/* Bison Grammar Scanner -*- C -*-
- Copyright (C) 2002, 2003, 2004, 2005, 2006 Free Software Foundation, Inc.
+ Copyright (C) 2002-2011 Free Software Foundation, Inc.
This file is part of Bison, the GNU Compiler Compiler.
- This program is free software; you can redistribute it and/or modify
+ This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
- the Free Software Foundation; either version 2 of the License, or
+ the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
- along with this program; if not, write to the Free Software
- Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
- 02110-1301 USA
-*/
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
-%option debug nodefault nounput noyywrap never-interactive
+%option debug nodefault noinput nounput noyywrap never-interactive
%option prefix="gram_" outfile="lex.yy.c"
%{
#undef gram_wrap
#define gram_wrap() 1
-#include "system.h"
+#define FLEX_PREFIX(Id) gram_ ## Id
+#include <src/flex-scanner.h>
+#include <src/complain.h>
+#include <src/files.h>
+#include <src/gram.h>
+#include <quotearg.h>
+#include <src/reader.h>
+#include <src/uniqstr.h>
+
+#include <ctype.h>
#include <mbswidth.h>
#include <quote.h>
-#include "complain.h"
-#include "files.h"
-#include "getargs.h"
-#include "gram.h"
-#include "quotearg.h"
-#include "reader.h"
-#include "uniqstr.h"
+#include <src/scan-gram.h>
+
+#define YY_DECL GRAM_LEX_DECL
#define YY_USER_INIT \
- do \
- { \
- scanner_cursor.file = current_file; \
- scanner_cursor.line = 1; \
- scanner_cursor.column = 1; \
- code_start = scanner_cursor; \
- } \
- while (0)
-
-/* Pacify "gcc -Wmissing-prototypes" when flex 2.5.31 is used. */
-int gram_get_lineno (void);
-FILE *gram_get_in (void);
-FILE *gram_get_out (void);
-int gram_get_leng (void);
-char *gram_get_text (void);
-void gram_set_lineno (int);
-void gram_set_in (FILE *);
-void gram_set_out (FILE *);
-int gram_get_debug (void);
-void gram_set_debug (int);
-int gram_lex_destroy (void);
+ code_start = scanner_cursor = loc->start; \
/* Location of scanner cursor. */
-boundary scanner_cursor;
+static boundary scanner_cursor;
-static void adjust_location (location *, char const *, size_t);
-#define YY_USER_ACTION adjust_location (loc, yytext, yyleng);
+#define YY_USER_ACTION location_compute (loc, &scanner_cursor, yytext, yyleng);
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)
-/* OBSTACK_FOR_STRING -- Used to store all the characters that we need to
- keep (to construct ID, STRINGS etc.). Use the following macros to
- use it.
+#define RETURN_PERCENT_FLAG(Value) \
+ RETURN_VALUE(PERCENT_FLAG, uniqstr, uniqstr_new (Value))
- Use STRING_GROW to append what has just been matched, and
- STRING_FINISH to end the string (it puts the ending 0).
- STRING_FINISH also stores this string in LAST_STRING, which can be
- used, and which is used by STRING_FREE to free the last string. */
+#define RETURN_VALUE(Token, Field, Value) \
+ do { \
+ val->Field = Value; \
+ return Token; \
+ } while (0)
-static struct obstack obstack_for_string;
+#define ROLLBACK_CURRENT_TOKEN \
+ do { \
+ scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0); \
+ yyless (0); \
+ } while (0)
/* A string representing the most recently saved token. */
-char *last_string;
-
-/* The location of the most recently saved token, if it was a
- BRACED_CODE token; otherwise, this has an unspecified value. */
-location last_braced_code_loc;
-
-#define STRING_GROW \
- obstack_grow (&obstack_for_string, yytext, yyleng)
-
-#define STRING_FINISH \
- do { \
- obstack_1grow (&obstack_for_string, '\0'); \
- last_string = obstack_finish (&obstack_for_string); \
- } while (0)
+static char *last_string;
-#define STRING_FREE \
- obstack_free (&obstack_for_string, 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
-scanner_last_string_free (void)
+gram_scanner_last_string_free (void)
{
STRING_FREE;
}
-/* Within well-formed rules, RULE_LENGTH is the number of values in
- the current rule so far, which says where to find `$0' with respect
- to the top of the stack. It is not the same as the rule->length in
- the case of mid rule actions.
-
- Outside of well-formed rules, RULE_LENGTH has an undefined value. */
-static int rule_length;
-
-static void rule_length_overflow (location) __attribute__ ((__noreturn__));
-
-/* Increment the rule length by one, checking for overflow. */
-static inline void
-increment_rule_length (location loc)
-{
- rule_length++;
-
- /* Don't allow rule_length == INT_MAX, since that might cause
- confusion with strtol if INT_MAX == LONG_MAX. */
- if (rule_length == INT_MAX)
- rule_length_overflow (loc);
-}
-
-static void handle_dollar (int token_type, char *cp, location loc);
-static void handle_at (int token_type, char *cp, location loc);
static void handle_syncline (char *, location);
static unsigned long int scan_integer (char const *p, int base, location loc);
static int convert_ucn_to_byte (char const *hex_text);
static void unexpected_newline (boundary, char const *);
%}
-%x SC_COMMENT SC_LINE_COMMENT SC_YACC_COMMENT
-%x SC_STRING SC_CHARACTER
-%x SC_AFTER_IDENTIFIER
+ /* A C-like comment in directives/rules. */
+%x SC_YACC_COMMENT
+ /* Strings and characters in directives/rules. */
%x SC_ESCAPED_STRING SC_ESCAPED_CHARACTER
-%x SC_PRE_CODE SC_BRACED_CODE SC_PROLOGUE SC_EPILOGUE
+ /* 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
+
+ /* Four types of user code:
+ - 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); */
+%x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE SC_PREDICATE
+ /* C and C++ comments in code. */
+%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_]
-id {letter}({letter}|[0-9])*
-directive %{letter}({letter}|[0-9]|-)*
+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 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. */
%%
%{
- /* 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);
- /* Token type to return, when applicable. */
- int token_type IF_LINT (= 0);
-
/* Location of most recent identifier, when applicable. */
location id_loc IF_LINT (= empty_location);
| Scanning white space. |
`-----------------------*/
-<INITIAL,SC_AFTER_IDENTIFIER,SC_PRE_CODE>
+<INITIAL,SC_AFTER_IDENTIFIER,SC_BRACKETED_ID,SC_RETURN_BRACKETED_ID>
{
/* Comments and white space. */
"," warn_at (*loc, _("stray `,' treated as white space"));
/*----------------------------.
| 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. */
+
<INITIAL>
{
- "%binary" return PERCENT_NONASSOC;
- "%debug" return PERCENT_DEBUG;
- "%default"[-_]"prec" return PERCENT_DEFAULT_PREC;
- "%define" return PERCENT_DEFINE;
- "%defines" return PERCENT_DEFINES;
- "%destructor" token_type = PERCENT_DESTRUCTOR; BEGIN SC_PRE_CODE;
- "%dprec" return PERCENT_DPREC;
- "%error"[-_]"verbose" return PERCENT_ERROR_VERBOSE;
- "%expect" return PERCENT_EXPECT;
- "%expect"[-_]"rr" return PERCENT_EXPECT_RR;
- "%file-prefix" return PERCENT_FILE_PREFIX;
+ "%binary" return PERCENT_NONASSOC;
+ "%code" return PERCENT_CODE;
+ "%debug" RETURN_PERCENT_FLAG("parse.trace");
+ "%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;
+ "%expect" return PERCENT_EXPECT;
+ "%expect"[-_]"rr" return PERCENT_EXPECT_RR;
+ "%file-prefix" return PERCENT_FILE_PREFIX;
"%fixed"[-_]"output"[-_]"files" return PERCENT_YACC;
- "%initial-action" token_type = PERCENT_INITIAL_ACTION; BEGIN SC_PRE_CODE;
- "%glr-parser" return PERCENT_GLR_PARSER;
- "%left" return PERCENT_LEFT;
- "%lex-param" token_type = PERCENT_LEX_PARAM; BEGIN SC_PRE_CODE;
- "%locations" return PERCENT_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;
- "%nonassoc" return PERCENT_NONASSOC;
- "%nondeterministic-parser" return PERCENT_NONDETERMINISTIC_PARSER;
- "%nterm" return PERCENT_NTERM;
- "%output" return PERCENT_OUTPUT;
- "%parse-param" token_type = PERCENT_PARSE_PARAM; BEGIN SC_PRE_CODE;
- "%prec" rule_length--; return PERCENT_PREC;
- "%printer" token_type = PERCENT_PRINTER; BEGIN SC_PRE_CODE;
- "%pure"[-_]"parser" return PERCENT_PURE_PARSER;
- "%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;
- "%type" return PERCENT_TYPE;
- "%union" token_type = PERCENT_UNION; BEGIN SC_PRE_CODE;
- "%verbose" return PERCENT_VERBOSE;
- "%yacc" return PERCENT_YACC;
+ "%initial-action" return PERCENT_INITIAL_ACTION;
+ "%glr-parser" return PERCENT_GLR_PARSER;
+ "%language" return PERCENT_LANGUAGE;
+ "%left" return PERCENT_LEFT;
+ "%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;
+ "%nonassoc" return PERCENT_NONASSOC;
+ "%nondeterministic-parser" return PERCENT_NONDETERMINISTIC_PARSER;
+ "%nterm" return PERCENT_NTERM;
+ "%output" return PERCENT_OUTPUT;
+ "%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_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;
+ "%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));
}
"=" return EQUAL;
- "|" rule_length = 0; return PIPE;
+ "|" return PIPE;
";" return SEMICOLON;
{id} {
- val->symbol = symbol_get (yytext, *loc);
+ val->uniqstr = uniqstr_new (yytext);
id_loc = *loc;
- increment_rule_length (*loc);
+ bracketed_id_str = NULL;
BEGIN SC_AFTER_IDENTIFIER;
}
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;
/* Code in between braces. */
"{" {
- if (current_rule->action)
- grammar_midrule_action ();
STRING_GROW;
- token_type = BRACED_CODE;
- braces_level = 0;
+ nesting = 0;
code_start = loc->start;
BEGIN SC_BRACED_CODE;
}
+ /* Semantic predicate. */
+ "%?"[ \f\n\t\v]*"{" {
+ nesting = 0;
+ code_start = loc->start;
+ BEGIN SC_PREDICATE;
+ }
+
/* 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;
}
"%%" {
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));
}
}
+ /*--------------------------------------------------------------.
+ | Supporting \0 complexifies our implementation for no expected |
+ | added value. |
+ `--------------------------------------------------------------*/
+
+<SC_ESCAPED_CHARACTER,SC_ESCAPED_STRING,SC_TAG>
+{
+ \0 complain_at (*loc, _("invalid null character"));
+}
+
+
/*-----------------------------------------------------------------.
| Scanning after an identifier, checking whether a colon is next. |
`-----------------------------------------------------------------*/
<SC_AFTER_IDENTIFIER>
{
+ "[" {
+ 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;
+ }
+ }
":" {
- rule_length = 0;
+ 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;
}
<<EOF>> {
+ BEGIN (bracketed_id_str ? SC_RETURN_BRACKETED_ID : INITIAL);
*loc = id_loc;
- BEGIN INITIAL;
return ID;
}
}
+ /*--------------------------------.
+ | Scanning bracketed identifiers. |
+ `--------------------------------*/
+
+<SC_BRACKETED_ID>
+{
+ {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));
+ }
+ <<EOF>> {
+ BEGIN bracketed_id_context_state;
+ unexpected_eof (bracketed_id_start, "]");
+ }
+}
+
+<SC_RETURN_BRACKETED_ID>
+{
+ . {
+ 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. |
<SC_ESCAPED_STRING>
{
- "\"" {
+ "\""|"\n" {
+ if (yytext[0] == '\n')
+ unexpected_newline (token_start, "\"");
+ STRING_FINISH;
+ loc->start = token_start;
+ val->chars = last_string;
+ BEGIN INITIAL;
+ return STRING;
+ }
+ <<EOF>> {
+ unexpected_eof (token_start, "\"");
STRING_FINISH;
loc->start = token_start;
val->chars = last_string;
- increment_rule_length (*loc);
BEGIN INITIAL;
return STRING;
}
- \n unexpected_newline (token_start, "\""); BEGIN INITIAL;
- <<EOF>> unexpected_eof (token_start, "\""); BEGIN INITIAL;
}
/*----------------------------------------------------------.
<SC_ESCAPED_CHARACTER>
{
- "'" {
- unsigned char last_string_1;
- STRING_GROW;
+ "'"|"\n" {
STRING_FINISH;
loc->start = token_start;
- val->symbol = symbol_get (quotearg_style (escape_quoting_style,
- last_string),
- *loc);
- symbol_class_set (val->symbol, token_sym, *loc, false);
- last_string_1 = last_string[1];
- symbol_user_token_number_set (val->symbol, last_string_1, *loc);
+ 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;
- increment_rule_length (*loc);
BEGIN INITIAL;
- return ID;
+ return CHAR;
+ }
+ <<EOF>> {
+ 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, "'");
+ STRING_FREE;
+ BEGIN INITIAL;
+ return CHAR;
}
- \n unexpected_newline (token_start, "'"); BEGIN INITIAL;
- <<EOF>> unexpected_eof (token_start, "'"); BEGIN INITIAL;
}
-<SC_ESCAPED_CHARACTER,SC_ESCAPED_STRING>
+ /*-----------------------------------------------------------.
+ | Scanning a Bison nested tag. The initial angle bracket is |
+ | already eaten. |
+ `-----------------------------------------------------------*/
+
+<SC_TAG>
{
- \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;
+ <<EOF>> {
+ 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. |
{
\\[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);
}
\\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);
}
\\(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);
}
}
<SC_CHARACTER,SC_STRING>
{
- {splice}|\\{splice}[^\n$@\[\]] STRING_GROW;
+ {splice}|\\{splice}[^\n\[\]] STRING_GROW;
}
<SC_CHARACTER>
| Strings, comments etc. can be found in user code. |
`---------------------------------------------------*/
-<SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
+<SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE,SC_PREDICATE>
{
"'" {
STRING_GROW;
}
- /*---------------------------------------------------------------.
- | Scanning after %union etc., possibly followed by white space. |
- | For %union only, allow arbitrary C code to appear before the |
- | following brace, as an extension to POSIX. |
- `---------------------------------------------------------------*/
-<SC_PRE_CODE>
+ /*-----------------------------------------------------------.
+ | Scanning some code in braces (actions, predicates). The |
+ | initial "{" is already eaten. |
+ `-----------------------------------------------------------*/
+
+<SC_BRACED_CODE,SC_PREDICATE>
{
- . {
- bool valid = yytext[0] == '{' || token_type == PERCENT_UNION;
- scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0);
- yyless (0);
+ "{"|"<"{splice}"%" STRING_GROW; nesting++;
+ "%"{splice}">" STRING_GROW; nesting--;
- if (valid)
- {
- braces_level = -1;
- code_start = loc->start;
- BEGIN SC_BRACED_CODE;
- }
- else
- {
- complain_at (*loc, _("missing `{' in %s"),
- token_name (token_type));
- obstack_sgrow (&obstack_for_string, "{}");
- STRING_FINISH;
- val->chars = last_string;
- BEGIN INITIAL;
- return token_type;
- }
- }
+ /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
+ (as `<' `<%'). */
+ "<"{splice}"<" STRING_GROW;
- <<EOF>> unexpected_eof (scanner_cursor, "{}"); BEGIN INITIAL;
+ <<EOF>> {
+ 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;
+ }
}
-
- /*---------------------------------------------------------------.
- | Scanning some code in braces (%union and actions). The initial |
- | "{" is already eaten. |
- `---------------------------------------------------------------*/
-
<SC_BRACED_CODE>
{
- "{"|"<"{splice}"%" STRING_GROW; braces_level++;
- "%"{splice}">" STRING_GROW; braces_level--;
"}" {
- bool outer_brace = --braces_level < 0;
-
- /* As an undocumented Bison extension, append `;' before the last
- brace in braced code, so that the user code can omit trailing
- `;'. But do not append `;' if emulating Yacc, since Yacc does
- not append one.
-
- FIXME: Bison should warn if a semicolon seems to be necessary
- here, and should omit the semicolon if it seems unnecessary
- (e.g., after ';', '{', or '}', each followed by comments or
- white space). Such a warning shouldn't depend on --yacc; it
- should depend on a new --pedantic option, which would cause
- Bison to warn if it detects an extension to POSIX. --pedantic
- should also diagnose other Bison extensions like %yacc.
- Perhaps there should also be a GCC-style --pedantic-errors
- option, so that such warnings are diagnosed as errors. */
- if (outer_brace && token_type == BRACED_CODE && ! yacc_flag)
- obstack_1grow (&obstack_for_string, ';');
-
obstack_1grow (&obstack_for_string, '}');
- if (outer_brace)
+ --nesting;
+ if (nesting < 0)
{
STRING_FINISH;
loc->start = code_start;
- val->chars = last_string;
- increment_rule_length (*loc);
- last_braced_code_loc = *loc;
+ val->code = last_string;
BEGIN INITIAL;
- return token_type;
+ return BRACED_CODE;
}
}
+}
- /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
- (as `<' `<%'). */
- "<"{splice}"<" STRING_GROW;
-
- "$"("<"{tag}">")?(-?[0-9]+|"$") handle_dollar (token_type, yytext, *loc);
- "@"(-?[0-9]+|"$") handle_at (token_type, yytext, *loc);
-
- "$" {
- warn_at (*loc, _("stray `$'"));
- obstack_sgrow (&obstack_for_string, "$][");
- }
- "@" {
- warn_at (*loc, _("stray `@'"));
- obstack_sgrow (&obstack_for_string, "@@");
+<SC_PREDICATE>
+{
+ "}" {
+ --nesting;
+ if (nesting < 0)
+ {
+ STRING_FINISH;
+ loc->start = code_start;
+ val->code = last_string;
+ BEGIN INITIAL;
+ return BRACED_PREDICATE;
+ }
+ else
+ obstack_1grow (&obstack_for_string, '}');
}
-
- <<EOF>> unexpected_eof (code_start, "}"); BEGIN INITIAL;
}
-
/*--------------------------------------------------------------.
| Scanning some prologue: from "%{" (already scanned) to "%}". |
`--------------------------------------------------------------*/
return PROLOGUE;
}
- <<EOF>> unexpected_eof (code_start, "%}"); BEGIN INITIAL;
+ <<EOF>> {
+ unexpected_eof (code_start, "%}");
+ STRING_FINISH;
+ loc->start = code_start;
+ val->chars = last_string;
+ BEGIN INITIAL;
+ return PROLOGUE;
+ }
}
}
- /*-----------------------------------------.
- | Escape M4 quoting characters in C code. |
- `-----------------------------------------*/
-
-<SC_COMMENT,SC_LINE_COMMENT,SC_STRING,SC_CHARACTER,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
-{
- \$ obstack_sgrow (&obstack_for_string, "$][");
- \@ obstack_sgrow (&obstack_for_string, "@@");
- \[ obstack_sgrow (&obstack_for_string, "@{");
- \] obstack_sgrow (&obstack_for_string, "@}");
-}
-
-
/*-----------------------------------------------------.
| By default, grow the string obstack with the input. |
`-----------------------------------------------------*/
-<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE,SC_STRING,SC_CHARACTER,SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>. |
-<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>\n STRING_GROW;
+<SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PREDICATE,SC_PROLOGUE,SC_EPILOGUE,SC_STRING,SC_CHARACTER,SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>. |
+ <SC_COMMENT,SC_LINE_COMMENT,SC_BRACED_CODE,SC_PREDICATE,SC_PROLOGUE,SC_EPILOGUE>\n STRING_GROW;
%%
-/* Keeps track of the maximum number of semantic values to the left of
- a handle (those referenced by $0, $-1, etc.) are required by the
- semantic actions of this grammar. */
-int max_left_semantic_context = 0;
-
-/* If BUF is null, add BUFSIZE (which in this case must be less than
- INT_MAX) to COLUMN; otherwise, add mbsnwidth (BUF, BUFSIZE, 0) to
- COLUMN. If an overflow occurs, or might occur but is undetectable,
- return INT_MAX. Assume COLUMN is nonnegative. */
-
-static inline int
-add_column_width (int column, char const *buf, size_t bufsize)
-{
- size_t width;
- unsigned int remaining_columns = INT_MAX - column;
-
- if (buf)
- {
- if (INT_MAX / 2 <= bufsize)
- return INT_MAX;
- width = mbsnwidth (buf, bufsize, 0);
- }
- else
- width = bufsize;
-
- return width <= remaining_columns ? column + width : INT_MAX;
-}
-
-/* Set *LOC and adjust scanner cursor to account for token TOKEN of
- size SIZE. */
-
-static void
-adjust_location (location *loc, char const *token, size_t size)
-{
- int line = scanner_cursor.line;
- int column = scanner_cursor.column;
- char const *p0 = token;
- char const *p = token;
- char const *lim = token + size;
-
- loc->start = scanner_cursor;
-
- for (p = token; p < lim; p++)
- switch (*p)
- {
- case '\n':
- line += line < INT_MAX;
- column = 1;
- p0 = p + 1;
- break;
-
- case '\t':
- column = add_column_width (column, p0, p - p0);
- column = add_column_width (column, NULL, 8 - ((column - 1) & 7));
- p0 = p + 1;
- break;
-
- default:
- break;
- }
-
- scanner_cursor.line = line;
- scanner_cursor.column = column = add_column_width (column, p0, p - p0);
-
- loc->end = scanner_cursor;
-
- if (line == INT_MAX && loc->start.line != INT_MAX)
- warn_at (*loc, _("line number overflow"));
- if (column == INT_MAX && loc->start.column != INT_MAX)
- warn_at (*loc, _("column number overflow"));
-}
-
-
/* Read bytes from FP into buffer BUF of size SIZE. Return the
number of bytes read. Remove '\r' from input, treating \r\n
and isolated \r as \n. */
}
-/*------------------------------------------------------------------.
-| TEXT is pointing to a wannabee semantic value (i.e., a `$'). |
-| |
-| Possible inputs: $[<TYPENAME>]($|integer) |
-| |
-| Output to OBSTACK_FOR_STRING a reference to this semantic value. |
-`------------------------------------------------------------------*/
-
-static inline bool
-handle_action_dollar (char *text, location loc)
-{
- const char *type_name = NULL;
- char *cp = text + 1;
-
- if (! current_rule)
- return false;
-
- /* Get the type name if explicit. */
- if (*cp == '<')
- {
- type_name = ++cp;
- while (*cp != '>')
- ++cp;
- *cp = '\0';
- ++cp;
- }
-
- if (*cp == '$')
- {
- if (!type_name)
- type_name = symbol_list_n_type_name_get (current_rule, loc, 0);
- if (!type_name && typed)
- complain_at (loc, _("$$ of `%s' has no declared type"),
- current_rule->sym->tag);
- if (!type_name)
- type_name = "";
- obstack_fgrow1 (&obstack_for_string,
- "]b4_lhs_value([%s])[", type_name);
- current_rule->used = true;
- }
- else
- {
- long int num = strtol (cp, NULL, 10);
-
- if (1 - INT_MAX + rule_length <= num && num <= rule_length)
- {
- int n = num;
- if (max_left_semantic_context < 1 - n)
- max_left_semantic_context = 1 - n;
- if (!type_name && 0 < n)
- type_name = symbol_list_n_type_name_get (current_rule, loc, n);
- if (!type_name && typed)
- complain_at (loc, _("$%d of `%s' has no declared type"),
- n, current_rule->sym->tag);
- if (!type_name)
- type_name = "";
- obstack_fgrow3 (&obstack_for_string,
- "]b4_rhs_value(%d, %d, [%s])[",
- rule_length, n, type_name);
- symbol_list_n_used_set (current_rule, n, true);
- }
- else
- complain_at (loc, _("integer out of range: %s"), quote (text));
- }
-
- return true;
-}
-
-
-/*----------------------------------------------------------------.
-| Map `$?' onto the proper M4 symbol, depending on its TOKEN_TYPE |
-| (are we in an action?). |
-`----------------------------------------------------------------*/
-
-static void
-handle_dollar (int token_type, char *text, location loc)
-{
- switch (token_type)
- {
- case BRACED_CODE:
- if (handle_action_dollar (text, loc))
- return;
- break;
-
- case PERCENT_DESTRUCTOR:
- case PERCENT_INITIAL_ACTION:
- case PERCENT_PRINTER:
- if (text[1] == '$')
- {
- obstack_sgrow (&obstack_for_string, "]b4_dollar_dollar[");
- return;
- }
- break;
-
- default:
- break;
- }
-
- complain_at (loc, _("invalid value: %s"), quote (text));
-}
-
-
-/*------------------------------------------------------.
-| TEXT is a location token (i.e., a `@...'). Output to |
-| OBSTACK_FOR_STRING a reference to this location. |
-`------------------------------------------------------*/
-
-static inline bool
-handle_action_at (char *text, location loc)
-{
- char *cp = text + 1;
- locations_flag = true;
-
- if (! current_rule)
- return false;
-
- if (*cp == '$')
- obstack_sgrow (&obstack_for_string, "]b4_lhs_location[");
- else
- {
- long int num = strtol (cp, NULL, 10);
-
- if (1 - INT_MAX + rule_length <= num && num <= rule_length)
- {
- int n = num;
- obstack_fgrow2 (&obstack_for_string, "]b4_rhs_location(%d, %d)[",
- rule_length, n);
- }
- else
- complain_at (loc, _("integer out of range: %s"), quote (text));
- }
-
- return true;
-}
-
-
-/*----------------------------------------------------------------.
-| Map `@?' onto the proper M4 symbol, depending on its TOKEN_TYPE |
-| (are we in an action?). |
-`----------------------------------------------------------------*/
-
-static void
-handle_at (int token_type, char *text, location loc)
-{
- switch (token_type)
- {
- case BRACED_CODE:
- handle_action_at (text, loc);
- return;
-
- case PERCENT_INITIAL_ACTION:
- case PERCENT_DESTRUCTOR:
- case PERCENT_PRINTER:
- if (text[1] == '$')
- {
- obstack_sgrow (&obstack_for_string, "]b4_at_dollar[");
- return;
- }
- break;
-
- default:
- break;
- }
-
- complain_at (loc, _("invalid value: %s"), quote (text));
-}
-
/*------------------------------------------------------.
| Scan NUMBER for a base-BASE integer at location LOC. |
{
char *after_num;
unsigned long int lineno = strtoul (args, &after_num, 10);
- char *file = strchr (after_num, '"') + 1;
- *strchr (file, '"') = '\0';
+ char *file = mbschr (after_num, '"') + 1;
+ *mbschr (file, '"') = '\0';
if (INT_MAX <= lineno)
{
warn_at (loc, _("line number overflow"));
lineno = INT_MAX;
}
- scanner_cursor.file = current_file = uniqstr_new (file);
- scanner_cursor.line = lineno;
- scanner_cursor.column = 1;
-}
-
-
-/*---------------------------------.
-| Report a rule that is too long. |
-`---------------------------------*/
-
-static void
-rule_length_overflow (location loc)
-{
- fatal_at (loc, _("rule is too long"));
+ current_file = uniqstr_new (file);
+ boundary_set (&scanner_cursor, current_file, lineno, 1);
}
`-------------------------*/
void
-scanner_initialize (void)
+gram_scanner_initialize (void)
{
obstack_init (&obstack_for_string);
}
`-----------------------------------------------*/
void
-scanner_free (void)
+gram_scanner_free (void)
{
obstack_free (&obstack_for_string, 0);
/* Reclaim Flex's buffers. */
- yy_delete_buffer (YY_CURRENT_BUFFER);
+ yylex_destroy ();
}
projects / bison.git / blobdiff