X-Git-Url: http://nsz.repo.hu/git/?a=blobdiff_plain;f=token.c;h=f278a6f55cd884a87d2489b71f44a7e6bb4e6e6e;hb=8845feb009950d2c76a394660a5bd81a0be0f2a6;hp=c2d5345f0922bb8ee34d90530bdc8883a9cda8c4;hpb=0e0faae013a4302a0cc4f6e465301a82483e0a52;p=cparser diff --git a/token.c b/token.c index c2d5345..f278a6f 100644 --- a/token.c +++ b/token.c @@ -1,54 +1,71 @@ +/* + * This file is part of cparser. + * Copyright (C) 2012 Matthias Braun + */ #include #include "token_t.h" +#include "symbol_t.h" #include #include +#include #include "symbol.h" #include "lang_features.h" #include "adt/array.h" +#include "adt/error.h" +#include "adt/util.h" -static symbol_t *token_symbols[T_LAST_TOKEN]; +symbol_t *token_symbols[T_LAST_TOKEN]; -source_position_t builtin_source_position = { "", 0 }; +const position_t builtin_position = { "", 0, 0, true }; -void init_tokens(void) -{ - symbol_t *symbol; - int last_id = -2; +static token_kind_t last_id; - memset(token_symbols, 0, T_LAST_TOKEN * sizeof(token_symbols[0])); +static symbol_t *intern_register_token(token_kind_t id, const char *string) +{ + assert(id < T_LAST_TOKEN); + symbol_t *symbol = symbol_table_insert(string); + if (token_symbols[id] == NULL) + 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; \ - 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; + } +} -#define TS(x,str,val) \ - assert(T_##x >= 0 && T_##x < T_LAST_TOKEN); \ - symbol = symbol_table_insert(str); \ - token_symbols[T_##x] = symbol; +static void register_pp_token(pp_token_kind_t const id, char const *const string) +{ + assert(id < TP_LAST_TOKEN); + symbol_t *const symbol = symbol_table_insert(string); + symbol->pp_ID = id; +} -#include "tokens.inc" +void init_tokens(void) +{ + static bool tokens_initialized = false; + if (tokens_initialized) + return; + tokens_initialized = true; -#undef TS -#undef T + memset(token_symbols, 0, T_LAST_TOKEN * sizeof(token_symbols[0])); -#define T(mode,x,str,val) \ - assert(TP_##x >= 0 && TP_##x < TP_LAST_TOKEN); \ - symbol = symbol_table_insert(str); \ - symbol->pp_ID = TP_##x; +#define T(mode,x,str,val) register_token(mode, x, str); +#include "tokens.inc" +#undef T +#define T(token) register_pp_token(TP_##token, #token); #include "tokens_preprocessor.inc" - #undef T } @@ -56,52 +73,92 @@ 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) { - 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]; - if(symbol != NULL) { - fprintf(f, "'%s'", symbol->string); - } else { - if(token_type >= 0 && token_type < 256) { - fprintf(f, "'%c'", token_type); - return; - } - fputs("unknown token", f); + fputs(token_symbols[token_kind]->string, f); +} + +char const *get_string_encoding_prefix(string_encoding_t const enc) +{ + switch (enc) { + case STRING_ENCODING_CHAR: return ""; + case STRING_ENCODING_CHAR16: return "u"; + case STRING_ENCODING_CHAR32: return "U"; + case STRING_ENCODING_UTF8: return "u8"; + case STRING_ENCODING_WIDE: return "L"; } + panic("invalid string encoding"); } void print_token(FILE *f, const token_t *token) { - switch(token->type) { + char delim = '\''; + char const *enc = ""; + char const *val; + switch (token->kind) { case T_IDENTIFIER: - fprintf(f, "symbol '%s'", token->v.symbol->string); - break; - case T_INTEGER: - fprintf(f, "integer number %lld", token->v.intvalue); - break; - case T_FLOATINGPOINT: - fprintf(f, "floatingpointer number %LF", token->v.floatvalue); + case T_MACRO_PARAMETER: + case T_UNKNOWN_CHAR: + val = token->base.symbol->string; break; + case T_STRING_LITERAL: - fprintf(f, "string '%s'", token->v.string.begin); /* TODO suboptimal */ + delim = '"'; + /* FALLTHROUGH */ + case T_CHARACTER_CONSTANT: + enc = get_string_encoding_prefix(token->literal.string.encoding); + /* FALLTHROUGH */ + case T_NUMBER: + val = token->literal.string.begin; break; + + default: { + char const *kind = (token->base.symbol ? token->base.symbol : token_symbols[token->kind])->string; + fprintf(f, "'%s'", kind); + return; + } + } + fprintf(f, "%s %s%c%s%c", token_symbols[token->kind]->string, enc, delim, val, delim); +} + +bool tokens_would_paste(token_kind_t token1, token_kind_t token2) +{ + char const c = 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 == T_NUMBER; + case '#': return c == '#' || c == '%'; + case T_GREATERGREATER: return c == '='; + case T_LESSLESS: return c == '='; + case '^': return c == '='; + case '!': return c == '='; + + case T_IDENTIFIER: + return token2 == T_CHARACTER_CONSTANT || + token2 == T_IDENTIFIER || + token2 == T_NUMBER || + token2 == T_STRING_LITERAL; /* L */ + + case T_NUMBER: + return token2 == T_IDENTIFIER || token2 == T_NUMBER || + token2 == '.' || token2 == '+' || token2 == '-'; + default: - print_token_type(f, (token_type_t)token->type); - break; + return false; } }