X-Git-Url: http://nsz.repo.hu/git/?a=blobdiff_plain;f=token.c;h=c9b0de859dfae002c8b8d6a37f16139866cd362c;hb=d2bf8b03b1aec5141800a300c62595930023ad79;hp=1fbaafd0410933055172733dd48c3096de58d2ab;hpb=2beaa4f65961fe297663e1cec9e5632b7f3e1cba;p=cparser diff --git a/token.c b/token.c index 1fbaafd..c9b0de8 100644 --- a/token.c +++ b/token.c @@ -28,61 +28,70 @@ #include "symbol.h" #include "lang_features.h" #include "adt/array.h" +#include "adt/util.h" static symbol_t *token_symbols[T_LAST_TOKEN]; static symbol_t *pp_token_symbols[TP_LAST_TOKEN]; -const source_position_t builtin_source_position = { "", 0 }; +const source_position_t builtin_source_position = { "", 0, 0, true }; -void init_tokens(void) +static token_kind_t last_id; + +static symbol_t *intern_register_token(token_kind_t id, const char *string) { - symbol_t *symbol; - int last_id = -2; + assert(id < T_LAST_TOKEN); + symbol_t *symbol = symbol_table_insert(string); + if (token_symbols[id] == NULL) + token_symbols[id] = symbol; + return symbol; +} - memset(token_symbols, 0, T_LAST_TOKEN * sizeof(token_symbols[0])); - memset(pp_token_symbols, 0, TP_LAST_TOKEN * sizeof(pp_token_symbols[0])); +static symbol_t *intern_register_pp_token(preprocessor_token_kind_t id, const char *string) +{ + assert(id < TP_LAST_TOKEN); + symbol_t *symbol = symbol_table_insert(string); + if (pp_token_symbols[id] == NULL) + pp_token_symbols[id] = symbol; + return symbol; +} -#define T(mode,x,str,val) \ - if (T_##x > 255) { \ - assert(T_##x >= last_id); \ - last_id = T_##x; \ - } \ - if (c_mode & (mode)) { \ - assert(T_##x >= 0 && T_##x < T_LAST_TOKEN); \ - symbol = symbol_table_insert(str); \ - symbol->ID = T_##x; \ - if (token_symbols[T_##x] == NULL) \ - token_symbols[T_##x] = symbol; \ +static void register_token(unsigned mode, token_kind_t id, const char *string) +{ + if (id > 255) { + assert(id >= last_id); + last_id = id; + } + if (c_mode & mode) { + symbol_t *symbol = intern_register_token(id, string); + symbol->ID = id; } +} + +static void register_pp_token(unsigned mode, preprocessor_token_kind_t id, + const char *string) +{ + if (! (c_mode & mode)) + return; -#define TS(x,str,val) \ - assert(T_##x >= 0 && T_##x < T_LAST_TOKEN); \ - symbol = symbol_table_insert(str); \ - if (token_symbols[T_##x] == NULL) \ - token_symbols[T_##x] = symbol; \ + symbol_t *symbol = intern_register_pp_token(id, string); + symbol->pp_ID = id; +} -#include "tokens.inc" +void init_tokens(void) +{ + memset(token_symbols, 0, T_LAST_TOKEN * sizeof(token_symbols[0])); + memset(pp_token_symbols, 0, TP_LAST_TOKEN * sizeof(pp_token_symbols[0])); +#define T(mode,x,str,val) register_token(mode, T_##x, str); +#define TS(x,str,val) intern_register_token(T_##x, str); +#include "tokens.inc" #undef TS #undef T -#define T(mode,x,str,val) \ - if (c_mode & (mode)) { \ - assert(TP_##x >= 0 && TP_##x < TP_LAST_TOKEN); \ - symbol = symbol_table_insert(str); \ - symbol->pp_ID = TP_##x; \ - if (pp_token_symbols[TP_##x] == NULL) \ - pp_token_symbols[TP_##x] = symbol; \ - } - -#define TS(x,str,val) \ - assert(TP_##x >= 0 && TP_##x < T_LAST_TOKEN); \ - symbol = symbol_table_insert(str); \ - if (pp_token_symbols[TP_##x] == NULL) \ - pp_token_symbols[TP_##x] = symbol; - +#define T(mode,x,str,val) register_pp_token(mode, TP_##x, str); +#define TS(x,str,val) intern_register_pp_token(TP_##x, str); #include "tokens_preprocessor.inc" - +#undef TS #undef T } @@ -90,40 +99,30 @@ void exit_tokens(void) { } -void print_token_type(FILE *f, token_type_t token_type) +void print_token_kind(FILE *f, token_kind_t token_kind) { - if(token_type == T_EOF) { + if(token_kind == T_EOF) { fputs("end of file", f); return; } - if(token_type == T_ERROR) { - fputs("error", f); - return; - } - int token_symbols_len = T_LAST_TOKEN; - if(token_type < 0 || token_type >= token_symbols_len) { + if (token_kind >= lengthof(token_symbols)) { fputs("invalid token", f); return; } - const symbol_t *symbol = token_symbols[token_type]; + const symbol_t *symbol = token_symbols[token_kind]; if(symbol != NULL) { fputs(symbol->string, f); } else { - if(token_type >= 0 && token_type < 256) { - fputc(token_type, f); + if (token_kind < 256) { + fputc(token_kind, f); return; } fputs("unknown token", f); } } -symbol_t *get_token_symbol(const token_t *token) -{ - return token_symbols[token->type]; -} - static void print_stringrep(const string_t *string, FILE *f) { for (size_t i = 0; i < string->size; ++i) { @@ -133,65 +132,63 @@ static void print_stringrep(const string_t *string, FILE *f) void print_token(FILE *f, const token_t *token) { - switch(token->type) { + switch(token->kind) { case T_IDENTIFIER: - fprintf(f, "identifier '%s'", token->symbol->string); + fprintf(f, "identifier '%s'", token->base.symbol->string); break; case T_INTEGER: - case T_INTEGER_OCTAL: - case T_INTEGER_HEXADECIMAL: case T_FLOATINGPOINT: - case T_FLOATINGPOINT_HEXADECIMAL: - print_token_type(f, (token_type_t)token->type); + print_token_kind(f, (token_kind_t)token->kind); fputs(" '", f); - print_stringrep(&token->literal, f); - if (token->symbol != NULL) - fputs(token->symbol->string, f); + print_stringrep(&token->number.number, f); + if (token->number.suffix.size > 0) + print_stringrep(&token->number.suffix, f); fputc('\'', f); break; case T_WIDE_STRING_LITERAL: case T_STRING_LITERAL: - print_token_type(f, (token_type_t)token->type); - fprintf(f, " \"%s\"", token->literal.begin); + print_token_kind(f, (token_kind_t)token->kind); + fprintf(f, " \"%s\"", token->string.string.begin); break; case T_CHARACTER_CONSTANT: case T_WIDE_CHARACTER_CONSTANT: - print_token_type(f, (token_type_t)token->type); + print_token_kind(f, (token_kind_t)token->kind); fputs(" \'", f); - print_stringrep(&token->literal, f); + print_stringrep(&token->string.string, f); fputs("'", f); break; + default: - fputc('\'', f); - print_token_type(f, (token_type_t)token->type); - fputc('\'', f); + if (token->base.symbol) { + fprintf(f, "'%s'", token->base.symbol->string); + } else { + fputc('\'', f); + print_token_kind(f, (token_kind_t)token->kind); + fputc('\'', f); + } break; } } -void print_pp_token_type(FILE *f, int token_type) +void print_pp_token_kind(FILE *f, int token_kind) { - if (token_type == TP_EOF) { + if (token_kind == TP_EOF) { fputs("end of file", f); return; } - if (token_type == TP_ERROR) { - fputs("error", f); - return; - } int token_symbols_len = TP_LAST_TOKEN; - if (token_type < 0 || token_type >= token_symbols_len) { + if (token_kind < 0 || token_kind >= token_symbols_len) { fputs("invalid token", f); return; } - const symbol_t *symbol = pp_token_symbols[token_type]; + const symbol_t *symbol = pp_token_symbols[token_kind]; if (symbol != NULL) { fputs(symbol->string, f); } else { - if(token_type >= 0 && token_type < 256) { - fputc(token_type, f); + if(token_kind >= 0 && token_kind < 256) { + fputc(token_kind, f); return; } fputs("unknown token", f); @@ -200,18 +197,54 @@ void print_pp_token_type(FILE *f, int token_type) void print_pp_token(FILE *f, const token_t *token) { - switch((preprocessor_token_type_t) token->type) { + switch((preprocessor_token_kind_t) token->kind) { case TP_IDENTIFIER: - fprintf(f, "identifier '%s'", token->symbol->string); + fprintf(f, "identifier '%s'", token->base.symbol->string); break; case TP_NUMBER: - fprintf(f, "number '%s'", token->literal.begin); + fprintf(f, "number '%s'", token->number.number.begin); break; case TP_STRING_LITERAL: - fprintf(f, "string \"%s\"", token->literal.begin); + fprintf(f, "string \"%s\"", token->string.string.begin); break; default: - print_pp_token_type(f, (preprocessor_token_type_t) token->type); + print_pp_token_kind(f, (preprocessor_token_kind_t) token->kind); break; } } + +bool tokens_would_paste(preprocessor_token_kind_t token1, + preprocessor_token_kind_t token2) +{ + char c = token2 < 256 ? (char) token2 : pp_token_symbols[token2]->string[0]; + + switch (token1) { + case '>': return c == '>' || c == '='; + case '<': return c == '<' || c == '=' || c == '%' || c == ':'; + case '+': return c == '+' || c == '='; + case '-': return c == '-' || c == '>'; + case '/': return c == '/' || c == '=' || c == '*'; + case '%': return c == ':' || c == '=' || c == '>'; + case '&': return c == '&' || c == '='; + case '|': return c == '|' || c == '='; + case ':': return c == ':' || c == '>'; + case '*': return c == '*' || c == '='; + case '.': return c == '.' || c == '%' || token2 == TP_NUMBER; + case '#': return c == '#' || c == '%'; + case TP_GREATERGREATER: return c == '='; + case TP_LESSLESS: return c == '='; + case '^': return c == '='; + case '!': return c == '='; + case TP_IDENTIFIER: + return token2 == TP_IDENTIFIER || token2 == TP_NUMBER || + token2 == TP_CHARACTER_CONSTANT || + token2 == TP_WIDE_CHARACTER_CONSTANT || + token2 == TP_WIDE_STRING_LITERAL || + token2 == TP_STRING_LITERAL; /* L */ + case TP_NUMBER: + return token2 == TP_NUMBER || token2 == TP_IDENTIFIER || + token2 == '.' || token2 == '+' || token2 == '-'; + default: + return false; + } +}