X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/d78f0ac9d8c8830542faf9d00d0b6ef652dda45e..a7706735d143e3d7ab255e162499584e5337ba72:/src/scan-gram.l diff --git a/src/scan-gram.l b/src/scan-gram.l index 7e4c1f10..aff07697 100644 --- a/src/scan-gram.l +++ b/src/scan-gram.l @@ -1,7 +1,7 @@ /* Bison Grammar Scanner -*- C -*- - Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008 Free Software Foundation, - Inc. + Copyright (C) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009 + Free Software Foundation, Inc. This file is part of Bison, the GNU Compiler Compiler. @@ -18,7 +18,7 @@ You should have received a copy of the GNU General Public License along with this program. If not, see . */ -%option debug nodefault nounput noyywrap never-interactive +%option debug nodefault noinput nounput noyywrap never-interactive %option prefix="gram_" outfile="lex.yy.c" %{ @@ -28,19 +28,20 @@ #define gram_wrap() 1 #define FLEX_PREFIX(Id) gram_ ## Id -#include "flex-scanner.h" +#include -#include "complain.h" -#include "files.h" -#include "gram.h" -#include "quotearg.h" -#include "reader.h" -#include "uniqstr.h" +#include +#include +#include +#include +#include +#include +#include #include #include -#include "scan-gram.h" +#include #define YY_DECL GRAM_LEX_DECL @@ -55,9 +56,33 @@ static boundary scanner_cursor; static size_t no_cr_read (FILE *, char *, size_t); #define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size)) +#define RETURN_PERCENT_PARAM(Value) \ + RETURN_VALUE(PERCENT_PARAM, param, param_ ## Value) + +#define RETURN_PERCENT_FLAG(Value) \ + RETURN_VALUE(PERCENT_FLAG, uniqstr, uniqstr_new (Value)) + +#define RETURN_VALUE(Token, Field, Value) \ + do { \ + val->Field = Value; \ + return Token; \ + } while (0) + +#define ROLLBACK_CURRENT_TOKEN \ + do { \ + scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0); \ + yyless (0); \ + } while (0) + /* A string representing the most recently saved token. */ static char *last_string; +/* Bracketed identifier. */ +static uniqstr bracketed_id_str = 0; +static location bracketed_id_loc; +static boundary bracketed_id_start; +static int bracketed_id_context_state = 0; + void gram_scanner_last_string_free (void) { @@ -78,6 +103,8 @@ static void unexpected_newline (boundary, char const *); /* A identifier was just read in directives/rules. Special state to capture the sequence `identifier :'. */ %x SC_AFTER_IDENTIFIER + /* A complex tag, with nested angles brackets. */ +%x SC_TAG /* Three types of user code: - prologue (code between `%{' `%}' in the first section, before %%); @@ -88,16 +115,20 @@ static void unexpected_newline (boundary, char const *); %x SC_COMMENT SC_LINE_COMMENT /* Strings and characters in code. */ %x SC_STRING SC_CHARACTER + /* Bracketed identifiers support. */ +%x SC_BRACKETED_ID SC_RETURN_BRACKETED_ID -letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] +letter [-.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] id {letter}({letter}|[0-9])* -directive %{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 and newline, as this simplifies our implementation. */ -tag [^\0\n>]+ + 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. */ @@ -105,8 +136,9 @@ splice (\\[ \f\t\v]*\n)* %% %{ - /* Nesting level of the current code in braces. */ - int braces_level IF_LINT (= 0); + /* Nesting level. Either for nested braces, or nested angle brackets + (but not mixed). */ + int nesting IF_LINT (= 0); /* Parent context state, when applicable. */ int context_state IF_LINT (= 0); @@ -129,7 +161,7 @@ splice (\\[ \f\t\v]*\n)* | Scanning white space. | `-----------------------*/ - + { /* Comments and white space. */ "," warn_at (*loc, _("stray `,' treated as white space")); @@ -152,11 +184,18 @@ splice (\\[ \f\t\v]*\n)* /*----------------------------. | Scanning Bison directives. | `----------------------------*/ + + /* For directives that are also command line options, the regex must be + "%..." + after "[-_]"s are removed, and the directive must match the --long + option name, with a single string argument. Otherwise, add exceptions + to ../build-aux/cross-options.pl. */ + { "%binary" return PERCENT_NONASSOC; "%code" return PERCENT_CODE; - "%debug" return PERCENT_DEBUG; + "%debug" RETURN_PERCENT_FLAG("parse.trace"); "%default"[-_]"prec" return PERCENT_DEFAULT_PREC; "%define" return PERCENT_DEFINE; "%defines" return PERCENT_DEFINES; @@ -171,8 +210,8 @@ splice (\\[ \f\t\v]*\n)* "%glr-parser" return PERCENT_GLR_PARSER; "%language" return PERCENT_LANGUAGE; "%left" return PERCENT_LEFT; - "%lex-param" return PERCENT_LEX_PARAM; - "%locations" return PERCENT_LOCATIONS; + "%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; @@ -181,11 +220,12 @@ splice (\\[ \f\t\v]*\n)* "%nondeterministic-parser" return PERCENT_NONDETERMINISTIC_PARSER; "%nterm" return PERCENT_NTERM; "%output" return PERCENT_OUTPUT; - "%parse-param" return PERCENT_PARSE_PARAM; + "%param" RETURN_PERCENT_PARAM(both); + "%parse-param" RETURN_PERCENT_PARAM(parse); "%prec" return PERCENT_PREC; "%precedence" return PERCENT_PRECEDENCE; "%printer" return PERCENT_PRINTER; - "%pure"[-_]"parser" return PERCENT_PURE_PARSER; + "%pure"[-_]"parser" RETURN_PERCENT_FLAG("api.pure"); "%require" return PERCENT_REQUIRE; "%right" return PERCENT_RIGHT; "%skeleton" return PERCENT_SKELETON; @@ -205,12 +245,11 @@ splice (\\[ \f\t\v]*\n)* "=" return EQUAL; "|" return PIPE; ";" return SEMICOLON; - "<*>" return TYPE_TAG_ANY; - "<>" return TYPE_TAG_NONE; {id} { val->uniqstr = uniqstr_new (yytext); id_loc = *loc; + bracketed_id_str = NULL; BEGIN SC_AFTER_IDENTIFIER; } @@ -223,8 +262,14 @@ splice (\\[ \f\t\v]*\n)* return INT; } - /* Characters. We don't check there is only one. */ - "'" STRING_GROW; token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER; + /* 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)); + } + + /* Characters. */ + "'" token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER; /* Strings. */ "\"" token_start = loc->start; BEGIN SC_ESCAPED_STRING; @@ -235,18 +280,25 @@ splice (\\[ \f\t\v]*\n)* /* Code in between braces. */ "{" { STRING_GROW; - braces_level = 0; + nesting = 0; code_start = loc->start; BEGIN SC_BRACED_CODE; } /* 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 TYPE; + return TAG; + } + "<" { + nesting = 0; + token_start = loc->start; + BEGIN SC_TAG; } "%%" { @@ -256,6 +308,13 @@ splice (\\[ \f\t\v]*\n)* return PERCENT_PERCENT; } + "[" { + bracketed_id_str = NULL; + bracketed_id_start = loc->start; + bracketed_id_context_state = YY_START; + BEGIN SC_BRACKETED_ID; + } + . { complain_at (*loc, _("invalid character: %s"), quote (yytext)); } @@ -267,31 +326,111 @@ splice (\\[ \f\t\v]*\n)* } + /*--------------------------------------------------------------. + | Supporting \0 complexifies our implementation for no expected | + | added value. | + `--------------------------------------------------------------*/ + + +{ + \0 complain_at (*loc, _("invalid null character")); +} + + /*-----------------------------------------------------------------. | Scanning after an identifier, checking whether a colon is next. | `-----------------------------------------------------------------*/ { + "[" { + if (bracketed_id_str) + { + ROLLBACK_CURRENT_TOKEN; + BEGIN SC_RETURN_BRACKETED_ID; + *loc = id_loc; + return ID; + } + else + { + bracketed_id_start = loc->start; + bracketed_id_context_state = YY_START; + BEGIN SC_BRACKETED_ID; + } + } ":" { + BEGIN (bracketed_id_str ? SC_RETURN_BRACKETED_ID : INITIAL); *loc = id_loc; - BEGIN INITIAL; return ID_COLON; } . { - scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0); - yyless (0); + ROLLBACK_CURRENT_TOKEN; + BEGIN (bracketed_id_str ? SC_RETURN_BRACKETED_ID : INITIAL); *loc = id_loc; - BEGIN INITIAL; return ID; } <> { + BEGIN (bracketed_id_str ? SC_RETURN_BRACKETED_ID : INITIAL); *loc = id_loc; - BEGIN INITIAL; return ID; } } + /*--------------------------------. + | Scanning bracketed identifiers. | + `--------------------------------*/ + + +{ + {id} { + if (bracketed_id_str) + { + complain_at (*loc, _("unexpected identifier in bracketed name: %s"), + quote (yytext)); + } + else + { + bracketed_id_str = uniqstr_new (yytext); + bracketed_id_loc = *loc; + } + } + "]" { + BEGIN bracketed_id_context_state; + if (bracketed_id_str) + { + if (INITIAL == bracketed_id_context_state) + { + val->uniqstr = bracketed_id_str; + bracketed_id_str = 0; + *loc = bracketed_id_loc; + return BRACKETED_ID; + } + } + else + complain_at (*loc, _("an identifier expected")); + } + . { + complain_at (*loc, _("invalid character in bracketed name: %s"), + quote (yytext)); + } + <> { + BEGIN bracketed_id_context_state; + unexpected_eof (bracketed_id_start, "]"); + } +} + + +{ + . { + ROLLBACK_CURRENT_TOKEN; + val->uniqstr = bracketed_id_str; + bracketed_id_str = 0; + *loc = bracketed_id_loc; + BEGIN INITIAL; + return BRACKETED_ID; + } +} + /*---------------------------------------------------------------. | Scanning a Yacc comment. The initial `/ *' is already eaten. | @@ -362,35 +501,82 @@ splice (\\[ \f\t\v]*\n)* { "'"|"\n" { - if (yytext[0] == '\n') - unexpected_newline (token_start, "'"); - STRING_GROW; STRING_FINISH; loc->start = token_start; - val->character = last_string[1]; + 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")); + } + if (yytext[0] == '\n') + unexpected_newline (token_start, "'"); STRING_FREE; BEGIN INITIAL; return CHAR; } <> { - unexpected_eof (token_start, "'"); STRING_FINISH; loc->start = token_start; - if (strlen(last_string) > 1) - val->character = last_string[1]; - else - val->character = last_string[0]; + 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, "'"); STRING_FREE; BEGIN INITIAL; return CHAR; } } - + /*-----------------------------------------------------------. + | Scanning a Bison nested tag. The initial angle bracket is | + | already eaten. | + `-----------------------------------------------------------*/ + + { - \0 complain_at (*loc, _("invalid null character")); -} + ">" { + --nesting; + if (nesting < 0) + { + STRING_FINISH; + loc->start = token_start; + val->uniqstr = uniqstr_new (last_string); + STRING_FREE; + BEGIN INITIAL; + return TAG; + } + 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; + } +} /*----------------------------. | Decode escaped characters. | @@ -400,10 +586,9 @@ splice (\\[ \f\t\v]*\n)* { \\[0-7]{1,3} { unsigned long int c = strtoul (yytext + 1, NULL, 8); - if (UCHAR_MAX < c) - complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext)); - else if (! c) - complain_at (*loc, _("invalid null character: %s"), quote (yytext)); + if (!c || UCHAR_MAX < c) + complain_at (*loc, _("invalid number after \\-escape: %s"), + yytext+1); else obstack_1grow (&obstack_for_string, c); } @@ -411,10 +596,9 @@ splice (\\[ \f\t\v]*\n)* \\x[0-9abcdefABCDEF]+ { verify (UCHAR_MAX < ULONG_MAX); unsigned long int c = strtoul (yytext + 2, NULL, 16); - if (UCHAR_MAX < c) - complain_at (*loc, _("invalid escape sequence: %s"), quote (yytext)); - else if (! c) - complain_at (*loc, _("invalid null character: %s"), quote (yytext)); + if (!c || UCHAR_MAX < c) + complain_at (*loc, _("invalid number after \\-escape: %s"), + yytext+1); else obstack_1grow (&obstack_for_string, c); } @@ -432,16 +616,20 @@ 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 escape sequence: %s"), quote (yytext)); - else if (! c) - complain_at (*loc, _("invalid null character: %s"), quote (yytext)); + if (c <= 0) + complain_at (*loc, _("invalid number after \\-escape: %s"), + yytext+1); else obstack_1grow (&obstack_for_string, c); } \\(.|\n) { - complain_at (*loc, _("unrecognized escape sequence: %s"), quote (yytext)); - STRING_GROW; + char const *p = yytext + 1; + /* Quote only if escaping won't make the character visible. */ + if (isspace ((unsigned char) *p) && 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); } } @@ -509,13 +697,13 @@ splice (\\[ \f\t\v]*\n)* { - "{"|"<"{splice}"%" STRING_GROW; braces_level++; - "%"{splice}">" STRING_GROW; braces_level--; + "{"|"<"{splice}"%" STRING_GROW; nesting++; + "%"{splice}">" STRING_GROW; nesting--; "}" { obstack_1grow (&obstack_for_string, '}'); - --braces_level; - if (braces_level < 0) + --nesting; + if (nesting < 0) { STRING_FINISH; loc->start = code_start;