X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=blobdiff_plain;f=toke.c;h=d5e8fb7a8eee076f3f8a4957064b767495fb968a;hb=e5e20432e9e03bcbcca94d9d3a505629a0b07975;hp=63408570dd6b50758dd316aeac9e77ca8b25b750;hpb=4eb8286e4c6ac94fdae21a64c54a936bc836983e;p=p5sagit%2Fp5-mst-13.2.git diff --git a/toke.c b/toke.c index 6340857..d5e8fb7 100644 --- a/toke.c +++ b/toke.c @@ -11,55 +11,26 @@ * "It all comes from here, the stench and the peril." --Frodo */ +/* + * This file is the lexer 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"; +static void restore_rsfp(pTHXo_ void *f); +static void restore_expect(pTHXo_ void *e); +static void restore_lex_expect(pTHXo_ void *e); + #define UTF (PL_hints & HINT_UTF8) /* * Note: we try to be careful never to call the isXXX_utf8() functions @@ -78,7 +49,8 @@ static char ident_too_long[] = "Identifier too long"; * 1999-02-27 mjd-perl-patch@plover.com */ #define isCONTROLVAR(x) (isUPPER(x) || strchr("[\\]^_?", (x))) -/* The following are arranged oddly so that the guard on the switch statement +/* 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). */ @@ -121,17 +93,47 @@ int* yychar_pointer = NULL; # define yylval (*yylval_pointer) # define yychar (*yychar_pointer) # define PERL_YYLEX_PARAM yylval_pointer,yychar_pointer -#else -# define PERL_YYLEX_PARAM +# 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 + * lexer 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 : pattern-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)) @@ -171,8 +173,15 @@ int* yychar_pointer = NULL; /* 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++; @@ -185,33 +194,53 @@ ao(int toketype) return toketype; } +/* + * S_no_op + * When Perl expects an operator and finds something else, no_op + * prints the warning. It always prints " 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"); + 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_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 if (s <= oldbp) - warn("\t(Missing operator before end of line?)\n"); 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; @@ -239,27 +268,41 @@ missingterm(char *s) 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 + */ + 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. + */ + STATIC void -depcom(void) +S_depcom(pTHX) { deprecate("comma-less variable list"); } +/* + * experimental text filters for win32 carriage-returns, utf16-to-utf8 and + * utf16-to-utf8-reversed. + */ + #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) @@ -268,10 +311,8 @@ win32_textfilter(int idx, SV *sv, int 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) { @@ -286,7 +327,7 @@ utf16_textfilter(int idx, SV *sv, int maxlen) } 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) { @@ -300,10 +341,14 @@ utf16rev_textfilter(int idx, SV *sv, int maxlen) return count; } -#endif +/* + * Perl_lex_start + * Initialize variables. Uses the Perl save_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; @@ -363,44 +408,34 @@ lex_start(SV *line) 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 + * Finalizer for lexing operations. Must be called when the parser is + * done with the lexer. + */ + 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 "foo.pm" + * If so, it sets the current line number and file to the values in the comment. + */ STATIC void -incline(char *s) +S_incline(pTHX_ char *s) { dTHR; char *t; @@ -440,8 +475,14 @@ incline(char *s) 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) { @@ -455,6 +496,8 @@ skipspace(register char *s) if (*s++ == '\n' && PL_in_eval && !PL_rsfp) incline(s); } + + /* comment */ if (s < PL_bufend && *s == '#') { while (s < PL_bufend && *s != '\n') s++; @@ -466,9 +509,19 @@ skipspace(register char *s) } } } + + /* 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; - if ((s = filter_gets(PL_linestr, PL_rsfp, (prevlen = SvCUR(PL_linestr)))) == Nullch) { + + /* 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)" : @@ -478,8 +531,19 @@ skipspace(register char *s) } else sv_setpv(PL_linestr,";"); - PL_oldoldbufptr = PL_oldbufptr = PL_bufptr = s = PL_linestart = SvPVX(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()) @@ -489,10 +553,16 @@ skipspace(register char *s) 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); @@ -503,11 +573,21 @@ skipspace(register char *s) } } +/* + * 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; @@ -516,19 +596,28 @@ check_uni(void) { 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; @@ -546,10 +635,23 @@ uni(I32 f, char *s) #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; @@ -569,8 +671,17 @@ lop(I32 f, expectation x, char *s) return LSTOP; } +/* + * S_force_next + * When the lexer realizes it knows the next token (for instance, + * it is reordering tokens for the parser) then it can call S_force_next + * to know what token to return the next time the lexer is called. Caller + * will need to 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++; @@ -581,8 +692,24 @@ force_next(I32 type) } } +/* + * 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" lexer 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; @@ -611,8 +738,17 @@ force_word(register char *start, int token, int check_keyword, int allow_pack, i return s; } +/* + * S_force_ident + * Called when the lexer wants $foo *foo &foo etc, but the program + * text only contains the "foo" portion. The first argument is a pointer + * to the "foo", and the second argument is the type symbol to prefix. + * Forces the next token to be a "WORD". + * Creates the symbol if it didn't already exist (via gv_fetchpv()). + */ + STATIC void -force_ident(register char *s, int kind) +S_force_ident(pTHX_ register char *s, int kind) { if (s && *s) { OP* o = (OP*)newSVOP(OP_CONST, 0, newSVpv(s,0)); @@ -634,8 +770,13 @@ force_ident(register char *s, int kind) } } +/* + * S_force_version + * Forces the next token to be a version number. + */ + STATIC char * -force_version(char *s) +S_force_version(pTHX_ char *s) { OP *version = Nullop; @@ -661,8 +802,16 @@ force_version(char *s) return (s); } +/* + * S_tokeq + * Tokenize a quoted string passed in as an SV. It finds the next + * chunk, up to end of string or a backslash. It may make a new + * SV containing that chunk (if HINT_NEW_STRING is on). It also + * turns \\ into \. + */ + STATIC SV * -tokeq(SV *sv) +S_tokeq(pTHX_ SV *sv) { register char *s; register char *send; @@ -683,7 +832,7 @@ tokeq(SV *sv) goto finish; d = s; if ( PL_hints & HINT_NEW_STRING ) - pv = sv_2mortal(newSVpv(SvPVX(pv), len)); + pv = sv_2mortal(newSVpvn(SvPVX(pv), len)); while (s < send) { if (*s == '\\') { if (s + 1 < send && (s[1] == '\\')) @@ -699,8 +848,40 @@ tokeq(SV *sv) return sv; } +/* + * Now come three functions related to double-quote context, + * S_sublex_start, S_sublex_push, and S_sublex_done. They're used when + * converting things like "\u\Lgnat" into ucfirst(lc("gnat")). They + * interact with PL_lex_state, and create fake ( ... ) argument lists + * to handle functions and concatenation. + * They assume that whoever calls them will be setting up a fake + * join call, because each subthing puts a ',' after it. This lets + * "lower \luPpEr" + * become + * join($, , '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; @@ -719,7 +900,7 @@ sublex_start(void) SV *nsv; p = SvPV(sv, len); - nsv = newSVpv(p, len); + nsv = newSVpvn(p, len); SvREFCNT_dec(sv); sv = nsv; } @@ -743,8 +924,16 @@ sublex_start(void) 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; @@ -770,7 +959,8 @@ sublex_push(void) PL_linestr = PL_lex_stuff; PL_lex_stuff = Nullsv; - PL_bufend = PL_bufptr = PL_oldbufptr = PL_oldoldbufptr = PL_linestart = SvPVX(PL_linestr); + PL_bufend = PL_bufptr = PL_oldbufptr = PL_oldoldbufptr = PL_linestart + = SvPVX(PL_linestr); PL_bufend += SvCUR(PL_linestr); SAVEFREESV(PL_linestr); @@ -796,21 +986,26 @@ sublex_push(void) 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; } if (PL_lex_casemods) { /* oops, we've got some unbalanced parens */ PL_lex_state = LEX_INTERPCASEMOD; - return yylex(PERL_YYLEX_PARAM); + 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; @@ -823,7 +1018,7 @@ sublex_done(void) 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 @@ -920,7 +1115,7 @@ sublex_done(void) */ 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 */ @@ -928,15 +1123,14 @@ scan_const(char *start) 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 - ? (PL_sublex_info.sub_op->op_private & (PL_lex_repl ? OPpTRANS_FROM_UTF : OPpTRANS_TO_UTF)) + 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 = + char *leaveit = /* set of acceptably-backslashed characters */ PL_lex_inpat ? "\\.^$@AGZdDwWsSbBpPXC+*?|()-nrtfeaxcz0123456789[{]} \t\n\r\f\v#" : ""; @@ -951,8 +1145,8 @@ scan_const(char *start) I32 max; /* last character in range */ i = d - SvPVX(sv); /* remember current offset */ - SvGROW(sv, SvLEN(sv) + 256); /* expand the sv -- there'll never be more'n 256 chars in a range for it to grow by */ - d = SvPVX(sv) + i; /* restore d after the grow potentially has changed the ptr */ + SvGROW(sv, SvLEN(sv) + 256); /* never more than 256 chars in a range */ + d = SvPVX(sv) + i; /* refresh d after realloc */ d -= 2; /* eat the first char and the - */ min = (U8)*d; /* first char in range */ @@ -1077,7 +1271,7 @@ scan_const(char *start) { 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; } @@ -1102,7 +1296,7 @@ scan_const(char *start) { dTHR; if (ckWARN(WARN_UNSAFE) && isALPHA(*s)) - warner(WARN_UNSAFE, + Perl_warner(aTHX_ WARN_UNSAFE, "Unrecognized escape \\%c passed through", *s); /* default action is to copy the quoted character */ @@ -1130,14 +1324,13 @@ scan_const(char *start) 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); @@ -1150,7 +1343,7 @@ scan_const(char *start) 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); } @@ -1190,12 +1383,21 @@ scan_const(char *start) 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++; @@ -1232,9 +1434,29 @@ scan_const(char *start) return s; } +/* S_intuit_more + * Returns TRUE if there's more to the expression (e.g., a subscript), + * FALSE otherwise. + * + * 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 + */ + /* This is the one truly awful dwimmer necessary to conflate C and sed. */ + STATIC int -intuit_more(register char *s) +S_intuit_more(pTHX_ register char *s) { if (PL_lex_brackets) return TRUE; @@ -1268,6 +1490,7 @@ intuit_more(register char *s) 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; @@ -1363,8 +1586,29 @@ intuit_more(register char *s) 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]; @@ -1387,6 +1631,11 @@ intuit_method(char *start, GV *gv) 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; @@ -1411,7 +1660,7 @@ intuit_method(char *start, GV *gv) 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); @@ -1422,8 +1671,15 @@ intuit_method(char *start, GV *gv) 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"); @@ -1454,7 +1710,7 @@ incl_perldb(void) */ 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 */ PL_filter_debug = atoi((char*)datasv); @@ -1465,12 +1721,14 @@ filter_add(filter_t funcp, SV *datasv) 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 */ +#ifdef DEBUGGING if (PL_filter_debug) { STRLEN n_a; - warn("filter_add func %p (%s)", funcp, SvPV(datasv, 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); @@ -1479,26 +1737,29 @@ filter_add(filter_t funcp, SV *datasv) /* Delete most recently added instance of this filter function. */ void -filter_del(filter_t funcp) +Perl_filter_del(pTHX_ filter_t funcp) { +#ifdef DEBUGGING if (PL_filter_debug) - warn("filter_del func %p", funcp); + 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 */ @@ -1511,8 +1772,10 @@ filter_read(int idx, SV *buf_sv, int maxlen) 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. */ +#ifdef DEBUGGING if (PL_filter_debug) - warn("filter_read %d: from rsfp\n", idx); + Perl_warn(aTHX_ "filter_read %d: from rsfp\n", idx); +#endif /* DEBUGGING */ if (maxlen) { /* Want a block */ int len ; @@ -1540,25 +1803,29 @@ filter_read(int idx, SV *buf_sv, int maxlen) } /* Skip this filter slot if filter has been deleted */ if ( (datasv = FILTER_DATA(idx)) == &PL_sv_undef){ +#ifdef DEBUGGING if (PL_filter_debug) - warn("filter_read %d: skipped (filter deleted)\n", idx); + 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); +#ifdef DEBUGGING if (PL_filter_debug) { STRLEN n_a; - warn("filter_read %d: via function %p (%s)\n", + 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) { @@ -1609,7 +1876,12 @@ filter_gets(register SV *sv, register PerlIO *fp, STRLEN append) if we already built the token before, use it. */ -int yylex(PERL_YYLEX_PARAM_DECL) +int +#ifdef USE_PURE_BISON +Perl_yylex(pTHX_ YYSTYPE *lvalp, int *lcharp) +#else +Perl_yylex(pTHX) +#endif { dTHR; register char *s; @@ -1638,7 +1910,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) */ if (PL_in_my) { if (strchr(PL_tokenbuf,':')) - yyerror(form(PL_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); @@ -1681,7 +1953,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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); } } @@ -1701,7 +1973,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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)); } @@ -1741,7 +2013,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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') { @@ -1761,7 +2033,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (PL_bufptr != PL_bufend) PL_bufptr += 2; PL_lex_state = LEX_INTERPCONCAT; - return yylex(PERL_YYLEX_PARAM); + return yylex(); } else { s = PL_bufptr + 1; @@ -1796,7 +2068,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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) { @@ -1805,7 +2077,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) Aop(OP_CONCAT); } else - return yylex(PERL_YYLEX_PARAM); + return yylex(); } case LEX_INTERPPUSH: @@ -1838,7 +2110,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) s = PL_bufptr; Aop(OP_CONCAT); } - return yylex(PERL_YYLEX_PARAM); + return yylex(); case LEX_INTERPENDMAYBE: if (intuit_more(PL_bufptr)) { @@ -1854,17 +2126,17 @@ int yylex(PERL_YYLEX_PARAM_DECL) return ')'; } if (PL_lex_inwhat == OP_SUBST && PL_linestr == PL_lex_repl - && SvCOMPILED(PL_lex_repl)) + && SvEVALED(PL_lex_repl)) { if (PL_bufptr != PL_bufend) - croak("Bad evalled substitution pattern"); + 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(); @@ -1894,11 +2166,11 @@ int yylex(PERL_YYLEX_PARAM_DECL) Aop(OP_CONCAT); else { PL_bufptr = s; - return yylex(PERL_YYLEX_PARAM); + return yylex(); } } - return yylex(PERL_YYLEX_PARAM); + return yylex(); case LEX_FORMLINE: PL_lex_state = LEX_NORMAL; s = scan_formline(PL_bufptr); @@ -1919,7 +2191,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) default: if (isIDFIRST_lazy(s)) goto keylookup; - croak("Unrecognized character \\x%02X", *s & 255); + Perl_croak(aTHX_ "Unrecognized character \\x%02X", *s & 255); case 4: case 26: goto fake_eof; /* emulate EOF on ^D or ^Z */ @@ -1928,7 +2200,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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) @@ -1961,20 +2233,20 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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 @@ -2138,7 +2410,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) newargv = PL_origargv; newargv[0] = ipath; PerlProc_execv(ipath, newargv); - croak("Can't exec %s", ipath); + Perl_croak(aTHX_ "Can't exec %s", ipath); } if (d) { U32 oldpdb = PL_perldb; @@ -2153,7 +2425,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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); @@ -2178,13 +2450,13 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (PL_lex_formbrack && PL_lex_brackets <= PL_lex_formbrack) { PL_bufptr = s; PL_lex_state = LEX_FORMLINE; - return yylex(PERL_YYLEX_PARAM); + return yylex(); } 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: @@ -2202,7 +2474,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (PL_lex_formbrack && PL_lex_brackets <= PL_lex_formbrack) { PL_bufptr = s; PL_lex_state = LEX_FORMLINE; - return yylex(PERL_YYLEX_PARAM); + return yylex(); } } else { @@ -2254,7 +2526,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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; } } @@ -2372,7 +2644,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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) { @@ -2513,7 +2785,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) || (*t == '=' && t[1] == '>'))) OPERATOR(HASHBRACK); if (PL_expect == XREF) - PL_expect = XSTATE; /* was XTERM, trying XSTATE */ + PL_expect = XTERM; else { PL_lex_brackstack[PL_lex_brackets-1] = XSTATE; PL_expect = XSTATE; @@ -2529,7 +2801,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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) @@ -2539,7 +2811,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (PL_lex_fakebrack) { PL_lex_state = LEX_INTERPEND; PL_bufptr = s; - return yylex(PERL_YYLEX_PARAM); /* ignore fake brackets */ + return yylex(); /* ignore fake brackets */ } if (*s == '-' && s[1] == '>') PL_lex_state = LEX_INTERPENDMAYBE; @@ -2550,7 +2822,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (PL_lex_brackets < PL_lex_fakebrack) { PL_bufptr = s; PL_lex_fakebrack = 0; - return yylex(PERL_YYLEX_PARAM); /* ignore fake brackets */ + return yylex(); /* ignore fake brackets */ } force_next('}'); TOKEN(';'); @@ -2563,7 +2835,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (PL_expect == XOPERATOR) { if (ckWARN(WARN_SEMICOLON) && isIDFIRST_lazy(s) && PL_bufptr == PL_linestart) { PL_curcop->cop_line--; - warner(WARN_SEMICOLON, PL_warn_nosemi); + Perl_warner(aTHX_ WARN_SEMICOLON, PL_warn_nosemi); PL_curcop->cop_line++; } BAop(OP_BIT_AND); @@ -2596,7 +2868,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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') ) @@ -2692,11 +2964,11 @@ int yylex(PERL_YYLEX_PARAM_DECL) } if (s[1] == '#' && (isIDFIRST_lazy(s+2) || strchr("{$:+-", s[2]))) { - if (PL_expect == XOPERATOR) - no_op("Array length", PL_bufptr); 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; @@ -2704,10 +2976,11 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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"); @@ -2723,6 +2996,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) } d = s; + tmp = (I32)*s; if (PL_lex_state == LEX_NORMAL) s = skipspace(s); @@ -2738,7 +3012,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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); } @@ -2755,8 +3029,8 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (isIDFIRST_lazy(t)) { t = scan_word(t, tmpbuf, sizeof tmpbuf, TRUE, &len); for (; isSPACE(*t); t++) ; - if (*t == ';' && perl_get_cv(tmpbuf, FALSE)) - warner(WARN_SYNTAX, + if (*t == ';' && get_cv(tmpbuf, FALSE)) + Perl_warner(aTHX_ WARN_SYNTAX, "You need to quote \"%s\"", tmpbuf); } } @@ -2764,7 +3038,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) } 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; @@ -2835,7 +3109,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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); } @@ -2949,7 +3223,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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); @@ -3068,7 +3342,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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 &"); } @@ -3089,7 +3363,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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; } @@ -3097,7 +3371,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (PL_expect == XOPERATOR) { if (PL_bufptr == PL_linestart) { PL_curcop->cop_line--; - warner(WARN_SEMICOLON, PL_warn_nosemi); + Perl_warner(aTHX_ WARN_SEMICOLON, PL_warn_nosemi); PL_curcop->cop_line++; } else @@ -3112,7 +3386,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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; @@ -3129,7 +3403,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) /* 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 @@ -3152,11 +3426,8 @@ int yylex(PERL_YYLEX_PARAM_DECL) PL_oldoldbufptr < PL_bufptr && (PL_oldoldbufptr == PL_last_lop || PL_oldoldbufptr == PL_last_uni) && /* NO SKIPSPACE BEFORE HERE! */ - (PL_expect == XREF - || ((PL_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 == '('; @@ -3172,8 +3443,10 @@ int yylex(PERL_YYLEX_PARAM_DECL) /* (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; } @@ -3186,11 +3459,8 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (*s == '(') { CLINE; if (gv && GvCVu(gv)) { - CV *cv; - if ((cv = GvCV(gv)) && SvPOK(cv)) - PL_last_proto = SvPV((SV*)cv, n_a); for (d = s + 1; *d == ' ' || *d == '\t'; d++) ; - if (*d == ')' && (sv = cv_const_sv(cv))) { + if (*d == ')' && (sv = cv_const_sv(GvCV(gv)))) { s = d + 1; goto its_constant; } @@ -3199,7 +3469,6 @@ int yylex(PERL_YYLEX_PARAM_DECL) PL_expect = XOPERATOR; force_next(WORD); yylval.ival = 0; - PL_last_lop_op = OP_ENTERSUB; TOKEN('&'); } @@ -3220,11 +3489,10 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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))) { @@ -3238,60 +3506,51 @@ int yylex(PERL_YYLEX_PARAM_DECL) /* 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 && - !(PL_last_lop_op == OP_ENTERSUB - && PL_last_proto - && PL_last_proto[PL_last_proto[0] == ';' ? 1 : 0] == '*')) - { - 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, PL_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); @@ -3304,7 +3563,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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__: @@ -3323,7 +3582,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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(); @@ -3538,7 +3797,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) p += 2; p = skipspace(p); if (isIDFIRST_lazy(p)) - croak("Missing $ on loop variable"); + Perl_croak(aTHX_ "Missing $ on loop variable"); } OPERATOR(FOR); @@ -3725,8 +3984,8 @@ int yylex(PERL_YYLEX_PARAM_DECL) TERM(sublex_start()); case KEY_map: - LOP(OP_MAPSTART, XREF); - + LOP(OP_MAPSTART, *s == '(' ? XTERM : XREF); + case KEY_mkdir: LOP(OP_MKDIR,XTERM); @@ -3781,9 +4040,10 @@ int yylex(PERL_YYLEX_PARAM_DECL) char *t; 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); @@ -3856,12 +4116,12 @@ int yylex(PERL_YYLEX_PARAM_DECL) if (!warned && ckWARN(WARN_SYNTAX)) { for (; !isSPACE(*d) && len; --len, ++d) { if (*d == ',') { - warner(WARN_SYNTAX, + Perl_warner(aTHX_ WARN_SYNTAX, "Possible attempt to separate words with commas"); ++warned; } else if (*d == '#') { - warner(WARN_SYNTAX, + Perl_warner(aTHX_ WARN_SYNTAX, "Possible attempt to put comments in qw() list"); ++warned; } @@ -4061,7 +4321,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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); @@ -4131,7 +4391,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) 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); @@ -4323,7 +4583,7 @@ int yylex(PERL_YYLEX_PARAM_DECL) } I32 -keyword(register char *d, I32 len) +Perl_keyword(pTHX_ register char *d, I32 len) { switch (*d) { case '_': @@ -4446,7 +4706,7 @@ keyword(register char *d, I32 len) 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; @@ -4942,7 +5202,7 @@ keyword(register char *d, I32 len) } STATIC void -checkcomma(register char *s, char *name, char *what) +S_checkcomma(pTHX_ register char *s, char *name, char *what) { char *w; @@ -4959,7 +5219,7 @@ checkcomma(register char *s, char *name, char *what) 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)) @@ -4977,17 +5237,17 @@ checkcomma(register char *s, char *name, char *what) 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 */ @@ -5011,7 +5271,7 @@ new_constant(char *s, STRLEN len, char *key, SV *sv, SV *pv, char *type) 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 @@ -5029,7 +5289,7 @@ new_constant(char *s, STRLEN len, char *key, SV *sv, SV *pv, char *type) 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); @@ -5038,8 +5298,8 @@ new_constant(char *s, STRLEN len, char *key, SV *sv, SV *pv, char *type) PUSHs(cv); PUTBACK; - if (PL_op = pp_entersub(ARGS)) - CALLRUNOPS(); + if (PL_op = Perl_pp_entersub(aTHX)) + CALLRUNOPS(aTHX); LEAVE; SPAGAIN; @@ -5057,13 +5317,13 @@ new_constant(char *s, STRLEN len, char *key, SV *sv, SV *pv, char *type) } 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); + Perl_croak(aTHX_ ident_too_long); if (isALNUM(*s)) /* UTF handled below */ *d++ = *s++; else if (*s == '\'' && allow_package && isIDFIRST_lazy(s+1)) { @@ -5080,7 +5340,7 @@ scan_word(register char *s, char *dest, STRLEN destlen, int allow_package, STRLE 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; @@ -5094,7 +5354,7 @@ scan_word(register char *s, char *dest, STRLEN destlen, int allow_package, STRLE } 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; @@ -5110,14 +5370,14 @@ scan_ident(register char *s, register char *send, char *dest, STRLEN destlen, I3 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); + Perl_croak(aTHX_ ident_too_long); if (isALNUM(*s)) /* UTF handled below */ *d++ = *s++; else if (*s == '\'' && isIDFIRST_lazy(s+1)) { @@ -5134,7 +5394,7 @@ scan_ident(register char *s, register char *send, char *dest, STRLEN destlen, I3 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; @@ -5195,7 +5455,7 @@ scan_ident(register char *s, register char *send, char *dest, STRLEN destlen, I3 while ((isALNUM(*s) || *s == ':') && d < e) *d++ = *s++; if (d >= e) - croak(ident_too_long); + Perl_croak(aTHX_ ident_too_long); } *d = '\0'; while (s < send && (*s == ' ' || *s == '\t')) s++; @@ -5203,7 +5463,7 @@ scan_ident(register char *s, register char *send, char *dest, STRLEN destlen, I3 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); } @@ -5223,7 +5483,7 @@ scan_ident(register char *s, register char *send, char *dest, STRLEN destlen, I3 *d++ = *s++; } if (d >= e) - croak(ident_too_long); + Perl_croak(aTHX_ ident_too_long); *d = '\0'; } if (*s == '}') { @@ -5235,9 +5495,9 @@ scan_ident(register char *s, register char *send, char *dest, STRLEN destlen, I3 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); } @@ -5253,7 +5513,8 @@ scan_ident(register char *s, register char *send, char *dest, STRLEN destlen, I3 return s; } -void pmflag(U16 *pmfl, int ch) +void +Perl_pmflag(pTHX_ U16 *pmfl, int ch) { if (ch == 'i') *pmfl |= PMf_FOLD; @@ -5272,7 +5533,7 @@ void pmflag(U16 *pmfl, int ch) } STATIC char * -scan_pat(char *start, I32 type) +S_scan_pat(pTHX_ char *start, I32 type) { PMOP *pm; char *s; @@ -5282,7 +5543,7 @@ scan_pat(char *start, I32 type) 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); @@ -5304,7 +5565,7 @@ scan_pat(char *start, I32 type) } STATIC char * -scan_subst(char *start) +S_scan_subst(pTHX_ char *start) { register char *s; register PMOP *pm; @@ -5319,7 +5580,7 @@ scan_subst(char *start) 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) @@ -5334,7 +5595,7 @@ scan_subst(char *start) 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 */ @@ -5356,13 +5617,13 @@ scan_subst(char *start) 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; } @@ -5374,7 +5635,7 @@ scan_subst(char *start) } STATIC char * -scan_trans(char *start) +S_scan_trans(pTHX_ char *start) { register char* s; OP *o; @@ -5392,7 +5653,7 @@ scan_trans(char *start) 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--; @@ -5405,7 +5666,7 @@ scan_trans(char *start) 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) { @@ -5441,7 +5702,7 @@ scan_trans(char *start) utf8 |= OPpTRANS_TO_UTF; break; default: - croak("Too many /C and /U options"); + Perl_croak(aTHX_ "Too many /C and /U options"); } } s++; @@ -5454,7 +5715,7 @@ scan_trans(char *start) } STATIC char * -scan_heredoc(register char *s) +S_scan_heredoc(pTHX_ register char *s) { dTHR; SV *herewas; @@ -5494,7 +5755,7 @@ scan_heredoc(register char *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; @@ -5524,9 +5785,9 @@ scan_heredoc(register char *s) #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); @@ -5664,25 +5925,29 @@ retval: */ 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++; @@ -5710,7 +5975,7 @@ scan_inputsymbol(char *start) set_csh(); s = scan_str(start); if (!s) - croak("Glob not terminated"); + Perl_croak(aTHX_ "Glob not terminated"); return s; } else { @@ -5800,7 +6065,7 @@ scan_inputsymbol(char *start) */ STATIC char * -scan_str(char *start) +S_scan_str(pTHX_ char *start) { dTHR; SV *sv; /* scalar value: string */ @@ -5987,13 +6252,13 @@ scan_str(char *start) */ 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 */ @@ -6003,7 +6268,7 @@ scan_num(char *start) 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, or a binary number. @@ -6020,9 +6285,22 @@ scan_num(char *start) we in octal/hex/binary?" indicator to disallow hex characters when in octal mode. */ - UV u; + dTHR; + NV n = 0.0; + UV u = 0; I32 shift; bool overflowed = FALSE; + static NV nvshift[5] = { 1.0, 2.0, 4.0, 8.0, 16.0 }; + static char* bases[5] = { "", "binary", "", "octal", + "hexadecimal" }; + static char* Bases[5] = { "", "Binary", "", "Octal", + "Hexadecimal" }; + static char *maxima[5] = { "", + "0b11111111111111111111111111111111", + "", + "0b37777777777", + "0xffffffff" }; + char *base, *Base, *max; /* check for hex */ if (s[1] == 'x') { @@ -6038,11 +6316,16 @@ scan_num(char *start) /* so it must be octal */ else shift = 3; - u = 0; + + base = bases[shift]; + Base = Bases[shift]; + max = maxima[shift]; /* 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 */ + /* x is used in the overflow test, + b is the digit we're adding on */ + UV x, b; switch (*s) { @@ -6058,17 +6341,17 @@ scan_num(char *start) /* 8 and 9 are not octal */ case '8': case '9': if (shift == 3) - yyerror("Illegal octal digit"); + yyerror(Perl_form(aTHX_ "Illegal octal digit '%c'", *s)); else if (shift == 1) - yyerror("Illegal binary digit"); + yyerror(Perl_form(aTHX_ "Illegal binary digit '%c'", *s)); /* FALL THROUGH */ /* octal digits */ case '2': case '3': case '4': case '5': case '6': case '7': if (shift == 1) - yyerror("Illegal binary digit"); + yyerror(Perl_form(aTHX_ "Illegal binary digit '%c'", *s)); /* FALL THROUGH */ case '0': case '1': @@ -6088,15 +6371,34 @@ scan_num(char *start) */ 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" - : ((shift == 3) ? "octal" : "binary")); - overflowed = TRUE; + if (!overflowed) { + x = u << shift; /* make room for the digit */ + + if ((x >> shift) != u + && !(PL_hints & HINT_NEW_BINARY)) { + dTHR; + overflowed = TRUE; + n = (NV) u; + if (ckWARN_d(WARN_UNSAFE)) + Perl_warner(aTHX_ ((shift == 3) ? + WARN_OCTAL : WARN_UNSAFE), + "Integer overflow in %s number", + base); + } else + u = x | b; /* add the digit to the end */ + } + if (overflowed) { + n *= nvshift[shift]; + /* If an NV has not enough bits in its + * mantissa to represent an UV this summing of + * small low-order numbers is a waste of time + * (because the NV cannot preserve the + * low-order bits anyway): we could just + * remember when did we overflow and in the + * end just multiply n by the right + * amount. */ + n += (NV) b; } - u = n | b; /* add the digit to the end */ break; } } @@ -6106,8 +6408,25 @@ scan_num(char *start) */ out: sv = NEWSV(92,0); - sv_setuv(sv, u); - if ( PL_hints & HINT_NEW_BINARY) + if (overflowed) { + dTHR; + if (ckWARN(WARN_UNSAFE) && (NV) n > 4294967295.0) + Perl_warner(aTHX_ WARN_UNSAFE, + "%s number > %s non-portable", + Base, max); + sv_setnv(sv, n); + } + else { +#if UV_SIZEOF > 4 + dTHR; + if (ckWARN(WARN_UNSAFE) && u > 0xffffffff) + Perl_warner(aTHX_ WARN_UNSAFE, + "%s number > %s non-portable", + Base, max); +#endif + sv_setuv(sv, u); + } + if (PL_hints & HINT_NEW_BINARY) sv = new_constant(start, s - start, "binary", sv, Nullsv, NULL); } break; @@ -6131,13 +6450,13 @@ scan_num(char *start) 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++; } @@ -6147,7 +6466,7 @@ scan_num(char *start) 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 @@ -6164,7 +6483,7 @@ scan_num(char *start) 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; } @@ -6185,7 +6504,7 @@ scan_num(char *start) /* read digits of exponent (no underbars :-) */ while (isDIGIT(*s)) { if (d >= e) - croak(number_too_long); + Perl_croak(aTHX_ number_too_long); *d++ = *s++; } } @@ -6195,9 +6514,8 @@ scan_num(char *start) /* 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 @@ -6210,7 +6528,7 @@ scan_num(char *start) 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); @@ -6228,12 +6546,12 @@ scan_num(char *start) } 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) { @@ -6302,7 +6620,7 @@ scan_formline(register char *s) } STATIC void -set_csh(void) +S_set_csh(pTHX) { #ifdef CSH if (!PL_cshlen) @@ -6311,7 +6629,7 @@ set_csh(void) } I32 -start_subparse(I32 is_format, U32 flags) +Perl_start_subparse(pTHX_ I32 is_format, U32 flags) { dTHR; I32 oldsavestack_ix = PL_savestack_ix; @@ -6346,7 +6664,7 @@ start_subparse(I32 is_format, U32 flags) 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 */ @@ -6368,18 +6686,18 @@ start_subparse(I32 is_format, U32 flags) } 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; @@ -6415,39 +6733,86 @@ yyerror(char *s) 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); +}