/* Bison Grammar Scanner -*- C -*-
- Copyright (C) 2002 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.
- 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., 59 Temple Place, Suite 330, Boston, MA
- 02111-1307 USA
-*/
+ along with this program. If not, see <http://www.gnu.org/licenses/>. */
-%option debug nodefault noyywrap nounput never-interactive stack
+%option debug nodefault nounput noyywrap never-interactive
%option prefix="gram_" outfile="lex.yy.c"
%{
-#include "system.h"
-#include "complain.h"
-#include "quote.h"
-#include "getargs.h"
-#include "gram.h"
-#include "reader.h"
-
-/* Each time we match a string, move the end cursor to its end. */
-#define YY_USER_INIT \
-do { \
- LOCATION_RESET (*yylloc); \
- yylloc->file = infile; \
- /* This is only to avoid GCC warnings. */ \
- if (yycontrol) {;}; \
-} while (0)
-
-#define YY_USER_ACTION LOCATION_COLUMNS (*yylloc, yyleng);
-#define YY_LINES LOCATION_LINES (*yylloc, yyleng);
-#define YY_STEP LOCATION_STEP (*yylloc);
-
-/* STRING_OBSTACK -- Used to store all the characters that we need to
- keep (to construct ID, STRINGS etc.). Use the following macros to
- use it.
-
- Use YY_OBS_GROW to append what has just been matched, and
- YY_OBS_FINISH to end the string (it puts the ending 0).
- YY_OBS_FINISH also stores this string in LAST_STRING, which can be
- used, and which is used by YY_OBS_FREE to free the last string. */
-
-static struct obstack string_obstack;
-char *last_string;
-
-#define YY_OBS_GROW \
- obstack_grow (&string_obstack, yytext, yyleng)
-
-#define YY_OBS_FINISH \
- do { \
- obstack_1grow (&string_obstack, '\0'); \
- last_string = obstack_finish (&string_obstack); \
- } while (0)
+/* Work around a bug in flex 2.5.31. See Debian bug 333231
+ <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=333231>. */
+#undef gram_wrap
+#define gram_wrap() 1
-#define YY_OBS_FREE \
- do { \
- obstack_free (&string_obstack, last_string); \
- } while (0)
+#define FLEX_PREFIX(Id) gram_ ## Id
+#include <src/flex-scanner.h>
-void
-scanner_last_string_free (void)
-{
- YY_OBS_FREE;
-}
+#include <src/complain.h>
+#include <src/files.h>
+#include <src/gram.h>
+#include <quotearg.h>
+#include <src/reader.h>
+#include <src/uniqstr.h>
+
+#include <mbswidth.h>
+#include <quote.h>
+#include <src/scan-gram.h>
-static int braces_level = 0;
-static int percent_percent_count = 0;
+#define YY_DECL GRAM_LEX_DECL
-/* 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.
+#define YY_USER_INIT \
+ code_start = scanner_cursor = loc->start; \
- Outside of well-formed rules, RULE_LENGTH has an undefined value. */
-static int rule_length;
+/* Location of scanner cursor. */
+static boundary scanner_cursor;
-static void handle_dollar PARAMS ((braced_code_t code_kind,
- char *cp, location_t location));
-static void handle_at PARAMS ((braced_code_t code_kind,
- char *cp, location_t location));
+#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_FLAG(Value) \
+ do { \
+ val->uniqstr = uniqstr_new (Value); \
+ return PERCENT_FLAG; \
+ } while (0)
+
+
+/* A string representing the most recently saved token. */
+static char *last_string;
+
+void
+gram_scanner_last_string_free (void)
+{
+ STRING_FREE;
+}
+
+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_eof (boundary, char const *);
+static void unexpected_newline (boundary, char const *);
%}
-%x SC_COMMENT
-%x SC_STRING SC_CHARACTER
+ /* 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_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
+
+ /* Three 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 %%). */
+%x SC_PROLOGUE SC_BRACED_CODE SC_EPILOGUE
+ /* C and C++ comments in code. */
+%x SC_COMMENT SC_LINE_COMMENT
+ /* Strings and characters in code. */
+%x SC_STRING SC_CHARACTER
-id [.a-zA-Z_][.a-zA-Z_0-9]*
-int [0-9]+
-eols (\n|\r|\n\r|\r\n)+
-blanks [ \t\f]+
+letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_]
+id {letter}({letter}|[0-9])*
+directive %{letter}({letter}|[0-9]|-)*
+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)*
%%
%{
- /* At each yylex invocation, mark the current position as the
- start of the next token. */
-#define TR_POS 0
-#if TR_POS
- fprintf (stderr, "FOO1: %p: ", yylloc);
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, "\n");
-#endif
- YY_STEP;
-#if TR_POS
- fprintf (stderr, "BAR1: ");
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, "\n");
-#endif
+ /* 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);
+
+ /* Location of most recent identifier, when applicable. */
+ location id_loc IF_LINT (= empty_location);
+
+ /* Where containing code started, when applicable. Its initial
+ value is relevant only when yylex is invoked in the SC_EPILOGUE
+ start condition. */
+ boundary code_start = scanner_cursor;
+
+ /* Where containing comment or string or character literal started,
+ when applicable. */
+ boundary token_start IF_LINT (= scanner_cursor);
%}
+ /*-----------------------.
+ | Scanning white space. |
+ `-----------------------*/
+
+<INITIAL,SC_AFTER_IDENTIFIER>
+{
+ /* Comments and white space. */
+ "," warn_at (*loc, _("stray `,' treated as white space"));
+ [ \f\n\t\v] |
+ "//".* ;
+ "/*" {
+ token_start = loc->start;
+ context_state = YY_START;
+ BEGIN SC_YACC_COMMENT;
+ }
+
+ /* #line directives are not documented, and may be withdrawn or
+ modified in future versions of Bison. */
+ ^"#line "{int}" \"".*"\"\n" {
+ handle_syncline (yytext + sizeof "#line " - 1, *loc);
+ }
+}
+
+
/*----------------------------.
| 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;
- "%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;
- "%file-prefix" return PERCENT_FILE_PREFIX;
+ "%binary" return PERCENT_NONASSOC;
+ "%code" return PERCENT_CODE;
+ "%debug" RETURN_PERCENT_FLAG("debug");
+ "%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_FLAG("error_verbose");
+ "%expect" return PERCENT_EXPECT;
+ "%expect"[-_]"rr" return PERCENT_EXPECT_RR;
+ "%file-prefix" return PERCENT_FILE_PREFIX;
"%fixed"[-_]"output"[-_]"files" return PERCENT_YACC;
- "%glr"[-_]"parser" return PERCENT_GLR_PARSER;
- "%left" return PERCENT_LEFT;
- "%locations" return PERCENT_LOCATIONS;
- "%merge" return PERCENT_MERGE;
- "%name"[-_]"prefix" return PERCENT_NAME_PREFIX;
- "%no"[-_]"lines" return PERCENT_NO_LINES;
- "%nonassoc" return PERCENT_NONASSOC;
- "%nterm" return PERCENT_NTERM;
- "%output" return PERCENT_OUTPUT;
- "%prec" { rule_length--; return PERCENT_PREC; }
- "%printer" return PERCENT_PRINTER;
- "%pure"[-_]"parser" return PERCENT_PURE_PARSER;
- "%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;
+ "%initial-action" return PERCENT_INITIAL_ACTION;
+ "%glr-parser" return PERCENT_GLR_PARSER;
+ "%language" return PERCENT_LANGUAGE;
+ "%left" return PERCENT_LEFT;
+ "%lex-param" return PERCENT_LEX_PARAM;
+ "%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" return PERCENT_PARSE_PARAM;
+ "%prec" return PERCENT_PREC;
+ "%precedence" return PERCENT_PRECEDENCE;
+ "%printer" return PERCENT_PRINTER;
+ "%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" 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 COLON; }
- "|" { rule_length = 0; return PIPE; }
+ "|" return PIPE;
";" return SEMICOLON;
- {eols} YY_LINES; YY_STEP;
- {blanks} YY_STEP;
- {id} {
- yylval->symbol = symbol_get (yytext, *yylloc);
- rule_length++;
- return ID;
+ {id} {
+ val->uniqstr = uniqstr_new (yytext);
+ id_loc = *loc;
+ BEGIN SC_AFTER_IDENTIFIER;
}
- {int} yylval->integer = strtol (yytext, 0, 10); return INT;
+ {int} {
+ val->integer = scan_integer (yytext, 10, *loc);
+ return INT;
+ }
+ 0[xX][0-9abcdefABCDEF]+ {
+ val->integer = scan_integer (yytext, 16, *loc);
+ return INT;
+ }
/* Characters. We don't check there is only one. */
- "'" YY_OBS_GROW; yy_push_state (SC_ESCAPED_CHARACTER);
+ "'" STRING_GROW; token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER;
/* Strings. */
- "\"" YY_OBS_GROW; yy_push_state (SC_ESCAPED_STRING);
-
- /* Comments. */
- "/*" yy_push_state (SC_COMMENT);
- "//".* YY_STEP;
+ "\"" token_start = loc->start; BEGIN SC_ESCAPED_STRING;
/* Prologue. */
- "%{" yy_push_state (SC_PROLOGUE);
+ "%{" code_start = loc->start; BEGIN SC_PROLOGUE;
/* Code in between braces. */
- "{" YY_OBS_GROW; ++braces_level; yy_push_state (SC_BRACED_CODE);
+ "{" {
+ STRING_GROW;
+ nesting = 0;
+ code_start = loc->start;
+ BEGIN SC_BRACED_CODE;
+ }
/* A type. */
- "<"[^>]+">" {
- obstack_grow (&string_obstack, yytext + 1, yyleng - 2);
- YY_OBS_FINISH;
- yylval->string = last_string;
- return 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;
+ BEGIN SC_TAG;
}
-
- "%%" {
+ "%%" {
+ static int percent_percent_count;
if (++percent_percent_count == 2)
- yy_push_state (SC_EPILOGUE);
+ BEGIN SC_EPILOGUE;
return PERCENT_PERCENT;
}
- . {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": invalid character: `%c'\n"), *yytext);
- YY_STEP;
+ . {
+ complain_at (*loc, _("invalid character: %s"), quote (yytext));
+ }
+
+ <<EOF>> {
+ loc->start = loc->end = scanner_cursor;
+ yyterminate ();
}
}
- /*------------------------------------------------------------.
- | Whatever the start condition (but those which correspond to |
- | entity `swallowed' by Bison: SC_ESCAPED_STRING and |
- | SC_ESCAPED_CHARACTER), no M4 character must escape as is. |
- `------------------------------------------------------------*/
+ /*--------------------------------------------------------------.
+ | Supporting \0 complexifies our implementation for no expected |
+ | added value. |
+ `--------------------------------------------------------------*/
-<SC_COMMENT,SC_STRING,SC_CHARACTER,SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
+<SC_ESCAPED_CHARACTER,SC_ESCAPED_STRING,SC_TAG>
{
- \[ if (YY_START != SC_COMMENT) obstack_sgrow (&string_obstack, "@<:@");
- \] if (YY_START != SC_COMMENT) obstack_sgrow (&string_obstack, "@:>@");
+ \0 complain_at (*loc, _("invalid null character"));
}
+ /*-----------------------------------------------------------------.
+ | Scanning after an identifier, checking whether a colon is next. |
+ `-----------------------------------------------------------------*/
+
+<SC_AFTER_IDENTIFIER>
+{
+ ":" {
+ *loc = id_loc;
+ BEGIN INITIAL;
+ return ID_COLON;
+ }
+ . {
+ scanner_cursor.column -= mbsnwidth (yytext, yyleng, 0);
+ yyless (0);
+ *loc = id_loc;
+ BEGIN INITIAL;
+ return ID;
+ }
+ <<EOF>> {
+ *loc = id_loc;
+ BEGIN INITIAL;
+ return ID;
+ }
+}
- /*-----------------------------------------------------------.
- | Scanning a C comment. The initial `/ *' is already eaten. |
- `-----------------------------------------------------------*/
+
+ /*---------------------------------------------------------------.
+ | Scanning a Yacc comment. The initial `/ *' is already eaten. |
+ `---------------------------------------------------------------*/
+
+<SC_YACC_COMMENT>
+{
+ "*/" BEGIN context_state;
+ .|\n ;
+ <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state;
+}
+
+
+ /*------------------------------------------------------------.
+ | Scanning a C comment. The initial `/ *' is already eaten. |
+ `------------------------------------------------------------*/
<SC_COMMENT>
{
- "*/" { /* End of the comment. */
- if (yy_top_state () == INITIAL)
- {
- YY_STEP;
- }
- else
- {
- YY_OBS_GROW;
- }
- yy_pop_state ();
- }
+ "*"{splice}"/" STRING_GROW; BEGIN context_state;
+ <<EOF>> unexpected_eof (token_start, "*/"); BEGIN context_state;
+}
- [^\[\]*\n\r]+ if (yy_top_state () != INITIAL) YY_OBS_GROW;
- {eols} if (yy_top_state () != INITIAL) YY_OBS_GROW; YY_LINES;
- . /* Stray `*'. */if (yy_top_state () != INITIAL) YY_OBS_GROW;
- <<EOF>> {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": unexpected end of file in a comment\n"));
- yy_pop_state ();
- }
+ /*--------------------------------------------------------------.
+ | Scanning a line comment. The initial `//' is already eaten. |
+ `--------------------------------------------------------------*/
+
+<SC_LINE_COMMENT>
+{
+ "\n" STRING_GROW; BEGIN context_state;
+ {splice} STRING_GROW;
+ <<EOF>> BEGIN context_state;
}
- /*----------------------------------------------------------------.
- | Scanning a C string, including its escapes. The initial `"' is |
- | already eaten. |
- `----------------------------------------------------------------*/
+ /*------------------------------------------------.
+ | Scanning a Bison string, including its escapes. |
+ | The initial quote is already eaten. |
+ `------------------------------------------------*/
<SC_ESCAPED_STRING>
{
- "\"" {
- assert (yy_top_state () == INITIAL);
- YY_OBS_GROW;
- YY_OBS_FINISH;
- yylval->string = last_string;
- yy_pop_state ();
- rule_length++;
+ "\""|"\n" {
+ if (yytext[0] == '\n')
+ unexpected_newline (token_start, "\"");
+ STRING_FINISH;
+ loc->start = token_start;
+ val->chars = last_string;
+ BEGIN INITIAL;
return STRING;
}
-
- [^\"\n\r\\]+ YY_OBS_GROW;
-
- {eols} obstack_1grow (&string_obstack, '\n'); YY_LINES;
-
<<EOF>> {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": unexpected end of file in a string\n"));
- assert (yy_top_state () == INITIAL);
- YY_OBS_FINISH;
- yylval->string = last_string;
- yy_pop_state ();
+ unexpected_eof (token_start, "\"");
+ STRING_FINISH;
+ loc->start = token_start;
+ val->chars = last_string;
+ BEGIN INITIAL;
return STRING;
}
}
- /*---------------------------------------------------------------.
- | Scanning a C character, decoding its escapes. The initial "'" |
- | is already eaten. |
- `---------------------------------------------------------------*/
+ /*----------------------------------------------------------.
+ | Scanning a Bison character literal, decoding its escapes. |
+ | The initial quote is already eaten. |
+ `----------------------------------------------------------*/
<SC_ESCAPED_CHARACTER>
{
- "'" {
- YY_OBS_GROW;
- assert (yy_top_state () == INITIAL);
- {
- YY_OBS_FINISH;
- yylval->symbol = symbol_get (last_string, *yylloc);
- symbol_class_set (yylval->symbol, token_sym, *yylloc);
- symbol_user_token_number_set (yylval->symbol,
- (unsigned char) last_string[1], *yylloc);
- YY_OBS_FREE;
- yy_pop_state ();
- rule_length++;
- return ID;
- }
+ "'"|"\n" {
+ if (yytext[0] == '\n')
+ unexpected_newline (token_start, "'");
+ STRING_GROW;
+ STRING_FINISH;
+ loc->start = token_start;
+ val->character = last_string[1];
+ STRING_FREE;
+ BEGIN INITIAL;
+ return CHAR;
+ }
+ <<EOF>> {
+ 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];
+ STRING_FREE;
+ BEGIN INITIAL;
+ return CHAR;
}
+}
- [^\n\r\\] YY_OBS_GROW;
+ /*-----------------------------------------------------------.
+ | Scanning a Bison nested tag. The initial angle bracket is |
+ | already eaten. |
+ `-----------------------------------------------------------*/
- {eols} obstack_1grow (&string_obstack, '\n'); YY_LINES;
+<SC_TAG>
+{
+ ">" {
+ --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>> {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": unexpected end of file in a character\n"));
- assert (yy_top_state () == INITIAL);
- YY_OBS_FINISH;
- yylval->string = last_string;
- yy_pop_state ();
- return CHARACTER;
+ 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. |
`----------------------------*/
<SC_ESCAPED_STRING,SC_ESCAPED_CHARACTER>
{
- \\[0-7]{3} {
- long c = strtol (yytext + 1, 0, 8);
- if (c > 255)
- {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": invalid escape: %s\n"), quote (yytext));
- YY_STEP;
- }
+ \\[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));
else
- obstack_1grow (&string_obstack, c);
+ obstack_1grow (&obstack_for_string, c);
}
- \\x[0-9a-fA-F]{2} {
- obstack_1grow (&string_obstack, strtol (yytext + 2, 0, 16));
+ \\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));
+ else
+ obstack_1grow (&obstack_for_string, c);
}
- \\a obstack_1grow (&string_obstack, '\a');
- \\b obstack_1grow (&string_obstack, '\b');
- \\f obstack_1grow (&string_obstack, '\f');
- \\n obstack_1grow (&string_obstack, '\n');
- \\r obstack_1grow (&string_obstack, '\r');
- \\t obstack_1grow (&string_obstack, '\t');
- \\v obstack_1grow (&string_obstack, '\v');
- \\[\\""''] obstack_1grow (&string_obstack, yytext[1]);
+ \\a obstack_1grow (&obstack_for_string, '\a');
+ \\b obstack_1grow (&obstack_for_string, '\b');
+ \\f obstack_1grow (&obstack_for_string, '\f');
+ \\n obstack_1grow (&obstack_for_string, '\n');
+ \\r obstack_1grow (&obstack_for_string, '\r');
+ \\t obstack_1grow (&obstack_for_string, '\t');
+ \\v obstack_1grow (&obstack_for_string, '\v');
+
+ /* \\[\"\'?\\] would be shorter, but it confuses xgettext. */
+ \\("\""|"'"|"?"|"\\") obstack_1grow (&obstack_for_string, yytext[1]);
+
+ \\(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));
+ else
+ obstack_1grow (&obstack_for_string, c);
+ }
\\(.|\n) {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": unrecognized escape: %s\n"), quote (yytext));
- YY_OBS_GROW;
+ complain_at (*loc, _("unrecognized escape sequence: %s"), quote (yytext));
+ STRING_GROW;
}
- /* FLex wants this rule, in case of a `\<<EOF>>'. */
- \\ YY_OBS_GROW;
}
+ /*--------------------------------------------.
+ | Scanning user-code characters and strings. |
+ `--------------------------------------------*/
- /*----------------------------------------------------------.
- | Scanning a C character without decoding its escapes. The |
- | initial "'" is already eaten. |
- `----------------------------------------------------------*/
+<SC_CHARACTER,SC_STRING>
+{
+ {splice}|\\{splice}[^\n\[\]] STRING_GROW;
+}
<SC_CHARACTER>
{
- "'" {
- YY_OBS_GROW;
- assert (yy_top_state () != INITIAL);
- yy_pop_state ();
- }
-
- [^\[\]\'\n\r\\]+ YY_OBS_GROW;
- \\(.|\n) YY_OBS_GROW;
- /* FLex wants this rule, in case of a `\<<EOF>>'. */
- \\ YY_OBS_GROW;
-
- {eols} YY_OBS_GROW; YY_LINES;
-
- <<EOF>> {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": unexpected end of file in a character\n"));
- assert (yy_top_state () != INITIAL);
- yy_pop_state ();
- }
+ "'" STRING_GROW; BEGIN context_state;
+ \n unexpected_newline (token_start, "'"); BEGIN context_state;
+ <<EOF>> unexpected_eof (token_start, "'"); BEGIN context_state;
}
-
- /*----------------------------------------------------------------.
- | Scanning a C string, without decoding its escapes. The initial |
- | `"' is already eaten. |
- `----------------------------------------------------------------*/
-
<SC_STRING>
{
- "\"" {
- assert (yy_top_state () != INITIAL);
- YY_OBS_GROW;
- yy_pop_state ();
- }
-
- [^\[\]\"\n\r\\]+ YY_OBS_GROW;
- \\(.|\n) YY_OBS_GROW;
- /* FLex wants this rule, in case of a `\<<EOF>>'. */
- \\ YY_OBS_GROW;
-
- {eols} YY_OBS_GROW; YY_LINES;
-
- <<EOF>> {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": unexpected end of file in a string\n"));
- assert (yy_top_state () != INITIAL);
- yy_pop_state ();
- }
+ "\"" STRING_GROW; BEGIN context_state;
+ \n unexpected_newline (token_start, "\""); BEGIN context_state;
+ <<EOF>> unexpected_eof (token_start, "\""); BEGIN context_state;
}
<SC_BRACED_CODE,SC_PROLOGUE,SC_EPILOGUE>
{
- /* Characters. We don't check there is only one. */
- "'" YY_OBS_GROW; yy_push_state (SC_CHARACTER);
-
- /* Strings. */
- "\"" YY_OBS_GROW; yy_push_state (SC_STRING);
-
- /* Comments. */
- "/*" YY_OBS_GROW; yy_push_state (SC_COMMENT);
- "//".* YY_OBS_GROW;
-
- /* Not comments. */
- "/" YY_OBS_GROW;
+ "'" {
+ STRING_GROW;
+ context_state = YY_START;
+ token_start = loc->start;
+ BEGIN SC_CHARACTER;
+ }
+ "\"" {
+ STRING_GROW;
+ context_state = YY_START;
+ token_start = loc->start;
+ BEGIN SC_STRING;
+ }
+ "/"{splice}"*" {
+ STRING_GROW;
+ context_state = YY_START;
+ token_start = loc->start;
+ BEGIN SC_COMMENT;
+ }
+ "/"{splice}"/" {
+ STRING_GROW;
+ context_state = YY_START;
+ BEGIN SC_LINE_COMMENT;
+ }
}
- /*---------------------------------------------------------------.
- | Scanning some code in braces (%union and actions). The initial |
- | "{" is already eaten. |
- `---------------------------------------------------------------*/
+
+ /*-----------------------------------------------------------.
+ | Scanning some code in braces (actions). The initial "{" is |
+ | already eaten. |
+ `-----------------------------------------------------------*/
<SC_BRACED_CODE>
{
+ "{"|"<"{splice}"%" STRING_GROW; nesting++;
+ "%"{splice}">" STRING_GROW; nesting--;
"}" {
- YY_OBS_GROW;
- if (--braces_level == 0)
+ obstack_1grow (&obstack_for_string, '}');
+
+ --nesting;
+ if (nesting < 0)
{
- yy_pop_state ();
- YY_OBS_FINISH;
- yylval->string = last_string;
- rule_length++;
+ STRING_FINISH;
+ loc->start = code_start;
+ val->code = last_string;
+ BEGIN INITIAL;
return BRACED_CODE;
}
}
- "{" YY_OBS_GROW; braces_level++;
-
- "$"("<"[^>]+">")?(-?[0-9]+|"$") { handle_dollar (current_braced_code,
- yytext, *yylloc); }
- "@"(-?[0-9]+|"$") { handle_at (current_braced_code,
- yytext, *yylloc); }
-
- [^$@\[\]/\'\"\{\}\n\r]+ YY_OBS_GROW;
- {eols} YY_OBS_GROW; YY_LINES;
-
- /* A lose $, or /, or etc. */
- . YY_OBS_GROW;
+ /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
+ (as `<' `<%'). */
+ "<"{splice}"<" STRING_GROW;
<<EOF>> {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": unexpected end of file in a braced code\n"));
- yy_pop_state ();
- YY_OBS_FINISH;
- yylval->string = last_string;
+ unexpected_eof (code_start, "}");
+ STRING_FINISH;
+ loc->start = code_start;
+ val->code = last_string;
+ BEGIN INITIAL;
return BRACED_CODE;
}
-
}
<SC_PROLOGUE>
{
"%}" {
- yy_pop_state ();
- YY_OBS_FINISH;
- yylval->string = last_string;
+ STRING_FINISH;
+ loc->start = code_start;
+ val->chars = last_string;
+ BEGIN INITIAL;
return PROLOGUE;
}
- [^%\[\]/\'\"\n\r]+ YY_OBS_GROW;
- "%" YY_OBS_GROW;
- {eols} YY_OBS_GROW; YY_LINES;
-
<<EOF>> {
- LOCATION_PRINT (stderr, *yylloc);
- fprintf (stderr, _(": unexpected end of file in a prologue\n"));
- yy_pop_state ();
- YY_OBS_FINISH;
- yylval->string = last_string;
+ unexpected_eof (code_start, "%}");
+ STRING_FINISH;
+ loc->start = code_start;
+ val->chars = last_string;
+ BEGIN INITIAL;
return PROLOGUE;
}
}
/*---------------------------------------------------------------.
| Scanning the epilogue (everything after the second "%%", which |
- | has already been eaten. |
+ | has already been eaten). |
`---------------------------------------------------------------*/
<SC_EPILOGUE>
{
- ([^\[\]]|{eols})+ YY_OBS_GROW;
-
<<EOF>> {
- yy_pop_state ();
- YY_OBS_FINISH;
- yylval->string = last_string;
+ STRING_FINISH;
+ loc->start = code_start;
+ val->chars = last_string;
+ BEGIN INITIAL;
return EPILOGUE;
}
}
-%%
+ /*-----------------------------------------------------.
+ | By default, grow the string obstack with the input. |
+ `-----------------------------------------------------*/
-/*------------------------------------------------------------------.
-| TEXT is pointing to a wannabee semantic value (i.e., a `$'). |
-| |
-| Possible inputs: $[<TYPENAME>]($|integer) |
-| |
-| Output to the STRING_OBSTACK a reference to this semantic value. |
-`------------------------------------------------------------------*/
+<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;
-static inline void
-handle_action_dollar (char *text, location_t location)
-{
- const char *type_name = NULL;
- char *cp = text + 1;
+%%
- /* Get the type name if explicit. */
- if (*cp == '<')
- {
- type_name = ++cp;
- while (*cp != '>')
- ++cp;
- *cp = '\0';
- ++cp;
- }
+/* 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. */
- if (*cp == '$')
- {
- if (!type_name)
- type_name = symbol_list_n_type_name_get (current_rule, location, 0);
- if (!type_name && typed)
- complain_at (location, _("$$ of `%s' has no declared type"),
- current_rule->sym->tag);
- if (!type_name)
- type_name = "";
- obstack_fgrow1 (&string_obstack,
- "]b4_lhs_value([%s])[", type_name);
- }
- else if (('0' <= *cp && *cp <= '9') || *cp == '-')
+static size_t
+no_cr_read (FILE *fp, char *buf, size_t size)
+{
+ size_t bytes_read = fread (buf, 1, size, fp);
+ if (bytes_read)
{
- int n = strtol (cp, &cp, 10);
-
- if (n > rule_length)
- complain_at (location, _("invalid value: %s%d"), "$", n);
- else
+ char *w = memchr (buf, '\r', bytes_read);
+ if (w)
{
- if (!type_name && n > 0)
- type_name = symbol_list_n_type_name_get (current_rule, location,
- n);
- if (!type_name && typed)
- complain_at (location, _("$%d of `%s' has no declared type"),
- n, current_rule->sym->tag);
- if (!type_name)
- type_name = "";
- obstack_fgrow3 (&string_obstack,
- "]b4_rhs_value([%d], [%d], [%s])[",
- rule_length, n, type_name);
+ char const *r = ++w;
+ char const *lim = buf + bytes_read;
+
+ for (;;)
+ {
+ /* Found an '\r'. Treat it like '\n', but ignore any
+ '\n' that immediately follows. */
+ w[-1] = '\n';
+ if (r == lim)
+ {
+ int ch = getc (fp);
+ if (ch != '\n' && ungetc (ch, fp) != ch)
+ break;
+ }
+ else if (*r == '\n')
+ r++;
+
+ /* Copy until the next '\r'. */
+ do
+ {
+ if (r == lim)
+ return w - buf;
+ }
+ while ((*w++ = *r++) != '\r');
+ }
+
+ return w - buf;
}
}
- else
+
+ return bytes_read;
+}
+
+
+
+/*------------------------------------------------------.
+| Scan NUMBER for a base-BASE integer at location LOC. |
+`------------------------------------------------------*/
+
+static unsigned long int
+scan_integer (char const *number, int base, location loc)
+{
+ verify (INT_MAX < ULONG_MAX);
+ unsigned long int num = strtoul (number, NULL, base);
+
+ if (INT_MAX < num)
{
- complain_at (location, _("%s is invalid"), quote (text));
+ complain_at (loc, _("integer out of range: %s"), quote (number));
+ num = INT_MAX;
}
+
+ return num;
}
-/*---------------------------------------------------------------.
-| TEXT is expexted tp be $$ in some code associated to a symbol: |
-| destructor or printer. |
-`---------------------------------------------------------------*/
+/*------------------------------------------------------------------.
+| Convert universal character name UCN to a single-byte character, |
+| and return that character. Return -1 if UCN does not correspond |
+| to a single-byte character. |
+`------------------------------------------------------------------*/
-static inline void
-handle_symbol_code_dollar (char *text, location_t location)
+static int
+convert_ucn_to_byte (char const *ucn)
{
- char *cp = text + 1;
- if (*cp == '$')
- obstack_sgrow (&string_obstack, "]b4_dollar_dollar[");
- else
- complain_at (location, _("%s is invalid"), quote (text));
+ verify (UCHAR_MAX <= INT_MAX);
+ unsigned long int code = strtoul (ucn + 2, NULL, 16);
+
+ /* FIXME: Currently we assume Unicode-compatible unibyte characters
+ on ASCII hosts (i.e., Latin-1 on hosts with 8-bit bytes). On
+ non-ASCII hosts we support only the portable C character set.
+ These limitations should be removed once we add support for
+ multibyte characters. */
+
+ if (UCHAR_MAX < code)
+ return -1;
+
+#if ! ('$' == 0x24 && '@' == 0x40 && '`' == 0x60 && '~' == 0x7e)
+ {
+ /* A non-ASCII host. Use CODE to index into a table of the C
+ basic execution character set, which is guaranteed to exist on
+ all Standard C platforms. This table also includes '$', '@',
+ and '`', which are not in the basic execution character set but
+ which are unibyte characters on all the platforms that we know
+ about. */
+ static signed char const table[] =
+ {
+ '\0', -1, -1, -1, -1, -1, -1, '\a',
+ '\b', '\t', '\n', '\v', '\f', '\r', -1, -1,
+ -1, -1, -1, -1, -1, -1, -1, -1,
+ -1, -1, -1, -1, -1, -1, -1, -1,
+ ' ', '!', '"', '#', '$', '%', '&', '\'',
+ '(', ')', '*', '+', ',', '-', '.', '/',
+ '0', '1', '2', '3', '4', '5', '6', '7',
+ '8', '9', ':', ';', '<', '=', '>', '?',
+ '@', 'A', 'B', 'C', 'D', 'E', 'F', 'G',
+ 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
+ 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
+ 'X', 'Y', 'Z', '[', '\\', ']', '^', '_',
+ '`', 'a', 'b', 'c', 'd', 'e', 'f', 'g',
+ 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o',
+ 'p', 'q', 'r', 's', 't', 'u', 'v', 'w',
+ 'x', 'y', 'z', '{', '|', '}', '~'
+ };
+
+ code = code < sizeof table ? table[code] : -1;
+ }
+#endif
+
+ return code;
}
-/*-----------------------------------------------------------------.
-| Dispatch onto handle_action_dollar, or handle_destructor_dollar, |
-| depending upon CODE_KIND. |
-`-----------------------------------------------------------------*/
+/*----------------------------------------------------------------.
+| Handle `#line INT "FILE"'. ARGS has already skipped `#line '. |
+`----------------------------------------------------------------*/
static void
-handle_dollar (braced_code_t braced_code_kind,
- char *text, location_t location)
+handle_syncline (char *args, location loc)
{
- switch (braced_code_kind)
+ char *after_num;
+ unsigned long int lineno = strtoul (args, &after_num, 10);
+ char *file = strchr (after_num, '"') + 1;
+ *strchr (file, '"') = '\0';
+ if (INT_MAX <= lineno)
{
- case action_braced_code:
- handle_action_dollar (text, location);
- break;
-
- case destructor_braced_code:
- case printer_braced_code:
- handle_symbol_code_dollar (text, location);
- break;
+ warn_at (loc, _("line number overflow"));
+ lineno = INT_MAX;
}
+ current_file = uniqstr_new (file);
+ boundary_set (&scanner_cursor, current_file, lineno, 1);
}
-/*------------------------------------------------------.
-| TEXT is a location token (i.e., a `@...'). Output to |
-| STRING_OBSTACK a reference to this location. |
-`------------------------------------------------------*/
+/*----------------------------------------------------------------.
+| 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. |
+`----------------------------------------------------------------*/
-static inline void
-handle_action_at (char *text, location_t location)
+static void
+unexpected_end (boundary start, char const *msgid, char const *token_end)
{
- char *cp = text + 1;
- locations_flag = 1;
-
- if (*cp == '$')
- {
- obstack_sgrow (&string_obstack, "]b4_lhs_location[");
- }
- else if (('0' <= *cp && *cp <= '9') || *cp == '-')
- {
- int n = strtol (cp, &cp, 10);
-
- if (n > rule_length)
- complain_at (location, _("invalid value: %s%d"), "@", n);
- else
- obstack_fgrow2 (&string_obstack, "]b4_rhs_location([%d], [%d])[",
- rule_length, n);
- }
- else
- {
- complain_at (location, _("%s is invalid"), quote (text));
- }
+ location loc;
+ loc.start = start;
+ loc.end = scanner_cursor;
+ complain_at (loc, _(msgid), token_end);
}
-/*---------------------------------------------------------------.
-| TEXT is expexted tp be @$ in some code associated to a symbol: |
-| destructor or printer. |
-`---------------------------------------------------------------*/
+/*------------------------------------------------------------------------.
+| 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. |
+`------------------------------------------------------------------------*/
-static inline void
-handle_symbol_code_at (char *text, location_t location)
+static void
+unexpected_eof (boundary start, char const *token_end)
{
- char *cp = text + 1;
- if (*cp == '$')
- obstack_sgrow (&string_obstack, "]b4_at_dollar[");
- else
- complain_at (location, _("%s is invalid"), quote (text));
+ unexpected_end (start, N_("missing `%s' at end of file"), token_end);
}
-/*-------------------------------------------------------------------.
-| Dispatch onto handle_action_at, or handle_destructor_at, depending |
-| upon CODE_KIND. |
-`-------------------------------------------------------------------*/
+/*----------------------------------------.
+| Likewise, but for unexpected newlines. |
+`----------------------------------------*/
static void
-handle_at (braced_code_t braced_code_kind,
- char *text, location_t location)
+unexpected_newline (boundary start, char const *token_end)
{
- switch (braced_code_kind)
- {
- case action_braced_code:
- handle_action_at (text, location);
- break;
-
- case destructor_braced_code:
- case printer_braced_code:
- handle_symbol_code_at (text, location);
- break;
- }
+ unexpected_end (start, N_("missing `%s' at end of line"), token_end);
}
`-------------------------*/
void
-scanner_initialize (void)
+gram_scanner_initialize (void)
{
- obstack_init (&string_obstack);
+ obstack_init (&obstack_for_string);
}
`-----------------------------------------------*/
void
-scanner_free (void)
+gram_scanner_free (void)
{
- obstack_free (&string_obstack, 0);
+ obstack_free (&obstack_for_string, 0);
/* Reclaim Flex's buffers. */
- yy_delete_buffer (YY_CURRENT_BUFFER);
+ yylex_destroy ();
}
projects / bison.git / blobdiff