X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/a706a1cc03800f9bedd3658c6e53f17fc1f5d350..ba60c39547a445dee3e07920931b4d7a81843868:/src/scan-gram.l?ds=inline
diff --git a/src/scan-gram.l b/src/scan-gram.l
index ad200267..15b6b31d 100644
--- a/src/scan-gram.l
+++ b/src/scan-gram.l
@@ -1,11 +1,12 @@
/* Bison Grammar Scanner -*- C -*-
- Copyright (C) 2002 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,
@@ -14,177 +15,120 @@
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 . */
-%option debug nodefault noyywrap never-interactive
+%option debug nodefault noinput nounput noyywrap never-interactive
%option prefix="gram_" outfile="lex.yy.c"
%{
-#include "system.h"
-#include "mbswidth.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 extend_location (yylloc, yytext, yyleng);
-#define YY_STEP LOCATION_STEP (*yylloc)
-
-#define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size))
+/* Work around a bug in flex 2.5.31. See Debian bug 333231
+ . */
+#undef gram_wrap
+#define gram_wrap() 1
+#define FLEX_PREFIX(Id) gram_ ## Id
+#include
-/* 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. */
-
-static size_t
-no_cr_read (FILE *fp, char *buf, size_t size)
-{
- size_t s = fread (buf, 1, size, fp);
- if (s)
- {
- char *w = memchr (buf, '\r', s);
- if (w)
- {
- char const *r = ++w;
- char const *lim = buf + s;
+#include
+#include
+#include
+#include
+#include
+#include
- 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++;
+#include
+#include
+#include
- /* Copy until the next '\r'. */
- do
- {
- if (r == lim)
- return w - buf;
- }
- while ((*w++ = *r++) != '\r');
- }
+#include
- return w - buf;
- }
- }
+#define YY_DECL GRAM_LEX_DECL
- return s;
-}
+#define YY_USER_INIT \
+ code_start = scanner_cursor = loc->start; \
+/* Location of scanner cursor. */
+static boundary scanner_cursor;
-/* Extend *LOC to account for token TOKEN of size SIZE. */
+#define YY_USER_ACTION location_compute (loc, &scanner_cursor, yytext, yyleng);
-static void
-extend_location (location_t *loc, char const *token, int size)
-{
- int line = loc->last_line;
- int column = loc->last_column;
- char const *p0 = token;
- char const *p = token;
- char const *lim = token + size;
-
- for (p = token; p < lim; p++)
- switch (*p)
- {
- case '\r':
- /* \r shouldn't survive no_cr_read. */
- abort ();
-
- case '\n':
- line++;
- column = 1;
- p0 = p + 1;
- break;
-
- case '\t':
- column += mbsnwidth (p0, p - p0, 0);
- column += 8 - ((column - 1) & 7);
- p0 = p + 1;
- break;
- }
-
- loc->last_line = line;
- loc->last_column = column + mbsnwidth (p0, p - p0, 0);
-}
-
-
-
-/* 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 size_t no_cr_read (FILE *, char *, size_t);
+#define YY_INPUT(buf, result, size) ((result) = no_cr_read (yyin, buf, size))
-static struct obstack string_obstack;
+#define RETURN_PERCENT_PARAM(Value) \
+ RETURN_VALUE(PERCENT_PARAM, param, param_ ## Value)
-#define YY_OBS_GROW \
- obstack_grow (&string_obstack, yytext, yyleng)
+#define RETURN_PERCENT_FLAG(Value) \
+ RETURN_VALUE(PERCENT_FLAG, uniqstr, uniqstr_new (Value))
-#define YY_OBS_FINISH \
- do { \
- obstack_1grow (&string_obstack, '\0'); \
- last_string = obstack_finish (&string_obstack); \
+#define RETURN_VALUE(Token, Field, Value) \
+ do { \
+ val->Field = Value; \
+ return Token; \
} while (0)
-#define YY_OBS_FREE \
- obstack_free (&string_obstack, last_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. */
+static char *last_string;
-/* 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.
+/* 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;
- Outside of well-formed rules, RULE_LENGTH has an undefined value. */
-static int rule_length;
+void
+gram_scanner_last_string_free (void)
+{
+ STRING_FREE;
+}
-static void handle_dollar (braced_code_t code_kind,
- char *cp, location_t location);
-static void handle_at (braced_code_t code_kind,
- char *cp, location_t location);
-static void handle_syncline (char *args, location_t *location);
+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_end_of_file (location_t, char const *);
+static void unexpected_eof (boundary, char const *);
+static void unexpected_newline (boundary, char const *);
%}
-%x SC_COMMENT SC_LINE_COMMENT SC_YACC_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
+
+ /* 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]|-)*
-int [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. */
@@ -192,117 +136,176 @@ 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);
- /* Scanner context when scanning C code. */
- int c_context IF_LINT (= 0);
+ /* Location of most recent identifier, when applicable. */
+ location id_loc IF_LINT (= empty_location);
- /* A string representing the most recently saved token. */
- char *last_string;
+ /* 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;
- /* At each yylex invocation, mark the current position as the
- start of the next token. */
- YY_STEP;
+ /* Where containing comment or string or character literal started,
+ when applicable. */
+ boundary token_start IF_LINT (= scanner_cursor);
%}
+ /*-----------------------.
+ | Scanning white space. |
+ `-----------------------*/
+
+
+{
+ /* 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. */
+
{
- "%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("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;
- "%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;
- "%parse-param" return PERCENT_PARSE_PARAM;
- "%prec" rule_length--; return PERCENT_PREC;
- "%printer" return PERCENT_PRINTER;
- "%pure"[-_]"parser" return PERCENT_PURE_PARSER;
- "%right" return PERCENT_RIGHT;
- "%lex-param" return PERCENT_LEX_PARAM;
- "%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 (*yylloc, _("invalid directive: %s"), quote (yytext));
- YY_STEP;
- }
-
- ^"#line "{int}" \""[^\"]*"\"\n" handle_syncline (yytext + strlen ("#line "), yylloc); YY_STEP;
+ "%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 COLON;
- "|" rule_length = 0; return PIPE;
- "," return COMMA;
+ "|" return PIPE;
";" return SEMICOLON;
- [ \f\n\t\v] YY_STEP;
-
- {id} {
- yylval->symbol = symbol_get (yytext, *yylloc);
- rule_length++;
- return ID;
+ {id} {
+ val->uniqstr = uniqstr_new (yytext);
+ id_loc = *loc;
+ bracketed_id_str = NULL;
+ BEGIN SC_AFTER_IDENTIFIER;
}
{int} {
- unsigned long num;
- errno = 0;
- num = strtoul (yytext, 0, 10);
- if (INT_MAX < num || errno)
- {
- complain_at (*yylloc, _("integer out of range: %s"), quote (yytext));
- num = INT_MAX;
- }
- yylval->integer = num;
+ 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; 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));
+ }
- /* Strings. */
- "\"" YY_OBS_GROW; BEGIN SC_ESCAPED_STRING;
+ /* Characters. */
+ "'" token_start = loc->start; BEGIN SC_ESCAPED_CHARACTER;
- /* Comments. */
- "/*" BEGIN SC_YACC_COMMENT;
- "//".* YY_STEP;
+ /* Strings. */
+ "\"" token_start = loc->start; BEGIN SC_ESCAPED_STRING;
/* Prologue. */
- "%{" BEGIN SC_PROLOGUE;
+ "%{" code_start = loc->start; BEGIN SC_PROLOGUE;
/* Code in between braces. */
- "{" YY_OBS_GROW; braces_level = 0; BEGIN SC_BRACED_CODE;
+ "{" {
+ STRING_GROW;
+ 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 (&string_obstack, yytext + 1, yyleng - 2);
- YY_OBS_FINISH;
- yylval->string = last_string;
- return TYPE;
+ 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;
}
"%%" {
@@ -312,9 +315,126 @@ 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));
+ }
+
+ <> {
+ loc->start = loc->end = scanner_cursor;
+ yyterminate ();
+ }
+}
+
+
+ /*--------------------------------------------------------------.
+ | 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;
+ return ID_COLON;
+ }
. {
- complain_at (*yylloc, _("invalid character: %s"), quote (yytext));
- YY_STEP;
+ ROLLBACK_CURRENT_TOKEN;
+ BEGIN (bracketed_id_str ? SC_RETURN_BRACKETED_ID : INITIAL);
+ *loc = id_loc;
+ return ID;
+ }
+ <> {
+ BEGIN (bracketed_id_str ? SC_RETURN_BRACKETED_ID : INITIAL);
+ *loc = id_loc;
+ 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;
}
}
@@ -325,13 +445,9 @@ splice (\\[ \f\t\v]*\n)*
{
- "*/" {
- YY_STEP;
- BEGIN INITIAL;
- }
-
+ "*/" BEGIN context_state;
.|\n ;
- <> unexpected_end_of_file (*yylloc, "*/");
+ <> unexpected_eof (token_start, "*/"); BEGIN context_state;
}
@@ -341,8 +457,8 @@ splice (\\[ \f\t\v]*\n)*
{
- "*"{splice}"/" YY_OBS_GROW; BEGIN c_context;
- <> unexpected_end_of_file (*yylloc, "*/");
+ "*"{splice}"/" STRING_GROW; BEGIN context_state;
+ <> unexpected_eof (token_start, "*/"); BEGIN context_state;
}
@@ -352,56 +468,122 @@ splice (\\[ \f\t\v]*\n)*
{
- "\n" YY_OBS_GROW; BEGIN c_context;
- {splice} YY_OBS_GROW;
- <> BEGIN c_context;
+ "\n" STRING_GROW; BEGIN context_state;
+ {splice} STRING_GROW;
+ <> 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. |
+ `------------------------------------------------*/
{
- "\"" {
- YY_OBS_GROW;
- YY_OBS_FINISH;
- yylval->string = last_string;
- 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;
+ }
+ <> {
+ unexpected_eof (token_start, "\"");
+ STRING_FINISH;
+ loc->start = token_start;
+ val->chars = last_string;
BEGIN INITIAL;
return STRING;
}
-
- .|\n YY_OBS_GROW;
- <> unexpected_end_of_file (*yylloc, "\"");
}
- /*---------------------------------------------------------------.
- | 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. |
+ `----------------------------------------------------------*/
{
- "'" {
- YY_OBS_GROW;
- 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;
- rule_length++;
+ "'"|"\n" {
+ 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"));
+ }
+ if (yytext[0] == '\n')
+ unexpected_newline (token_start, "'");
+ STRING_FREE;
BEGIN INITIAL;
- return ID;
+ 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, "'");
+ STRING_FREE;
+ BEGIN INITIAL;
+ return CHAR;
}
-
- .|\n YY_OBS_GROW;
- <> unexpected_end_of_file (*yylloc, "'");
}
+ /*-----------------------------------------------------------.
+ | Scanning a Bison nested tag. The initial angle bracket is |
+ | already eaten. |
+ `-----------------------------------------------------------*/
+
+
+{
+ ">" {
+ --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. |
@@ -410,83 +592,75 @@ splice (\\[ \f\t\v]*\n)*
{
\\[0-7]{1,3} {
- unsigned long c = strtoul (yytext + 1, 0, 8);
- if (UCHAR_MAX < c)
- {
- complain_at (*yylloc, _("invalid escape sequence: %s"),
- quote (yytext));
- YY_STEP;
- }
+ unsigned long int c = strtoul (yytext + 1, NULL, 8);
+ if (!c || UCHAR_MAX < c)
+ complain_at (*loc, _("invalid number after \\-escape: %s"),
+ yytext+1);
else
- obstack_1grow (&string_obstack, c);
+ obstack_1grow (&obstack_for_string, c);
}
- \\x[0-9a-fA-F]+ {
- unsigned long c;
- errno = 0;
- c = strtoul (yytext + 2, 0, 16);
- if (UCHAR_MAX < c || errno)
- {
- complain_at (*yylloc, _("invalid escape sequence: %s"),
- quote (yytext));
- YY_STEP;
- }
+ \\x[0-9abcdefABCDEF]+ {
+ 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"),
+ yytext+1);
else
- obstack_1grow (&string_obstack, 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]);
- \\(u|U[0-9a-fA-F]{4})[0-9a-fA-F]{4} {
+ obstack_1grow (&obstack_for_string, c);
+ }
+
+ \\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 (*yylloc, _("invalid escape sequence: %s"),
- quote (yytext));
- YY_STEP;
- }
+ if (c <= 0)
+ complain_at (*loc, _("invalid number after \\-escape: %s"),
+ yytext+1);
else
- obstack_1grow (&string_obstack, c);
+ obstack_1grow (&obstack_for_string, c);
}
\\(.|\n) {
- complain_at (*yylloc, _("unrecognized escape sequence: %s"),
- quote (yytext));
- YY_OBS_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);
}
}
+ /*--------------------------------------------.
+ | Scanning user-code characters and strings. |
+ `--------------------------------------------*/
- /*----------------------------------------------------------.
- | Scanning a C character without decoding its escapes. The |
- | initial "'" is already eaten. |
- `----------------------------------------------------------*/
+
+{
+ {splice}|\\{splice}[^\n\[\]] STRING_GROW;
+}
{
- "'" YY_OBS_GROW; BEGIN c_context;
- \\{splice}[^\[\]] YY_OBS_GROW;
- {splice} YY_OBS_GROW;
- <> unexpected_end_of_file (*yylloc, "'");
+ "'" STRING_GROW; BEGIN context_state;
+ \n unexpected_newline (token_start, "'"); BEGIN context_state;
+ <> unexpected_eof (token_start, "'"); BEGIN context_state;
}
-
- /*----------------------------------------------------------------.
- | Scanning a C string, without decoding its escapes. The initial |
- | `"' is already eaten. |
- `----------------------------------------------------------------*/
-
{
- "\"" YY_OBS_GROW; BEGIN c_context;
- \\{splice}[^\[\]] YY_OBS_GROW;
- {splice} YY_OBS_GROW;
- <> unexpected_end_of_file (*yylloc, "\"");
+ "\"" STRING_GROW; BEGIN context_state;
+ \n unexpected_newline (token_start, "\""); BEGIN context_state;
+ <> unexpected_eof (token_start, "\""); BEGIN context_state;
}
@@ -494,49 +668,93 @@ splice (\\[ \f\t\v]*\n)*
| Strings, comments etc. can be found in user code. |
`---------------------------------------------------*/
-
+
{
- "'" YY_OBS_GROW; c_context = YY_START; BEGIN SC_CHARACTER;
- "\"" YY_OBS_GROW; c_context = YY_START; BEGIN SC_STRING;
- "/"{splice}"*" YY_OBS_GROW; c_context = YY_START; BEGIN SC_COMMENT;
- "/"{splice}"/" YY_OBS_GROW; c_context = YY_START; BEGIN SC_LINE_COMMENT;
+ "'" {
+ 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, predicates). The |
+ | initial "{" is already eaten. |
+ `-----------------------------------------------------------*/
+
+
+{
+ "{"|"<"{splice}"%" STRING_GROW; nesting++;
+ "%"{splice}">" STRING_GROW; nesting--;
+
+ /* 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;
+ }
+}
{
- "{"|"<"{splice}"%" YY_OBS_GROW; braces_level++;
- "%"{splice}">" YY_OBS_GROW; braces_level--;
"}" {
- YY_OBS_GROW;
- braces_level--;
- if (braces_level < 0)
+ obstack_1grow (&obstack_for_string, '}');
+
+ --nesting;
+ if (nesting < 0)
{
- YY_OBS_FINISH;
- yylval->string = last_string;
- rule_length++;
+ STRING_FINISH;
+ loc->start = code_start;
+ val->code = last_string;
BEGIN INITIAL;
return BRACED_CODE;
}
}
-
- /* Tokenize `<<%' correctly (as `<<' `%') rather than incorrrectly
- (as `<' `<%'). */
- "<"{splice}"<" YY_OBS_GROW;
-
- "$"("<"{tag}">")?(-?[0-9]+|"$") { handle_dollar (current_braced_code,
- yytext, *yylloc); }
- "@"(-?[0-9]+|"$") { handle_at (current_braced_code,
- yytext, *yylloc); }
-
- <> unexpected_end_of_file (*yylloc, "}");
}
+
+{
+ "}" {
+ --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, '}');
+ }
+}
/*--------------------------------------------------------------.
| Scanning some prologue: from "%{" (already scanned) to "%}". |
@@ -545,13 +763,21 @@ splice (\\[ \f\t\v]*\n)*
{
"%}" {
- YY_OBS_FINISH;
- yylval->string = last_string;
+ STRING_FINISH;
+ loc->start = code_start;
+ val->chars = last_string;
BEGIN INITIAL;
return PROLOGUE;
}
- <> unexpected_end_of_file (*yylloc, "%}");
+ <> {
+ unexpected_eof (code_start, "%}");
+ STRING_FINISH;
+ loc->start = code_start;
+ val->chars = last_string;
+ BEGIN INITIAL;
+ return PROLOGUE;
+ }
}
@@ -563,200 +789,89 @@ splice (\\[ \f\t\v]*\n)*
{
<> {
- 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, escaping M4 |
- | quoting characters. |
- `----------------------------------------------------------------*/
-
-
-{
- \[ obstack_sgrow (&string_obstack, "@<:@");
- \] obstack_sgrow (&string_obstack, "@:>@");
- .|\n YY_OBS_GROW;
-}
+ /*-----------------------------------------------------.
+ | By default, grow the string obstack with the input. |
+ `-----------------------------------------------------*/
+. |
+ \n STRING_GROW;
%%
-/*------------------------------------------------------------------.
-| TEXT is pointing to a wannabee semantic value (i.e., a `$'). |
-| |
-| Possible inputs: $[]($|integer) |
-| |
-| Output to the STRING_OBSTACK a reference to this semantic value. |
-`------------------------------------------------------------------*/
+/* 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. */
-static inline void
-handle_action_dollar (char *text, location_t location)
+static size_t
+no_cr_read (FILE *fp, char *buf, size_t size)
{
- const char *type_name = NULL;
- char *cp = text + 1;
-
- /* Get the type name if explicit. */
- if (*cp == '<')
+ size_t bytes_read = fread (buf, 1, size, fp);
+ if (bytes_read)
{
- type_name = ++cp;
- while (*cp != '>')
- ++cp;
- *cp = '\0';
- ++cp;
- }
+ char *w = memchr (buf, '\r', bytes_read);
+ if (w)
+ {
+ char const *r = ++w;
+ char const *lim = buf + bytes_read;
- if (*cp == '$')
- {
- if (!type_name)
- type_name = symbol_list_n_type_name_get (current_rule, location, 0);
- if (!type_name && typed)
- complain_at (location, _("$$ of `%s' has no declared type"),
- current_rule->sym->tag);
- if (!type_name)
- type_name = "";
- obstack_fgrow1 (&string_obstack,
- "]b4_lhs_value([%s])[", type_name);
- }
- else
- {
- long num;
- errno = 0;
- num = strtol (cp, 0, 10);
+ for (;;)
+ {
+ /* Found an '\r'. Treat it like '\n', but ignore any
+ '\n' that immediately follows. */
+ w[-1] = '\n';
+ if (r == lim)
+ {
+ int ch = getc (fp);
+ if (ch != '\n' && ungetc (ch, fp) != ch)
+ break;
+ }
+ else if (*r == '\n')
+ r++;
- if (INT_MIN <= num && num <= rule_length && ! errno)
- {
- int n = num;
- if (!type_name && n > 0)
- type_name = symbol_list_n_type_name_get (current_rule, location,
- n);
- if (!type_name && typed)
- complain_at (location, _("$%d of `%s' has no declared type"),
- n, current_rule->sym->tag);
- if (!type_name)
- type_name = "";
- obstack_fgrow3 (&string_obstack,
- "]b4_rhs_value([%d], [%d], [%s])[",
- rule_length, n, type_name);
+ /* Copy until the next '\r'. */
+ do
+ {
+ if (r == lim)
+ return w - buf;
+ }
+ while ((*w++ = *r++) != '\r');
+ }
+
+ return w - buf;
}
- else
- complain_at (location, _("integer out of range: %s"), quote (text));
}
-}
-
-/*---------------------------------------------------------------.
-| TEXT is expected to be $$ in some code associated to a symbol: |
-| destructor or printer. |
-`---------------------------------------------------------------*/
-
-static inline void
-handle_symbol_code_dollar (char *text, location_t location)
-{
- char *cp = text + 1;
- if (*cp == '$')
- obstack_sgrow (&string_obstack, "]b4_dollar_dollar[");
- else
- complain_at (location, _("invalid value: %s"), quote (text));
+ return bytes_read;
}
-/*-----------------------------------------------------------------.
-| Dispatch onto handle_action_dollar, or handle_destructor_dollar, |
-| depending upon CODE_KIND. |
-`-----------------------------------------------------------------*/
-
-static void
-handle_dollar (braced_code_t braced_code_kind,
- char *text, location_t location)
-{
- switch (braced_code_kind)
- {
- case action_braced_code:
- handle_action_dollar (text, location);
- break;
-
- case destructor_braced_code:
- case printer_braced_code:
- handle_symbol_code_dollar (text, location);
- break;
- }
-}
-
/*------------------------------------------------------.
-| TEXT is a location token (i.e., a `@...'). Output to |
-| STRING_OBSTACK a reference to this location. |
+| Scan NUMBER for a base-BASE integer at location LOC. |
`------------------------------------------------------*/
-static inline void
-handle_action_at (char *text, location_t location)
+static unsigned long int
+scan_integer (char const *number, int base, location loc)
{
- char *cp = text + 1;
- locations_flag = 1;
+ verify (INT_MAX < ULONG_MAX);
+ unsigned long int num = strtoul (number, NULL, base);
- if (*cp == '$')
+ if (INT_MAX < num)
{
- obstack_sgrow (&string_obstack, "]b4_lhs_location[");
- }
- else
- {
- long num;
- errno = 0;
- num = strtol (cp, 0, 10);
-
- if (INT_MIN <= num && num <= rule_length && ! errno)
- {
- int n = num;
- obstack_fgrow2 (&string_obstack, "]b4_rhs_location([%d], [%d])[",
- rule_length, n);
- }
- else
- complain_at (location, _("integer out of range: %s"), quote (text));
+ complain_at (loc, _("integer out of range: %s"), quote (number));
+ num = INT_MAX;
}
-}
-
-
-/*---------------------------------------------------------------.
-| TEXT is expected to be @$ in some code associated to a symbol: |
-| destructor or printer. |
-`---------------------------------------------------------------*/
-static inline void
-handle_symbol_code_at (char *text, location_t location)
-{
- char *cp = text + 1;
- if (*cp == '$')
- obstack_sgrow (&string_obstack, "]b4_at_dollar[");
- else
- complain_at (location, _("invalid value: %s"), quote (text));
-}
-
-
-/*-------------------------------------------------------------------.
-| Dispatch onto handle_action_at, or handle_destructor_at, depending |
-| upon CODE_KIND. |
-`-------------------------------------------------------------------*/
-
-static void
-handle_at (braced_code_t braced_code_kind,
- char *text, location_t location)
-{
- 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;
- }
+ return num;
}
@@ -769,7 +884,8 @@ handle_at (braced_code_t braced_code_kind,
static int
convert_ucn_to_byte (char const *ucn)
{
- unsigned long code = strtoul (ucn + 2, 0, 16);
+ 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
@@ -821,34 +937,58 @@ convert_ucn_to_byte (char const *ucn)
`----------------------------------------------------------------*/
static void
-handle_syncline (char *args, location_t *location)
+handle_syncline (char *args, location loc)
{
- int lineno = strtol (args, &args, 10);
- const char *file = NULL;
- file = strchr (args, '"') + 1;
- *strchr (file, '"') = 0;
- /* FIXME: Leaking... Can't free, as some locations are still
- pointing to the old file name. */
- infile = xstrdup (file);
- location->file = infile;
- location->last_line = lineno;
+ char *after_num;
+ unsigned long int lineno = strtoul (args, &after_num, 10);
+ char *file = mbschr (after_num, '"') + 1;
+ *mbschr (file, '"') = '\0';
+ if (INT_MAX <= lineno)
+ {
+ warn_at (loc, _("line number overflow"));
+ lineno = INT_MAX;
+ }
+ current_file = uniqstr_new (file);
+ boundary_set (&scanner_cursor, current_file, lineno, 1);
}
-/*-------------------------------------------------------------.
-| Report an unexpected end of file at LOC. An end of file was |
-| encountered and the expected TOKEN_END was missing. After |
-| reporting the problem, pretend that TOKEN_END was found. |
-`-------------------------------------------------------------*/
+/*----------------------------------------------------------------.
+| 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 void
-unexpected_end_of_file (location_t loc, char const *token_end)
+unexpected_end (boundary start, char const *msgid, char const *token_end)
{
- size_t i;
+ location loc;
+ loc.start = start;
+ loc.end = scanner_cursor;
+ complain_at (loc, _(msgid), token_end);
+}
+
- complain_at (loc, _("missing `%s' at end of file"), token_end);
- for (i = strlen (token_end); i != 0; i--)
- unput (token_end[i - 1]);
+/*------------------------------------------------------------------------.
+| 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 void
+unexpected_eof (boundary start, char const *token_end)
+{
+ unexpected_end (start, N_("missing `%s' at end of file"), token_end);
+}
+
+
+/*----------------------------------------.
+| Likewise, but for unexpected newlines. |
+`----------------------------------------*/
+
+static void
+unexpected_newline (boundary start, char const *token_end)
+{
+ unexpected_end (start, N_("missing `%s' at end of line"), token_end);
}
@@ -857,9 +997,9 @@ unexpected_end_of_file (location_t loc, char const *token_end)
`-------------------------*/
void
-scanner_initialize (void)
+gram_scanner_initialize (void)
{
- obstack_init (&string_obstack);
+ obstack_init (&obstack_for_string);
}
@@ -868,9 +1008,9 @@ scanner_initialize (void)
`-----------------------------------------------*/
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 ();
}