* data/glr.c (YYERROR): Update definition.

(yyrecoverSyntaxError): Correct yyerrState logic. Correct comment.
Allow states with only a default reduction.

Fixes to avoid problem that $-N rules in GLR parsers can cause
buffer overruns, corrupting state.

* src/output.c (prepare_rules): Output max_left_semantic_context.
* src/reader.h (max_left_semantic_context): New
* src/scan-gram.l (max_left_semantic_context): Define.
(handle_action_dollar): Update max_left_semantic_context.
* data/glr.c (YYMAXLEFT): New.
(yydoAction): Increase yyrhsVals size.
(yyresolveAction): Ditto.

Fixes to problems with location handling in GLR parsers reported by
Frank Heckenbach (2003/06/05).

* data/glr.c (YYLTYPE): Make trivial if locations not used.
(YYRHSLOC): Add parentheses, make depend on whether locations used.
(YYLLOC_DEFAULT): Ditto.
(yyuserAction): Use YYLLOC_DEFAULT.
(yydoAction): Remove redundant code.

* tests/cxx-type.at: Exercise location information.
(yylex): Track locations.
(stmtMerge): Return value rather than printing.
This commit is contained in:
Paul Hilfinger
2003-06-10 02:44:58 +00:00
parent 144c1e767a
commit 25005f6ab0
6 changed files with 240 additions and 144 deletions

View File

@@ -24,7 +24,10 @@ AT_BANNER([[C++ Type Syntax (GLR).]])
# and with RESOLVE1 and RESOLVE2 as annotations on the conflicted rule for
# stmt. Then compile the result.
m4_define([_AT_TEST_GLR_CXXTYPES],
[AT_DATA_GRAMMAR([types.y],
[
AT_BISON_OPTION_PUSHDEFS([$1])
AT_DATA_GRAMMAR([types.y],
[[/* Simplified C++ Type and Expression Grammar. */
$1
@@ -32,24 +35,10 @@ $1
%{
#include <stdio.h>
#define YYSTYPE const char*
#define YYLTYPE int
]m4_bmatch([$2], [stmtMerge],
[ static YYSTYPE stmtMerge (YYSTYPE x0, YYSTYPE x1);])[
#define YYINITDEPTH 10
int yyerror (
#if YYPURE && YYLSP_NEEDED
YYLTYPE *yylocation,
#endif
const char *s
);
#if YYPURE
]m4_bmatch([$1], [location],
[ int yylex (YYSTYPE *lvalp, YYLTYPE *llocp);],
[ int yylex (YYSTYPE *lvalp);])[
#else
int yylex (void);
#endif
static char* format (const char*, ...);
%}
@@ -63,30 +52,35 @@ $1
%%
prog :
| prog stmt { printf ("\n"); }
| prog stmt {
]AT_LOCATION_IF([
printf ("%d.%d-%d.%d: ",
@2.first_line, @2.first_column,
@2.last_line, @2.last_column);])[
printf ("%s\n", ]$[2);
}
;
stmt : expr ';' $2
stmt : expr ';' $2 { $$ = ]$[1; }
| decl $3
| error ';'
| '@' { YYACCEPT; }
| error ';' { $$ = "<error>"; }
| '@' { YYACCEPT; }
;
expr : ID { printf ("%s ", ]$[1); }
| TYPENAME '(' expr ')'
{ printf ("%s <cast> ", ]$[1); }
| expr '+' expr { printf ("+ "); }
| expr '=' expr { printf ("= "); }
expr : ID
| TYPENAME '(' expr ')' { $$ = format ("<cast>(%s,%s)", ]$[3, ]$[1); }
| expr '+' expr { $$ = format ("+(%s,%s)", ]$[1, ]$[3); }
| expr '=' expr { $$ = format ("=(%s,%s)", ]$[1, ]$[3); }
;
decl : TYPENAME declarator ';'
{ printf ("%s <declare> ", ]$[1); }
{ $$ = format ("<declare>(%s,%s)", ]$[1, ]$[2); }
| TYPENAME declarator '=' expr ';'
{ printf ("%s <init-declare> ", ]$[1); }
{ $$ = format ("<init-declare>(%s,%s,%s)", ]$[1, ]$[2, ]$[4); }
;
declarator : ID { printf ("\"%s\" ", ]$[1); }
| '(' declarator ')'
declarator : ID
| '(' declarator ')' { $$ = ]$[2; }
;
%%
@@ -94,6 +88,7 @@ declarator : ID { printf ("\"%s\" ", ]$[1); }
#include <ctype.h>
#include <stdlib.h>
#include <string.h>
#include <stdarg.h>
int
main (int argc, char** argv)
@@ -105,23 +100,25 @@ main (int argc, char** argv)
exit (yyparse ());
}
#if YYPURE
int
]m4_bmatch([$1], [location],
[yylex (YYSTYPE *lvalp, YYLTYPE *llocp)],
[yylex (YYSTYPE *lvalp)])[
#if YYPURE && YYLSP_NEEDED
yylex (YYSTYPE *lvalp, YYLTYPE *llocp)
#elif YYPURE
yylex (YYSTYPE *lvalp)
#else
int
yylex ()
yylex ()
#endif
{
char buffer[256];
int c;
unsigned int i;
static int lineNum = 1;
static int colNum = 1;
#if YYPURE
# define yylloc (*llocp)
# define yylval (*lvalp)
]m4_bmatch([$1], [location],[ (void) llocp;])[
#endif
while (1)
@@ -131,28 +128,53 @@ yylex ()
{
case EOF:
return 0;
case ' ': case '\t': case '\n': case '\f':
case '\t':
colNum = 1 + ((colNum + 7) & ~7);
break;
case ' ': case '\f':
colNum += 1;
break;
case '\n':
lineNum += 1;
colNum = 1;
break;
default:
if (isalpha (c))
{
i = 0;
{
int tok;
#if YYLSP_NEEDED
yylloc.first_line = yylloc.last_line = lineNum;
yylloc.first_column = colNum;
#endif
if (isalpha (c))
{
i = 0;
do
{
buffer[i++] = c;
if (i == sizeof buffer - 1)
abort ();
c = getchar ();
}
while (isalnum (c) || c == '_');
do
{
buffer[i++] = c;
colNum += 1;
if (i == sizeof buffer - 1)
abort ();
c = getchar ();
}
while (isalnum (c) || c == '_');
ungetc (c, stdin);
buffer[i++] = 0;
yylval = strcpy (malloc (i), buffer);
return isupper ((unsigned char) buffer[0]) ? TYPENAME : ID;
}
return c;
ungetc (c, stdin);
buffer[i++] = 0;
tok = isupper ((unsigned char) buffer[0]) ? TYPENAME : ID;
yylval = strcpy (malloc (i), buffer);
}
else
{
colNum += 1;
tok = c;
yylval = "";
}
#if YYLSP_NEEDED
yylloc.last_column = colNum-1;
#endif
return tok;
}
}
}
}
@@ -160,28 +182,32 @@ yylex ()
int
yyerror (
#if YYPURE && YYLSP_NEEDED
YYLTYPE *yylocation,
YYLTYPE *llocp,
#endif
const char *s
)
{
#if YYPURE && YYLSP_NEEDED
(void) *yylocation;
#endif
fprintf (stderr, "%s\n", s);
return 0;
}
static char* format (const char* form, ...)
{
char buffer[1024];
va_list args;
va_start (args, form);
vsprintf (buffer, form, args);
va_end (args);
return strcpy (malloc (strlen (buffer) + 1), buffer);
}
]]
m4_bmatch([$2], [stmtMerge],
[[static YYSTYPE
stmtMerge (YYSTYPE x0, YYSTYPE x1)
{
/* Use the arguments. */
(void) x0;
(void) x1;
printf ("<OR> ");
return "";
return format ("<OR>(%s,%s)", x0, x1);
}
]])
)
@@ -214,30 +240,55 @@ This is total garbage, but it should be ignored.
AT_CHECK([bison -o types.c types.y], 0, [], ignore)
AT_COMPILE([types])
AT_BISON_OPTION_POPDEFS
])
m4_define([_AT_RESOLVED_GLR_OUTPUT],
[[z q +
"x" T <declare>
"x" y T <init-declare>
x y =
x T <cast> y +
"x" T <declare>
"y" z q + T <init-declare>
y
z q +
[[+(z,q)
<declare>(T,x)
<init-declare>(T,x,y)
=(x,y)
+(<cast>(x,T),y)
<declare>(T,x)
<init-declare>(T,y,+(z,q))
<error>
+(z,q)
]])
m4_define([_AT_RESOLVED_GLR_OUTPUT_WITH_LOC],
[[3.1-3.6: +(z,q)
5.1-5.4: <declare>(T,x)
7.1-7.8: <init-declare>(T,x,y)
9.1-9.6: =(x,y)
11.1-11.10: +(<cast>(x,T),y)
13.1-13.6: <declare>(T,x)
15.1-15.14: <init-declare>(T,y,+(z,q))
17.6-17.16: <error>
19.1-19.6: +(z,q)
]])
m4_define([_AT_AMBIG_GLR_OUTPUT],
[[z q +
"x" T <declare>
"x" y T <init-declare>
x y =
x T <cast> y +
"x" T <declare> x T <cast> <OR>
"y" z q + T <init-declare> y T <cast> z q + = <OR>
y
z q +
[[+(z,q)
<declare>(T,x)
<init-declare>(T,x,y)
=(x,y)
+(<cast>(x,T),y)
<OR>(<declare>(T,x),<cast>(x,T))
<OR>(<init-declare>(T,y,+(z,q)),=(<cast>(y,T),+(z,q)))
<error>
+(z,q)
]])
m4_define([_AT_AMBIG_GLR_OUTPUT_WITH_LOC],
[[3.1-3.6: +(z,q)
5.1-5.4: <declare>(T,x)
7.1-7.8: <init-declare>(T,x,y)
9.1-9.6: =(x,y)
11.1-11.10: +(<cast>(x,T),y)
13.1-13.6: <OR>(<declare>(T,x),<cast>(x,T))
15.1-15.14: <OR>(<init-declare>(T,y,+(z,q)),=(<cast>(y,T),+(z,q)))
17.6-17.16: <error>
19.1-19.6: +(z,q)
]])
m4_define([_AT_GLR_STDERR],
@@ -262,7 +313,7 @@ AT_CLEANUP
AT_SETUP([GLR: Resolve ambiguity, impure, locations])
_AT_TEST_GLR_CXXTYPES([%locations],[%dprec 1],[%dprec 2])
AT_PARSER_CHECK([[./types test-input | sed 's/ *$//']], 0,
_AT_RESOLVED_GLR_OUTPUT, _AT_GLR_STDERR)
_AT_RESOLVED_GLR_OUTPUT_WITH_LOC, _AT_GLR_STDERR)
AT_CLEANUP
AT_SETUP([GLR: Resolve ambiguity, pure, no locations])
@@ -276,7 +327,7 @@ AT_SETUP([GLR: Resolve ambiguity, pure, locations])
_AT_TEST_GLR_CXXTYPES([%pure-parser %locations],
[%dprec 1], [%dprec 2])
AT_PARSER_CHECK([[./types test-input | sed 's/ *$//']], 0,
_AT_RESOLVED_GLR_OUTPUT, _AT_GLR_STDERR)
_AT_RESOLVED_GLR_OUTPUT_WITH_LOC, _AT_GLR_STDERR)
AT_CLEANUP
AT_SETUP([GLR: Merge conflicting parses, impure, no locations])
@@ -290,7 +341,7 @@ AT_SETUP([GLR: Merge conflicting parses, impure, locations])
_AT_TEST_GLR_CXXTYPES([%locations],
[%merge <stmtMerge>], [%merge <stmtMerge>])
AT_PARSER_CHECK([[./types test-input | sed 's/ *$//']], 0,
_AT_AMBIG_GLR_OUTPUT, _AT_GLR_STDERR)
_AT_AMBIG_GLR_OUTPUT_WITH_LOC, _AT_GLR_STDERR)
AT_CLEANUP
AT_SETUP([GLR: Merge conflicting parses, pure, no locations])
@@ -303,7 +354,7 @@ AT_SETUP([GLR: Merge conflicting parses, pure, locations])
_AT_TEST_GLR_CXXTYPES([%pure-parser %locations],
[%merge <stmtMerge>],[%merge <stmtMerge>])
AT_PARSER_CHECK([[./types test-input | sed 's/ *$//']], 0,
_AT_AMBIG_GLR_OUTPUT, _AT_GLR_STDERR)
_AT_AMBIG_GLR_OUTPUT_WITH_LOC, _AT_GLR_STDERR)
AT_CLEANUP
AT_SETUP([GLR: Verbose messages, resolve ambiguity, impure, no locations])