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

feat 重构stream流API并适配lex_parse和lexer

Browse Source
This commit is contained in:
zzy
2025-12-08 23:04:11 +08:00
parent 1ab07a5815
commit 36bff64a91
17 changed files with 402 additions and 244 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
libs/lex_parser/include/lex_parser.h
View File

@@ -11,15 +11,16 @@ static inline cbool lex_parse_is_whitespace(int ch) {
return ch == ' ' || ch == '\t';
}
int lex_parse_char(core_stream_t *input, core_pos_t *pos);
cbool lex_parse_string(core_stream_t *input, core_pos_t *pos,
int lex_parse_char(core_probe_stream_t *input, core_pos_t *pos);
cbool lex_parse_string(core_probe_stream_t *input, core_pos_t *pos,
cstring_t *output);
cbool lex_parse_number(core_stream_t *input, core_pos_t *pos, usize *output);
cbool lex_parse_identifier(core_stream_t *input, core_pos_t *pos,
cbool lex_parse_number(core_probe_stream_t *input, core_pos_t *pos,
usize *output);
cbool lex_parse_identifier(core_probe_stream_t *input, core_pos_t *pos,
cstring_t *output);
void lex_parse_skip_endline(core_stream_t *input, core_pos_t *pos);
void lex_parse_skip_block_comment(core_stream_t *input, core_pos_t *pos);
void lex_parse_skip_line(core_stream_t *input, core_pos_t *pos);
void lex_parse_skip_whitespace(core_stream_t *input, core_pos_t *pos);
void lex_parse_skip_endline(core_probe_stream_t *input, core_pos_t *pos);
void lex_parse_skip_block_comment(core_probe_stream_t *input, core_pos_t *pos);
void lex_parse_skip_line(core_probe_stream_t *input, core_pos_t *pos);
void lex_parse_skip_whitespace(core_probe_stream_t *input, core_pos_t *pos);
#endif /* __SMCC_LEX_PARSER_H__ */

174
libs/lex_parser/src/lex_parser.c
View File

@@ -1,18 +1,18 @@
#include <lex_parser.h>
void lex_parse_skip_endline(core_stream_t *input, core_pos_t *pos) {
void lex_parse_skip_endline(core_probe_stream_t *input, core_pos_t *pos) {
Assert(input != null && pos != null);
core_stream_reset_char(input);
int ch = core_stream_peek_char(input);
core_probe_stream_reset(input);
int ch = core_probe_stream_peek(input);
if (ch == '\r') {
core_stream_next_char(input);
ch = core_stream_peek_char(input);
core_probe_stream_consume(input);
ch = core_probe_stream_peek(input);
if (ch == '\n') {
core_stream_next_char(input);
core_probe_stream_consume(input);
}
core_pos_next_line(pos);
} else if (ch == '\n') {
core_stream_next_char(input);
core_probe_stream_consume(input);
core_pos_next_line(pos);
} else {
LOG_WARN("not a newline character");
@@ -57,12 +57,12 @@ static inline int got_simple_escape(int ch) {
/* clang-format on */
}
void lex_parse_skip_line(core_stream_t *input, core_pos_t *pos) {
core_stream_t *stream = input;
void lex_parse_skip_line(core_probe_stream_t *input, core_pos_t *pos) {
core_probe_stream_t *stream = input;
Assert(stream != null && pos != null);
core_stream_reset_char(stream);
core_probe_stream_reset(stream);
while (1) {
int ch = core_stream_peek_char(stream);
int ch = core_probe_stream_peek(stream);
if (ch == core_stream_eof) {
return;
@@ -73,29 +73,29 @@ void lex_parse_skip_line(core_stream_t *input, core_pos_t *pos) {
lex_parse_skip_endline(stream, pos);
return;
} else {
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
}
}
}
void lex_parse_skip_block_comment(core_stream_t *input, core_pos_t *pos) {
core_stream_t *stream = input;
void lex_parse_skip_block_comment(core_probe_stream_t *input, core_pos_t *pos) {
core_probe_stream_t *stream = input;
Assert(stream != null && pos != null);
int ch;
core_stream_reset_char(stream);
ch = core_stream_next_char(stream);
core_probe_stream_reset(stream);
ch = core_probe_stream_consume(stream);
core_pos_next(pos);
// FIXME Assertion
Assert(ch == '/');
ch = core_stream_next_char(stream);
ch = core_probe_stream_consume(stream);
core_pos_next(pos);
Assert(ch == '*');
// all ready match `/*`
while (1) {
core_stream_reset_char(stream);
ch = core_stream_peek_char(stream);
core_probe_stream_reset(stream);
ch = core_probe_stream_peek(stream);
if (ch == core_stream_eof) {
LOG_WARN("Unterminated block comment");
@@ -106,12 +106,12 @@ void lex_parse_skip_block_comment(core_stream_t *input, core_pos_t *pos) {
lex_parse_skip_endline(stream, pos);
continue;
}
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
if (ch == '*') {
ch = core_stream_peek_char(stream);
ch = core_probe_stream_peek(stream);
if (ch == '/') {
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
return;
}
@@ -119,35 +119,35 @@ void lex_parse_skip_block_comment(core_stream_t *input, core_pos_t *pos) {
}
}
void lex_parse_skip_whitespace(core_stream_t *input, core_pos_t *pos) {
core_stream_t *stream = input;
void lex_parse_skip_whitespace(core_probe_stream_t *input, core_pos_t *pos) {
core_probe_stream_t *stream = input;
Assert(stream != null && pos != null);
core_stream_reset_char(stream);
core_probe_stream_reset(stream);
while (1) {
int ch = core_stream_peek_char(stream);
int ch = core_probe_stream_peek(stream);
if (!lex_parse_is_whitespace(ch)) {
return;
}
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
}
}
static inline cbool _lex_parse_uint(core_stream_t *input, core_pos_t *pos,
static inline cbool _lex_parse_uint(core_probe_stream_t *input, core_pos_t *pos,
int base, usize *output) {
Assert(input != null && pos != null);
if (input == null || pos == null) {
return false;
}
Assert(base == 2 || base == 8 || base == 10 || base == 16);
core_stream_reset_char(input);
core_probe_stream_reset(input);
int ch, tmp;
usize n = 0;
usize offset = pos->offset;
while (1) {
ch = core_stream_peek_char(input);
ch = core_probe_stream_peek(input);
if (ch == core_stream_eof) {
break;
@@ -166,7 +166,7 @@ static inline cbool _lex_parse_uint(core_stream_t *input, core_pos_t *pos,
return false;
}
core_stream_next_char(input);
core_probe_stream_consume(input);
core_pos_next(pos);
n = n * base + tmp;
// TODO number overflow
@@ -187,11 +187,11 @@ static inline cbool _lex_parse_uint(core_stream_t *input, core_pos_t *pos,
* @return int
* https://cppreference.cn/w/c/language/character_constant
*/
int lex_parse_char(core_stream_t *input, core_pos_t *pos) {
core_stream_t *stream = input;
int lex_parse_char(core_probe_stream_t *input, core_pos_t *pos) {
core_probe_stream_t *stream = input;
Assert(stream != null && pos != null);
core_stream_reset_char(stream);
int ch = core_stream_peek_char(stream);
core_probe_stream_reset(stream);
int ch = core_probe_stream_peek(stream);
int ret = core_stream_eof;
if (ch == core_stream_eof) {
@@ -201,17 +201,17 @@ int lex_parse_char(core_stream_t *input, core_pos_t *pos) {
LOG_WARN("Unexpected character '%c' at begin", ch);
goto ERR;
}
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
ch = core_stream_next_char(stream);
ch = core_probe_stream_consume(stream);
core_pos_next(pos);
if (ch == core_stream_eof) {
LOG_WARN("Unexpected EOF at middle");
goto ERR;
} else if (ch == '\\') {
ch = core_stream_next_char(stream);
ch = core_probe_stream_consume(stream);
core_pos_next(pos);
if (ch == '0') {
// 数字转义序列
@@ -237,7 +237,7 @@ int lex_parse_char(core_stream_t *input, core_pos_t *pos) {
} else {
ret = ch;
}
if ((ch = core_stream_next_char(stream)) != '\'') {
if ((ch = core_probe_stream_consume(stream)) != '\'') {
LOG_ERROR("Unclosed character literal '%c' at end, expect `'`", ch);
core_pos_next(pos);
goto ERR;
@@ -257,12 +257,12 @@ ERR:
* @return cbool
* https://cppreference.cn/w/c/language/string_literal
*/
cbool lex_parse_string(core_stream_t *input, core_pos_t *pos,
cbool lex_parse_string(core_probe_stream_t *input, core_pos_t *pos,
cstring_t *output) {
core_stream_t *stream = input;
core_probe_stream_t *stream = input;
Assert(stream != null && pos != null && output != null);
core_stream_reset_char(stream);
int ch = core_stream_peek_char(stream);
core_probe_stream_reset(stream);
int ch = core_probe_stream_peek(stream);
Assert(cstring_is_empty(output));
if (ch == core_stream_eof) {
@@ -272,12 +272,12 @@ cbool lex_parse_string(core_stream_t *input, core_pos_t *pos,
LOG_WARN("Unexpected character '%c' at begin", ch);
goto ERR;
}
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
cstring_t str = cstring_from_cstr("");
while (1) {
ch = core_stream_peek_char(stream);
ch = core_probe_stream_peek(stream);
if (ch == core_stream_eof) {
LOG_ERROR("Unexpected EOF at string literal");
@@ -287,8 +287,8 @@ cbool lex_parse_string(core_stream_t *input, core_pos_t *pos,
goto ERR;
} else if (ch == '\\') {
// TODO bad practice and maybe bugs here
core_stream_next_char(stream);
ch = core_stream_next_char(stream);
core_probe_stream_consume(stream);
ch = core_probe_stream_consume(stream);
int val = got_simple_escape(ch);
if (val == -1) {
LOG_ERROR("Invalid escape character it is \\%c [%d]", ch, ch);
@@ -297,12 +297,12 @@ cbool lex_parse_string(core_stream_t *input, core_pos_t *pos,
continue;
}
} else if (ch == '"') {
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
break;
}
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
cstring_push(&str, ch);
}
@@ -323,48 +323,72 @@ ERR:
* @return cbool
* https://cppreference.cn/w/c/language/integer_constant
*/
cbool lex_parse_number(core_stream_t *input, core_pos_t *pos, usize *output) {
core_stream_t *stream = input;
cbool lex_parse_number(core_probe_stream_t *input, core_pos_t *pos,
usize *output) {
core_probe_stream_t *stream = input;
Assert(stream != null && pos != null && output != null);
core_stream_reset_char(stream);
int ch = core_stream_peek_char(stream);
int base = 0;
core_probe_stream_reset(stream);
int ch = core_probe_stream_peek(stream);
int base = 10; // 默认十进制
if (ch == core_stream_eof) {
LOG_WARN("Unexpected EOF at begin");
goto ERR;
} else if (ch == '0') {
ch = core_stream_peek_char(stream);
}
if (ch == '0') {
// 消费 '0'
core_probe_stream_consume(stream);
core_pos_next(pos);
// 查看下一个字符
ch = core_probe_stream_peek(stream);
if (ch == 'x' || ch == 'X') {
// 十六进制
base = 16;
core_stream_next_char(stream);
core_pos_next(pos);
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
} else if (ch == 'b' || ch == 'B') {
// FIXME C23 external integer base
// 二进制 (C23扩展)
base = 2;
core_stream_next_char(stream);
core_pos_next(pos);
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
} else if (ch >= '0' && ch <= '7') {
// 八进制
base = 8;
core_stream_next_char(stream);
core_pos_next(pos);
} else if (ch == '9' || ch == '8') {
// 不消费，数字将由 _lex_parse_uint 处理
} else if (ch == '8' || ch == '9') {
LOG_ERROR("Invalid digit '%d' in octal literal", ch);
return false;
} else {
base = 10;
// 只是0，十进制
*output = 0;
return true;
}
} else {
} else if (ch >= '1' && ch <= '9') {
// 十进制，不消费，由 _lex_parse_uint 处理
base = 10;
} else {
// 无效的数字
return false;
}
// 解析整数部分
core_stream_reset_char(stream);
core_probe_stream_reset(stream);
usize n;
if (_lex_parse_uint(stream, pos, base, &n) == false) {
// 如果没有匹配任何数字，但输入是 '0'，已经处理过了
// 对于十进制数字，至少应该有一个数字
if (base == 10) {
// 单个数字的情况，例如 "1"
// 我们需要消费这个数字并返回它的值
if (ch >= '1' && ch <= '9') {
core_probe_stream_consume(stream);
core_pos_next(pos);
*output = ch - '0';
return true;
}
}
return false;
}
*output = n;
@@ -382,13 +406,13 @@ ERR:
* @return cbool
* https://cppreference.cn/w/c/language/identifier
*/
cbool lex_parse_identifier(core_stream_t *input, core_pos_t *pos,
cbool lex_parse_identifier(core_probe_stream_t *input, core_pos_t *pos,
cstring_t *output) {
Assert(input != null && pos != null && output != null);
Assert(cstring_is_empty(output));
core_stream_t *stream = input;
core_stream_reset_char(stream);
int ch = core_stream_peek_char(stream);
core_probe_stream_t *stream = input;
core_probe_stream_reset(stream);
int ch = core_probe_stream_peek(stream);
if (ch == core_stream_eof) {
LOG_WARN("Unexpected EOF at begin");
@@ -396,9 +420,9 @@ cbool lex_parse_identifier(core_stream_t *input, core_pos_t *pos,
(ch >= 'A' && ch <= 'Z')) {
while (1) {
cstring_push(output, ch);
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(pos);
ch = core_stream_peek_char(stream);
ch = core_probe_stream_peek(stream);
if ((ch >= 'a' && ch <= 'z') || (ch >= 'A' && ch <= 'Z') ||
(ch == '_') || (ch >= '0' && ch <= '9')) {
continue;

8
libs/lex_parser/tests/test_char.c
View File

@@ -5,9 +5,9 @@
cbool check_char(const char *str, int expect, int *output) {
log_set_level(&__default_logger_root, 0);
core_pos_t pos = core_pos_init();
core_mem_stream_t mem_stream;
core_stream_t *stream =
core_mem_stream_init(&mem_stream, str, smcc_strlen(str), false);
core_mem_probe_stream_t mem_stream;
core_probe_stream_t *stream =
core_mem_probe_stream_init(&mem_stream, str, smcc_strlen(str), false);
*output = lex_parse_char(stream, &pos);
return *output == expect;
}
@@ -57,4 +57,4 @@ TEST_LIST = {
{"test_escape_char", test_escape_char},
{"test_invalid_char", test_invalid_char},
{NULL, NULL},
};
};

6
libs/lex_parser/tests/test_identifier.c
View File

@@ -5,9 +5,9 @@
cbool check_identifier(const char *str, const char *expect, cstring_t *output) {
log_set_level(&__default_logger_root, 0);
core_pos_t pos = core_pos_init();
core_mem_stream_t mem_stream;
core_stream_t *stream =
core_mem_stream_init(&mem_stream, str, smcc_strlen(str), false);
core_mem_probe_stream_t mem_stream;
core_probe_stream_t *stream =
core_mem_probe_stream_init(&mem_stream, str, smcc_strlen(str), false);
cbool ret = lex_parse_identifier(stream, &pos, output);
if (ret && expect) {

7
libs/lex_parser/tests/test_number.c
View File

@@ -4,9 +4,9 @@ cbool check(const char *str, usize expect, usize *output) {
// TODO maybe have other logger
log_set_level(&__default_logger_root, 0);
core_pos_t pos = core_pos_init();
core_mem_stream_t mem_stream;
core_stream_t *stream =
core_mem_stream_init(&mem_stream, str, smcc_strlen(str), false);
core_mem_probe_stream_t mem_stream;
core_probe_stream_t *stream =
core_mem_probe_stream_init(&mem_stream, str, smcc_strlen(str), false);
return lex_parse_number(stream, &pos, output);
}
@@ -16,6 +16,7 @@ cbool check(const char *str, usize expect, usize *output) {
cbool ret = check(str, expect, &_output); \
TEST_CHECK(ret == true); \
TEST_CHECK(_output == expect); \
TEST_MSG("Produced: %llu", _output); \
} while (0)
#define CHECK_INVALID(str) \

9
libs/lex_parser/tests/test_skip_block_comment.c
View File

@@ -5,9 +5,9 @@
void check_skip_block_comment(const char *str, const char *expect_remaining) {
log_set_level(&__default_logger_root, 0);
core_pos_t pos = core_pos_init();
core_mem_stream_t mem_stream;
core_stream_t *stream =
core_mem_stream_init(&mem_stream, str, smcc_strlen(str), false);
core_mem_probe_stream_t mem_stream;
core_probe_stream_t *stream =
core_mem_probe_stream_init(&mem_stream, str, smcc_strlen(str), false);
lex_parse_skip_block_comment(stream, &pos);
@@ -15,7 +15,8 @@ void check_skip_block_comment(const char *str, const char *expect_remaining) {
char buffer[256] = {0};
int i = 0;
int ch;
while ((ch = core_stream_next_char(stream)) != core_stream_eof && i < 255) {
while ((ch = core_probe_stream_consume(stream)) != core_stream_eof &&
i < 255) {
buffer[i++] = (char)ch;
}

9
libs/lex_parser/tests/test_skip_line.c
View File

@@ -5,9 +5,9 @@
void check_skip_line(const char *str, const char *expect_remaining) {
log_set_level(&__default_logger_root, 0);
core_pos_t pos = core_pos_init();
core_mem_stream_t mem_stream;
core_stream_t *stream =
core_mem_stream_init(&mem_stream, str, smcc_strlen(str), false);
core_mem_probe_stream_t mem_stream;
core_probe_stream_t *stream =
core_mem_probe_stream_init(&mem_stream, str, smcc_strlen(str), false);
lex_parse_skip_line(stream, &pos);
@@ -15,7 +15,8 @@ void check_skip_line(const char *str, const char *expect_remaining) {
char buffer[256] = {0};
int i = 0;
int ch;
while ((ch = core_stream_next_char(stream)) != core_stream_eof && i < 255) {
while ((ch = core_probe_stream_consume(stream)) != core_stream_eof &&
i < 255) {
buffer[i++] = (char)ch;
}

6
libs/lex_parser/tests/test_string.c
View File

@@ -5,9 +5,9 @@
cbool check_string(const char *str, const char *expect, cstring_t *output) {
log_set_level(&__default_logger_root, 0);
core_pos_t pos = core_pos_init();
core_mem_stream_t mem_stream;
core_stream_t *stream =
core_mem_stream_init(&mem_stream, str, smcc_strlen(str), false);
core_mem_probe_stream_t mem_stream;
core_probe_stream_t *stream =
core_mem_probe_stream_init(&mem_stream, str, smcc_strlen(str), false);
cbool ret = lex_parse_string(stream, &pos, output);
if (ret && expect) {

4
libs/lexer/include/lexer.h
View File

@@ -21,7 +21,7 @@ typedef struct lexer_token {
* 封装词法分析所需的状态信息和缓冲区管理
*/
typedef struct cc_lexer {
core_stream_t *stream;
core_probe_stream_t *stream;
core_pos_t pos;
} smcc_lexer_t;
@@ -30,7 +30,7 @@ typedef struct cc_lexer {
* @param[out] lexer 要初始化的词法分析器实例
* @param[in] stream 输入流对象指针
*/
void lexer_init(smcc_lexer_t *lexer, core_stream_t *stream);
void lexer_init(smcc_lexer_t *lexer, core_probe_stream_t *stream);
/**
* @brief 获取原始token

82
libs/lexer/src/lexer.c
View File

@@ -75,7 +75,7 @@ static inline int keyword_cmp(const char *name, int len) {
return -1; // Not a keyword.
}
void lexer_init(smcc_lexer_t *lexer, core_stream_t *stream) {
void lexer_init(smcc_lexer_t *lexer, core_probe_stream_t *stream) {
lexer->stream = stream;
lexer->pos = core_pos_init();
// FIXME
@@ -86,9 +86,9 @@ void lexer_init(smcc_lexer_t *lexer, core_stream_t *stream) {
static void parse_line(smcc_lexer_t *lexer, lexer_tok_t *token) {
token->loc = lexer->pos;
core_stream_t *stream = lexer->stream;
core_stream_reset_char(stream);
int ch = core_stream_peek_char(stream);
core_probe_stream_t *stream = lexer->stream;
core_probe_stream_reset(stream);
int ch = core_probe_stream_next(stream);
usize n;
cstring_t str = cstring_new();
@@ -104,7 +104,7 @@ static void parse_line(smcc_lexer_t *lexer, lexer_tok_t *token) {
const char line[] = "line";
for (int i = 0; i < (int)sizeof(line); i++) {
ch = core_stream_next_char(stream);
ch = core_probe_stream_consume(stream);
core_pos_next(&lexer->pos);
if (ch != line[i]) {
LEX_WARN("Maroc does not support in lexer rather in preprocessor, "
@@ -118,12 +118,12 @@ static void parse_line(smcc_lexer_t *lexer, lexer_tok_t *token) {
goto SKIP_LINE;
}
if (core_stream_next_char(stream) != ' ') {
if (core_probe_stream_consume(stream) != ' ') {
lex_parse_skip_line(lexer->stream, &lexer->pos);
token->loc.line = token->value.n;
}
if (core_stream_peek_char(stream) != '"') {
if (core_probe_stream_next(stream) != '"') {
LEX_ERROR("Invalid `#` line");
goto SKIP_LINE;
}
@@ -149,26 +149,26 @@ ERR:
void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
token->loc = lexer->pos;
token->type = TOKEN_UNKNOWN;
core_stream_t *stream = lexer->stream;
core_probe_stream_t *stream = lexer->stream;
core_stream_reset_char(stream);
core_probe_stream_reset(stream);
token_type_t type = TOKEN_UNKNOWN;
int ch = core_stream_peek_char(stream);
int ch = core_probe_stream_next(stream);
// once step
switch (ch) {
case '=':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '=':
type = TOKEN_EQ;
goto double_char;
default:
core_stream_reset_char(stream), type = TOKEN_ASSIGN;
core_probe_stream_reset(stream), type = TOKEN_ASSIGN;
break;
}
break;
case '+':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '+':
type = TOKEN_ADD_ADD;
goto double_char;
@@ -176,12 +176,12 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_ASSIGN_ADD;
goto double_char;
default:
core_stream_reset_char(stream), type = TOKEN_ADD;
core_probe_stream_reset(stream), type = TOKEN_ADD;
break;
}
break;
case '-':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '-':
type = TOKEN_SUB_SUB;
goto double_char;
@@ -192,22 +192,22 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_DEREF;
goto double_char;
default:
core_stream_reset_char(stream), type = TOKEN_SUB;
core_probe_stream_reset(stream), type = TOKEN_SUB;
break;
}
break;
case '*':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '=':
type = TOKEN_ASSIGN_MUL;
goto double_char;
default:
core_stream_reset_char(stream), type = TOKEN_MUL;
core_probe_stream_reset(stream), type = TOKEN_MUL;
break;
}
break;
case '/':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '=':
type = TOKEN_ASSIGN_DIV;
goto double_char;
@@ -220,22 +220,22 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
token->type = TOKEN_BLOCK_COMMENT;
goto END;
default:
core_stream_reset_char(stream), type = TOKEN_DIV;
core_probe_stream_reset(stream), type = TOKEN_DIV;
break;
}
break;
case '%':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '=':
type = TOKEN_ASSIGN_MOD;
goto double_char;
default:
core_stream_reset_char(stream), type = TOKEN_MOD;
core_probe_stream_reset(stream), type = TOKEN_MOD;
break;
}
break;
case '&':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '&':
type = TOKEN_AND_AND;
goto double_char;
@@ -243,12 +243,12 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_ASSIGN_AND;
goto double_char;
default:
core_stream_reset_char(stream), type = TOKEN_AND;
core_probe_stream_reset(stream), type = TOKEN_AND;
break;
}
break;
case '|':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '|':
type = TOKEN_OR_OR;
goto double_char;
@@ -256,27 +256,27 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_ASSIGN_OR;
goto double_char;
default:
core_stream_reset_char(stream), type = TOKEN_OR;
core_probe_stream_reset(stream), type = TOKEN_OR;
break;
}
break;
case '^':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '=':
type = TOKEN_ASSIGN_XOR;
goto double_char;
default:
core_stream_reset_char(stream), type = TOKEN_XOR;
core_probe_stream_reset(stream), type = TOKEN_XOR;
break;
}
break;
case '<':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '=':
type = TOKEN_LE;
goto double_char;
case '<': {
if (core_stream_peek_char(stream) == '=') {
if (core_probe_stream_next(stream) == '=') {
type = TOKEN_ASSIGN_L_SH;
goto triple_char;
} else {
@@ -286,17 +286,17 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
break;
}
default:
core_stream_reset_char(stream), type = TOKEN_LT;
core_probe_stream_reset(stream), type = TOKEN_LT;
break;
}
break;
case '>':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '=':
type = TOKEN_GE;
goto double_char;
case '>': {
if (core_stream_peek_char(stream) == '=') {
if (core_probe_stream_next(stream) == '=') {
type = TOKEN_ASSIGN_R_SH;
goto triple_char;
} else {
@@ -306,7 +306,7 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
break;
}
default:
core_stream_reset_char(stream), type = TOKEN_GT;
core_probe_stream_reset(stream), type = TOKEN_GT;
break;
}
break;
@@ -314,12 +314,12 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_BIT_NOT;
break;
case '!':
switch (core_stream_peek_char(stream)) {
switch (core_probe_stream_next(stream)) {
case '=':
type = TOKEN_NEQ;
goto double_char;
default:
core_stream_reset_char(stream), type = TOKEN_NOT;
core_probe_stream_reset(stream), type = TOKEN_NOT;
break;
}
break;
@@ -351,8 +351,8 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_COLON;
break;
case '.':
if (core_stream_peek_char(stream) == '.' &&
core_stream_peek_char(stream) == '.') {
if (core_probe_stream_next(stream) == '.' &&
core_probe_stream_next(stream) == '.') {
type = TOKEN_ELLIPSIS;
goto triple_char;
}
@@ -452,13 +452,13 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
}
goto once_char;
triple_char:
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(&lexer->pos);
double_char:
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(&lexer->pos);
once_char:
core_stream_next_char(stream);
core_probe_stream_consume(stream);
core_pos_next(&lexer->pos);
token->type = type;
END:

8
libs/lexer/tests/test_parse.c
View File

@@ -8,10 +8,10 @@ static inline void test_lexer_string(const char *input,
token_type_t expected_type) {
smcc_lexer_t lexer;
lexer_tok_t token;
core_mem_stream_t stream;
core_mem_probe_stream_t stream;
lexer_init(&lexer,
core_mem_stream_init(&stream, input, strlen(input), false));
lexer_init(&lexer, core_mem_probe_stream_init(&stream, input, strlen(input),
false));
lexer_get_token(&lexer, &token);
TEST_CHECK(token.type == expected_type);
@@ -167,4 +167,4 @@ TEST_LIST = {{"operators", test_operators},
{"literals", test_literals},
{"edge_cases", test_edge_cases},
// {"error_handling", test_error_handling},
{NULL, NULL}};
{NULL, NULL}};

8
libs/lexer/tests/test_run.c
View File

@@ -63,9 +63,9 @@ int main(int argc, char *argv[]) {
}
smcc_lexer_t lexer;
core_mem_stream_t mem_stream = {0};
core_stream_t *stream =
core_mem_stream_init(&mem_stream, buffer, fsize, false);
core_mem_probe_stream_t mem_stream = {0};
core_probe_stream_t *stream =
core_mem_probe_stream_init(&mem_stream, buffer, fsize, false);
Assert(stream != null);
cstring_clear(&stream->name);
cstring_push_cstr(&stream->name, file_name, strlen(file_name));
@@ -78,7 +78,7 @@ int main(int argc, char *argv[]) {
break;
}
LOG_DEBUG("token `%s` at %s:%u:%u", get_tok_name(tok.type),
tok.loc.name, tok.loc.line, tok.loc.col);
cstring_as_cstr(&tok.loc.name), tok.loc.line, tok.loc.col);
Assert(tok.loc.offset <= fsize);
// LOG_DEBUG("%s", tok.val.str);
// printf("line: %d, column: %d, type: %3d, typename: %s\n",

2
runtime/libcore/include/core_str.h
View File

@@ -57,7 +57,7 @@ static inline void cstring_free(cstring_t *str) {
if (str == null) {
return;
}
if (str->data != null && str->cap != 0) {
if (str->cap != 0 && str->data != null) {
smcc_free(str->data);
str->data = null;
}

129
runtime/libcore/include/core_stream.h
View File

@@ -1,67 +1,120 @@
#ifndef __SMCC_CORE_STREAM_H__
#define __SMCC_CORE_STREAM_H__
#ifndef __SMCC_CORE_PROBE_STREAM_H__
#define __SMCC_CORE_PROBE_STREAM_H__
#include "core_impl.h"
#include "core_macro.h"
#include "core_mem.h"
#include "core_str.h"
struct core_stream;
typedef struct core_stream core_stream_t;
struct core_probe_stream;
typedef struct core_probe_stream core_probe_stream_t;
#define core_stream_eof (-1)
struct core_stream {
/**
* @brief 带探针的流接口
*
* 这个流提供了双指针机制：当前读取位置（头指针）和探针位置（尾指针）。
* 尾指针只能向前移动，用于查看而不消费。
* 头指针可以前进或单次后退，但不能一直后退到尾指针后面。
*/
struct core_probe_stream {
cstring_t name;
/// @brief 消费头指针处的字符（移动头指针）
int (*consume)(core_probe_stream_t *stream);
/// @brief 查看当前探针位置的字符，不移动任何指针
int (*peek)(core_probe_stream_t *stream);
/// @brief 移动探针位置并返回字符
int (*next)(core_probe_stream_t *stream);
/// @brief 移动头指针到探针位置
void (*sync)(core_probe_stream_t *stream);
/// @brief 重置探针位置到头指针位置
void (*reset)(core_probe_stream_t *stream);
/// @brief 回退一个字符（单次后退，头指针后退一步）
cbool (*back)(core_probe_stream_t *stream);
/// @brief 读取指定数量的字符到缓冲区
usize (*read_buf)(core_stream_t *stream, char *buffer, usize count);
usize (*read_buf)(core_probe_stream_t *stream, char *buffer, usize count);
/// @brief 获取下一个字符
int (*peek_char)(core_stream_t *stream);
/// @brief 检查是否到达流末尾
cbool (*is_at_end)(core_probe_stream_t *stream);
/// @brief 重置字符流位置
void (*reset_char)(core_stream_t *stream);
/// @brief 读取并消费下一个字符（移动流位置）
int (*next_char)(core_stream_t *stream);
/// @brief 释放资源
void (*free_stream)(core_stream_t *steam);
/// @brief 销毁流并释放资源
void (*destroy)(core_probe_stream_t *stream);
};
static inline usize core_stream_read_buf(core_stream_t *self, char *buffer,
usize count) {
static inline int core_probe_stream_consume(core_probe_stream_t *self) {
return self->consume(self);
}
static inline int core_probe_stream_peek(core_probe_stream_t *self) {
return self->peek(self);
}
static inline int core_probe_stream_next(core_probe_stream_t *self) {
return self->next(self);
}
static inline void core_probe_stream_sync(core_probe_stream_t *self) {
self->sync(self);
}
static inline cbool core_probe_stream_back(core_probe_stream_t *self) {
return self->back(self);
}
static inline void core_probe_stream_reset(core_probe_stream_t *self) {
self->reset(self);
}
static inline usize core_probe_stream_read_buf(core_probe_stream_t *self,
char *buffer, usize count) {
return self->read_buf(self, buffer, count);
}
static inline int core_stream_peek_char(core_stream_t *self) {
return self->peek_char(self);
static inline cbool core_probe_stream_is_at_end(core_probe_stream_t *self) {
return self->is_at_end(self);
}
static inline void core_stream_reset_char(core_stream_t *self) {
self->reset_char(self);
static inline cbool core_probe_stream_has_more(core_probe_stream_t *self) {
return !self->is_at_end(self);
}
static inline int core_stream_next_char(core_stream_t *self) {
return self->next_char(self);
static inline void core_probe_stream_destroy(core_probe_stream_t *self) {
self->destroy(self);
}
static inline void core_stream_free_stream(core_stream_t *self) {
self->free_stream(self);
}
#ifndef __SMCC_CORE_NO_MEM_STREAM__
typedef struct core_mem_stream {
core_stream_t stream;
#ifndef __SMCC_CORE_NO_MEM_PROBE_STREAM__
/**
* @brief 内存探针流结构
*/
typedef struct core_mem_probe_stream {
core_probe_stream_t stream;
const char *data;
usize data_length;
usize curr_pos;
usize peek_pos;
cbool owned;
} core_mem_stream_t;
core_stream_t *core_mem_stream_init(core_mem_stream_t *stream, const char *data,
usize length, cbool need_copy);
usize curr_pos; // 当前读取位置
usize probe_pos; // 探针位置（用于peek）
cbool owned; // 是否拥有数据（需要释放）
} core_mem_probe_stream_t;
/**
* @brief 初始化内存探针流
*
* @param stream 流结构指针
* @param data 数据指针
* @param length 数据长度
* @param need_copy 是否需要复制数据
* @return core_probe_stream_t* 成功返回流指针，失败返回NULL
*/
core_probe_stream_t *core_mem_probe_stream_init(core_mem_probe_stream_t *stream,
const char *data, usize length,
cbool need_copy);
#endif
#endif /* __SMCC_CORE_STREAM_H__ */
#endif /* __SMCC_CORE_PROBE_STREAM_H__ */

1
runtime/libcore/include/core_vec.h
View File

@@ -70,6 +70,7 @@
(vec).cap = cap; \
(vec).data = data; \
} \
Assert((vec).data != null); \
(vec).data[(vec).size++] = value; \
} while (0)

164
runtime/libcore/src/stream.c
View File

@@ -1,10 +1,82 @@
#include <core_log.h>
#include <core_stream.h>
// 内存流的具体实现结构
static usize read_buf(core_stream_t *_stream, char *buffer, usize count) {
Assert(buffer != null && buffer != null);
core_mem_stream_t *stream = (core_mem_stream_t *)_stream;
#ifndef __SMCC_CORE_NO_MEM_PROBE_STREAM__
static int mem_probe_stream_consume(core_probe_stream_t *_stream) {
Assert(_stream != null);
core_mem_probe_stream_t *stream = (core_mem_probe_stream_t *)_stream;
if (stream->curr_pos >= stream->data_length) {
return core_stream_eof;
}
unsigned char ch = stream->data[stream->curr_pos++];
// 如果探针位置落后于当前读取位置，则更新探针位置
if (stream->probe_pos < stream->curr_pos) {
stream->probe_pos = stream->curr_pos;
}
return (int)ch;
}
static int mem_probe_stream_peek(core_probe_stream_t *_stream) {
Assert(_stream != null);
core_mem_probe_stream_t *stream = (core_mem_probe_stream_t *)_stream;
if (stream->probe_pos >= stream->data_length) {
return core_stream_eof;
}
// 只查看而不移动探针位置
return (int)(unsigned char)stream->data[stream->probe_pos];
}
static int mem_probe_stream_next(core_probe_stream_t *_stream) {
Assert(_stream != null);
core_mem_probe_stream_t *stream = (core_mem_probe_stream_t *)_stream;
if (stream->probe_pos >= stream->data_length) {
return core_stream_eof;
}
// 返回探针位置的字符，并将探针位置向前移动
int ch = (int)(unsigned char)stream->data[stream->probe_pos];
stream->probe_pos++;
return ch;
}
static void mem_probe_stream_sync(core_probe_stream_t *_stream) {
Assert(_stream != null);
core_mem_probe_stream_t *stream = (core_mem_probe_stream_t *)_stream;
// 移动头指针到探针位置（消费已查看的字符）
if (stream->probe_pos > stream->curr_pos) {
stream->curr_pos = stream->probe_pos;
}
}
static cbool mem_probe_stream_back(core_probe_stream_t *_stream) {
Assert(_stream != null);
core_mem_probe_stream_t *stream = (core_mem_probe_stream_t *)_stream;
// 只能回退一个字符，且不能回退到探针位置之前
if (stream->curr_pos == 0 || stream->curr_pos <= stream->probe_pos) {
return false;
}
stream->curr_pos--;
return true;
}
static usize mem_probe_stream_read_buf(core_probe_stream_t *_stream, char *buffer,
usize count) {
Assert(_stream != null);
core_mem_probe_stream_t *stream = (core_mem_probe_stream_t *)_stream;
if (buffer == null) {
LOG_WARN("Buffer is null");
return 0;
}
usize remaining = stream->data_length - stream->curr_pos;
usize to_read = (remaining < count) ? remaining : count;
@@ -12,63 +84,48 @@ static usize read_buf(core_stream_t *_stream, char *buffer, usize count) {
if (to_read > 0) {
smcc_memcpy(buffer, stream->data + stream->curr_pos, to_read);
stream->curr_pos += to_read;
// 更新探针位置
if (stream->probe_pos < stream->curr_pos) {
stream->probe_pos = stream->curr_pos;
}
} else {
LOG_WARN("Reading past end of stream "
"[maybe count is too large or negative?]");
LOG_WARN("Reading past end of stream [maybe count is too large or "
"negative?]");
}
return to_read;
}
static int peek_char(core_stream_t *_stream) {
static void mem_probe_stream_reset(core_probe_stream_t *_stream) {
Assert(_stream != null);
core_mem_stream_t *stream = (core_mem_stream_t *)_stream;
core_mem_probe_stream_t *stream = (core_mem_probe_stream_t *)_stream;
// 如果已经到达末尾，返回EOF
if (stream->peek_pos >= stream->data_length) {
return core_stream_eof; // EOF
}
return (int)(unsigned char)stream->data[stream->peek_pos++];
// 重置探针位置到头指针位置
stream->probe_pos = stream->curr_pos;
}
static int next_char(core_stream_t *_stream) {
Assert(_stream != NULL);
core_mem_stream_t *stream = (core_mem_stream_t *)_stream;
// 如果已经到达末尾，返回EOF
if (stream->curr_pos >= stream->data_length) {
return core_stream_eof; // EOF
}
unsigned char ch = stream->data[stream->curr_pos++];
if (stream->peek_pos < stream->curr_pos) {
stream->peek_pos = stream->curr_pos;
}
return (int)ch;
}
static void reset_char(core_stream_t *_stream) {
Assert(_stream != NULL);
core_mem_stream_t *stream = (core_mem_stream_t *)_stream;
stream->peek_pos = stream->curr_pos;
}
static void free_stream(core_stream_t *_stream) {
static cbool mem_probe_stream_is_at_end(core_probe_stream_t *_stream) {
Assert(_stream != null);
core_mem_stream_t *stream = (core_mem_stream_t *)_stream;
core_mem_probe_stream_t *stream = (core_mem_probe_stream_t *)_stream;
return stream->curr_pos >= stream->data_length;
}
static void mem_probe_stream_destroy(core_probe_stream_t *_stream) {
Assert(_stream != null);
core_mem_probe_stream_t *stream = (core_mem_probe_stream_t *)_stream;
// FIXME maybe double free?
cstring_free(&stream->stream.name);
if (stream->owned) {
smcc_free((void *)stream->data);
stream->data = null;
}
}
core_stream_t *core_mem_stream_init(core_mem_stream_t *stream, const char *data,
usize length, cbool need_copy) {
core_probe_stream_t *core_mem_probe_stream_init(core_mem_probe_stream_t *stream,
const char *data, usize length,
cbool need_copy) {
if (stream == null || data == null) {
LOG_ERROR("param error");
return null;
@@ -94,15 +151,22 @@ core_stream_t *core_mem_stream_init(core_mem_stream_t *stream, const char *data,
}
stream->data_length = length;
stream->curr_pos = 0;
stream->peek_pos = 0;
stream->probe_pos = 0;
stream->stream.name = cstring_from_cstr("mem_stream");
stream->stream.name = cstring_from_cstr("mem_probe_stream");
stream->stream.read_buf = read_buf;
stream->stream.peek_char = peek_char;
stream->stream.next_char = next_char;
stream->stream.reset_char = reset_char;
stream->stream.free_stream = free_stream;
// 设置函数指针
stream->stream.consume = mem_probe_stream_consume;
stream->stream.peek = mem_probe_stream_peek;
stream->stream.next = mem_probe_stream_next;
stream->stream.sync = mem_probe_stream_sync;
stream->stream.back = mem_probe_stream_back;
stream->stream.read_buf = mem_probe_stream_read_buf;
stream->stream.reset = mem_probe_stream_reset;
stream->stream.is_at_end = mem_probe_stream_is_at_end;
stream->stream.destroy = mem_probe_stream_destroy;
return (void *)stream;
return (core_probe_stream_t *)stream;
}
#endif /* __SMCC_CORE_NO_MEM_PROBE_STREAM__ */

12
runtime/log/include/log.h
View File

@@ -16,6 +16,16 @@
#define log_exit exit
#endif
#ifdef __GNUC__ // GCC, Clang, ICC
#define __smcc_log_unreachable() (__builtin_unreachable())
#elif defined _MSC_VER // MSVC
#define __smcc_log_unreachable() (__assume(false))
#elif defined __SMCC_BUILT_IN__ // The SMCC (my compiler)
#define __smcc_log_unreachable() (__smcc_builtin_unreachable())
#else
#define __smcc_log_unreachable()
#endif
#ifndef log_snprintf
#define log_snprintf(...)
#warning "log_snprintf not defined"
@@ -172,6 +182,8 @@ void logger_destroy(logger_t *logger);
do { \
if (!(cond)) { \
LOG_FATAL(__VA_ARGS__); \
log_exit(1); \
__smcc_log_unreachable(); \
} \
} while (0)