ack/util/ceg/EM_parser/common/pars.g

330 lines
6.7 KiB
Plaintext
Raw Permalink Normal View History

1987-12-18 16:23:42 +00:00
/* This file contains the parser for the 'EM_table'.
* Every row in the table is converted to a C-function.
* The parser is a fairly simple program, for each row it prints a function-
* header and the following actions are printed as C-statements. A CALL is just
* copied, and an ASSEM_INSTR will be handed to the routine assemble().
* Assemble() will turn the assembly-string into a sequence of C-function calls.
* How this should be done is expressed in the 'as_table'.
*
* There are however some complicating factors:
* - A row in the 'EM_table' may contain CONDITION's.
* The parser supports this by outptutting some "if" and "else"'s.
*
* - The user is allowed to use some abbreviation in the 'EM_table', (s)he
* can use the DEF_C_INSTR's ( e.g. C_loe..).
* One such DEF_C_INSTR must be expanded in 3 or more C-functions.
* The solution is to copy the DEF_C_INSTR to a temporary file. Then
* this file will be processed as many times as needed to generate all
* the C-functions.
*
* - The assembler-instructions must be processed in blocks not one at a
* time, so buffering is needed.
*
* - When generating object-code, the user can make use of the dist()
* function. This function should return the number of bytes between
* the current instruction and a labelled one, both forward and backward
* references must be handled.
* This requires the saving of the generated function, and
* backpatching in a second phase.
*/
1987-11-25 11:19:38 +00:00
{
#include <system.h>
#include "decl.h"
#include "em.h"
1987-12-18 16:23:42 +00:00
1987-11-25 11:19:38 +00:00
extern char yytext[];
extern int yylineno;
extern int first_action, last_action, token;
t_token tok, stok;
int no_conversions = FALSE, library, segment = UNKNOWN;
File *outfile;
char *to_change;
}
1989-01-26 14:43:09 +00:00
%token C_INSTR, DEF_C_INSTR, CONDITION, ARROW,
1987-11-25 11:19:38 +00:00
CALL, ASSEM_INSTR, DEFAULT, ERROR;
%start table, table;
%start def_row, def_row;
%start c_table, c_table;
%lexical lex_analyzer ;
table : row*
;
row : C_INSTR { set_outfile( yytext); header( yytext);
set_segment(segment);
}
1987-11-25 11:19:38 +00:00
[ special | simple] { out( "}\n\n");}
| DEF_C_INSTR { init_defaults( yytext);}
[ Dspecial | Dsimple] { handle_defaults();}
;
1988-05-31 14:05:41 +00:00
special :
{ out( "if( 0 ) ;\n"); }
1987-11-25 11:19:38 +00:00
[ CONDITION { out( "else "); question( &yytext[0]);}
simple { out( "}\n");}
]*
DEFAULT { out( "else {\n");}
simple { out( "}\n");}
;
simple : ARROW { save_output();}
1987-11-25 11:19:38 +00:00
actionlist { back_patch();}
1988-05-31 14:05:41 +00:00
;
1987-11-25 11:19:38 +00:00
actionlist : { first_action = TRUE;}
[ action { first_action = FALSE;}
[ ';' action]*
]?
'.'
;
action : CALL { print_call( &yytext[0]);}
| as_block { last_action = ( token == '.');
do_block_assemble();}
;
as_block : { init_as_block();}
ASSEM_INSTR { save_as( &yytext[0]);}
more_as
;
more_as :
%if ( next_token() == ASSEM_INSTR)
';'
ASSEM_INSTR { save_as( &yytext[0]);}
more_as
|
;
1987-12-18 16:23:42 +00:00
/* Grammar rules for a default-instruction, just copy to a temporary file. */
1987-11-25 11:19:38 +00:00
Dspecial: CONDITION { out( " %s ", yytext);}
Dsimple
[ CONDITION { out( " %s ", yytext);}
Dsimple
]*
DEFAULT { out( " %s ", yytext);}
Dsimple
;
Dsimple : ARROW { out( "%s", yytext);}
Dactionlist
;
Dactionlist :
[ Daction
[ ';' { out( ";\n");}
Daction
]*
]?
'.' { out( ".\n");}
;
Daction : CALL { out( "%s", yytext);}
| ASSEM_INSTR { out( "\"%s\"", yytext);}
;
def_row : { set_segment(segment);}
[ special | simple] { out( "}\n\n");}
1987-11-25 11:19:38 +00:00
;
1987-12-18 16:23:42 +00:00
/* This is a grammar to handle the -c flag ( only one entry has changed).
* Skip all entries in the table until the one that has changed, then
* use the grammar rules for the normal case above.
*/
1987-11-25 11:19:38 +00:00
c_table : c_row*
;
1988-09-14 14:03:50 +00:00
c_row : %if ( to_change && strcmp( yytext, to_change) == 0)
C_INSTR { set_outfile( yytext); header( yytext);
set_segment(segment);
}
1988-09-14 14:03:50 +00:00
[ special | simple] { out( "}\n\n"); to_change = 0; }
1987-11-25 11:19:38 +00:00
| C_INSTR
[ c_special | c_simple]
| %if ( to_change && strcmp( yytext, to_change) == 0)
1987-11-25 11:19:38 +00:00
DEF_C_INSTR { init_defaults( yytext);}
[ Dspecial | Dsimple] { handle_defaults(); to_change = 0; }
1987-11-25 11:19:38 +00:00
| DEF_C_INSTR
[ c_special | c_simple]
;
c_special : CONDITION
c_simple
[ CONDITION
c_simple
]*
DEFAULT
c_simple
;
c_simple: ARROW
c_actionlist
;
c_actionlist :
[ c_action [ ';' c_action]* ]? '.'
;
c_action: CALL
| ASSEM_INSTR
;
1987-12-18 16:23:42 +00:00
1987-11-25 11:19:38 +00:00
{
int saved = 0, token;
1988-09-07 14:47:40 +00:00
int nerrors = 0;
1987-11-25 11:19:38 +00:00
LLmessage( inserted_token)
int inserted_token;
{
1988-09-07 14:47:40 +00:00
nerrors++;
1987-11-25 11:19:38 +00:00
if ( inserted_token == 0) {
fprint( STDERR, "EM_table : syntax error in line %d, >>",
yylineno);
print_token( LLsymb);
fprint( STDERR, "<< will be deleted!!\n");
}
else if ( inserted_token < 0) {
1988-09-07 14:47:40 +00:00
fprint(STDERR,"EM_table : syntax error in line %d, garbage at end of table\n",
1987-11-25 11:19:38 +00:00
yylineno);
}
else {
fprint( STDERR, "EM_table : syntax error in line %d, >>",
yylineno);
print_token( inserted_token);
fprint( STDERR, "<< will be inserted!!\n");
token = LLsymb;
saved = 1;
}
}
print_token( token)
int token;
{
switch ( token) {
case C_INSTR : fprint( STDERR, "C_INSTR %s", yytext);
break;
case ASSEM_INSTR : fprint( STDERR, "STRING %s", yytext);
break;
case CALL : fprint( STDERR, "CALL %s", yytext);
break;
case ARROW : fprint( STDERR, "==> ");
break;
case CONDITION: fprint( STDERR, "CONDITION %s", yytext);
break;
case DEFAULT : fprint( STDERR, "default ");
break;
case ERROR : fprint( STDERR, "unmatched %s", yytext);
break;
default : fprint( STDERR, " %c", token);
break;
}
}
1987-12-18 16:23:42 +00:00
/* The lexical analyzer.
* It uses mylex() to get the tokens.
* It supports the buffering of one token.
* If the curent token is a C_INSTR it will set the global variable
* 'C_instr_info' to point to the correct information ( arguments, header,
* etc.).
*/
1987-11-25 11:19:38 +00:00
int lex_analyzer()
{
extern char *next;
if ( saved) {
saved = 0;
}
else {
token = mylex();
*next = '\0';
}
if ( token == C_INSTR)
set_C_instr_info( yytext);
return( token);
}
int next_token()
{
extern char *next;
if ( !saved) {
token = mylex();
*next = '\0';
saved = 1;
}
return( token);
}
1987-12-18 16:23:42 +00:00
/******************************************************************************/
1987-11-25 11:19:38 +00:00
1987-12-18 16:23:42 +00:00
/* Handle arguments :
* -l : Library, generate one function per file.
* -c <name> : Change, only update the <name> file.
*/
1987-11-25 11:19:38 +00:00
main( argc, argv)
int argc;
char **argv;
{
outfile = STDOUT;
if ( argc > 1) {
if ( strcmp( argv[1], "-l") == 0)
library = TRUE;
else if( strcmp( argv[1], "-c") == 0) {
library = TRUE;
to_change = argv[2];
c_table();
1988-09-14 14:03:50 +00:00
if (to_change) {
fprint( STDERR, "No rule for %s\n", to_change);
exit( 1);
}
exit(nerrors);
1987-11-25 11:19:38 +00:00
}
}
else {
library = FALSE;
1987-12-18 16:23:42 +00:00
file_header();
1987-11-25 11:19:38 +00:00
}
1988-09-07 14:47:40 +00:00
table();
exit(nerrors);
1987-11-25 11:19:38 +00:00
}
}