zzy/scc
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat add func call and rewrite codes

Browse Source
This commit is contained in:
ZZY
2025-03-07 12:29:53 +08:00
parent 09299e339c
commit 95bf44eb3f
37 changed files with 3369 additions and 1063 deletions
Download Patch File Download Diff File Expand all files Collapse all files

76
ccompiler/frontend/lexer/lexer.c
View File

@@ -26,13 +26,15 @@ the distribution and installation instructions.
Chris Fraser / cwf@aya.yale.edu
David Hanson / drh@drhanson.net
*/
#define FRONTEND_IMPLEMENTATION
#include "../frontend.h"
#include "token.h"
#include "lexer.h"
static const struct {
const char* name;
enum CSTD_KEYWORD std_type;
enum TokenType tok;
tok_type_t tok;
} keywords[] = {
#define X(name, std_type, tok, ...) { #name, std_type, tok },
KEYWORD_TABLE
@@ -72,7 +74,7 @@ static inline int keyword_cmp(const char* name, int len) {
return -1; // Not a keyword.
}
void init_lexer(struct Lexer* lexer, const char* file_name, void* stream, lexer_sread_fn sread)
void init_lexer(lexer_t* lexer, const char* file_name, void* stream, lexer_sread_fn sread)
{
lexer->cur_ptr = lexer->end_ptr = (unsigned char*)&(lexer->buffer);
lexer->index = 1;
@@ -86,12 +88,12 @@ void init_lexer(struct Lexer* lexer, const char* file_name, void* stream, lexer_
}
}
static void flush_buffer(struct Lexer* lexer) {
static void flush_buffer(lexer_t* lexer) {
int num = lexer->end_ptr - lexer->cur_ptr;
for (int i = 0; i < num; i++) {
lexer->buffer[i] = lexer->cur_ptr[i];
}
lexer->cur_ptr = lexer->buffer;
lexer->cur_ptr = (unsigned char*)lexer->buffer;
int read_size = LEXER_BUFFER_SIZE - num;
// TODO size_t to int maybe lose precision
@@ -109,7 +111,7 @@ static void flush_buffer(struct Lexer* lexer) {
}
}
static void goto_newline(struct Lexer* lexer) {
static void goto_newline(lexer_t* lexer) {
do {
if (lexer->cur_ptr == lexer->end_ptr) {
flush_buffer(lexer);
@@ -119,7 +121,7 @@ static void goto_newline(struct Lexer* lexer) {
} while (*lexer->cur_ptr != '\n' && *lexer->cur_ptr != '\0');
}
static void goto_block_comment(struct Lexer* lexer) {
static void goto_block_comment(lexer_t* lexer) {
while (1) {
if (lexer->end_ptr - lexer->cur_ptr < 2) {
flush_buffer(lexer);
@@ -155,7 +157,7 @@ static char got_slash(unsigned char* peek) {
}
}
static void parse_char_literal(struct Lexer* lexer, struct Token* token) {
static void parse_char_literal(lexer_t* lexer, tok_t* token) {
char val = 0;
unsigned char* peek = lexer->cur_ptr + 1;
if (*peek == '\\') {
@@ -166,16 +168,16 @@ static void parse_char_literal(struct Lexer* lexer, struct Token* token) {
}
if (*peek != '\'') error("Unclosed character literal");
token->constant.ch = val;
token->val.ch = val;
lexer->cur_ptr = peek + 1;
token->constant.have = 1;
token->val.have = 1;
token->type = TOKEN_CHAR_LITERAL;
}
static void parse_string_literal(struct Lexer* lexer, struct Token* token) {
static void parse_string_literal(lexer_t* lexer, tok_t* token) {
unsigned char* peek = lexer->cur_ptr + 1;
// TODO string literal size check
char* dest = token->constant.str = xmalloc(LEXER_MAX_TOKEN_SIZE + 1);
char* dest = token->val.str = xmalloc(LEXER_MAX_TOKEN_SIZE + 1);
int len = 0;
while (*peek != '"') {
@@ -191,12 +193,12 @@ static void parse_string_literal(struct Lexer* lexer, struct Token* token) {
}
dest[len] = '\0';
lexer->cur_ptr = peek + 1;
token->constant.have = 1;
token->val.have = 1;
token->type = TOKEN_STRING_LITERAL;
}
// FIXME it write by AI maybe error
static void parse_number(struct Lexer* lexer, struct Token* token) {
static void parse_number(lexer_t* lexer, tok_t* token) {
unsigned char* peek = lexer->cur_ptr;
int base = 10;
int is_float = 0;
@@ -255,12 +257,12 @@ static void parse_number(struct Lexer* lexer, struct Token* token) {
if ((*peek == 'e' || *peek == 'E') && base == 10) {
is_float = 1;
peek++;
int exp_sign = 1;
// int exp_sign = 1;
int exponent = 0;
if (*peek == '+') peek++;
else if (*peek == '-') {
exp_sign = -1;
// exp_sign = -1;
peek++;
}
@@ -273,19 +275,19 @@ static void parse_number(struct Lexer* lexer, struct Token* token) {
// 存储结果
lexer->cur_ptr = peek;
token->constant.have = 1;
token->val.have = 1;
if (is_float) {
token->constant.d = float_val;
token->val.d = float_val;
token->type = TOKEN_FLOAT_LITERAL;
} else {
token->constant.ll = int_val;
token->val.ll = int_val;
token->type = TOKEN_INT_LITERAL;
}
}
#define GOT_ONE_TOKEN_BUF_SIZE 64
// /zh/c/language/operator_arithmetic.html
void get_token(struct Lexer* lexer, struct Token* token) {
void get_token(lexer_t* lexer, tok_t* token) {
// 需要保证缓冲区始终可读
if (lexer->end_ptr - lexer->cur_ptr < GOT_ONE_TOKEN_BUF_SIZE) {
flush_buffer(lexer);
@@ -305,8 +307,8 @@ void get_token(struct Lexer* lexer, struct Token* token) {
token->type = TOKEN_FLUSH;
}
enum TokenType tok = TOKEN_INIT;
struct TokenConstant constant;
tok_type_t tok = TOKEN_INIT;
tok_val_t constant;
constant.have = 0;
// once step
@@ -392,7 +394,7 @@ void get_token(struct Lexer* lexer, struct Token* token) {
switch (*peek++) {
case '=': tok = TOKEN_NEQ; break;
default: peek--, tok = TOKEN_NOT; break;
}
} break;
case '[':
tok = TOKEN_L_BRACKET; break;
case ']':
@@ -454,7 +456,7 @@ void get_token(struct Lexer* lexer, struct Token* token) {
case 'S': case 'T': case 'U': case 'V': case 'W': case 'X':case 'Y': case 'Z':
case '_':
// TOKEN_IDENT
if (*peek == 'L' && *peek == '\'' || *peek == 'L' && *peek == '"') {
if ((*peek == 'L' && *peek == '\'') || (*peek == 'L' && *peek == '"')) {
error("unsupport wide-character char literal by `L` format");
}
while (1) {
@@ -469,18 +471,18 @@ void get_token(struct Lexer* lexer, struct Token* token) {
break;
}
int res = keyword_cmp(lexer->cur_ptr, peek - (lexer->cur_ptr));
int res = keyword_cmp((const char*)lexer->cur_ptr, peek - (lexer->cur_ptr));
if (res == -1) {
int strlen = peek - lexer->cur_ptr;
unsigned char* str = xmalloc(strlen + 1);
constant.have = 1;
constant.str = str;
constant.str = (char*)str;
for (int i = 0; i < strlen; i++) {
str[i] = lexer->cur_ptr[i];
}
str[strlen] = '\0';
constant.have = 1;
constant.str = str;
constant.str = (char*)str;
tok = TOKEN_IDENT; break;
} else {
tok = keywords[res].tok; break;
@@ -492,32 +494,16 @@ void get_token(struct Lexer* lexer, struct Token* token) {
lexer->cur_ptr = peek;
END:
token->constant = constant;
token->val = constant;
token->type = tok;
}
// get_token maybe got invalid (with parser)
void get_valid_token(struct Lexer* lexer, struct Token* token) {
enum TokenType type;
void get_valid_token(lexer_t* lexer, tok_t* token) {
tok_type_t type;
do {
get_token(lexer, token);
type = token->type;
} while (type == TOKEN_FLUSH || type == TOKEN_LINE_COMMENT || type == TOKEN_BLOCK_COMMENT);
}
// 生成字符串映射（根据需求选择#str或#name）
static const char* token_strings[] = {
// 普通token使用#str
#define X(str, tok) [tok] = #str,
TOKEN_TABLE
#undef X
// 关键字使用#name
#define X(name, std, tok) [tok] = #name,
KEYWORD_TABLE
#undef X
};
const char* get_token_name(enum TokenType type) {
return token_strings[type];
}