1

/*

     2

 *  matiec - a compiler for the programming languages defined in IEC 61131-3

     3

 *

     4

 *  Copyright (C) 2003-2011  Mario de Sousa (msousa@fe.up.pt)

     5

 *

     6

 *  This program is free software: you can redistribute it and/or modify

     7

 *  it under the terms of the GNU General Public License as published by

     8

 *  the Free Software Foundation, either version 3 of the License, or

     9

 *  (at your option) any later version.

    10

 *

    11

 *  This program is distributed in the hope that it will be useful,

    12

 *  but WITHOUT ANY WARRANTY; without even the implied warranty of

    13

 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the

    14

 *  GNU General Public License for more details.

    15

 *

    16

 *  You should have received a copy of the GNU General Public License

    17

 *  along with this program.  If not, see <http://www.gnu.org/licenses/>.

    18

 *

    19

 *

    20

 * This code is made available on the understanding that it will not be

    21

 * used in safety-critical situations without a full and competent review.

    22

 */

    23

    24

/*

    25

 * An IEC 61131-3 compiler.

    26

 *

    27

 * Based on the

    28

 * FINAL DRAFT - IEC 61131-3, 2nd Ed. (2001-12-10)

    29

 *

    30

 */

    31

    32

/*

    33

 * Stage 2

    34

 * =======

    35

 *

    36

 *  This file contains the syntax definition of the textual

    37

 *  languages IL and ST, as well as the textual version of SFC.

    38

 *  The syntax parser, comprising the 2nd stage of the overall 

    39

 *  compiler, is generated by runing bison on this file.

    40

 */

    41

    42

    43

    44

    45

/**********************************************************************/

    46

/**********************************************************************/

    47

/**********************************************************************/

    48

/**********************************************************************/

    49

/*******                                                        *******/

    50

/*******   The following syntax does not have any conflicts.    *******/

    51

/*******                                                        *******/

    52

/*******   P L E A S E    K E E P    I T    T H A T    W A Y !  *******/

    53

/*******   ===================================================  *******/

    54

/*******                                                        *******/

    55

/**********************************************************************/

    56

/**********************************************************************/

    57

/**********************************************************************/

    58

/**********************************************************************/

    59

    60

    61

    62

    63

%{

    64

#include <string.h>	/* required for strdup()  */

    65

    66

    67

/* declare the token parser generated by flex... */

    68

int yylex(void);

    69

    70

/* declare the error handler defined at the end of this file */

    71

void yyerror (const char *error_msg);

    72

    73

/* produce a more verbose parsing error message */

    74

#define YYERROR_VERBOSE

    75

    76

/* Include debuging code.

    77

 * Printing of debug info must then be activated by setting

    78

 * the variable yydebug to 1.

    79

 */

    80

#define YYDEBUG 0

    81

    82

    83

/* file with declaration of absyntax classes... */

    84

#include "../absyntax/absyntax.hh"

    85

    86

/* file with declaration of token constants. Generated by bison! */

    87

#include "iec.y.hh"

    88

    89

/* The interface through which bison and flex interact. */

    90

#include "stage1_2_priv.hh"

    91

    92

    93

/* an ugly hack!!

    94

 * We will probably not need it when we decide

    95

 *  to cut down the abstract syntax down to size.

    96

 *  We keep it as it is until we get to write

    97

 *  stages 3 and 4 of the compiler. Who knows,

    98

 *  we might just find out that we really do need

    99

 *  the abstract syntax tree to stay as it is

   100

 *  afterall!

   101

 */

   102

/* for each element <elem> in list_c * <list>

   103

 * execute the code <code>

   104

 */

   105

#define FOR_EACH_ELEMENT(elem, list, code) {		\

   106

  symbol_c *elem;					\

   107

  for(int i = 0; i < list->n; i++) {			\

   108

    elem = list->elements[i];				\

   109

    code;						\

   110

  }							\

   111

}

   112

   113

   114

   115

/* Macros used to pass the line and column locations when

   116

 * creating a new object for the abstract syntax tree.

   117

 */

   118

#define locloc(foo) foo.first_line, foo.first_column, foo.last_line, foo.last_column

   119

#define   locf(foo) foo.first_line, foo.first_column

   120

#define   locl(foo) foo.last_line,  foo.last_column

   121

   122

   123

   124

/* A macro for printing out internal parser errors... */

   125

#define ERROR error_exit(__FILE__,__LINE__)

   126

/* function defined in main.cc */

   127

extern void error_exit(const char *file_name, int line_no);

   128

   129

   130

   131

/*************************/

   132

/* global variables...   */

   133

/*************************/

   134

/* NOTE: For some strange reason bison ver 2.3 is including these declarations

   135

 *       in the iec.y.hh file, which is in turn included by flex.

   136

 *       We cannot therefore define any variables over here, but merely declare 

   137

 *       their existance (otherwise we get errors when linking the code, since we

   138

 *       would get a new variable defined each time iec.y.hh is included!).

   139

 *       Even though the variables are declared 'extern' over here, they will in

   140

 *       fact be defined towards the end of this same file (i.e. in the prologue)

   141

 */

   142

   143

   144

/* NOTE: These variable are really parameters we would like the stage2__ function to pass

   145

 *       to the yyparse() function. However, the yyparse() function is created automatically

   146

 *       by bison, so we cannot add parameters to this function. The only other

   147

 *       option is to use global variables! yuck!

   148

 */

   149

   150

/* A global flag used to tell the parser if overloaded funtions should be allowed.

   151

 * The IEC 61131-3 standard allows overloaded funtions in the standard library,

   152

 * but disallows them in user code...

   153

 */

   154

extern bool allow_function_overloading;

   155

   156

/* A global flag used to tell the parser whether to include the full variable location

   157

 * when printing out error messages...

   158

 */

   159

extern bool full_token_loc;

   160

   161

/* A pointer to the root of the parsing tree that will be generated 

   162

 * by bison.

   163

 */

   164

extern symbol_c *tree_root;

   165

   166

   167

   168

/************************/

   169

/* forward declarations */

   170

/************************/

   171

/* The functions declared here are defined at the end of this file... */

   172

   173

/* Convert an il_operator_c into an identifier_c */

   174

symbol_c *il_operator_c_2_identifier_c(symbol_c *il_operator);

   175

   176

/* return if current token is a syntax element */

   177

/* ERROR_CHECK_BEGIN */

   178

bool is_current_syntax_token();

   179

/* ERROR_CHECK_END */

   180

   181

/* print an error message */

   182

void print_err_msg(int first_line,

   183

                   int first_column,

   184

                   int last_line,

   185

                   int last_column,

   186

                   const char *additional_error_msg);

   187

%}

   188

   189

   190

   191

   192

// %glr-parser

   193

// %expect-rr 1

   194

   195

   196

   197

   198

%union {

   199

    symbol_c 	*leaf;

   200

    list_c	*list;

   201

    char 	*ID;	/* token value */

   202

}

   203

   204

/*

   205

 TODO: DO we need to define a destructor do free

   206

       memory when recovering from errors, or do the

   207

       class destructors already handle this?

   208

       Following is example on how to define

   209

       detructors, using the syntax:

   210

       %destructor { CODE } SYMBOLS

   211

%union

   212

      {

   213

        char *string;

   214

      }

   215

      %token <string> STRING

   216

      %type  <string> string

   217

      %destructor { free ($$); } STRING string

   218

*/

   219

   220

   221

   222

   223

/*************************************/

   224

/* Prelimenary helpful constructs... */

   225

/*************************************/

   226

/* A token used to identify the very end of the input file

   227

 * after all includes have already been processed.

   228

 *

   229

 * Flex automatically returns the token with value 0

   230

 * at the end of the file. We therefore specify here

   231

 * a token with that exact same value here, so we can use it

   232

 * to detect the very end of the input files.

   233

 */

   234

%token END_OF_INPUT 0

   235

   236

/* A bogus token that, in principle, flex MUST NEVER generate */

   237

/* USE 1:

   238

 * ======

   239

 * This token is currently also being used as the default

   240

 * initialisation value of the token_id member in

   241

 * the symbol_c base class.

   242

 *

   243

 * USE 2

   244

 * =====

   245

 * This token may also be used in the future to remove

   246

 * mysterious reduce/reduce conflicts due to the fact

   247

 * that our grammar may not be LALR(1) but merely LR(1).

   248

 * This means that bison cannot handle it without some

   249

 * caoxing from ourselves. We will then need this token

   250

 * to do the coaxing...

   251

 */

   252

%token BOGUS_TOKEN_ID

   253

   254

%type <leaf>	start

   255

   256

%type <leaf>	any_identifier

   257

   258

%token <ID>	prev_declared_variable_name_token

   259

%token <ID>	prev_declared_direct_variable_token

   260

%token <ID>	prev_declared_fb_name_token

   261

%type <leaf>	prev_declared_variable_name

   262

%type <leaf>	prev_declared_direct_variable

   263

%type <leaf>	prev_declared_fb_name

   264

   265

%token  <ID>	prev_declared_simple_type_name_token

   266

%token  <ID>	prev_declared_subrange_type_name_token

   267

%token  <ID>	prev_declared_enumerated_type_name_token

   268

%token  <ID>	prev_declared_array_type_name_token

   269

%token  <ID>	prev_declared_structure_type_name_token

   270

%token  <ID>	prev_declared_string_type_name_token

   271

   272

%type  <leaf>	prev_declared_simple_type_name

   273

%type  <leaf>	prev_declared_subrange_type_name

   274

%type  <leaf>	prev_declared_enumerated_type_name

   275

%type  <leaf>	prev_declared_array_type_name

   276

%type  <leaf>	prev_declared_structure_type_name

   277

%type  <leaf>	prev_declared_string_type_name

   278

   279

%token <ID>	prev_declared_derived_function_name_token

   280

%token <ID>	prev_declared_derived_function_block_name_token

   281

%token <ID>	prev_declared_program_type_name_token

   282

%type  <leaf>	prev_declared_derived_function_name

   283

%type  <leaf>	prev_declared_derived_function_block_name

   284

%type  <leaf>	prev_declared_program_type_name

   285

   286

   287

   288

   289

/**********************************************************************************/

   290

/* B XXX - Things that are missing from the standard, but should have been there! */

   291

/**********************************************************************************/

   292

   293

/* Pragmas that our compiler will accept.

   294

 * See the comment in iec.flex for why these pragmas exist. 

   295

 */

   296

%token          disable_code_generation_pragma_token

   297

%token          enable_code_generation_pragma_token

   298

%type <leaf>	disable_code_generation_pragma

   299

%type <leaf>	enable_code_generation_pragma

   300

   301

   302

/* All other pragmas that we do not support... */

   303

/* In most stage 4, the text inside the pragmas will simply be copied to the output file.

   304

 * This allows us to insert C code (if using stage 4 generating C code) 

   305

 * inside/interningled with the IEC 61131-3 code!

   306

 */

   307

%token <ID>	pragma_token

   308

%type <leaf>	pragma

   309

   310

/* The joining of all previous pragmas, i.e. any possible pragma */

   311

%type <leaf>	any_pragma

   312

   313

   314

/* Where do these tokens belong?? They are missing from the standard! */

   315

/* NOTE: There are other tokens related to these 'EN' ENO', that are also 

   316

 * missing from the standard. However, their location in the annex B is 

   317

 * relatively obvious, so they have been inserted in what seems to us their 

   318

 * correct place in order to ease understanding of the parser...

   319

 */

   320

%token	EN

   321

%token	ENO

   322

%type <leaf>	en_identifier

   323

%type <leaf>	eno_identifier

   324

   325

   326

   327

   328

/***************************/

   329

/* B 0 - Programming Model */

   330

/***************************/

   331

%type <list>	library

   332

%type <leaf>	library_element_declaration

   333

   334

   335

/*******************************************/

   336

/* B 1.1 - Letters, digits and identifiers */

   337

/*******************************************/

   338

/* Done totally within flex...

   339

  letter

   340

  digit

   341

  octal_digit

   342

  hex_digit

   343

*/

   344

%token <ID>	identifier_token

   345

%type  <leaf>	identifier

   346

   347

/*********************/

   348

/* B 1.2 - Constants */

   349

/*********************/

   350

%type <leaf>	constant

   351

%type <leaf>	non_negative_constant

   352

   353

/******************************/

   354

/* B 1.2.1 - Numeric Literals */

   355

/******************************/

   356

/* Done totally within flex...

   357

  bit

   358

*/

   359

%type  <leaf> numeric_literal

   360

%type  <leaf> integer_literal

   361

%type  <leaf> signed_integer

   362

%token <ID>   integer_token

   363

%type  <leaf> integer

   364

%token <ID>   binary_integer_token

   365

%type  <leaf> binary_integer

   366

%token <ID>   octal_integer_token

   367

%type  <leaf> octal_integer

   368

%token <ID>   hex_integer_token

   369

%type  <leaf> hex_integer

   370

%token <ID>   real_token

   371

%type  <leaf> real

   372

%type  <leaf> signed_real

   373

%type  <leaf> real_literal

   374

// %type  <leaf> exponent

   375

%type  <leaf> bit_string_literal

   376

%type  <leaf> boolean_literal

   377

   378

%token safeboolean_true_literal_token

   379

%token safeboolean_false_literal_token

   380

%token boolean_true_literal_token

   381

%token boolean_false_literal_token

   382

   383

%token FALSE

   384

%token TRUE

   385

   386

   387

/*******************************/

   388

/* B 1.2.2 - Character Strings */

   389

/*******************************/

   390

%token <ID>   single_byte_character_string_token

   391

%token <ID>   double_byte_character_string_token

   392

   393

%type  <leaf> character_string

   394

%type  <leaf> single_byte_character_string

   395

%type  <leaf> double_byte_character_string

   396

   397

   398

/***************************/

   399

/* B 1.2.3 - Time Literals */

   400

/***************************/

   401

%type  <leaf> time_literal

   402

   403

   404

/************************/

   405

/* B 1.2.3.1 - Duration */

   406

/************************/

   407

%type  <leaf>	duration

   408

%type  <leaf>	interval

   409

%type  <leaf>	days

   410

%type  <leaf>	fixed_point

   411

%type  <leaf>	hours

   412

%type  <leaf>	minutes

   413

%type  <leaf>	seconds

   414

%type  <leaf>	milliseconds

   415

   416

%type  <leaf>	integer_d

   417

%type  <leaf>	integer_h

   418

%type  <leaf>	integer_m

   419

%type  <leaf>	integer_s

   420

%type  <leaf>	integer_ms

   421

%type  <leaf>	fixed_point_d

   422

%type  <leaf>	fixed_point_h

   423

%type  <leaf>	fixed_point_m

   424

%type  <leaf>	fixed_point_s

   425

%type  <leaf>	fixed_point_ms

   426

   427

%token <ID>	fixed_point_token

   428

%token <ID>	fixed_point_d_token

   429

%token <ID>	integer_d_token

   430

%token <ID>	fixed_point_h_token

   431

%token <ID>	integer_h_token

   432

%token <ID>	fixed_point_m_token

   433

%token <ID>	integer_m_token

   434

%token <ID>	fixed_point_s_token

   435

%token <ID>	integer_s_token

   436

%token <ID>	fixed_point_ms_token

   437

%token <ID>	integer_ms_token

   438

   439

// %token TIME

   440

%token T_SHARP

   441

   442

   443

/************************************/

   444

/* B 1.2.3.2 - Time of day and Date */

   445

/************************************/

   446

%type  <leaf>	time_of_day

   447

%type  <leaf>	daytime

   448

%type  <leaf>	day_hour

   449

%type  <leaf>	day_minute

   450

%type  <leaf>	day_second

   451

%type  <leaf>	date

   452

%type  <leaf>	date_literal

   453

%type  <leaf>	year

   454

%type  <leaf>	month

   455

%type  <leaf>	day

   456

%type  <leaf>	date_and_time

   457

   458

// %token TIME_OF_DAY

   459

// %token DATE

   460

%token D_SHARP

   461

// %token DATE_AND_TIME

   462

   463

   464

/**********************/

   465

/* B 1.3 - Data Types */

   466

/**********************/

   467

/* Strangely, the following symbol does seem to be required! */

   468

// %type  <leaf> data_type_name

   469

%type  <leaf> non_generic_type_name

   470

   471

   472

/***********************************/

   473

/* B 1.3.1 - Elementary Data Types */

   474

/***********************************/

   475

/* NOTES:

   476

 *

   477

 *    - To make the definition of bit_string_literal more

   478

 *      concise, it is useful to use an extra non-terminal

   479

 *      symbol (i.e. a grouping or construct) that groups the

   480

 *      following elements (BYTE, WORD, DWORD, LWORD).

   481

 *      Note that the definition of bit_string_type_name

   482

 *      (according to the spec) includes the above elements

   483

 *      and an extra BOOL.

   484

 *      We could use an extra construct with the first four

   485

 *      elements to be used solely in the definition of

   486

 *      bit_string_literal, but with the objective of not

   487

 *      having to replicate the actions (if we ever need

   488

 *      to change them, they would need to be changed in both

   489

 *      bit_string_type_name and the extra grouping), we

   490

 *      have re-defined bit_string_type_name as only including

   491

 *      the first four elements.

   492

 *      In order to have our parser implement the specification

   493

 *      correctly we have augmented every occurence of

   494

 *      bit_string_type_name in other rules with the BOOL

   495

 *      token. Since bit_string_type_name only appears in

   496

 *      the rule for elementary_type_name, this does not

   497

 *      seem to be a big concession to make!

   498

 *

   499

 *    - We have added a helper symbol to concentrate the

   500

 *      instantiation of STRING and WSTRING into a single

   501

 *      location (elementary_string_type_name).

   502

 *      These two elements show up in several other rules,

   503

 *      but we want to create the equivalent abstract syntax

   504

 *      in a single location of this file, in order to make

   505

 *      possible future changes easier to edit...

   506

 */

   507

%type  <leaf>	elementary_type_name

   508

%type  <leaf>	numeric_type_name

   509

%type  <leaf>	integer_type_name

   510

%type  <leaf>	signed_integer_type_name

   511

%type  <leaf>	unsigned_integer_type_name

   512

%type  <leaf>	real_type_name

   513

%type  <leaf>	date_type_name