRETERMBLOCK : beginning a non-code-defining {} block (eg, hash ref)
+ * PREREF : *EXPR where EXPR is not a simple identifier
+ * TERM : expression term
+ * LOOPX : loop exiting command (goto, last, dump, etc)
+ * FTST : file test operator
+ * FUN0 : zero-argument function
+ * FUN1 : not used
+ * BOop : bitwise or or xor
+ * BAop : bitwise and
+ * SHop : shift operator
+ * PWop : power operator
+ * PMop : matching operator
+ * Aop : addition-level operator
+ * Mop : multiplication-level operator
+ * Eop : equality-testing operator
+ * Rop : relational operator <= != gt
+ *
+ * Also see LOP and lop() below.
+ */
+
#define TOKEN(retval) return (PL_bufptr = s,(int)retval)
#define OPERATOR(retval) return (PL_expect = XTERM,PL_bufptr = s,(int)retval)
#define AOPERATOR(retval) return ao((PL_expect = XTERM,PL_bufptr = s,(int)retval))
/* grandfather return to old style */
#define OLDLOP(f) return(yylval.ival=f,PL_expect = XTERM,PL_bufptr = s,(int)LSTOP)
+/*
+ * S_ao
+ *
+ * This subroutine detects &&= and ||= and turns an ANDAND or OROR
+ * into an OP_ANDASSIGN or OP_ORASSIGN
+ */
+
STATIC int
-ao(int toketype)
+S_ao(pTHX_ int toketype)
{
if (*PL_bufptr == '=') {
PL_bufptr++;
return toketype;
}
+/*
+ * S_no_op
+ * When Perl expects an operator and finds something else, no_op
+ * prints the warning. It always prints "<something> found where
+ * operator expected. It prints "Missing semicolon on previous line?"
+ * if the surprise occurs at the start of the line. "do you need to
+ * predeclare ..." is printed out for code like "sub bar; foo bar $x"
+ * where the compiler doesn't know if foo is a method call or a function.
+ * It prints "Missing operator before end of line" if there's nothing
+ * after the missing operator, or "... before <...>" if there is something
+ * after the missing operator.
+ */
+
STATIC void
-no_op(char *what, char *s)
+S_no_op(pTHX_ char *what, char *s)
{
char *oldbp = PL_bufptr;
bool is_first = (PL_oldbufptr == PL_linestart);
+ assert(s >= oldbp);
PL_bufptr = s;
- yywarn(form("%s found where operator expected", what));
+ yywarn(Perl_form(aTHX_ "%s found where operator expected", what));
if (is_first)
- warn("\t(Missing semicolon on previous line?)\n");
- else if (PL_oldoldbufptr && isIDFIRST(*PL_oldoldbufptr)) {
+ Perl_warn(aTHX_ "\t(Missing semicolon on previous line?)\n");
+ else if (PL_oldoldbufptr && isIDFIRST_lazy(PL_oldoldbufptr)) {
char *t;
- for (t = PL_oldoldbufptr; *t && (isALNUM(*t) || *t == ':'); t++) ;
+ for (t = PL_oldoldbufptr; *t && (isALNUM_lazy(t) || *t == ':'); t++) ;
if (t < PL_bufptr && isSPACE(*t))
- warn("\t(Do you need to predeclare %.*s?)\n",
+ Perl_warn(aTHX_ "\t(Do you need to predeclare %.*s?)\n",
t - PL_oldoldbufptr, PL_oldoldbufptr);
-
}
else
- warn("\t(Missing operator before %.*s?)\n", s - oldbp, oldbp);
+ Perl_warn(aTHX_ "\t(Missing operator before %.*s?)\n", s - oldbp, oldbp);
PL_bufptr = oldbp;
}
+/*
+ * S_missingterm
+ * Complain about missing quote/regexp/heredoc terminator.
+ * If it's called with (char *)NULL then it cauterizes the line buffer.
+ * If we're in a delimited string and the delimiter is a control
+ * character, it's reformatted into a two-char sequence like ^C.
+ * This is fatal.
+ */
+
STATIC void
-missingterm(char *s)
+S_missingterm(pTHX_ char *s)
{
char tmpbuf[3];
char q;
s = tmpbuf;
}
q = strchr(s,'"') ? '\'' : '"';
- croak("Can't find string terminator %c%s%c anywhere before EOF",q,s,q);
+ Perl_croak(aTHX_ "Can't find string terminator %c%s%c anywhere before EOF",q,s,q);
}
+/*
+ * Perl_deprecate
+ * Warns that something is deprecated. Duh.
+ */
+
void
-deprecate(char *s)
+Perl_deprecate(pTHX_ char *s)
{
dTHR;
if (ckWARN(WARN_DEPRECATED))
- warner(WARN_DEPRECATED, "Use of %s is deprecated", s);
+ Perl_warner(aTHX_ WARN_DEPRECATED, "Use of %s is deprecated", s);
}
+/*
+ * depcom
+ * Deprecate a comma-less variable list. Called from three places
+ * in the tokenizer.
+ */
+
STATIC void
-depcom(void)
+S_depcom(pTHX)
{
deprecate("comma-less variable list");
}
+/*
+ * text filters for win32 carriage-returns, utf16-to-utf8 and
+ * utf16-to-utf8-reversed, whatever that is.
+ */
+
#ifdef WIN32
STATIC I32
-win32_textfilter(int idx, SV *sv, int maxlen)
+S_win32_textfilter(pTHX_ int idx, SV *sv, int maxlen)
{
I32 count = FILTER_READ(idx+1, sv, maxlen);
if (count > 0 && !maxlen)
}
#endif
-#ifndef PERL_OBJECT
-
STATIC I32
-utf16_textfilter(int idx, SV *sv, int maxlen)
+S_utf16_textfilter(pTHX_ int idx, SV *sv, int maxlen)
{
I32 count = FILTER_READ(idx+1, sv, maxlen);
if (count) {
}
STATIC I32
-utf16rev_textfilter(int idx, SV *sv, int maxlen)
+S_utf16rev_textfilter(pTHX_ int idx, SV *sv, int maxlen)
{
I32 count = FILTER_READ(idx+1, sv, maxlen);
if (count) {
return count;
}
-#endif
+/*
+ * Perl_lex_start
+ * Initialize variables. Called by perl.c. It uses the Perl stack
+ * to save its state (for recursive calls to the parser).
+ */
void
-lex_start(SV *line)
+Perl_lex_start(pTHX_ SV *line)
{
dTHR;
char *s;
PL_oldoldbufptr = PL_oldbufptr = PL_bufptr = PL_linestart = SvPVX(PL_linestr);
PL_bufend = PL_bufptr + SvCUR(PL_linestr);
SvREFCNT_dec(PL_rs);
- PL_rs = newSVpv("\n", 1);
+ PL_rs = newSVpvn("\n", 1);
PL_rsfp = 0;
}
+/*
+ * Perl_lex_end
+ * Tidy up. Called from pp_ctl.c in the sv_compile_2op(), doeval(),
+ * and pp_leaveeval() subroutines.
+ */
+
void
-lex_end(void)
+Perl_lex_end(pTHX)
{
PL_doextract = FALSE;
}
-STATIC void
-restore_rsfp(void *f)
-{
- PerlIO *fp = (PerlIO*)f;
-
- if (PL_rsfp == PerlIO_stdin())
- PerlIO_clearerr(PL_rsfp);
- else if (PL_rsfp && (PL_rsfp != fp))
- PerlIO_close(PL_rsfp);
- PL_rsfp = fp;
-}
-
-STATIC void
-restore_expect(void *e)
-{
- /* a safe way to store a small integer in a pointer */
- PL_expect = (expectation)((char *)e - PL_tokenbuf);
-}
-
-STATIC void
-restore_lex_expect(void *e)
-{
- /* a safe way to store a small integer in a pointer */
- PL_lex_expect = (expectation)((char *)e - PL_tokenbuf);
-}
+/*
+ * S_incline
+ * This subroutine has nothing to do with tilting, whether at windmills
+ * or pinball tables. Its name is short for "increment line". It
+ * increments the current line number in PL_curcop->cop_line and checks
+ * to see whether the line starts with a comment of the form
+ * # line 500
+ * If so, it sets the current line number to the number in the comment.
+ */
STATIC void
-incline(char *s)
+S_incline(pTHX_ char *s)
{
dTHR;
char *t;
PL_curcop->cop_line = atoi(n)-1;
}
+/*
+ * S_skipspace
+ * Called to gobble the appropriate amount and type of whitespace.
+ * Skips comments as well.
+ */
+
STATIC char *
-skipspace(register char *s)
+S_skipspace(pTHX_ register char *s)
{
dTHR;
if (PL_lex_formbrack && PL_lex_brackets <= PL_lex_formbrack) {
}
for (;;) {
STRLEN prevlen;
- while (s < PL_bufend && isSPACE(*s))
- s++;
+ while (s < PL_bufend && isSPACE(*s)) {
+ if (*s++ == '\n' && PL_in_eval && !PL_rsfp)
+ incline(s);
+ }
+
+ /* comment */
if (s < PL_bufend && *s == '#') {
while (s < PL_bufend && *s != '\n')
s++;
- if (s < PL_bufend)
+ if (s < PL_bufend) {
s++;
+ if (PL_in_eval && !PL_rsfp) {
+ incline(s);
+ continue;
+ }
+ }
}
+
+ /* only continue to recharge the buffer if we're at the end
+ * of the buffer, we're not reading from a source filter, and
+ * we're in normal lexing mode
+ */
if (s < PL_bufend || !PL_rsfp || PL_lex_state != LEX_NORMAL)
return s;
+
+ /* try to recharge the buffer */
if ((s = filter_gets(PL_linestr, PL_rsfp, (prevlen = SvCUR(PL_linestr)))) == Nullch) {
+ /* end of file. Add on the -p or -n magic */
if (PL_minus_n || PL_minus_p) {
sv_setpv(PL_linestr,PL_minus_p ?
";}continue{print or die qq(-p destination: $!\\n)" :
}
else
sv_setpv(PL_linestr,";");
+
+ /* reset variables for next time we lex */
PL_oldoldbufptr = PL_oldbufptr = PL_bufptr = s = PL_linestart = SvPVX(PL_linestr);
PL_bufend = SvPVX(PL_linestr) + SvCUR(PL_linestr);
+
+ /* Close the filehandle. Could be from -P preprocessor,
+ * STDIN, or a regular file. If we were reading code from
+ * STDIN (because the commandline held no -e or filename)
+ * then we don't close it, we reset it so the code can
+ * read from STDIN too.
+ */
+
if (PL_preprocess && !PL_in_eval)
(void)PerlProc_pclose(PL_rsfp);
else if ((PerlIO*)PL_rsfp == PerlIO_stdin())
PL_rsfp = Nullfp;
return s;
}
+
+ /* not at end of file, so we only read another line */
PL_linestart = PL_bufptr = s + prevlen;
PL_bufend = s + SvCUR(PL_linestr);
s = PL_bufptr;
incline(s);
+
+ /* debugger active and we're not compiling the debugger code,
+ * so store the line into the debugger's array of lines
+ */
if (PERLDB_LINE && PL_curstash != PL_debstash) {
SV *sv = NEWSV(85,0);
}
}
+/*
+ * S_check_uni
+ * Check the unary operators to ensure there's no ambiguity in how they're
+ * used. An ambiguous piece of code would be:
+ * rand + 5
+ * This doesn't mean rand() + 5. Because rand() is a unary operator,
+ * the +5 is its argument.
+ */
+
STATIC void
-check_uni(void) {
+S_check_uni(pTHX)
+{
char *s;
- char ch;
char *t;
+ dTHR;
if (PL_oldoldbufptr != PL_last_uni)
return;
while (isSPACE(*PL_last_uni))
PL_last_uni++;
- for (s = PL_last_uni; isALNUM(*s) || *s == '-'; s++) ;
+ for (s = PL_last_uni; isALNUM_lazy(s) || *s == '-'; s++) ;
if ((t = strchr(s, '(')) && t < PL_bufptr)
return;
- ch = *s;
- *s = '\0';
- warn("Warning: Use of \"%s\" without parens is ambiguous", PL_last_uni);
- *s = ch;
+ if (ckWARN_d(WARN_AMBIGUOUS)){
+ char ch = *s;
+ *s = '\0';
+ Perl_warner(aTHX_ WARN_AMBIGUOUS,
+ "Warning: Use of \"%s\" without parens is ambiguous",
+ PL_last_uni);
+ *s = ch;
+ }
}
+/* workaround to replace the UNI() macro with a function. Only the
+ * hints/uts.sh file mentions this. Other comments elsewhere in the
+ * source indicate Microport Unix might need it too.
+ */
+
#ifdef CRIPPLED_CC
#undef UNI
#define UNI(f) return uni(f,s)
STATIC int
-uni(I32 f, char *s)
+S_uni(pTHX_ I32 f, char *s)
{
yylval.ival = f;
PL_expect = XTERM;
#endif /* CRIPPLED_CC */
+/*
+ * LOP : macro to build a list operator. Its behaviour has been replaced
+ * with a subroutine, S_lop() for which LOP is just another name.
+ */
+
#define LOP(f,x) return lop(f,x,s)
+/*
+ * S_lop
+ * Build a list operator (or something that might be one). The rules:
+ * - if we have a next token, then it's a list operator [why?]
+ * - if the next thing is an opening paren, then it's a function
+ * - else it's a list operator
+ */
+
STATIC I32
-lop(I32 f, expectation x, char *s)
+S_lop(pTHX_ I32 f, expectation x, char *s)
{
dTHR;
yylval.ival = f;
return LSTOP;
}
+/*
+ * S_force_next
+ * When the tokenizer realizes it knows the next token (for instance,
+ * it is reordering tokens for the parser) then it can call S_force_next
+ * to make the current token be the next one. It will also set
+ * PL_nextval, and possibly PL_expect to ensure the lexer handles the
+ * token correctly.
+ */
+
STATIC void
-force_next(I32 type)
+S_force_next(pTHX_ I32 type)
{
PL_nexttype[PL_nexttoke] = type;
PL_nexttoke++;
}
}
+/*
+ * S_force_word
+ * When the lexer knows the next thing is a word (for instance, it has
+ * just seen -> and it knows that the next char is a word char, then
+ * it calls S_force_word to stick the next word into the PL_next lookahead.
+ *
+ * Arguments:
+ * char *start : start of the buffer
+ * int token : PL_next will be this type of bare word (e.g., METHOD,WORD)
+ * int check_keyword : if true, Perl checks to make sure the word isn't
+ * a keyword (do this if the word is a label, e.g. goto FOO)
+ * int allow_pack : if true, : characters will also be allowed (require,
+ * use, etc. do this)
+ * int allow_initial_tick : used by the "sub" tokenizer only.
+ */
+
STATIC char *
-force_word(register char *start, int token, int check_keyword, int allow_pack, int allow_initial_tick)
+S_force_word(pTHX_ register char *start, int token, int check_keyword, int allow_pack, int allow_initial_tick)
{
register char *s;
STRLEN len;
start = skipspace(start);
s = start;
- if (isIDFIRST(*s) ||
+ if (isIDFIRST_lazy(s) ||
(allow_pack && *s == ':') ||
(allow_initial_tick && *s == '\'') )
{
PL_expect = XTERM;
else {
PL_expect = XOPERATOR;
- force_next(')');
- force_next('(');
}
}
PL_nextval[PL_nexttoke].opval = (OP*)newSVOP(OP_CONST,0, newSVpv(PL_tokenbuf,0));
return s;
}
+/*
+ * S_force_ident
+ * Called when the tokenizer wants $foo *foo &foo etc, but the program
+ * text only contains the "foo" portion. The first argument is a pointer
+ * to the "foo", and the second argument is the type symbol to prefix.
+ * Forces the next token to be a "WORD".
+ * Creates the symbol if it didn't already exist (through the gv_fetchpv
+ * call).
+ */
+
STATIC void
-force_ident(register char *s, int kind)
+S_force_ident(pTHX_ register char *s, int kind)
{
if (s && *s) {
OP* o = (OP*)newSVOP(OP_CONST, 0, newSVpv(s,0));
}
}
+/*
+ * S_force_version
+ * Forces the next token to be a version number.
+ */
+
STATIC char *
-force_version(char *s)
+S_force_version(pTHX_ char *s)
{
OP *version = Nullop;
return (s);
}
+/*
+ * S_tokeq
+ * Tokenize a quoted string passed in as an SV. It finds the next
+ * chunk, up to end of string or a backslash. It may make a new
+ * SV containing that chunk (if HINT_NEW_STRING is on). It also
+ * turns \\ into \.
+ */
+
STATIC SV *
-tokeq(SV *sv)
+S_tokeq(pTHX_ SV *sv)
{
register char *s;
register char *send;
goto finish;
d = s;
if ( PL_hints & HINT_NEW_STRING )
- pv = sv_2mortal(newSVpv(SvPVX(pv), len));
+ pv = sv_2mortal(newSVpvn(SvPVX(pv), len));
while (s < send) {
if (*s == '\\') {
if (s + 1 < send && (s[1] == '\\'))
return sv;
}
+/*
+ * Now come three functions related to double-quote context,
+ * S_sublex_start, S_sublex_push, and S_sublex_done. They're used when
+ * converting things like "\u\Lgnat" into ucfirst(lc("gnat")). They
+ * interact with PL_lex_state, and create fake ( ... ) argument lists
+ * to handle functions and concatenation.
+ * They assume that whoever calls them will be setting up a fake
+ * join call, because each subthing puts a ',' after it. This lets
+ * "lower \luPpEr"
+ * become
+ * join($, , 'lower ', lcfirst( 'uPpEr', ) ,)
+ *
+ * (I'm not sure whether the spurious commas at the end of lcfirst's
+ * arguments and join's arguments are created or not).
+ */
+
+/*
+ * S_sublex_start
+ * Assumes that yylval.ival is the op we're creating (e.g. OP_LCFIRST).
+ *
+ * Pattern matching will set PL_lex_op to the pattern-matching op to
+ * make (we return THING if yylval.ival is OP_NULL, PMFUNC otherwise).
+ *
+ * OP_CONST and OP_READLINE are easy--just make the new op and return.
+ *
+ * Everything else becomes a FUNC.
+ *
+ * Sets PL_lex_state to LEX_INTERPPUSH unless (ival was OP_NULL or we
+ * had an OP_CONST or OP_READLINE). This just sets us up for a
+ * call to S_sublex_push().
+ */
+
STATIC I32
-sublex_start(void)
+S_sublex_start(pTHX)
{
register I32 op_type = yylval.ival;
SV *nsv;
p = SvPV(sv, len);
- nsv = newSVpv(p, len);
+ nsv = newSVpvn(p, len);
SvREFCNT_dec(sv);
sv = nsv;
}
return FUNC;
}
+/*
+ * S_sublex_push
+ * Create a new scope to save the lexing state. The scope will be
+ * ended in S_sublex_done. Returns a '(', starting the function arguments
+ * to the uc, lc, etc. found before.
+ * Sets PL_lex_state to LEX_INTERPCONCAT.
+ */
+
STATIC I32
-sublex_push(void)
+S_sublex_push(pTHX)
{
dTHR;
ENTER;
return '(';
}
+/*
+ * S_sublex_done
+ * Restores lexer state after a S_sublex_push.
+ */
+
STATIC I32
-sublex_done(void)
+S_sublex_done(pTHX)
{
if (!PL_lex_starts++) {
PL_expect = XOPERATOR;
- yylval.opval = (OP*)newSVOP(OP_CONST, 0, newSVpv("",0));
+ yylval.opval = (OP*)newSVOP(OP_CONST, 0, newSVpvn("",0));
return THING;
}
return yylex();
}
- /* Is there a right-hand side to take care of? */
+ /* Is there a right-hand side to take care of? (s//RHS/ or tr//RHS/) */
if (PL_lex_repl && (PL_lex_inwhat == OP_SUBST || PL_lex_inwhat == OP_TRANS)) {
PL_linestr = PL_lex_repl;
PL_lex_inpat = 0;
PL_lex_casemods = 0;
*PL_lex_casestack = '\0';
PL_lex_starts = 0;
- if (SvCOMPILED(PL_lex_repl)) {
+ if (SvEVALED(PL_lex_repl)) {
PL_lex_state = LEX_INTERPNORMAL;
PL_lex_starts++;
+ /* we don't clear PL_lex_repl here, so that we can check later
+ whether this is an evalled subst; that means we rely on the
+ logic to ensure sublex_done() is called again only via the
+ branch (in yylex()) that clears PL_lex_repl, else we'll loop */
}
- else
+ else {
PL_lex_state = LEX_INTERPCONCAT;
- PL_lex_repl = Nullsv;
+ PL_lex_repl = Nullsv;
+ }
return ',';
}
else {
*/
STATIC char *
-scan_const(char *start)
+S_scan_const(pTHX_ char *start)
{
register char *send = PL_bufend; /* end of the constant */
SV *sv = NEWSV(93, send - start); /* sv for the constant */
register char *d = SvPVX(sv); /* destination for copies */
bool dorange = FALSE; /* are we in a translit range? */
I32 len; /* ? */
- I32 utf = PL_lex_inwhat == OP_TRANS
+ I32 utf = (PL_lex_inwhat == OP_TRANS && PL_sublex_info.sub_op)
? (PL_sublex_info.sub_op->op_private & (OPpTRANS_FROM_UTF|OPpTRANS_TO_UTF))
: UTF;
- I32 thisutf = PL_lex_inwhat == OP_TRANS
+ I32 thisutf = (PL_lex_inwhat == OP_TRANS && PL_sublex_info.sub_op)
? (PL_sublex_info.sub_op->op_private & (PL_lex_repl ? OPpTRANS_FROM_UTF : OPpTRANS_TO_UTF))
: UTF;
-
/* leaveit is the set of acceptably-backslashed characters */
char *leaveit =
PL_lex_inpat
}
/* check for embedded arrays (@foo, @:foo, @'foo, @{foo}, @$foo) */
- else if (*s == '@' && s[1] && (isALNUM(s[1]) || strchr(":'{$", s[1])))
+ else if (*s == '@' && s[1] && (isALNUM_lazy(s+1) || strchr(":'{$", s[1])))
break;
/* check for embedded scalars. only stop if we're sure it's a
s++;
/* some backslashes we leave behind */
- if (*s && strchr(leaveit, *s)) {
+ if (*leaveit && *s && strchr(leaveit, *s)) {
*d++ = '\\';
*d++ = *s++;
continue;
{
dTHR; /* only for ckWARN */
if (ckWARN(WARN_SYNTAX))
- warner(WARN_SYNTAX, "\\%c better written as $%c", *s, *s);
+ Perl_warner(aTHX_ WARN_SYNTAX, "\\%c better written as $%c", *s, *s);
*--s = '$';
break;
}
continue;
}
/* FALL THROUGH */
- /* default action is to copy the quoted character */
default:
- *d++ = *s++;
- continue;
+ {
+ dTHR;
+ if (ckWARN(WARN_UNSAFE) && isALPHA(*s))
+ Perl_warner(aTHX_ WARN_UNSAFE,
+ "Unrecognized escape \\%c passed through",
+ *s);
+ /* default action is to copy the quoted character */
+ *d++ = *s++;
+ continue;
+ }
/* \132 indicates an octal constant */
case '0': case '1': case '2': case '3':
if (!utf) {
dTHR;
if (ckWARN(WARN_UTF8))
- warner(WARN_UTF8,
+ Perl_warner(aTHX_ WARN_UTF8,
"Use of \\x{} without utf8 declaration");
}
/* note: utf always shorter than hex */
d = (char*)uv_to_utf8((U8*)d,
scan_hex(s + 1, e - s - 1, &len));
s = e + 1;
-
}
else {
UV uv = (UV)scan_hex(s, 2, &len);
if (uv >= 127 && UTF) {
dTHR;
if (ckWARN(WARN_UTF8))
- warner(WARN_UTF8,
+ Perl_warner(aTHX_ WARN_UTF8,
"\\x%.*s will produce malformed UTF-8 character; use \\x{%.*s} for that",
len,s,len,s);
}
case 't':
*d++ = '\t';
break;
+#ifdef EBCDIC
+ case 'e':
+ *d++ = '\047'; /* CP 1047 */
+ break;
+ case 'a':
+ *d++ = '\057'; /* CP 1047 */
+ break;
+#else
case 'e':
*d++ = '\033';
break;
case 'a':
*d++ = '\007';
break;
+#endif
} /* end switch */
s++;
return s;
}
-/* This is the one truly awful dwimmer necessary to conflate C and sed. */
+/* S_intuit_more
+ * Returns TRUE if there's more to the expression (e.g., a subscript),
+ * FALSE otherwise.
+ * This is the one truly awful dwimmer necessary to conflate C and sed.
+ *
+ * It deals with "$foo[3]" and /$foo[3]/ and /$foo[0123456789$]+/
+ *
+ * ->[ and ->{ return TRUE
+ * { and [ outside a pattern are always subscripts, so return TRUE
+ * if we're outside a pattern and it's not { or [, then return FALSE
+ * if we're in a pattern and the first char is a {
+ * {4,5} (any digits around the comma) returns FALSE
+ * if we're in a pattern and the first char is a [
+ * [] returns FALSE
+ * [SOMETHING] has a funky algorithm to decide whether it's a
+ * character class or not. It has to deal with things like
+ * /$foo[-3]/ and /$foo[$bar]/ as well as /$foo[$\d]+/
+ * anything else returns TRUE
+ */
+
STATIC int
-intuit_more(register char *s)
+S_intuit_more(pTHX_ register char *s)
{
if (PL_lex_brackets)
return TRUE;
if (*s == ']' || *s == '^')
return FALSE;
else {
+ /* this is terrifying, and it works */
int weight = 2; /* let's weigh the evidence */
char seen[256];
unsigned char un_char = 255, last_un_char;
case '&':
case '$':
weight -= seen[un_char] * 10;
- if (isALNUM(s[1])) {
+ if (isALNUM_lazy(s+1)) {
scan_ident(s, send, tmpbuf, sizeof tmpbuf, FALSE);
if ((int)strlen(tmpbuf) > 1 && gv_fetchpv(tmpbuf,FALSE, SVt_PV))
weight -= 100;
return TRUE;
}
+/*
+ * S_intuit_method
+ *
+ * Does all the checking to disambiguate
+ * foo bar
+ * between foo(bar) and bar->foo. Returns 0 if not a method, otherwise
+ * FUNCMETH (bar->foo(args)) or METHOD (bar->foo args).
+ *
+ * First argument is the stuff after the first token, e.g. "bar".
+ *
+ * Not a method if bar is a filehandle.
+ * Not a method if foo is a subroutine prototyped to take a filehandle.
+ * Not a method if it's really "Foo $bar"
+ * Method if it's "foo $bar"
+ * Not a method if it's really "print foo $bar"
+ * Method if it's really "foo package::" (interpreted as package->foo)
+ * Not a method if bar is known to be a subroutne ("sub bar; foo bar")
+ * Not a method if bar is a filehandle or package, but is quotd with
+ * =>
+ */
+
STATIC int
-intuit_method(char *start, GV *gv)
+S_intuit_method(pTHX_ char *start, GV *gv)
{
char *s = start + (*start == '$');
char tmpbuf[sizeof PL_tokenbuf];
gv = 0;
}
s = scan_word(s, tmpbuf, sizeof tmpbuf, TRUE, &len);
+ /* start is the beginning of the possible filehandle/object,
+ * and s is the end of it
+ * tmpbuf is a copy of it
+ */
+
if (*start == '$') {
if (gv || PL_last_lop_op == OP_PRINT || isUPPER(*PL_tokenbuf))
return 0;
return 0; /* no assumptions -- "=>" quotes bearword */
bare_package:
PL_nextval[PL_nexttoke].opval = (OP*)newSVOP(OP_CONST, 0,
- newSVpv(tmpbuf,0));
+ newSVpvn(tmpbuf,len));
PL_nextval[PL_nexttoke].opval->op_private = OPpCONST_BARE;
PL_expect = XTERM;
force_next(WORD);
return 0;
}
+/*
+ * S_incl_perldb
+ * Return a string of Perl code to load the debugger. If PERL5DB
+ * is set, it will return the contents of that, otherwise a
+ * compile-time require of perl5db.pl.
+ */
+
STATIC char*
-incl_perldb(void)
+S_incl_perldb(pTHX)
{
if (PL_perldb) {
char *pdb = PerlEnv_getenv("PERL5DB");
* Note that IoTOP_NAME, IoFMT_NAME, IoBOTTOM_NAME, if set for
* private use must be set using malloc'd pointers.
*/
-static int filter_debug = 0;
SV *
-filter_add(filter_t funcp, SV *datasv)
+Perl_filter_add(pTHX_ filter_t funcp, SV *datasv)
{
if (!funcp){ /* temporary handy debugging hack to be deleted */
- filter_debug = atoi((char*)datasv);
+ PL_filter_debug = atoi((char*)datasv);
return NULL;
}
if (!PL_rsfp_filters)
if (!datasv)
datasv = NEWSV(255,0);
if (!SvUPGRADE(datasv, SVt_PVIO))
- die("Can't upgrade filter_add data to SVt_PVIO");
+ Perl_die(aTHX_ "Can't upgrade filter_add data to SVt_PVIO");
IoDIRP(datasv) = (DIR*)funcp; /* stash funcp into spare field */
- if (filter_debug)
- warn("filter_add func %p (%s)", funcp, SvPV(datasv,PL_na));
+#ifdef DEBUGGING
+ if (PL_filter_debug) {
+ STRLEN n_a;
+ Perl_warn(aTHX_ "filter_add func %p (%s)", funcp, SvPV(datasv, n_a));
+ }
+#endif /* DEBUGGING */
av_unshift(PL_rsfp_filters, 1);
av_store(PL_rsfp_filters, 0, datasv) ;
return(datasv);
/* Delete most recently added instance of this filter function. */
void
-filter_del(filter_t funcp)
+Perl_filter_del(pTHX_ filter_t funcp)
{
- if (filter_debug)
- warn("filter_del func %p", funcp);
+#ifdef DEBUGGING
+ if (PL_filter_debug)
+ Perl_warn(aTHX_ "filter_del func %p", funcp);
+#endif /* DEBUGGING */
if (!PL_rsfp_filters || AvFILLp(PL_rsfp_filters)<0)
return;
/* if filter is on top of stack (usual case) just pop it off */
if (IoDIRP(FILTER_DATA(AvFILLp(PL_rsfp_filters))) == (DIR*)funcp){
+ IoDIRP(FILTER_DATA(AvFILLp(PL_rsfp_filters))) = NULL;
sv_free(av_pop(PL_rsfp_filters));
return;
}
/* we need to search for the correct entry and clear it */
- die("filter_del can only delete in reverse order (currently)");
+ Perl_die(aTHX_ "filter_del can only delete in reverse order (currently)");
}
/* Invoke the n'th filter function for the current rsfp. */
I32
-filter_read(int idx, SV *buf_sv, int maxlen)
+Perl_filter_read(pTHX_ int idx, SV *buf_sv, int maxlen)
/* 0 = read one text line */
if (idx > AvFILLp(PL_rsfp_filters)){ /* Any more filters? */
/* Provide a default input filter to make life easy. */
/* Note that we append to the line. This is handy. */
- if (filter_debug)
- warn("filter_read %d: from rsfp\n", idx);
+#ifdef DEBUGGING
+ if (PL_filter_debug)
+ Perl_warn(aTHX_ "filter_read %d: from rsfp\n", idx);
+#endif /* DEBUGGING */
if (maxlen) {
/* Want a block */
int len ;
}
/* Skip this filter slot if filter has been deleted */
if ( (datasv = FILTER_DATA(idx)) == &PL_sv_undef){
- if (filter_debug)
- warn("filter_read %d: skipped (filter deleted)\n", idx);
+#ifdef DEBUGGING
+ if (PL_filter_debug)
+ Perl_warn(aTHX_ "filter_read %d: skipped (filter deleted)\n", idx);
+#endif /* DEBUGGING */
return FILTER_READ(idx+1, buf_sv, maxlen); /* recurse */
}
/* Get function pointer hidden within datasv */
funcp = (filter_t)IoDIRP(datasv);
- if (filter_debug)
- warn("filter_read %d: via function %p (%s)\n",
- idx, funcp, SvPV(datasv,PL_na));
+#ifdef DEBUGGING
+ if (PL_filter_debug) {
+ STRLEN n_a;
+ Perl_warn(aTHX_ "filter_read %d: via function %p (%s)\n",
+ idx, funcp, SvPV(datasv,n_a));
+ }
+#endif /* DEBUGGING */
/* Call function. The function is expected to */
/* call "FILTER_READ(idx+1, buf_sv)" first. */
/* Return: <0:error, =0:eof, >0:not eof */
- return (*funcp)(PERL_OBJECT_THIS_ idx, buf_sv, maxlen);
+ return (*funcp)(aTHXo_ idx, buf_sv, maxlen);
}
STATIC char *
-filter_gets(register SV *sv, register PerlIO *fp, STRLEN append)
+S_filter_gets(pTHX_ register SV *sv, register PerlIO *fp, STRLEN append)
{
#ifdef WIN32FILTER
if (!PL_rsfp_filters) {
*/
int
-yylex(void)
+#ifdef USE_PURE_BISON
+Perl_yylex(pTHX_ YYSTYPE *lvalp, int *lcharp)
+#else
+Perl_yylex(pTHX)
+#endif
{
dTHR;
register char *s;
GV *gv = Nullgv;
GV **gvp = 0;
+#ifdef USE_PURE_BISON
+ yylval_pointer = lvalp;
+ yychar_pointer = lcharp;
+#endif
+
/* check if there's an identifier for us to look at */
if (PL_pending_ident) {
/* pit holds the identifier we read and pending_ident is reset */
*/
if (PL_in_my) {
if (strchr(PL_tokenbuf,':'))
- croak(no_myglob,PL_tokenbuf);
+ yyerror(Perl_form(aTHX_ PL_no_myglob,PL_tokenbuf));
yylval.opval = newOP(OP_PADANY, 0);
yylval.opval->op_targ = pad_allocmy(PL_tokenbuf);
d++)
{
if (strnEQ(d,"<=>",3) || strnEQ(d,"cmp",3)) {
- croak("Can't use \"my %s\" in sort comparison",
+ Perl_croak(aTHX_ "Can't use \"my %s\" in sort comparison",
PL_tokenbuf);
}
}
if (pit == '@' && PL_lex_state != LEX_NORMAL && !PL_lex_brackets) {
GV *gv = gv_fetchpv(PL_tokenbuf+1, FALSE, SVt_PVAV);
if (!gv || ((PL_tokenbuf[0] == '@') ? !GvAV(gv) : !GvHV(gv)))
- yyerror(form("In string, %s now must be written as \\%s",
+ yyerror(Perl_form(aTHX_ "In string, %s now must be written as \\%s",
PL_tokenbuf, PL_tokenbuf));
}
case LEX_INTERPCASEMOD:
#ifdef DEBUGGING
if (PL_bufptr != PL_bufend && *PL_bufptr != '\\')
- croak("panic: INTERPCASEMOD");
+ Perl_croak(aTHX_ "panic: INTERPCASEMOD");
#endif
/* handle \E or end of string */
if (PL_bufptr == PL_bufend || PL_bufptr[1] == 'E') {
else if (*s == 'Q')
PL_nextval[PL_nexttoke].ival = OP_QUOTEMETA;
else
- croak("panic: yylex");
+ Perl_croak(aTHX_ "panic: yylex");
PL_bufptr = s + 1;
force_next(FUNC);
if (PL_lex_starts) {
PL_lex_state = LEX_INTERPCONCAT;
return ')';
}
+ if (PL_lex_inwhat == OP_SUBST && PL_linestr == PL_lex_repl
+ && SvEVALED(PL_lex_repl))
+ {
+ if (PL_bufptr != PL_bufend)
+ Perl_croak(aTHX_ "Bad evalled substitution pattern");
+ PL_lex_repl = Nullsv;
+ }
/* FALLTHROUGH */
case LEX_INTERPCONCAT:
#ifdef DEBUGGING
if (PL_lex_brackets)
- croak("panic: INTERPCONCAT");
+ Perl_croak(aTHX_ "panic: INTERPCONCAT");
#endif
if (PL_bufptr == PL_bufend)
return sublex_done();
retry:
switch (*s) {
default:
- /*
- * Note: we try to be careful never to call the isXXX_utf8() functions unless we're
- * pretty sure we've seen the beginning of a UTF-8 character (that is, the two high
- * bits are set). Otherwise we risk loading in the heavy-duty SWASHINIT and SWASHGET
- * routines unnecessarily. You will see this not just here but throughout this file.
- */
- if (UTF && (*s & 0xc0) == 0x80) {
- if (isIDFIRST_utf8((U8*)s))
- goto keylookup;
- }
- croak("Unrecognized character \\x%02X", *s & 255);
+ if (isIDFIRST_lazy(s))
+ goto keylookup;
+ Perl_croak(aTHX_ "Unrecognized character \\x%02X", *s & 255);
case 4:
case 26:
goto fake_eof; /* emulate EOF on ^D or ^Z */
PL_last_uni = 0;
PL_last_lop = 0;
if (PL_lex_brackets)
- yyerror("Missing right bracket");
+ yyerror("Missing right curly or square bracket");
TOKEN(0);
}
if (s++ < PL_bufend)
if (PL_minus_F) {
if (strchr("/'\"", *PL_splitstr)
&& strchr(PL_splitstr + 1, *PL_splitstr))
- sv_catpvf(PL_linestr, "@F=split(%s);", PL_splitstr);
+ Perl_sv_catpvf(aTHX_ PL_linestr, "@F=split(%s);", PL_splitstr);
else {
char delim;
s = "'~#\200\1'"; /* surely one char is unused...*/
while (s[1] && strchr(PL_splitstr, *s)) s++;
delim = *s;
- sv_catpvf(PL_linestr, "@F=split(%s%c",
+ Perl_sv_catpvf(aTHX_ PL_linestr, "@F=split(%s%c",
"q" + (delim == '\''), delim);
for (s = PL_splitstr; *s; s++) {
if (*s == '\\')
sv_catpvn(PL_linestr, "\\", 1);
sv_catpvn(PL_linestr, s, 1);
}
- sv_catpvf(PL_linestr, "%c);", delim);
+ Perl_sv_catpvf(aTHX_ PL_linestr, "%c);", delim);
}
}
else
else
newargv = PL_origargv;
newargv[0] = ipath;
- execv(ipath, newargv);
- croak("Can't exec %s", ipath);
+ PerlProc_execv(ipath, newargv);
+ Perl_croak(aTHX_ "Can't exec %s", ipath);
}
if (d) {
U32 oldpdb = PL_perldb;
if (*d == 'M' || *d == 'm') {
char *m = d;
while (*d && !isSPACE(*d)) d++;
- croak("Too late for \"-%.*s\" option",
+ Perl_croak(aTHX_ "Too late for \"-%.*s\" option",
(int)(d - m), m);
}
d = moreswitches(d);
goto retry;
case '\r':
#ifdef PERL_STRICT_CR
- warn("Illegal character \\%03o (carriage return)", '\r');
- croak(
+ Perl_warn(aTHX_ "Illegal character \\%03o (carriage return)", '\r');
+ Perl_croak(aTHX_
"(Maybe you didn't strip carriage returns after a network transfer?)\n");
#endif
case ' ': case '\t': case '\f': case 013:
case 'A': gv_fetchpv("\024",TRUE, SVt_PV); FTST(OP_FTATIME);
case 'C': gv_fetchpv("\024",TRUE, SVt_PV); FTST(OP_FTCTIME);
default:
- croak("Unrecognized file test: -%c", (int)tmp);
+ Perl_croak(aTHX_ "Unrecognized file test: -%c", (int)tmp);
break;
}
}
else if (*s == '>') {
s++;
s = skipspace(s);
- if (isIDFIRST(*s)) {
+ if (isIDFIRST_lazy(s)) {
s = force_word(s,METHOD,FALSE,TRUE,FALSE);
TOKEN(ARROW);
}
case ']':
s++;
if (PL_lex_brackets <= 0)
- yyerror("Unmatched right bracket");
+ yyerror("Unmatched right square bracket");
else
--PL_lex_brackets;
if (PL_lex_state == LEX_INTERPNORMAL) {
while (d < PL_bufend && (*d == ' ' || *d == '\t'))
d++;
}
- if (d < PL_bufend && isIDFIRST(*d)) {
+ if (d < PL_bufend && isIDFIRST_lazy(d)) {
d = scan_word(d, PL_tokenbuf + 1, sizeof PL_tokenbuf - 1,
FALSE, &len);
while (d < PL_bufend && (*d == ' ' || *d == '\t'))
}
t++;
}
- else if (isALPHA(*s)) {
- for (t++; t < PL_bufend && isALNUM(*t); t++) ;
+ else if (isIDFIRST_lazy(s)) {
+ for (t++; t < PL_bufend && isALNUM_lazy(t); t++) ;
}
while (t < PL_bufend && isSPACE(*t))
t++;
rightbracket:
s++;
if (PL_lex_brackets <= 0)
- yyerror("Unmatched right bracket");
+ yyerror("Unmatched right curly bracket");
else
PL_expect = (expectation)PL_lex_brackstack[--PL_lex_brackets];
if (PL_lex_brackets < PL_lex_formbrack)
if (PL_lex_fakebrack) {
PL_lex_state = LEX_INTERPEND;
PL_bufptr = s;
- return yylex(); /* ignore fake brackets */
+ return yylex(); /* ignore fake brackets */
}
if (*s == '-' && s[1] == '>')
PL_lex_state = LEX_INTERPENDMAYBE;
AOPERATOR(ANDAND);
s--;
if (PL_expect == XOPERATOR) {
- if (ckWARN(WARN_SEMICOLON) && isALPHA(*s) && PL_bufptr == PL_linestart) {
+ if (ckWARN(WARN_SEMICOLON) && isIDFIRST_lazy(s) && PL_bufptr == PL_linestart) {
PL_curcop->cop_line--;
- warner(WARN_SEMICOLON, warn_nosemi);
+ Perl_warner(aTHX_ WARN_SEMICOLON, PL_warn_nosemi);
PL_curcop->cop_line++;
}
BAop(OP_BIT_AND);
if (tmp == '~')
PMop(OP_MATCH);
if (ckWARN(WARN_SYNTAX) && tmp && isSPACE(*s) && strchr("+-*/%.^&|<",tmp))
- warner(WARN_SYNTAX, "Reversed %c= operator",(int)tmp);
+ Perl_warner(aTHX_ WARN_SYNTAX, "Reversed %c= operator",(int)tmp);
s--;
if (PL_expect == XSTATE && isALPHA(tmp) &&
(s == PL_linestart+1 || s[-2] == '\n') )
}
}
- if (s[1] == '#' && (isALPHA(s[2]) || strchr("_{$:+-", s[2]))) {
- if (PL_expect == XOPERATOR)
- no_op("Array length", PL_bufptr);
+ if (s[1] == '#' && (isIDFIRST_lazy(s+2) || strchr("{$:+-", s[2]))) {
PL_tokenbuf[0] = '@';
- s = scan_ident(s + 1, PL_bufend, PL_tokenbuf + 1, sizeof PL_tokenbuf - 1,
- FALSE);
+ s = scan_ident(s + 1, PL_bufend, PL_tokenbuf + 1,
+ sizeof PL_tokenbuf - 1, FALSE);
+ if (PL_expect == XOPERATOR)
+ no_op("Array length", s);
if (!PL_tokenbuf[1])
PREREF(DOLSHARP);
PL_expect = XOPERATOR;
TOKEN(DOLSHARP);
}
- if (PL_expect == XOPERATOR)
- no_op("Scalar", PL_bufptr);
PL_tokenbuf[0] = '$';
- s = scan_ident(s, PL_bufend, PL_tokenbuf + 1, sizeof PL_tokenbuf - 1, FALSE);
+ s = scan_ident(s, PL_bufend, PL_tokenbuf + 1,
+ sizeof PL_tokenbuf - 1, FALSE);
+ if (PL_expect == XOPERATOR)
+ no_op("Scalar", s);
if (!PL_tokenbuf[1]) {
if (s == PL_bufend)
yyerror("Final $ should be \\$ or $name");
}
d = s;
+ tmp = (I32)*s;
if (PL_lex_state == LEX_NORMAL)
s = skipspace(s);
PL_tokenbuf[0] = '@';
if (ckWARN(WARN_SYNTAX)) {
for(t = s + 1;
- isSPACE(*t) || isALNUM(*t) || *t == '$';
+ isSPACE(*t) || isALNUM_lazy(t) || *t == '$';
t++) ;
if (*t++ == ',') {
PL_bufptr = skipspace(PL_bufptr);
while (t < PL_bufend && *t != ']')
t++;
- warner(WARN_SYNTAX,
+ Perl_warner(aTHX_ WARN_SYNTAX,
"Multidimensional syntax %.*s not supported",
(t - PL_bufptr) + 1, PL_bufptr);
}
char tmpbuf[sizeof PL_tokenbuf];
STRLEN len;
for (t++; isSPACE(*t); t++) ;
- if (isIDFIRST(*t)) {
+ if (isIDFIRST_lazy(t)) {
t = scan_word(t, tmpbuf, sizeof tmpbuf, TRUE, &len);
- if (*t != '(' && perl_get_cv(tmpbuf, FALSE))
- warner(WARN_SYNTAX,
+ for (; isSPACE(*t); t++) ;
+ if (*t == ';' && get_cv(tmpbuf, FALSE))
+ Perl_warner(aTHX_ WARN_SYNTAX,
"You need to quote \"%s\"", tmpbuf);
}
}
}
PL_expect = XOPERATOR;
- if (PL_lex_state == LEX_NORMAL && isSPACE(*d)) {
+ if (PL_lex_state == LEX_NORMAL && isSPACE((char)tmp)) {
bool islop = (PL_last_lop == PL_oldoldbufptr);
if (!islop || PL_last_lop_op == OP_GREPSTART)
PL_expect = XOPERATOR;
else if (strchr("$@\"'`q", *s))
PL_expect = XTERM; /* e.g. print $fh "foo" */
- else if (strchr("&*<%", *s) && isIDFIRST(s[1]))
+ else if (strchr("&*<%", *s) && isIDFIRST_lazy(s+1))
PL_expect = XTERM; /* e.g. print $fh &sub */
- else if (isIDFIRST(*s)) {
+ else if (isIDFIRST_lazy(s)) {
char tmpbuf[sizeof PL_tokenbuf];
scan_word(s, tmpbuf, sizeof tmpbuf, TRUE, &len);
if (tmp = keyword(tmpbuf, len)) {
PL_expect = XTERM; /* e.g. print $fh 3 */
else if (*s == '.' && isDIGIT(s[1]))
PL_expect = XTERM; /* e.g. print $fh .3 */
- else if (strchr("/?-+", *s) && !isSPACE(s[1]))
+ else if (strchr("/?-+", *s) && !isSPACE(s[1]) && s[1] != '=')
PL_expect = XTERM; /* e.g. print $fh -1 */
- else if (*s == '<' && s[1] == '<' && !isSPACE(s[2]))
+ else if (*s == '<' && s[1] == '<' && !isSPACE(s[2]) && s[2] != '=')
PL_expect = XTERM; /* print $fh <<"EOF" */
}
PL_pending_ident = '$';
if (ckWARN(WARN_SYNTAX)) {
if (*s == '[' || *s == '{') {
char *t = s + 1;
- while (*t && (isALNUM(*t) || strchr(" \t$#+-'\"", *t)))
+ while (*t && (isALNUM_lazy(t) || strchr(" \t$#+-'\"", *t)))
t++;
if (*t == '}' || *t == ']') {
t++;
PL_bufptr = skipspace(PL_bufptr);
- warner(WARN_SYNTAX,
+ Perl_warner(aTHX_ WARN_SYNTAX,
"Scalar value %.*s better written as $%.*s",
t-PL_bufptr, PL_bufptr, t-PL_bufptr-1, PL_bufptr+1);
}
/* Disable warning on "study /blah/" */
if (PL_oldoldbufptr == PL_last_uni
&& (*PL_last_uni != 's' || s - PL_last_uni < 5
- || memNE(PL_last_uni, "study", 5) || isALNUM(PL_last_uni[5])))
+ || memNE(PL_last_uni, "study", 5) || isALNUM_lazy(PL_last_uni+5)))
check_uni();
s = scan_pat(s,OP_MATCH);
TERM(sublex_start());
case '\\':
s++;
if (ckWARN(WARN_SYNTAX) && PL_lex_inwhat && isDIGIT(*s))
- warner(WARN_SYNTAX,"Can't use \\%c to mean $%c in expression",
+ Perl_warner(aTHX_ WARN_SYNTAX,"Can't use \\%c to mean $%c in expression",
*s, *s);
if (PL_expect == XOPERATOR)
no_op("Backslash",s);
case 'z': case 'Z':
keylookup: {
+ STRLEN n_a;
gv = Nullgv;
gvp = 0;
gvp = 0;
if (ckWARN(WARN_AMBIGUOUS) && hgv
&& tmp != KEY_x && tmp != KEY_CORE) /* never ambiguous */
- warner(WARN_AMBIGUOUS,
+ Perl_warner(aTHX_ WARN_AMBIGUOUS,
"Ambiguous call resolved as CORE::%s(), %s",
GvENAME(hgv), "qualify as such or use &");
}
s = scan_word(s, PL_tokenbuf + len, sizeof PL_tokenbuf - len,
TRUE, &morelen);
if (!morelen)
- croak("Bad name after %s%s", PL_tokenbuf,
+ Perl_croak(aTHX_ "Bad name after %s%s", PL_tokenbuf,
*s == '\'' ? "'" : "::");
len += morelen;
}
if (PL_expect == XOPERATOR) {
if (PL_bufptr == PL_linestart) {
PL_curcop->cop_line--;
- warner(WARN_SEMICOLON, warn_nosemi);
+ Perl_warner(aTHX_ WARN_SEMICOLON, PL_warn_nosemi);
PL_curcop->cop_line++;
}
else
PL_tokenbuf[len - 2] == ':' && PL_tokenbuf[len - 1] == ':')
{
if (ckWARN(WARN_UNSAFE) && ! gv_fetchpv(PL_tokenbuf, FALSE, SVt_PVHV))
- warner(WARN_UNSAFE,
+ Perl_warner(aTHX_ WARN_UNSAFE,
"Bareword \"%s\" refers to nonexistent package",
PL_tokenbuf);
len -= 2;
/* if we saw a global override before, get the right name */
if (gvp) {
- sv = newSVpv("CORE::GLOBAL::",14);
+ sv = newSVpvn("CORE::GLOBAL::",14);
sv_catpv(sv,PL_tokenbuf);
}
else
PL_oldoldbufptr < PL_bufptr &&
(PL_oldoldbufptr == PL_last_lop || PL_oldoldbufptr == PL_last_uni) &&
/* NO SKIPSPACE BEFORE HERE! */
- (PL_expect == XREF
- || ((opargs[PL_last_lop_op] >> OASHIFT)& 7) == OA_FILEREF
- || (PL_last_lop_op == OP_ENTERSUB
- && PL_last_proto
- && PL_last_proto[PL_last_proto[0] == ';' ? 1 : 0] == '*')) )
+ (PL_expect == XREF ||
+ ((PL_opargs[PL_last_lop_op] >> OASHIFT)& 7) == OA_FILEREF))
{
bool immediate_paren = *s == '(';
/* Two barewords in a row may indicate method call. */
- if ((isALPHA(*s) || *s == '$') && (tmp=intuit_method(s,gv)))
+ if ((isIDFIRST_lazy(s) || *s == '$') && (tmp=intuit_method(s,gv)))
return tmp;
/* If not a declared subroutine, it's an indirect object. */
/* (But it's an indir obj regardless for sort.) */
if ((PL_last_lop_op == OP_SORT ||
- (!immediate_paren && (!gv || !GvCVu(gv))) ) &&
- (PL_last_lop_op != OP_MAPSTART && PL_last_lop_op != OP_GREPSTART)){
+ (!immediate_paren && (!gv || !GvCVu(gv)))) &&
+ (PL_last_lop_op != OP_MAPSTART &&
+ PL_last_lop_op != OP_GREPSTART))
+ {
PL_expect = (PL_last_lop == PL_oldoldbufptr) ? XTERM : XOPERATOR;
goto bareword;
}
/* If followed by a bareword, see if it looks like indir obj. */
- if ((isALPHA(*s) || *s == '$') && (tmp = intuit_method(s,gv)))
+ if ((isIDFIRST_lazy(s) || *s == '$') && (tmp = intuit_method(s,gv)))
return tmp;
/* Not a method, so call it a subroutine (if defined) */
if (gv && GvCVu(gv)) {
CV* cv;
- if (lastchar == '-')
- warn("Ambiguous use of -%s resolved as -&%s()",
+ if (lastchar == '-' && ckWARN_d(WARN_AMBIGUOUS))
+ Perl_warner(aTHX_ WARN_AMBIGUOUS,
+ "Ambiguous use of -%s resolved as -&%s()",
PL_tokenbuf, PL_tokenbuf);
- PL_last_lop = PL_oldbufptr;
- PL_last_lop_op = OP_ENTERSUB;
/* Check for a constant sub */
cv = GvCV(gv);
if ((sv = cv_const_sv(cv))) {
/* Resolve to GV now. */
op_free(yylval.opval);
yylval.opval = newCVREF(0, newGVOP(OP_GV, 0, gv));
+ yylval.opval->op_private |= OPpENTERSUB_NOPAREN;
+ PL_last_lop = PL_oldbufptr;
+ PL_last_lop_op = OP_ENTERSUB;
/* Is there a prototype? */
if (SvPOK(cv)) {
STRLEN len;
- PL_last_proto = SvPV((SV*)cv, len);
+ char *proto = SvPV((SV*)cv, len);
if (!len)
TERM(FUNC0SUB);
- if (strEQ(PL_last_proto, "$"))
+ if (strEQ(proto, "$"))
OPERATOR(UNIOPSUB);
- if (*PL_last_proto == '&' && *s == '{') {
+ if (*proto == '&' && *s == '{') {
sv_setpv(PL_subname,"__ANON__");
PREBLOCK(LSTOPSUB);
}
- } else
- PL_last_proto = NULL;
+ }
PL_nextval[PL_nexttoke].opval = yylval.opval;
PL_expect = XTERM;
force_next(WORD);
TOKEN(NOAMP);
}
- if (PL_hints & HINT_STRICT_SUBS &&
- lastchar != '-' &&
- strnNE(s,"->",2) &&
- PL_last_lop_op != OP_TRUNCATE && /* S/F prototype in opcode.pl */
- PL_last_lop_op != OP_ACCEPT &&
- PL_last_lop_op != OP_PIPE_OP &&
- PL_last_lop_op != OP_SOCKPAIR)
- {
- warn(
- "Bareword \"%s\" not allowed while \"strict subs\" in use",
- PL_tokenbuf);
- ++PL_error_count;
- }
-
/* Call it a bare word */
- bareword:
- if (ckWARN(WARN_RESERVED)) {
- if (lastchar != '-') {
- for (d = PL_tokenbuf; *d && isLOWER(*d); d++) ;
- if (!*d)
- warner(WARN_RESERVED, warn_reserved, PL_tokenbuf);
+ if (PL_hints & HINT_STRICT_SUBS)
+ yylval.opval->op_private |= OPpCONST_STRICT;
+ else {
+ bareword:
+ if (ckWARN(WARN_RESERVED)) {
+ if (lastchar != '-') {
+ for (d = PL_tokenbuf; *d && isLOWER(*d); d++) ;
+ if (!*d)
+ Perl_warner(aTHX_ WARN_RESERVED, PL_warn_reserved,
+ PL_tokenbuf);
+ }
}
}
safe_bareword:
- if (lastchar && strchr("*%&", lastchar)) {
- warn("Operator or semicolon missing before %c%s",
+ if (lastchar && strchr("*%&", lastchar) && ckWARN_d(WARN_AMBIGUOUS)) {
+ Perl_warner(aTHX_ WARN_AMBIGUOUS,
+ "Operator or semicolon missing before %c%s",
lastchar, PL_tokenbuf);
- warn("Ambiguous use of %c resolved as operator %c",
+ Perl_warner(aTHX_ WARN_AMBIGUOUS,
+ "Ambiguous use of %c resolved as operator %c",
lastchar, lastchar);
}
TOKEN(WORD);
case KEY___LINE__:
yylval.opval = (OP*)newSVOP(OP_CONST, 0,
- newSVpvf("%ld", (long)PL_curcop->cop_line));
+ Perl_newSVpvf(aTHX_ "%ld", (long)PL_curcop->cop_line));
TERM(THING);
case KEY___PACKAGE__:
char *pname = "main";
if (PL_tokenbuf[2] == 'D')
pname = HvNAME(PL_curstash ? PL_curstash : PL_defstash);
- gv = gv_fetchpv(form("%s::DATA", pname), TRUE, SVt_PVIO);
+ gv = gv_fetchpv(Perl_form(aTHX_ "%s::DATA", pname), TRUE, SVt_PVIO);
GvMULTI_on(gv);
if (!GvIO(gv))
GvIOp(gv) = newIO();
case KEY_foreach:
yylval.ival = PL_curcop->cop_line;
s = skipspace(s);
- if (PL_expect == XSTATE && isIDFIRST(*s)) {
+ if (PL_expect == XSTATE && isIDFIRST_lazy(s)) {
char *p = s;
if ((PL_bufend - p) >= 3 &&
strnEQ(p, "my", 2) && isSPACE(*(p + 2)))
p += 2;
p = skipspace(p);
- if (isIDFIRST(*p))
- croak("Missing $ on loop variable");
+ if (isIDFIRST_lazy(p))
+ Perl_croak(aTHX_ "Missing $ on loop variable");
}
OPERATOR(FOR);
TERM(sublex_start());
case KEY_map:
- LOP(OP_MAPSTART,XREF);
-
+ LOP(OP_MAPSTART, *s == '(' ? XTERM : XREF);
+
case KEY_mkdir:
LOP(OP_MKDIR,XTERM);
case KEY_my:
PL_in_my = TRUE;
s = skipspace(s);
- if (isIDFIRST(*s)) {
+ if (isIDFIRST_lazy(s)) {
s = scan_word(s, PL_tokenbuf, sizeof PL_tokenbuf, TRUE, &len);
PL_in_my_stash = gv_stashpv(PL_tokenbuf, FALSE);
if (!PL_in_my_stash) {
case KEY_open:
s = skipspace(s);
- if (isIDFIRST(*s)) {
+ if (isIDFIRST_lazy(s)) {
char *t;
- for (d = s; isALNUM(*d); d++) ;
+ for (d = s; isALNUM_lazy(d); d++) ;
t = skipspace(d);
- if (strchr("|&*+-=!?:.", *t))
- warn("Precedence problem: open %.*s should be open(%.*s)",
- d-s,s, d-s,s);
+ if (strchr("|&*+-=!?:.", *t) && ckWARN_d(WARN_AMBIGUOUS))
+ Perl_warner(aTHX_ WARN_AMBIGUOUS,
+ "Precedence problem: open %.*s should be open(%.*s)",
+ d-s,s, d-s,s);
}
LOP(OP_OPEN,XTERM);
s = scan_str(s);
if (!s)
missingterm((char*)0);
- if (ckWARN(WARN_SYNTAX) && SvLEN(PL_lex_stuff)) {
+ force_next(')');
+ if (SvCUR(PL_lex_stuff)) {
+ OP *words = Nullop;
+ int warned = 0;
d = SvPV_force(PL_lex_stuff, len);
- for (; len; --len, ++d) {
- if (*d == ',') {
- warner(WARN_SYNTAX,
- "Possible attempt to separate words with commas");
- break;
- }
- if (*d == '#') {
- warner(WARN_SYNTAX,
- "Possible attempt to put comments in qw() list");
- break;
+ while (len) {
+ for (; isSPACE(*d) && len; --len, ++d) ;
+ if (len) {
+ char *b = d;
+ if (!warned && ckWARN(WARN_SYNTAX)) {
+ for (; !isSPACE(*d) && len; --len, ++d) {
+ if (*d == ',') {
+ Perl_warner(aTHX_ WARN_SYNTAX,
+ "Possible attempt to separate words with commas");
+ ++warned;
+ }
+ else if (*d == '#') {
+ Perl_warner(aTHX_ WARN_SYNTAX,
+ "Possible attempt to put comments in qw() list");
+ ++warned;
+ }
+ }
+ }
+ else {
+ for (; !isSPACE(*d) && len; --len, ++d) ;
+ }
+ words = append_elem(OP_LIST, words,
+ newSVOP(OP_CONST, 0, newSVpvn(b, d-b)));
}
}
+ if (words) {
+ PL_nextval[PL_nexttoke].opval = words;
+ force_next(THING);
+ }
}
- force_next(')');
- PL_nextval[PL_nexttoke].opval = (OP*)newSVOP(OP_CONST, 0, tokeq(PL_lex_stuff));
+ if (PL_lex_stuff)
+ SvREFCNT_dec(PL_lex_stuff);
PL_lex_stuff = Nullsv;
- force_next(THING);
- force_next(',');
- PL_nextval[PL_nexttoke].opval = (OP*)newSVOP(OP_CONST, 0, newSVpv(" ",1));
- force_next(THING);
- force_next('(');
- yylval.ival = OP_SPLIT;
- CLINE;
PL_expect = XTERM;
- PL_bufptr = s;
- PL_last_lop = PL_oldbufptr;
- PL_last_lop_op = OP_SPLIT;
- return FUNC;
+ TOKEN('(');
case KEY_qq:
s = scan_str(s);
case KEY_require:
*PL_tokenbuf = '\0';
s = force_word(s,WORD,TRUE,TRUE,FALSE);
- if (isIDFIRST(*PL_tokenbuf))
+ if (isIDFIRST_lazy(PL_tokenbuf))
gv_stashpvn(PL_tokenbuf, strlen(PL_tokenbuf), TRUE);
else if (*s == '<')
yyerror("<> should be quotes");
checkcomma(s,PL_tokenbuf,"subroutine name");
s = skipspace(s);
if (*s == ';' || *s == ')') /* probably a close */
- croak("sort is now a reserved word");
+ Perl_croak(aTHX_ "sort is now a reserved word");
PL_expect = XTERM;
s = force_word(s,WORD,TRUE,TRUE,FALSE);
LOP(OP_SORT,XREF);
really_sub:
s = skipspace(s);
- if (isIDFIRST(*s) || *s == '\'' || *s == ':') {
+ if (isIDFIRST_lazy(s) || *s == '\'' || *s == ':') {
char tmpbuf[sizeof PL_tokenbuf];
PL_expect = XBLOCK;
d = scan_word(s, tmpbuf, sizeof tmpbuf, TRUE, &len);
if (PL_lex_stuff)
SvREFCNT_dec(PL_lex_stuff);
PL_lex_stuff = Nullsv;
- croak("Prototype not terminated");
+ Perl_croak(aTHX_ "Prototype not terminated");
}
/* strip spaces */
d = SvPVX(PL_lex_stuff);
PL_lex_stuff = Nullsv;
}
- if (*SvPV(PL_subname,PL_na) == '?') {
+ if (*SvPV(PL_subname,n_a) == '?') {
sv_setpv(PL_subname,"__ANON__");
TOKEN(ANONSUB);
}
}
I32
-keyword(register char *d, I32 len)
+Perl_keyword(pTHX_ register char *d, I32 len)
{
switch (*d) {
case '_':
break;
case 6:
if (strEQ(d,"exists")) return KEY_exists;
- if (strEQ(d,"elseif")) warn("elseif should be elsif");
+ if (strEQ(d,"elseif")) Perl_warn(aTHX_ "elseif should be elsif");
break;
case 8:
if (strEQ(d,"endgrent")) return -KEY_endgrent;
}
STATIC void
-checkcomma(register char *s, char *name, char *what)
+S_checkcomma(pTHX_ register char *s, char *name, char *what)
{
char *w;
if (*w)
for (; *w && isSPACE(*w); w++) ;
if (!*w || !strchr(";|})]oaiuw!=", *w)) /* an advisory hack only... */
- warner(WARN_SYNTAX, "%s (...) interpreted as function",name);
+ Perl_warner(aTHX_ WARN_SYNTAX, "%s (...) interpreted as function",name);
}
}
while (s < PL_bufend && isSPACE(*s))
s++;
while (s < PL_bufend && isSPACE(*s))
s++;
- if (isIDFIRST(*s)) {
+ if (isIDFIRST_lazy(s)) {
w = s++;
- while (isALNUM(*s))
+ while (isALNUM_lazy(s))
s++;
while (s < PL_bufend && isSPACE(*s))
s++;
if (*s == ',') {
int kw;
*s = '\0';
- kw = keyword(w, s - w) || perl_get_cv(w, FALSE) != 0;
+ kw = keyword(w, s - w) || get_cv(w, FALSE) != 0;
*s = ',';
if (kw)
return;
- croak("No comma allowed after %s", what);
+ Perl_croak(aTHX_ "No comma allowed after %s", what);
}
}
}
STATIC SV *
-new_constant(char *s, STRLEN len, char *key, SV *sv, SV *pv, char *type)
+S_new_constant(pTHX_ char *s, STRLEN len, char *key, SV *sv, SV *pv, char *type)
{
dSP;
HV *table = GvHV(PL_hintgv); /* ^H */
bool oldcatch = CATCH_GET;
SV **cvp;
SV *cv, *typesv;
- char buf[128];
if (!table) {
yyerror("%^H is not defined");
}
cvp = hv_fetch(table, key, strlen(key), FALSE);
if (!cvp || !SvOK(*cvp)) {
+ char buf[128];
sprintf(buf,"$^H{%s} is not defined", key);
yyerror(buf);
return sv;
sv_2mortal(sv); /* Parent created it permanently */
cv = *cvp;
if (!pv)
- pv = sv_2mortal(newSVpv(s, len));
+ pv = sv_2mortal(newSVpvn(s, len));
if (type)
typesv = sv_2mortal(newSVpv(type, 0));
else
if (PERLDB_SUB && PL_curstash != PL_debstash)
PL_op->op_private |= OPpENTERSUB_DB;
PUTBACK;
- pp_pushmark(ARGS);
+ Perl_pp_pushmark(aTHX);
EXTEND(sp, 4);
PUSHs(pv);
PUSHs(cv);
PUTBACK;
- if (PL_op = pp_entersub(ARGS))
- CALLRUNOPS();
+ if (PL_op = Perl_pp_entersub(aTHX))
+ CALLRUNOPS(aTHX);
LEAVE;
SPAGAIN;
POPSTACK;
if (!SvOK(res)) {
+ char buf[128];
sprintf(buf,"Call to &{$^H{%s}} did not return a defined value", key);
yyerror(buf);
}
}
STATIC char *
-scan_word(register char *s, char *dest, STRLEN destlen, int allow_package, STRLEN *slp)
+S_scan_word(pTHX_ register char *s, char *dest, STRLEN destlen, int allow_package, STRLEN *slp)
{
register char *d = dest;
register char *e = d + destlen - 3; /* two-character token, ending NUL */
for (;;) {
if (d >= e)
- croak(ident_too_long);
- if (isALNUM(*s))
+ Perl_croak(aTHX_ ident_too_long);
+ if (isALNUM(*s)) /* UTF handled below */
*d++ = *s++;
- else if (*s == '\'' && allow_package && isIDFIRST(s[1])) {
+ else if (*s == '\'' && allow_package && isIDFIRST_lazy(s+1)) {
*d++ = ':';
*d++ = ':';
s++;
*d++ = *s++;
*d++ = *s++;
}
- else if (UTF && (*s & 0xc0) == 0x80 && isALNUM_utf8((U8*)s)) {
+ else if (UTF && *(U8*)s >= 0xc0 && isALNUM_utf8((U8*)s)) {
char *t = s + UTF8SKIP(s);
while (*t & 0x80 && is_utf8_mark((U8*)t))
t += UTF8SKIP(t);
if (d + (t - s) > e)
- croak(ident_too_long);
+ Perl_croak(aTHX_ ident_too_long);
Copy(s, d, t - s, char);
d += t - s;
s = t;
}
STATIC char *
-scan_ident(register char *s, register char *send, char *dest, STRLEN destlen, I32 ck_uni)
+S_scan_ident(pTHX_ register char *s, register char *send, char *dest, STRLEN destlen, I32 ck_uni)
{
register char *d;
register char *e;
if (isDIGIT(*s)) {
while (isDIGIT(*s)) {
if (d >= e)
- croak(ident_too_long);
+ Perl_croak(aTHX_ ident_too_long);
*d++ = *s++;
}
}
else {
for (;;) {
if (d >= e)
- croak(ident_too_long);
- if (isALNUM(*s))
+ Perl_croak(aTHX_ ident_too_long);
+ if (isALNUM(*s)) /* UTF handled below */
*d++ = *s++;
- else if (*s == '\'' && isIDFIRST(s[1])) {
+ else if (*s == '\'' && isIDFIRST_lazy(s+1)) {
*d++ = ':';
*d++ = ':';
s++;
*d++ = *s++;
*d++ = *s++;
}
- else if (UTF && (*s & 0xc0) == 0x80 && isALNUM_utf8((U8*)s)) {
+ else if (UTF && *(U8*)s >= 0xc0 && isALNUM_utf8((U8*)s)) {
char *t = s + UTF8SKIP(s);
while (*t & 0x80 && is_utf8_mark((U8*)t))
t += UTF8SKIP(t);
if (d + (t - s) > e)
- croak(ident_too_long);
+ Perl_croak(aTHX_ ident_too_long);
Copy(s, d, t - s, char);
d += t - s;
s = t;
return s;
}
if (*s == '$' && s[1] &&
- (isALNUM(s[1]) || strchr("${", s[1]) || strnEQ(s+1,"::",2)) )
+ (isALNUM_lazy(s+1) || strchr("${", s[1]) || strnEQ(s+1,"::",2)) )
{
return s;
}
if (s < send)
*d = *s++;
d[1] = '\0';
- if (*d == '^' && *s && (isUPPER(*s) || strchr("[\\]^_?", *s))) {
+ if (*d == '^' && *s && isCONTROLVAR(*s)) {
*d = toCTRL(*s);
s++;
}
}
}
}
- if (isIDFIRST(*d) || (UTF && (*d & 0xc0) == 0x80 && isIDFIRST_utf8((U8*)d))) {
+ if (isIDFIRST_lazy(d)) {
d++;
if (UTF) {
e = s;
- while (e < send && (isALNUM(*e) || ((*e & 0xc0) == 0x80 && isALNUM_utf8((U8*)e)) || *e == ':')) {
+ while (e < send && isALNUM_lazy(e) || *e == ':') {
e += UTF8SKIP(e);
while (e < send && *e & 0x80 && is_utf8_mark((U8*)e))
e += UTF8SKIP(e);
s = e;
}
else {
- while (isALNUM(*s) || *s == ':')
+ while ((isALNUM(*s) || *s == ':') && d < e)
*d++ = *s++;
+ if (d >= e)
+ Perl_croak(aTHX_ ident_too_long);
}
*d = '\0';
while (s < send && (*s == ' ' || *s == '\t')) s++;
dTHR; /* only for ckWARN */
if (ckWARN(WARN_AMBIGUOUS) && keyword(dest, d - dest)) {
char *brack = *s == '[' ? "[...]" : "{...}";
- warner(WARN_AMBIGUOUS,
+ Perl_warner(aTHX_ WARN_AMBIGUOUS,
"Ambiguous use of %c{%s%s} resolved to %c%s%s",
funny, dest, brack, funny, dest, brack);
}
PL_lex_brackstack[PL_lex_brackets++] = XOPERATOR;
return s;
}
+ }
+ /* Handle extended ${^Foo} variables
+ * 1999-02-27 mjd-perl-patch@plover.com */
+ else if (!isALNUM(*d) && !isPRINT(*d) /* isCTRL(d) */
+ && isALNUM(*s))
+ {
+ d++;
+ while (isALNUM(*s) && d < e) {
+ *d++ = *s++;
+ }
+ if (d >= e)
+ Perl_croak(aTHX_ ident_too_long);
+ *d = '\0';
}
if (*s == '}') {
s++;
if (PL_lex_state == LEX_NORMAL) {
dTHR; /* only for ckWARN */
if (ckWARN(WARN_AMBIGUOUS) &&
- (keyword(dest, d - dest) || perl_get_cv(dest, FALSE)))
+ (keyword(dest, d - dest) || get_cv(dest, FALSE)))
{
- warner(WARN_AMBIGUOUS,
+ Perl_warner(aTHX_ WARN_AMBIGUOUS,
"Ambiguous use of %c{%s} resolved to %c%s",
funny, dest, funny, dest);
}
return s;
}
-void pmflag(U16 *pmfl, int ch)
+void
+Perl_pmflag(pTHX_ U16 *pmfl, int ch)
{
if (ch == 'i')
*pmfl |= PMf_FOLD;
}
STATIC char *
-scan_pat(char *start, I32 type)
+S_scan_pat(pTHX_ char *start, I32 type)
{
PMOP *pm;
char *s;
if (PL_lex_stuff)
SvREFCNT_dec(PL_lex_stuff);
PL_lex_stuff = Nullsv;
- croak("Search pattern not terminated");
+ Perl_croak(aTHX_ "Search pattern not terminated");
}
pm = (PMOP*)newPMOP(type, 0);
}
STATIC char *
-scan_subst(char *start)
+S_scan_subst(pTHX_ char *start)
{
register char *s;
register PMOP *pm;
if (PL_lex_stuff)
SvREFCNT_dec(PL_lex_stuff);
PL_lex_stuff = Nullsv;
- croak("Substitution pattern not terminated");
+ Perl_croak(aTHX_ "Substitution pattern not terminated");
}
if (s[-1] == PL_multi_open)
if (PL_lex_repl)
SvREFCNT_dec(PL_lex_repl);
PL_lex_repl = Nullsv;
- croak("Substitution replacement not terminated");
+ Perl_croak(aTHX_ "Substitution replacement not terminated");
}
PL_multi_start = first_start; /* so whole substitution is taken together */
if (es) {
SV *repl;
+ PL_sublex_info.super_bufptr = s;
+ PL_sublex_info.super_bufend = PL_bufend;
+ PL_multi_end = 0;
pm->op_pmflags |= PMf_EVAL;
- repl = newSVpv("",0);
+ repl = newSVpvn("",0);
while (es-- > 0)
sv_catpv(repl, es ? "eval " : "do ");
sv_catpvn(repl, "{ ", 2);
sv_catsv(repl, PL_lex_repl);
sv_catpvn(repl, " };", 2);
- SvCOMPILED_on(repl);
+ SvEVALED_on(repl);
SvREFCNT_dec(PL_lex_repl);
PL_lex_repl = repl;
}
}
STATIC char *
-scan_trans(char *start)
+S_scan_trans(pTHX_ char *start)
{
register char* s;
OP *o;
if (PL_lex_stuff)
SvREFCNT_dec(PL_lex_stuff);
PL_lex_stuff = Nullsv;
- croak("Transliteration pattern not terminated");
+ Perl_croak(aTHX_ "Transliteration pattern not terminated");
}
if (s[-1] == PL_multi_open)
s--;
if (PL_lex_repl)
SvREFCNT_dec(PL_lex_repl);
PL_lex_repl = Nullsv;
- croak("Transliteration replacement not terminated");
+ Perl_croak(aTHX_ "Transliteration replacement not terminated");
}
if (UTF) {
utf8 |= OPpTRANS_TO_UTF;
break;
default:
- croak("Too many /C and /U options");
+ Perl_croak(aTHX_ "Too many /C and /U options");
}
}
s++;
}
STATIC char *
-scan_heredoc(register char *s)
+S_scan_heredoc(pTHX_ register char *s)
{
dTHR;
SV *herewas;
s++, term = '\'';
else
term = '"';
- if (!isALNUM(*s))
+ if (!isALNUM_lazy(s))
deprecate("bare << to mean <<\"\"");
- for (; isALNUM(*s); s++) {
+ for (; isALNUM_lazy(s); s++) {
if (d < e)
*d++ = *s;
}
}
if (d >= PL_tokenbuf + sizeof PL_tokenbuf - 1)
- croak("Delimiter for here document is too long");
+ Perl_croak(aTHX_ "Delimiter for here document is too long");
*d++ = '\n';
*d = '\0';
len = d - PL_tokenbuf;
#endif
d = "\n";
if (outer || !(d=ninstr(s,PL_bufend,d,d+1)))
- herewas = newSVpv(s,PL_bufend-s);
+ herewas = newSVpvn(s,PL_bufend-s);
else
- s--, herewas = newSVpv(s,d-s);
+ s--, herewas = newSVpvn(s,d-s);
s += SvCUR(herewas);
tmpstr = NEWSV(87,79);
PL_multi_start = PL_curcop->cop_line;
PL_multi_open = PL_multi_close = '<';
term = *PL_tokenbuf;
- if (!outer) {
+ if (PL_lex_inwhat == OP_SUBST && PL_in_eval && !PL_rsfp) {
+ char *bufptr = PL_sublex_info.super_bufptr;
+ char *bufend = PL_sublex_info.super_bufend;
+ char *olds = s - SvCUR(herewas);
+ s = strchr(bufptr, '\n');
+ if (!s)
+ s = bufend;
+ d = s;
+ while (s < bufend &&
+ (*s != term || memNE(s,PL_tokenbuf,len)) ) {
+ if (*s++ == '\n')
+ PL_curcop->cop_line++;
+ }
+ if (s >= bufend) {
+ PL_curcop->cop_line = PL_multi_start;
+ missingterm(PL_tokenbuf);
+ }
+ sv_setpvn(herewas,bufptr,d-bufptr+1);
+ sv_setpvn(tmpstr,d+1,s-d);
+ s += len - 1;
+ sv_catpvn(herewas,s,bufend-s);
+ (void)strcpy(bufptr,SvPVX(herewas));
+
+ s = olds;
+ goto retval;
+ }
+ else if (!outer) {
d = s;
while (s < PL_bufend &&
(*s != term || memNE(s,PL_tokenbuf,len)) ) {
sv_catsv(tmpstr,PL_linestr);
}
}
- PL_multi_end = PL_curcop->cop_line;
s++;
+retval:
+ PL_multi_end = PL_curcop->cop_line;
if (SvCUR(tmpstr) + 5 < SvLEN(tmpstr)) {
SvLEN_set(tmpstr, SvCUR(tmpstr) + 1);
Renew(SvPVX(tmpstr), SvLEN(tmpstr), char);
*/
STATIC char *
-scan_inputsymbol(char *start)
+S_scan_inputsymbol(pTHX_ char *start)
{
register char *s = start; /* current position in buffer */
register char *d;
register char *e;
+ char *end;
I32 len;
d = PL_tokenbuf; /* start of temp holding space */
e = PL_tokenbuf + sizeof PL_tokenbuf; /* end of temp holding space */
- s = delimcpy(d, e, s + 1, PL_bufend, '>', &len); /* extract until > */
+ end = strchr(s, '\n');
+ if (!end)
+ end = PL_bufend;
+ s = delimcpy(d, e, s + 1, end, '>', &len); /* extract until > */
/* die if we didn't have space for the contents of the <>,
- or if it didn't end
+ or if it didn't end, or if we see a newline
*/
if (len >= sizeof PL_tokenbuf)
- croak("Excessively long <> operator");
- if (s >= PL_bufend)
- croak("Unterminated <> operator");
+ Perl_croak(aTHX_ "Excessively long <> operator");
+ if (s >= end)
+ Perl_croak(aTHX_ "Unterminated <> operator");
s++;
if (*d == '$' && d[1]) d++;
/* allow <Pkg'VALUE> or <Pkg::VALUE> */
- while (*d && (isALNUM(*d) || *d == '\'' || *d == ':'))
+ while (*d && (isALNUM_lazy(d) || *d == '\'' || *d == ':'))
d++;
/* If we've tried to read what we allow filehandles to look like, and
set_csh();
s = scan_str(start);
if (!s)
- croak("Glob not terminated");
+ Perl_croak(aTHX_ "Glob not terminated");
return s;
}
else {
if ((tmp = pad_findmy(d)) != NOT_IN_PAD) {
OP *o = newOP(OP_PADSV, 0);
o->op_targ = tmp;
- PL_lex_op = (OP*)newUNOP(OP_READLINE, 0, newUNOP(OP_RV2GV, 0, o));
+ PL_lex_op = (OP*)newUNOP(OP_READLINE, 0, o);
}
else {
GV *gv = gv_fetchpv(d+1,TRUE, SVt_PV);
PL_lex_op = (OP*)newUNOP(OP_READLINE, 0,
- newUNOP(OP_RV2GV, 0,
newUNOP(OP_RV2SV, 0,
- newGVOP(OP_GV, 0, gv))));
+ newGVOP(OP_GV, 0, gv)));
}
- /* we created the ops in lex_op, so make yylval.ival a null op */
+ PL_lex_op->op_flags |= OPf_SPECIAL;
+ /* we created the ops in PL_lex_op, so make yylval.ival a null op */
yylval.ival = OP_NULL;
}
*/
STATIC char *
-scan_str(char *start)
+S_scan_str(pTHX_ char *start)
{
dTHR;
SV *sv; /* scalar value: string */
Read a number in any of the formats that Perl accepts:
- 0(x[0-7A-F]+)|([0-7]+)
+ 0(x[0-7A-F]+)|([0-7]+)|(b[01])
[\d_]+(\.[\d_]*)?[Ee](\d+)
Underbars (_) are allowed in decimal numbers. If -w is on,
*/
char *
-scan_num(char *start)
+Perl_scan_num(pTHX_ char *start)
{
register char *s = start; /* current position in buffer */
register char *d; /* destination in temp buffer */
register char *e; /* end of temp buffer */
I32 tryiv; /* used to see if it can be an int */
- double value; /* number read, as a double */
+ NV value; /* number read, as a double */
SV *sv; /* place to put the converted number */
I32 floatit; /* boolean: int or float? */
char *lastub = 0; /* position of last underbar */
switch (*s) {
default:
- croak("panic: scan_num");
+ Perl_croak(aTHX_ "panic: scan_num");
/* if it starts with a 0, it could be an octal number, a decimal in
- 0.13 disguise, or a hexadecimal number.
+ 0.13 disguise, or a hexadecimal number, or a binary number.
*/
case '0':
{
/* variables:
u holds the "number so far"
- shift the power of 2 of the base (hex == 4, octal == 3)
+ shift the power of 2 of the base
+ (hex == 4, octal == 3, binary == 1)
overflowed was the number more than we can hold?
Shift is used when we add a digit. It also serves as an "are
- we in octal or hex?" indicator to disallow hex characters when
- in octal mode.
+ we in octal/hex/binary?" indicator to disallow hex characters
+ when in octal mode.
*/
+ dTHR;
UV u;
I32 shift;
- bool overflowed = FALSE;
/* check for hex */
if (s[1] == 'x') {
shift = 4;
s += 2;
+ } else if (s[1] == 'b') {
+ shift = 1;
+ s += 2;
}
/* check for a decimal in disguise */
else if (s[1] == '.')
shift = 3;
u = 0;
- /* read the rest of the octal number */
+ /* read the rest of the number */
for (;;) {
UV n, b; /* n is used in the overflow test, b is the digit we're adding on */
/* 8 and 9 are not octal */
case '8': case '9':
- if (shift != 4)
- yyerror("Illegal octal digit");
+ if (shift == 3)
+ yyerror(Perl_form(aTHX_ "Illegal octal digit '%c'", *s));
+ else
+ if (shift == 1)
+ yyerror(Perl_form(aTHX_ "Illegal binary digit '%c'", *s));
/* FALL THROUGH */
/* octal digits */
- case '0': case '1': case '2': case '3': case '4':
+ case '2': case '3': case '4':
case '5': case '6': case '7':
+ if (shift == 1)
+ yyerror(Perl_form(aTHX_ "Illegal binary digit '%c'", *s));
+ /* FALL THROUGH */
+
+ case '0': case '1':
b = *s++ & 15; /* ASCII digit -> value of digit */
goto digit;
digit:
n = u << shift; /* make room for the digit */
- if (!overflowed && (n >> shift) != u
- && !(PL_hints & HINT_NEW_BINARY)) {
- warn("Integer overflow in %s number",
- (shift == 4) ? "hex" : "octal");
- overflowed = TRUE;
+ if ((n >> shift) != u
+ && !(PL_hints & HINT_NEW_BINARY))
+ {
+ Perl_croak(aTHX_
+ "Integer overflow in %s number",
+ (shift == 4) ? "hexadecimal"
+ : ((shift == 3) ? "octal" : "binary"));
}
u = n | b; /* add the digit to the end */
break;
if (*s == '_') {
dTHR; /* only for ckWARN */
if (ckWARN(WARN_SYNTAX) && lastub && s - lastub != 3)
- warner(WARN_SYNTAX, "Misplaced _ in number");
+ Perl_warner(aTHX_ WARN_SYNTAX, "Misplaced _ in number");
lastub = ++s;
}
else {
/* check for end of fixed-length buffer */
if (d >= e)
- croak(number_too_long);
+ Perl_croak(aTHX_ number_too_long);
/* if we're ok, copy the character */
*d++ = *s++;
}
if (lastub && s - lastub != 3) {
dTHR;
if (ckWARN(WARN_SYNTAX))
- warner(WARN_SYNTAX, "Misplaced _ in number");
+ Perl_warner(aTHX_ WARN_SYNTAX, "Misplaced _ in number");
}
/* read a decimal portion if there is one. avoid
for (; isDIGIT(*s) || *s == '_'; s++) {
/* fixed length buffer check */
if (d >= e)
- croak(number_too_long);
+ Perl_croak(aTHX_ number_too_long);
if (*s != '_')
*d++ = *s;
}
/* read digits of exponent (no underbars :-) */
while (isDIGIT(*s)) {
if (d >= e)
- croak(number_too_long);
+ Perl_croak(aTHX_ number_too_long);
*d++ = *s++;
}
}
/* make an sv from the string */
sv = NEWSV(92,0);
- /* reset numeric locale in case we were earlier left in Swaziland */
- SET_NUMERIC_STANDARD();
- value = atof(PL_tokenbuf);
+
+ value = Atof(PL_tokenbuf);
/*
See if we can make do with an integer value without loss of
conversion at all.
*/
tryiv = I_V(value);
- if (!floatit && (double)tryiv == value)
+ if (!floatit && (NV)tryiv == value)
sv_setiv(sv, tryiv);
else
sv_setnv(sv, value);
}
STATIC char *
-scan_formline(register char *s)
+S_scan_formline(pTHX_ register char *s)
{
dTHR;
register char *eol;
register char *t;
- SV *stuff = newSVpv("",0);
+ SV *stuff = newSVpvn("",0);
bool needargs = FALSE;
while (!needargs) {
#else
for (t = s+1;*t == ' ' || *t == '\t' || *t == '\r'; t++) ;
#endif
- if (*t == '\n')
+ if (*t == '\n' || t == PL_bufend)
break;
}
if (PL_in_eval && !PL_rsfp) {
}
STATIC void
-set_csh(void)
+S_set_csh(pTHX)
{
#ifdef CSH
if (!PL_cshlen)
}
I32
-start_subparse(I32 is_format, U32 flags)
+Perl_start_subparse(pTHX_ I32 is_format, U32 flags)
{
dTHR;
I32 oldsavestack_ix = PL_savestack_ix;
PL_padix = 0;
PL_subline = PL_curcop->cop_line;
#ifdef USE_THREADS
- av_store(PL_comppad_name, 0, newSVpv("@_", 2));
+ av_store(PL_comppad_name, 0, newSVpvn("@_", 2));
PL_curpad[0] = (SV*)newAV();
SvPADMY_on(PL_curpad[0]); /* XXX Needed? */
#endif /* USE_THREADS */
}
int
-yywarn(char *s)
+Perl_yywarn(pTHX_ char *s)
{
dTHR;
--PL_error_count;
- PL_in_eval |= 2;
+ PL_in_eval |= EVAL_WARNONLY;
yyerror(s);
- PL_in_eval &= ~2;
+ PL_in_eval &= ~EVAL_WARNONLY;
return 0;
}
int
-yyerror(char *s)
+Perl_yyerror(pTHX_ char *s)
{
dTHR;
char *where = NULL;
where = "within string";
}
else {
- SV *where_sv = sv_2mortal(newSVpv("next char ", 0));
+ SV *where_sv = sv_2mortal(newSVpvn("next char ", 10));
if (yychar < 32)
- sv_catpvf(where_sv, "^%c", toCTRL(yychar));
+ Perl_sv_catpvf(aTHX_ where_sv, "^%c", toCTRL(yychar));
else if (isPRINT_LC(yychar))
- sv_catpvf(where_sv, "%c", yychar);
+ Perl_sv_catpvf(aTHX_ where_sv, "%c", yychar);
else
- sv_catpvf(where_sv, "\\%03o", yychar & 255);
+ Perl_sv_catpvf(aTHX_ where_sv, "\\%03o", yychar & 255);
where = SvPVX(where_sv);
}
msg = sv_2mortal(newSVpv(s, 0));
- sv_catpvf(msg, " at %_ line %ld, ",
+ Perl_sv_catpvf(aTHX_ msg, " at %_ line %ld, ",
GvSV(PL_curcop->cop_filegv), (long)PL_curcop->cop_line);
if (context)
- sv_catpvf(msg, "near \"%.*s\"\n", contlen, context);
+ Perl_sv_catpvf(aTHX_ msg, "near \"%.*s\"\n", contlen, context);
else
- sv_catpvf(msg, "%s\n", where);
+ Perl_sv_catpvf(aTHX_ msg, "%s\n", where);
if (PL_multi_start < PL_multi_end && (U32)(PL_curcop->cop_line - PL_multi_end) <= 1) {
- sv_catpvf(msg,
+ Perl_sv_catpvf(aTHX_ msg,
" (Might be a runaway multi-line %c%c string starting on line %ld)\n",
(int)PL_multi_open,(int)PL_multi_close,(long)PL_multi_start);
PL_multi_end = 0;
}
- if (PL_in_eval & 2)
- warn("%_", msg);
+ if (PL_in_eval & EVAL_WARNONLY)
+ Perl_warn(aTHX_ "%_", msg);
else if (PL_in_eval)
sv_catsv(ERRSV, msg);
else
PerlIO_write(PerlIO_stderr(), SvPVX(msg), SvCUR(msg));
if (++PL_error_count >= 10)
- croak("%_ has too many errors.\n", GvSV(PL_curcop->cop_filegv));
+ Perl_croak(aTHX_ "%_ has too many errors.\n", GvSV(PL_curcop->cop_filegv));
PL_in_my = 0;
PL_in_my_stash = Nullhv;
return 0;
}
+#ifdef PERL_OBJECT
+#define NO_XSLOCKS
+#include "XSUB.h"
+#endif
+
+/*
+ * restore_rsfp
+ * Restore a source filter.
+ */
+
+static void
+restore_rsfp(pTHXo_ void *f)
+{
+ PerlIO *fp = (PerlIO*)f;
+
+ if (PL_rsfp == PerlIO_stdin())
+ PerlIO_clearerr(PL_rsfp);
+ else if (PL_rsfp && (PL_rsfp != fp))
+ PerlIO_close(PL_rsfp);
+ PL_rsfp = fp;
+}
+
+/*
+ * restore_expect
+ * Restores the state of PL_expect when the lexing that begun with a
+ * start_lex() call has ended.
+ */
+
+static void
+restore_expect(pTHXo_ void *e)
+{
+ /* a safe way to store a small integer in a pointer */
+ PL_expect = (expectation)((char *)e - PL_tokenbuf);
+}
+
+/*
+ * restore_lex_expect
+ * Restores the state of PL_lex_expect when the lexing that begun with a
+ * start_lex() call has ended.
+ */
+
+static void
+restore_lex_expect(pTHXo_ void *e)
+{
+ /* a safe way to store a small integer in a pointer */
+ PL_lex_expect = (expectation)((char *)e - PL_tokenbuf);
+}
https://perl5.git.perl.org / perl5.git / blobdiff