X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/da5462d42a4270b1b09ced415e09bdec229074ff..026816664ff8283a55f91915843a8ff0ac5cf86c:/src/scan-gram.l?ds=sidebyside diff --git a/src/scan-gram.l b/src/scan-gram.l index 1833b9a9..555e6952 100644 --- a/src/scan-gram.l +++ b/src/scan-gram.l @@ -1,6 +1,6 @@ /* Bison Grammar Scanner -*- C -*- - Copyright (C) 2002-2012 Free Software Foundation, Inc. + Copyright (C) 2002-2013 Free Software Foundation, Inc. This file is part of Bison, the GNU Compiler Compiler. @@ -17,7 +17,7 @@ You should have received a copy of the GNU General Public License along with this program. If not, see . */ -%option debug nodefault noinput nounput noyywrap never-interactive +%option debug nodefault noinput noyywrap never-interactive %option prefix="gram_" outfile="lex.yy.c" %{ @@ -31,12 +31,13 @@ #include #include +#include #include #include #include #include -#include +#include #include #include @@ -73,6 +74,15 @@ static size_t no_cr_read (FILE *, char *, size_t); yyless (0); \ } while (0) +#define DEPRECATED(Msg) \ + do { \ + size_t i; \ + deprecated_directive (loc, yytext, Msg); \ + scanner_cursor.column -= mbsnwidth (Msg, strlen (Msg), 0); \ + for (i = strlen (Msg); i != 0; --i) \ + unput (Msg[i - 1]); \ + } while (0) + /* A string representing the most recently saved token. */ static char *last_string; @@ -100,16 +110,21 @@ static void unexpected_newline (boundary, char const *); /* Strings and characters in directives/rules. */ %x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER /* A identifier was just read in directives/rules. Special state - to capture the sequence `identifier :'. */ + to capture the sequence 'identifier :'. */ %x SC_AFTER_IDENTIFIER - /* A complex tag, with nested angles brackets. */ + + /* POSIX says that a tag must be both an id and a C union member, but + historically almost any character is allowed in a tag. We + disallow NUL, as this simplifies our implementation. We match + angle brackets in nested pairs: several languages use them for + generics/template types. */ %x SC_TAG /* Four types of user code: - - prologue (code between `%{' `%}' in the first section, before %%); + - prologue (code between '%{' '%}' in the first section, before %%); - actions, printers, union, etc, (between braced in the middle section); - epilogue (everything after the second %%). - - predicate (code between `%?{' and `{' in middle section); */ + - predicate (code between '%?{' and '{' in middle section); */ %x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE SC_PREDICATE /* C and C++ comments in code. */ %x SC_COMMENT SC_LINE_COMMENT @@ -119,21 +134,18 @@ static void unexpected_newline (boundary, char const *); %x SC_BRACKETED_ID SC_RETURN_BRACKETED_ID letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] +notletter [^.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]{-}[%\{] id {letter}({letter}|[-0-9])* -directive %{id} int [0-9]+ -/* POSIX says that a tag must be both an id and a C union member, but - historically almost any character is allowed in a tag. We disallow - NUL, as this simplifies our implementation. We disallow angle - bracket to match them in nested pairs: several languages use them - for generics/template types. */ -tag [^\0<>]+ - /* Zero or more instances of backslash-newline. Following GCC, allow white space between the backslash and the newline. */ splice (\\[ \f\t\v]*\n)* +/* An equal sign, with optional leading whitespaces. This is used in some + deprecated constructs. */ +eqopt ([[:space:]]*=)? + %% %{ /* Nesting level. Either for nested braces, or nested angle brackets @@ -164,9 +176,11 @@ splice (\\[ \f\t\v]*\n)* { /* Comments and white space. */ - "," warn_at (*loc, _("stray `,' treated as white space")); + "," { + complain (loc, Wother, _("stray ',' treated as white space")); + } [ \f\n\t\v] | - "//".* ; + "//".* continue; "/*" { token_start = loc->start; context_state = YY_START; @@ -175,7 +189,7 @@ splice (\\[ \f\t\v]*\n)* /* #line directives are not documented, and may be withdrawn or modified in future versions of Bison. */ - ^"#line "{int}" \"".*"\"\n" { + ^"#line "{int}(" \"".*"\"")?"\n" { handle_syncline (yytext + sizeof "#line " - 1, *loc); } } @@ -196,16 +210,17 @@ splice (\\[ \f\t\v]*\n)* "%binary" return PERCENT_NONASSOC; "%code" return PERCENT_CODE; "%debug" RETURN_PERCENT_FLAG("parse.trace"); - "%default"[-_]"prec" return PERCENT_DEFAULT_PREC; + "%default-prec" return PERCENT_DEFAULT_PREC; "%define" return PERCENT_DEFINE; "%defines" return PERCENT_DEFINES; "%destructor" return PERCENT_DESTRUCTOR; "%dprec" return PERCENT_DPREC; - "%error"[-_]"verbose" return PERCENT_ERROR_VERBOSE; + "%empty" return PERCENT_EMPTY; + "%error-verbose" return PERCENT_ERROR_VERBOSE; "%expect" return PERCENT_EXPECT; - "%expect"[-_]"rr" return PERCENT_EXPECT_RR; + "%expect-rr" return PERCENT_EXPECT_RR; "%file-prefix" return PERCENT_FILE_PREFIX; - "%fixed"[-_]"output"[-_]"files" return PERCENT_YACC; + "%fixed-output-files" return PERCENT_YACC; "%initial-action" return PERCENT_INITIAL_ACTION; "%glr-parser" return PERCENT_GLR_PARSER; "%language" return PERCENT_LANGUAGE; @@ -213,9 +228,9 @@ splice (\\[ \f\t\v]*\n)* "%lex-param" RETURN_PERCENT_PARAM(lex); "%locations" RETURN_PERCENT_FLAG("locations"); "%merge" return PERCENT_MERGE; - "%name"[-_]"prefix" return PERCENT_NAME_PREFIX; - "%no"[-_]"default"[-_]"prec" return PERCENT_NO_DEFAULT_PREC; - "%no"[-_]"lines" return PERCENT_NO_LINES; + "%name-prefix" return PERCENT_NAME_PREFIX; + "%no-default-prec" return PERCENT_NO_DEFAULT_PREC; + "%no-lines" return PERCENT_NO_LINES; "%nonassoc" return PERCENT_NONASSOC; "%nondeterministic-parser" return PERCENT_NONDETERMINISTIC_PARSER; "%nterm" return PERCENT_NTERM; @@ -225,21 +240,34 @@ splice (\\[ \f\t\v]*\n)* "%prec" return PERCENT_PREC; "%precedence" return PERCENT_PRECEDENCE; "%printer" return PERCENT_PRINTER; - "%pure"[-_]"parser" RETURN_PERCENT_FLAG("api.pure"); + "%pure-parser" RETURN_PERCENT_FLAG("api.pure"); "%require" return PERCENT_REQUIRE; "%right" return PERCENT_RIGHT; "%skeleton" return PERCENT_SKELETON; "%start" return PERCENT_START; "%term" return PERCENT_TOKEN; "%token" return PERCENT_TOKEN; - "%token"[-_]"table" return PERCENT_TOKEN_TABLE; + "%token-table" return PERCENT_TOKEN_TABLE; "%type" return PERCENT_TYPE; "%union" return PERCENT_UNION; "%verbose" return PERCENT_VERBOSE; "%yacc" return PERCENT_YACC; - {directive} { - complain_at (*loc, _("invalid directive: %s"), quote (yytext)); + /* deprecated */ + "%default"[-_]"prec" DEPRECATED("%default-prec"); + "%error"[-_]"verbose" DEPRECATED("%define parse.error verbose"); + "%expect"[-_]"rr" DEPRECATED("%expect-rr"); + "%file-prefix"{eqopt} DEPRECATED("%file-prefix"); + "%fixed"[-_]"output"[-_]"files" DEPRECATED("%fixed-output-files"); + "%name"[-_]"prefix"{eqopt} DEPRECATED("%name-prefix"); + "%no"[-_]"default"[-_]"prec" DEPRECATED("%no-default-prec"); + "%no"[-_]"lines" DEPRECATED("%no-lines"); + "%output"{eqopt} DEPRECATED("%output"); + "%pure"[-_]"parser" DEPRECATED("%pure-parser"); + "%token"[-_]"table" DEPRECATED("%token-table"); + + "%"{id} { + complain (loc, complaint, _("invalid directive: %s"), quote (yytext)); } "=" return EQUAL; @@ -265,7 +293,7 @@ splice (\\[ \f\t\v]*\n)* /* Identifiers may not start with a digit. Yet, don't silently accept "1FOO" as "1 FOO". */ {int}{id} { - complain_at (*loc, _("invalid identifier: %s"), quote (yytext)); + complain (loc, complaint, _("invalid identifier: %s"), quote (yytext)); } /* Characters. */ @@ -295,13 +323,6 @@ splice (\\[ \f\t\v]*\n)* /* A type. */ "<*>" return TAG_ANY; "<>" return TAG_NONE; - "<"{tag}">" { - obstack_grow (&obstack_for_string, yytext + 1, yyleng - 2); - STRING_FINISH; - val->uniqstr = uniqstr_new (last_string); - STRING_FREE; - return TAG; - } "<" { nesting = 0; token_start = loc->start; @@ -322,8 +343,10 @@ splice (\\[ \f\t\v]*\n)* BEGIN SC_BRACKETED_ID; } - . { - complain_at (*loc, _("invalid character: %s"), quote (yytext)); + [^\[%A-Za-z0-9_<>{}\"\'*;|=/, \f\n\t\v]+|. { + complain (loc, complaint, "%s: %s", + ngettext ("invalid character", "invalid characters", yyleng), + quote_mem (yytext, yyleng)); } <> { @@ -340,7 +363,7 @@ splice (\\[ \f\t\v]*\n)* { - \0 complain_at (*loc, _("invalid null character")); + \0 complain (loc, complaint, _("invalid null character")); } @@ -392,8 +415,9 @@ splice (\\[ \f\t\v]*\n)* {id} { if (bracketed_id_str) { - complain_at (*loc, _("unexpected identifier in bracketed name: %s"), - quote (yytext)); + complain (loc, complaint, + _("unexpected identifier in bracketed name: %s"), + quote (yytext)); } else { @@ -414,12 +438,16 @@ splice (\\[ \f\t\v]*\n)* } } else - complain_at (*loc, _("an identifier expected")); + complain (loc, complaint, _("an identifier expected")); } - . { - complain_at (*loc, _("invalid character in bracketed name: %s"), - quote (yytext)); + + [^\].A-Za-z0-9_/ \f\n\t\v]+|. { + complain (loc, complaint, "%s: %s", + ngettext ("invalid character in bracketed name", + "invalid characters in bracketed name", yyleng), + quote_mem (yytext, yyleng)); } + <> { BEGIN bracketed_id_context_state; unexpected_eof (bracketed_id_start, "]"); @@ -440,19 +468,19 @@ splice (\\[ \f\t\v]*\n)* /*---------------------------------------------------------------. - | Scanning a Yacc comment. The initial `/ *' is already eaten. | + | Scanning a Yacc comment. The initial '/ *' is already eaten. | `---------------------------------------------------------------*/ { "*/" BEGIN context_state; - .|\n ; + .|\n continue; <> unexpected_eof (token_start, "*/"); BEGIN context_state; } /*------------------------------------------------------------. - | Scanning a C comment. The initial `/ *' is already eaten. | + | Scanning a C comment. The initial '/ *' is already eaten. | `------------------------------------------------------------*/ @@ -463,7 +491,7 @@ splice (\\[ \f\t\v]*\n)* /*--------------------------------------------------------------. - | Scanning a line comment. The initial `//' is already eaten. | + | Scanning a line comment. The initial '//' is already eaten. | `--------------------------------------------------------------*/ @@ -481,23 +509,15 @@ splice (\\[ \f\t\v]*\n)* { - "\""|"\n" { - if (yytext[0] == '\n') - unexpected_newline (token_start, "\""); - STRING_FINISH; - loc->start = token_start; - val->chars = last_string; - BEGIN INITIAL; - return STRING; - } - <> { - unexpected_eof (token_start, "\""); + "\"" { STRING_FINISH; loc->start = token_start; - val->chars = last_string; + val->code = last_string; BEGIN INITIAL; return STRING; } + <> unexpected_eof (token_start, "\""); + "\n" unexpected_newline (token_start, "\""); } /*----------------------------------------------------------. @@ -507,53 +527,34 @@ splice (\\[ \f\t\v]*\n)* { - "'"|"\n" { + "'" { STRING_FINISH; loc->start = token_start; val->character = last_string[0]; + + /* FIXME: Eventually, make these errors. */ + if (last_string[0] == '\0') { - /* FIXME: Eventually, make these errors. */ - if (last_string[0] == '\0') - { - warn_at (*loc, _("empty character literal")); - /* '\0' seems dangerous even if we are about to complain. */ - val->character = '\''; - } - else if (last_string[1] != '\0') - warn_at (*loc, _("extra characters in character literal")); + complain (loc, Wother, _("empty character literal")); + /* '\0' seems dangerous even if we are about to complain. */ + val->character = '\''; } - if (yytext[0] == '\n') - unexpected_newline (token_start, "'"); - STRING_FREE; - BEGIN INITIAL; - return CHAR; - } - <> { - STRING_FINISH; - loc->start = token_start; - val->character = last_string[0]; - { - /* FIXME: Eventually, make these errors. */ - if (last_string[0] == '\0') - { - warn_at (*loc, _("empty character literal")); - /* '\0' seems dangerous even if we are about to complain. */ - val->character = '\''; - } - else if (last_string[1] != '\0') - warn_at (*loc, _("extra characters in character literal")); - } - unexpected_eof (token_start, "'"); + else if (last_string[1] != '\0') + complain (loc, Wother, + _("extra characters in character literal")); STRING_FREE; BEGIN INITIAL; return CHAR; } + "\n" unexpected_newline (token_start, "'"); + <> unexpected_eof (token_start, "'"); } - /*-----------------------------------------------------------. - | Scanning a Bison nested tag. The initial angle bracket is | - | already eaten. | - `-----------------------------------------------------------*/ + + + /*--------------------------------------------------------------. + | Scanning a tag. The initial angle bracket is already eaten. | + `--------------------------------------------------------------*/ { @@ -571,18 +572,10 @@ splice (\\[ \f\t\v]*\n)* STRING_GROW; } - [^<>]+ STRING_GROW; + ([^<>]|->)+ STRING_GROW; "<"+ STRING_GROW; nesting += yyleng; - <> { - unexpected_eof (token_start, ">"); - STRING_FINISH; - loc->start = token_start; - val->uniqstr = uniqstr_new (last_string); - STRING_FREE; - BEGIN INITIAL; - return TAG; - } + <> unexpected_eof (token_start, ">"); } /*----------------------------. @@ -594,7 +587,7 @@ splice (\\[ \f\t\v]*\n)* \\[0-7]{1,3} { unsigned long int c = strtoul (yytext + 1, NULL, 8); if (!c || UCHAR_MAX < c) - complain_at (*loc, _("invalid number after \\-escape: %s"), + complain (loc, complaint, _("invalid number after \\-escape: %s"), yytext+1); else obstack_1grow (&obstack_for_string, c); @@ -604,7 +597,7 @@ splice (\\[ \f\t\v]*\n)* verify (UCHAR_MAX < ULONG_MAX); unsigned long int c = strtoul (yytext + 2, NULL, 16); if (!c || UCHAR_MAX < c) - complain_at (*loc, _("invalid number after \\-escape: %s"), + complain (loc, complaint, _("invalid number after \\-escape: %s"), yytext+1); else obstack_1grow (&obstack_for_string, c); @@ -624,7 +617,7 @@ splice (\\[ \f\t\v]*\n)* \\(u|U[0-9abcdefABCDEF]{4})[0-9abcdefABCDEF]{4} { int c = convert_ucn_to_byte (yytext); if (c <= 0) - complain_at (*loc, _("invalid number after \\-escape: %s"), + complain (loc, complaint, _("invalid number after \\-escape: %s"), yytext+1); else obstack_1grow (&obstack_for_string, c); @@ -632,11 +625,12 @@ splice (\\[ \f\t\v]*\n)* \\(.|\n) { char const *p = yytext + 1; /* Quote only if escaping won't make the character visible. */ - if (isspace ((unsigned char) *p) && isprint ((unsigned char) *p)) + if (c_isspace ((unsigned char) *p) && c_isprint ((unsigned char) *p)) p = quote (p); else p = quotearg_style_mem (escape_quoting_style, p, 1); - complain_at (*loc, _("invalid character after \\-escape: %s"), p); + complain (loc, complaint, _("invalid character after \\-escape: %s"), + p); } } @@ -652,15 +646,15 @@ splice (\\[ \f\t\v]*\n)* { "'" STRING_GROW; BEGIN context_state; - \n unexpected_newline (token_start, "'"); BEGIN context_state; - <> unexpected_eof (token_start, "'"); BEGIN context_state; + \n unexpected_newline (token_start, "'"); + <> unexpected_eof (token_start, "'"); } { "\"" STRING_GROW; BEGIN context_state; - \n unexpected_newline (token_start, "\""); BEGIN context_state; - <> unexpected_eof (token_start, "\""); BEGIN context_state; + \n unexpected_newline (token_start, "\""); + <> unexpected_eof (token_start, "\""); } @@ -707,19 +701,11 @@ splice (\\[ \f\t\v]*\n)* "{"|"<"{splice}"%" STRING_GROW; nesting++; "%"{splice}">" STRING_GROW; nesting--; - /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly - (as `<' `<%'). */ + /* Tokenize '<<%' correctly (as '<<' '%') rather than incorrrectly + (as '<' '<%'). */ "<"{splice}"<" STRING_GROW; - <> { - int token = (YY_START == SC_BRACED_CODE) ? BRACED_CODE : BRACED_PREDICATE; - unexpected_eof (code_start, "}"); - STRING_FINISH; - loc->start = code_start; - val->code = last_string; - BEGIN INITIAL; - return token; - } + <> unexpected_eof (code_start, "}"); } @@ -765,19 +751,12 @@ splice (\\[ \f\t\v]*\n)* "%}" { STRING_FINISH; loc->start = code_start; - val->chars = last_string; + val->code = last_string; BEGIN INITIAL; return PROLOGUE; } - <> { - unexpected_eof (code_start, "%}"); - STRING_FINISH; - loc->start = code_start; - val->chars = last_string; - BEGIN INITIAL; - return PROLOGUE; - } + <> unexpected_eof (code_start, "%}"); } @@ -791,7 +770,7 @@ splice (\\[ \f\t\v]*\n)* <> { STRING_FINISH; loc->start = code_start; - val->chars = last_string; + val->code = last_string; BEGIN INITIAL; return EPILOGUE; } @@ -867,7 +846,8 @@ scan_integer (char const *number, int base, location loc) if (INT_MAX < num) { - complain_at (loc, _("integer out of range: %s"), quote (number)); + complain (&loc, complaint, _("integer out of range: %s"), + quote (number)); num = INT_MAX; } @@ -932,31 +912,35 @@ convert_ucn_to_byte (char const *ucn) } -/*----------------------------------------------------------------. -| Handle `#line INT "FILE"'. ARGS has already skipped `#line '. | -`----------------------------------------------------------------*/ +/*---------------------------------------------------------------------. +| Handle '#line INT( "FILE")?\n'. ARGS has already skipped '#line '. | +`---------------------------------------------------------------------*/ static void handle_syncline (char *args, location loc) { - char *after_num; - unsigned long int lineno = strtoul (args, &after_num, 10); - char *file = mbschr (after_num, '"') + 1; - *mbschr (file, '"') = '\0'; + char *file; + unsigned long int lineno = strtoul (args, &file, 10); if (INT_MAX <= lineno) { - warn_at (loc, _("line number overflow")); + complain (&loc, Wother, _("line number overflow")); lineno = INT_MAX; } - current_file = uniqstr_new (file); + + file = strchr (file, '"'); + if (file) + { + *strchr (file + 1, '"') = '\0'; + current_file = uniqstr_new (file + 1); + } boundary_set (&scanner_cursor, current_file, lineno, 1); } /*----------------------------------------------------------------. | For a token or comment starting at START, report message MSGID, | -| which should say that an end marker was found before | -| the expected TOKEN_END. | +| which should say that an end marker was found before the | +| expected TOKEN_END. Then, pretend that TOKEN_END was found. | `----------------------------------------------------------------*/ static void @@ -965,19 +949,33 @@ unexpected_end (boundary start, char const *msgid, char const *token_end) location loc; loc.start = start; loc.end = scanner_cursor; - complain_at (loc, _(msgid), token_end); + size_t i = strlen (token_end); + +/* Adjust scanner cursor so that any later message does not count + the characters about to be inserted. */ + scanner_cursor.column -= i; + + while (i != 0) + unput (token_end[--i]); + + token_end = quote (token_end); + /* Instead of '\'', display "'". */ + if (STREQ (token_end, "'\\''")) + token_end = "\"'\""; + complain (&loc, complaint, _(msgid), token_end); } /*------------------------------------------------------------------------. | Report an unexpected EOF in a token or comment starting at START. | | An end of file was encountered and the expected TOKEN_END was missing. | +| After reporting the problem, pretend that TOKEN_END was found. | `------------------------------------------------------------------------*/ static void unexpected_eof (boundary start, char const *token_end) { - unexpected_end (start, N_("missing `%s' at end of file"), token_end); + unexpected_end (start, N_("missing %s at end of file"), token_end); } @@ -988,7 +986,7 @@ unexpected_eof (boundary start, char const *token_end) static void unexpected_newline (boundary start, char const *token_end) { - unexpected_end (start, N_("missing `%s' at end of line"), token_end); + unexpected_end (start, N_("missing %s at end of line"), token_end); }