gmqcc/lexer.h

254 lines
7.4 KiB
C
Raw Normal View History

2012-07-16 11:59:10 +00:00
#ifndef GMQCC_LEXER_HDR_
#define GMQCC_LEXER_HDR_
typedef struct token_s token;
#include "ast.h"
struct token_s {
int ttype;
char *value;
2012-07-16 11:59:10 +00:00
union {
vector v;
int i;
double f;
int t; /* type */
} constval;
#if 0
2012-07-16 11:59:10 +00:00
struct token_s *next;
struct token_s *prev;
#endif
2012-07-16 11:59:10 +00:00
lex_ctx ctx;
};
#if 0
2012-07-16 11:59:10 +00:00
token* token_new();
void token_delete(token*);
token* token_copy(const token *cp);
void token_delete_all(token *t);
token* token_copy_all(const token *cp);
#endif
2012-07-16 11:59:10 +00:00
/* Lexer
*
*/
enum {
/* Other tokens which we can return: */
TOKEN_NONE = 0,
TOKEN_START = 128,
TOKEN_IDENT,
TOKEN_TYPENAME,
TOKEN_OPERATOR,
TOKEN_KEYWORD, /* loop */
TOKEN_DOTS, /* 3 dots, ... */
2012-07-16 11:59:10 +00:00
TOKEN_STRINGCONST, /* not the typename but an actual "string" */
TOKEN_CHARCONST,
TOKEN_VECTORCONST,
TOKEN_INTCONST,
TOKEN_FLOATCONST,
2012-11-01 22:22:58 +00:00
TOKEN_WHITE,
TOKEN_EOL,
2012-07-16 11:59:10 +00:00
TOKEN_EOF,
/* We use '< TOKEN_ERROR', so TOKEN_FATAL must come after it and any
* other error related tokens as well
*/
TOKEN_ERROR,
TOKEN_FATAL /* internal error, eg out of memory */
};
typedef struct {
char *name;
int value;
} frame_macro;
2012-07-16 11:59:10 +00:00
typedef struct {
FILE *file;
2012-11-11 09:27:09 +00:00
const char *open_string;
size_t open_string_length;
size_t open_string_pos;
2012-07-16 11:59:10 +00:00
char *name;
size_t line;
size_t sline; /* line at the start of a token */
char peek[256];
size_t peekpos;
bool eof;
token tok; /* not a pointer anymore */
2012-07-16 11:59:10 +00:00
struct {
bool noops;
bool nodigraphs; /* used when lexing string constants */
2012-11-01 22:22:58 +00:00
bool preprocessing; /* whitespace and EOLs become actual tokens */
bool mergelines; /* backslash at the end of a line escapes the newline */
2012-07-16 11:59:10 +00:00
} flags;
int framevalue;
frame_macro *frames;
char *modelname;
size_t push_line;
2012-07-16 11:59:10 +00:00
} lex_file;
lex_file* lex_open (const char *file);
2012-11-11 09:27:09 +00:00
lex_file* lex_open_string(const char *str, size_t len, const char *name);
2012-07-16 11:59:10 +00:00
void lex_close(lex_file *lex);
int lex_do (lex_file *lex);
void lex_cleanup(void);
2012-07-16 11:59:10 +00:00
/* Parser
*
*/
enum {
ASSOC_LEFT,
ASSOC_RIGHT
};
#define OP_SUFFIX 1
#define OP_PREFIX 2
typedef struct {
const char *op;
unsigned int operands;
2012-07-20 12:04:52 +00:00
unsigned int id;
2012-07-16 11:59:10 +00:00
unsigned int assoc;
unsigned int prec;
unsigned int flags;
} oper_info;
#define opid1(a) (a)
#define opid2(a,b) ((a<<8)|b)
#define opid3(a,b,c) ((a<<16)|(b<<8)|c)
static const oper_info c_operators[] = {
{ "(", 0, opid1('('), ASSOC_LEFT, 99, OP_PREFIX}, /* paren expression - non function call */
{ "++", 1, opid3('S','+','+'), ASSOC_LEFT, 16, OP_SUFFIX},
{ "--", 1, opid3('S','-','-'), ASSOC_LEFT, 16, OP_SUFFIX},
2012-07-16 11:59:10 +00:00
{ ".", 2, opid1('.'), ASSOC_LEFT, 15, 0 },
{ "(", 0, opid1('('), ASSOC_LEFT, 15, 0 }, /* function call */
{ "[", 2, opid1('['), ASSOC_LEFT, 15, 0 }, /* array subscript */
2012-07-16 11:59:10 +00:00
2012-08-16 12:24:12 +00:00
{ "!", 1, opid2('!', 'P'), ASSOC_RIGHT, 14, OP_PREFIX },
{ "~", 1, opid2('~', 'P'), ASSOC_RIGHT, 14, OP_PREFIX },
{ "+", 1, opid2('+','P'), ASSOC_RIGHT, 14, OP_PREFIX },
{ "-", 1, opid2('-','P'), ASSOC_RIGHT, 14, OP_PREFIX },
{ "++", 1, opid3('+','+','P'), ASSOC_RIGHT, 14, OP_PREFIX },
{ "--", 1, opid3('-','-','P'), ASSOC_RIGHT, 14, OP_PREFIX },
/* { "&", 1, opid2('&','P'), ASSOC_RIGHT, 14, OP_PREFIX }, */
2012-07-16 11:59:10 +00:00
{ "*", 2, opid1('*'), ASSOC_LEFT, 13, 0 },
{ "/", 2, opid1('/'), ASSOC_LEFT, 13, 0 },
{ "%", 2, opid1('%'), ASSOC_LEFT, 13, 0 },
2012-07-16 11:59:10 +00:00
{ "+", 2, opid1('+'), ASSOC_LEFT, 12, 0 },
{ "-", 2, opid1('-'), ASSOC_LEFT, 12, 0 },
2012-07-16 11:59:10 +00:00
{ "<<", 2, opid2('<','<'), ASSOC_LEFT, 11, 0 },
{ ">>", 2, opid2('>','>'), ASSOC_LEFT, 11, 0 },
2012-07-16 11:59:10 +00:00
{ "<", 2, opid1('<'), ASSOC_LEFT, 10, 0 },
{ ">", 2, opid1('>'), ASSOC_LEFT, 10, 0 },
{ "<=", 2, opid2('<','='), ASSOC_LEFT, 10, 0 },
{ ">=", 2, opid2('>','='), ASSOC_LEFT, 10, 0 },
2012-07-16 11:59:10 +00:00
{ "==", 2, opid2('=','='), ASSOC_LEFT, 9, 0 },
{ "!=", 2, opid2('!','='), ASSOC_LEFT, 9, 0 },
2012-07-16 11:59:10 +00:00
{ "&", 2, opid1('&'), ASSOC_LEFT, 8, 0 },
2012-07-16 11:59:10 +00:00
{ "^", 2, opid1('^'), ASSOC_LEFT, 7, 0 },
2012-07-16 11:59:10 +00:00
{ "|", 2, opid1('|'), ASSOC_LEFT, 6, 0 },
2012-07-16 11:59:10 +00:00
{ "&&", 2, opid2('&','&'), ASSOC_LEFT, 5, 0 },
2012-07-16 11:59:10 +00:00
{ "||", 2, opid2('|','|'), ASSOC_LEFT, 4, 0 },
2012-07-16 11:59:10 +00:00
{ "?", 3, opid2('?',':'), ASSOC_RIGHT, 3, 0 },
2012-11-21 19:36:42 +00:00
{ ":", 3, opid2(':','?'), ASSOC_RIGHT, 3, 0 },
2012-07-16 11:59:10 +00:00
{ "=", 2, opid1('='), ASSOC_RIGHT, 2, 0 },
{ "+=", 2, opid2('+','='), ASSOC_RIGHT, 2, 0 },
{ "-=", 2, opid2('-','='), ASSOC_RIGHT, 2, 0 },
{ "*=", 2, opid2('*','='), ASSOC_RIGHT, 2, 0 },
{ "/=", 2, opid2('/','='), ASSOC_RIGHT, 2, 0 },
{ "%=", 2, opid2('%','='), ASSOC_RIGHT, 2, 0 },
{ ">>=", 2, opid3('>','>','='), ASSOC_RIGHT, 2, 0 },
{ "<<=", 2, opid3('<','<','='), ASSOC_RIGHT, 2, 0 },
{ "&=", 2, opid2('&','='), ASSOC_RIGHT, 2, 0 },
{ "^=", 2, opid2('^','='), ASSOC_RIGHT, 2, 0 },
{ "|=", 2, opid2('|','='), ASSOC_RIGHT, 2, 0 },
2012-07-27 10:53:15 +00:00
{ ",", 2, opid1(','), ASSOC_LEFT, 1, 0 }
2012-07-16 11:59:10 +00:00
};
static const size_t c_operator_count = (sizeof(c_operators) / sizeof(c_operators[0]));
static const oper_info qcc_operators[] = {
{ "(", 0, opid1('('), ASSOC_LEFT, 99, OP_PREFIX}, /* paren expression - non function call */
{ ".", 2, opid1('.'), ASSOC_LEFT, 15, 0 },
{ "(", 0, opid1('('), ASSOC_LEFT, 15, 0 }, /* function call */
{ "[", 2, opid1('['), ASSOC_LEFT, 15, 0 }, /* array subscript */
{ "!", 1, opid2('!', 'P'), ASSOC_RIGHT, 14, OP_PREFIX },
{ "+", 1, opid2('+','P'), ASSOC_RIGHT, 14, OP_PREFIX },
{ "-", 1, opid2('-','P'), ASSOC_RIGHT, 14, OP_PREFIX },
{ "*", 2, opid1('*'), ASSOC_LEFT, 13, 0 },
{ "/", 2, opid1('/'), ASSOC_LEFT, 13, 0 },
{ "&", 2, opid1('&'), ASSOC_LEFT, 13, 0 },
{ "|", 2, opid1('|'), ASSOC_LEFT, 13, 0 },
{ "+", 2, opid1('+'), ASSOC_LEFT, 12, 0 },
{ "-", 2, opid1('-'), ASSOC_LEFT, 12, 0 },
{ "<", 2, opid1('<'), ASSOC_LEFT, 10, 0 },
{ ">", 2, opid1('>'), ASSOC_LEFT, 10, 0 },
{ "<=", 2, opid2('<','='), ASSOC_LEFT, 10, 0 },
{ ">=", 2, opid2('>','='), ASSOC_LEFT, 10, 0 },
{ "==", 2, opid2('=','='), ASSOC_LEFT, 10, 0 },
{ "!=", 2, opid2('!','='), ASSOC_LEFT, 10, 0 },
{ "=", 2, opid1('='), ASSOC_RIGHT, 8, 0 },
{ "+=", 2, opid2('+','='), ASSOC_RIGHT, 8, 0 },
{ "-=", 2, opid2('-','='), ASSOC_RIGHT, 8, 0 },
{ "*=", 2, opid2('*','='), ASSOC_RIGHT, 8, 0 },
{ "/=", 2, opid2('/','='), ASSOC_RIGHT, 8, 0 },
{ "%=", 2, opid2('%','='), ASSOC_RIGHT, 8, 0 },
{ "&=", 2, opid2('&','='), ASSOC_RIGHT, 8, 0 },
{ "|=", 2, opid2('|','='), ASSOC_RIGHT, 8, 0 },
{ "&&", 2, opid2('&','&'), ASSOC_LEFT, 5, 0 },
{ "||", 2, opid2('|','|'), ASSOC_LEFT, 5, 0 },
2012-11-21 19:36:42 +00:00
{ ",", 2, opid1(','), ASSOC_LEFT, 2, 0 },
{ "?", 3, opid2('?',':'), ASSOC_RIGHT, 1, 0 },
2012-11-21 19:36:42 +00:00
{ ":", 3, opid2(':','?'), ASSOC_RIGHT, 1, 0 }
};
static const size_t qcc_operator_count = (sizeof(qcc_operators) / sizeof(qcc_operators[0]));
extern const oper_info *operators;
extern size_t operator_count;
2012-07-16 11:59:10 +00:00
void lexerror(lex_file*, const char *fmt, ...);
#endif