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

feat(lex_parser): 初始化词法解析器模块

Browse Source 
新增词法解析器库 `smcc_lex_parser`，包含基础的词法规则解析功能：
- 支持字符、字符串、数字、标识符的解析
- 支持跳过注释、空白符、行尾等辅助函数
- 提供对应的单元测试用例，覆盖各类合法与非法输入情况

该模块依赖 `libcore`，并被 `smcc_lex` 模块引用以支持更上层的词法分析逻辑。
This commit is contained in:
zzy
2025-11-23 22:53:46 +08:00
parent 67af0c6bf2
commit 871d031ceb
18 changed files with 996 additions and 392 deletions
Download Patch File Download Diff File Expand all files Collapse all files

475
libs/lexer/src/lexer.c
View File

@@ -26,6 +26,7 @@ the distribution and installation instructions.
Chris Fraser / cwf@aya.yale.edu
David Hanson / drh@drhanson.net
*/
#include <lex_parser.h>
#include <lexer.h>
#include <lexer_log.h>
@@ -76,303 +77,21 @@ static inline int keyword_cmp(const char *name, int len) {
void lexer_init(smcc_lexer_t *lexer, core_stream_t *stream) {
lexer->stream = stream;
lexer->pos = (lexer_loc_t){
.name = cstring_as_cstr(&stream->name),
.name_len = cstring_len(&stream->name),
.line = 1,
.column = 1,
.offset = 0,
};
lexer->pos = core_pos_init();
// FIXME
lexer->pos.name = cstring_from_cstr(cstring_as_cstr(&stream->name));
}
#define stream_reset_char(stream) ((stream)->reset_char(stream))
#define stream_next_char(stream) ((stream)->next_char(stream))
#define stream_peek_char(stream) ((stream)->peek_char(stream))
#define lexer_next_pos(lexer) ((lexer)->pos.column++, (lexer)->pos.offset++)
#define lexer_next_line(lexer) ((lexer)->pos.line++, (lexer)->pos.column = 1)
#define set_err_token(token) ((token)->type = TOKEN_UNKNOWN)
static void skip_newline(smcc_lexer_t *lexer, lexer_tok_t *token) {
core_stream_t *stream = lexer->stream;
token->type = TOKEN_LINE_COMMENT;
// 循环直到遇到换行符或文件结束
while (1) {
int ch = stream_next_char(stream);
if (ch == core_stream_eof) {
// 到达文件末尾，直接返回
return;
}
// 更新位置信息
lexer_next_pos(lexer);
if (ch == '\n') {
// 遇到换行符，增加行号并重置列号
lexer_next_line(lexer);
return;
}
}
}
static void skip_block_comment(smcc_lexer_t *lexer, lexer_tok_t *token) {
core_stream_t *stream = lexer->stream;
token->type = TOKEN_BLOCK_COMMENT;
int ch;
stream_reset_char(stream);
ch = stream_next_char(stream);
lexer_next_pos(lexer);
// FIXME Assertion
Assert(ch == '/');
ch = stream_next_char(stream);
lexer_next_pos(lexer);
Assert(ch == '*');
// 我们已经识别了 "/*"，现在需要找到 "*/"
while (1) {
ch = stream_next_char(stream);
lexer_next_pos(lexer);
if (ch == core_stream_eof) {
// 未闭合的块注释
LEX_WARN("Unterminated block comment");
return;
}
// LEX_ERROR("%c", ch);
// 更新位置信息
if (ch == '\n') {
lexer_next_line(lexer);
} else if (ch == '*') {
// 查看下一个字符是否是 '/'
int next_ch = stream_peek_char(stream);
if (next_ch == '/') {
// 消费 '/' 字符
stream_next_char(stream);
// 更新位置信息
lexer_next_pos(lexer);
// 成功找到注释结束标记
return;
}
}
}
}
// TODO escape character not enough
static inline int got_slash(int peek) {
switch (peek) {
case '\\':
return '\\';
case '\'':
return '\'';
case '\"':
return '\"';
case '\?':
return '\?';
case '0':
return '\0';
case 'b':
return '\b';
case 'f':
return '\f';
case 'n':
return '\n';
case 'r':
return '\r';
case 't':
return '\t';
case 'v':
return '\v';
default:
break;
}
return -1;
}
static void parse_char(smcc_lexer_t *lexer, lexer_tok_t *token) {
token->loc = lexer->pos;
token->type = TOKEN_CHAR_LITERAL;
core_stream_t *stream = lexer->stream;
stream_reset_char(stream);
int ch = stream_peek_char(stream);
if (ch == core_stream_eof) {
LEX_WARN("Unexpected EOF at begin");
goto ERR;
} else if (ch != '\'') {
LEX_WARN("Unexpected character '%c' at begin", ch);
goto ERR;
}
stream_next_char(stream);
lexer_next_pos(lexer);
ch = stream_next_char(stream);
lexer_next_pos(lexer);
if (ch == core_stream_eof) {
LEX_WARN("Unexpected EOF at middle");
goto ERR;
} else if (ch == '\\') {
ch = stream_next_char(stream);
lexer_next_pos(lexer);
if ((ch = got_slash(ch)) == -1) {
LEX_ERROR("Invalid escape character");
// TODO 特殊情况处理
goto ERR;
}
token->value.ch = ch;
} else {
token->value.ch = ch;
}
if ((ch = stream_next_char(stream)) != '\'') {
LEX_ERROR("Unclosed character literal '%c' at end, expect `'`", ch);
lexer_next_pos(lexer);
goto ERR;
}
return;
ERR:
set_err_token(token);
}
static void parse_string(smcc_lexer_t *lexer, lexer_tok_t *token) {
token->loc = lexer->pos;
token->type = TOKEN_STRING_LITERAL;
core_stream_t *stream = lexer->stream;
stream_reset_char(stream);
int ch = stream_peek_char(stream);
if (ch == core_stream_eof) {
LEX_WARN("Unexpected EOF at begin");
goto ERR;
} else if (ch != '"') {
LEX_WARN("Unexpected character '%c' at begin", ch);
goto ERR;
}
stream_next_char(stream);
lexer_next_pos(lexer);
int base = 0;
cstring_t str = cstring_new();
while (1) {
ch = stream_peek_char(stream);
if (ch == core_stream_eof) {
LEX_ERROR("Unexpected EOF at string literal");
break;
} else if (ch == '\n') {
LEX_ERROR("Unexpected newline at string literal");
break;
} else if (ch == '\\') {
// TODO bad practice and maybe bugs here
stream_next_char(stream);
ch = stream_next_char(stream);
int val = got_slash(ch);
if (val == -1) {
LEX_ERROR("Invalid escape character it is \\%c [%d]", ch, ch);
} else {
cstring_push(&str, val);
continue;
}
} else if (ch == '"') {
stream_next_char(stream);
lexer_next_pos(lexer);
break;
}
stream_next_char(stream);
lexer_next_pos(lexer);
cstring_push(&str, ch);
}
token->value.cstr.data = (char *)cstring_as_cstr(&str);
token->value.cstr.len = cstring_len(&str);
return;
ERR:
set_err_token(token);
}
static void parse_number(smcc_lexer_t *lexer, lexer_tok_t *token) {
token->loc = lexer->pos;
core_stream_t *stream = lexer->stream;
stream_reset_char(stream);
int ch = stream_peek_char(stream);
int base = 0;
if (ch == core_stream_eof) {
LEX_WARN("Unexpected EOF at begin");
goto ERR;
} else if (ch == '0') {
ch = stream_peek_char(stream);
if (ch == 'x' || ch == 'X') {
base = 16;
stream_next_char(stream);
lexer_next_pos(lexer);
stream_next_char(stream);
lexer_next_pos(lexer);
} else if (ch == 'b' || ch == 'B') {
// FIXME C23 external integer base
base = 2;
stream_next_char(stream);
lexer_next_pos(lexer);
stream_next_char(stream);
lexer_next_pos(lexer);
} else if (ch >= '0' && ch <= '7') {
base = 8;
stream_next_char(stream);
lexer_next_pos(lexer);
} else {
base = 10;
}
} else {
base = 10;
}
// 解析整数部分
stream_reset_char(stream);
int tmp = 0;
token->value.n = 0;
while (1) {
ch = stream_peek_char(stream);
if (ch == core_stream_eof) {
break;
} else if (ch >= 'a' && ch <= 'z') {
tmp = ch - 'a' + 10;
} else if (ch >= 'A' && ch <= 'Z') {
tmp = ch - 'A' + 10;
} else if (ch >= '0' && ch <= '9') {
tmp = ch - '0';
} else {
break;
}
if (tmp >= base) {
LOG_ERROR("Invalid digit");
break;
}
stream_next_char(stream);
lexer_next_pos(lexer);
token->value.n = token->value.n * base + tmp;
// TODO number overflow
}
token->type = TOKEN_INT_LITERAL;
return;
ERR:
set_err_token(token);
}
static void parse_line(smcc_lexer_t *lexer, lexer_tok_t *token) {
token->loc = lexer->pos;
core_stream_t *stream = lexer->stream;
stream_reset_char(stream);
int ch = stream_peek_char(stream);
core_stream_reset_char(stream);
int ch = core_stream_peek_char(stream);
usize n;
cstring_t str = cstring_new();
if (ch == core_stream_eof) {
LEX_WARN("Unexpected EOF at begin");
@@ -384,9 +103,9 @@ static void parse_line(smcc_lexer_t *lexer, lexer_tok_t *token) {
const char line[] = "line";
for (int i = 0; i < sizeof(line); i++) {
ch = stream_next_char(stream);
lexer_next_pos(lexer);
for (int i = 0; i < (int)sizeof(line); i++) {
ch = core_stream_next_char(stream);
core_pos_next(&lexer->pos);
if (ch != line[i]) {
LEX_WARN("Maroc does not support in lexer rather in preprocessor, "
"it will be ignored");
@@ -394,38 +113,36 @@ static void parse_line(smcc_lexer_t *lexer, lexer_tok_t *token) {
}
}
parse_number(lexer, token);
if (token->type != TOKEN_INT_LITERAL) {
if (lex_parse_number(lexer->stream, &lexer->pos, &n) == false) {
LEX_ERROR("Invalid line number");
goto SKIP_LINE;
}
if (stream_next_char(stream) != ' ') {
skip_newline(lexer, token);
if (core_stream_next_char(stream) != ' ') {
lex_parse_skip_line(lexer->stream, &lexer->pos);
token->loc.line = token->value.n;
}
if (stream_peek_char(stream) != '"') {
if (core_stream_peek_char(stream) != '"') {
LEX_ERROR("Invalid `#` line");
goto SKIP_LINE;
}
parse_string(lexer, token);
if (token->type != TOKEN_STRING_LITERAL) {
if (lex_parse_string(lexer->stream, &lexer->pos, &str) == false) {
LEX_ERROR("Invalid filename");
goto SKIP_LINE;
}
skip_newline(lexer, token);
token->loc.line = token->value.n;
lex_parse_skip_line(lexer->stream, &lexer->pos);
token->loc.line = n;
// FIXME memory leak
token->loc.name = cstring_as_cstr((const cstring_t *)&token->value.cstr);
token->loc.name_len = cstring_len((const cstring_t *)&token->value.cstr);
token->loc.name = cstring_from_cstr(cstring_as_cstr(&str));
cstring_free(&str);
return;
SKIP_LINE:
skip_newline(lexer, token);
lex_parse_skip_line(lexer->stream, &lexer->pos);
ERR:
set_err_token(token);
cstring_free(&str);
}
// /zh/c/language/operator_arithmetic.html
@@ -434,24 +151,24 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
token->type = TOKEN_UNKNOWN;
core_stream_t *stream = lexer->stream;
stream_reset_char(stream);
core_stream_reset_char(stream);
token_type_t type = TOKEN_UNKNOWN;
int ch = stream_peek_char(stream);
int ch = core_stream_peek_char(stream);
// once step
switch (ch) {
case '=':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '=':
type = TOKEN_EQ;
goto double_char;
default:
stream_reset_char(stream), type = TOKEN_ASSIGN;
core_stream_reset_char(stream), type = TOKEN_ASSIGN;
break;
}
break;
case '+':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '+':
type = TOKEN_ADD_ADD;
goto double_char;
@@ -459,12 +176,12 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_ASSIGN_ADD;
goto double_char;
default:
stream_reset_char(stream), type = TOKEN_ADD;
core_stream_reset_char(stream), type = TOKEN_ADD;
break;
}
break;
case '-':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '-':
type = TOKEN_SUB_SUB;
goto double_char;
@@ -475,48 +192,50 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_DEREF;
goto double_char;
default:
stream_reset_char(stream), type = TOKEN_SUB;
core_stream_reset_char(stream), type = TOKEN_SUB;
break;
}
break;
case '*':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '=':
type = TOKEN_ASSIGN_MUL;
goto double_char;
default:
stream_reset_char(stream), type = TOKEN_MUL;
core_stream_reset_char(stream), type = TOKEN_MUL;
break;
}
break;
case '/':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '=':
type = TOKEN_ASSIGN_DIV;
goto double_char;
case '/':
skip_newline(lexer, token);
lex_parse_skip_line(lexer->stream, &lexer->pos);
token->type = TOKEN_LINE_COMMENT;
goto END;
case '*':
skip_block_comment(lexer, token);
lex_parse_skip_block_comment(lexer->stream, &lexer->pos);
token->type = TOKEN_BLOCK_COMMENT;
goto END;
default:
stream_reset_char(stream), type = TOKEN_DIV;
core_stream_reset_char(stream), type = TOKEN_DIV;
break;
}
break;
case '%':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '=':
type = TOKEN_ASSIGN_MOD;
goto double_char;
default:
stream_reset_char(stream), type = TOKEN_MOD;
core_stream_reset_char(stream), type = TOKEN_MOD;
break;
}
break;
case '&':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '&':
type = TOKEN_AND_AND;
goto double_char;
@@ -524,12 +243,12 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_ASSIGN_AND;
goto double_char;
default:
stream_reset_char(stream), type = TOKEN_AND;
core_stream_reset_char(stream), type = TOKEN_AND;
break;
}
break;
case '|':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '|':
type = TOKEN_OR_OR;
goto double_char;
@@ -537,27 +256,27 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_ASSIGN_OR;
goto double_char;
default:
stream_reset_char(stream), type = TOKEN_OR;
core_stream_reset_char(stream), type = TOKEN_OR;
break;
}
break;
case '^':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '=':
type = TOKEN_ASSIGN_XOR;
goto double_char;
default:
stream_reset_char(stream), type = TOKEN_XOR;
core_stream_reset_char(stream), type = TOKEN_XOR;
break;
}
break;
case '<':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '=':
type = TOKEN_LE;
goto double_char;
case '<': {
if (stream_peek_char(stream) == '=') {
if (core_stream_peek_char(stream) == '=') {
type = TOKEN_ASSIGN_L_SH;
goto triple_char;
} else {
@@ -567,17 +286,17 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
break;
}
default:
stream_reset_char(stream), type = TOKEN_LT;
core_stream_reset_char(stream), type = TOKEN_LT;
break;
}
break;
case '>':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '=':
type = TOKEN_GE;
goto double_char;
case '>': {
if (stream_peek_char(stream) == '=') {
if (core_stream_peek_char(stream) == '=') {
type = TOKEN_ASSIGN_R_SH;
goto triple_char;
} else {
@@ -587,7 +306,7 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
break;
}
default:
stream_reset_char(stream), type = TOKEN_GT;
core_stream_reset_char(stream), type = TOKEN_GT;
break;
}
break;
@@ -595,12 +314,12 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_BIT_NOT;
break;
case '!':
switch (stream_peek_char(stream)) {
switch (core_stream_peek_char(stream)) {
case '=':
type = TOKEN_NEQ;
goto double_char;
default:
stream_reset_char(stream), type = TOKEN_NOT;
core_stream_reset_char(stream), type = TOKEN_NOT;
break;
}
break;
@@ -632,8 +351,8 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_COLON;
break;
case '.':
if (stream_peek_char(stream) == '.' &&
stream_peek_char(stream) == '.') {
if (core_stream_peek_char(stream) == '.' &&
core_stream_peek_char(stream) == '.') {
type = TOKEN_ELLIPSIS;
goto triple_char;
}
@@ -643,17 +362,14 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = TOKEN_COND;
break;
case '\v':
case '\r':
case '\f':
case ' ':
case '\t':
type = TOKEN_BLANK;
break;
case '\r':
case '\n':
// you need to flush a newline or blank
stream_next_char(stream);
lexer_next_line(lexer);
// FIXME some error
lex_parse_skip_endline(lexer->stream, &lexer->pos);
token->type = TOKEN_BLANK;
goto END;
case '#':
@@ -665,17 +381,45 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
// EOF
type = TOKEN_EOF;
break;
case '\'':
parse_char(lexer, token);
case '\'': {
token->loc = lexer->pos;
token->type = TOKEN_CHAR_LITERAL;
int ch = lex_parse_char(lexer->stream, &lexer->pos);
if (ch == core_stream_eof) {
LEX_ERROR("Unexpected character literal");
token->type = TOKEN_UNKNOWN;
} else {
token->value.ch = ch;
}
goto END;
case '"':
parse_string(lexer, token);
}
case '"': {
token->loc = lexer->pos;
token->type = TOKEN_STRING_LITERAL;
cstring_t output = cstring_new();
if (lex_parse_string(lexer->stream, &lexer->pos, &output) == true) {
token->value.cstr.data = cstring_as_cstr(&output);
token->value.cstr.len = cstring_len(&output);
} else {
LEX_ERROR("Unexpected string literal");
token->type = TOKEN_UNKNOWN;
}
goto END;
}
/* clang-format off */
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
/* clang-format on */
parse_number(lexer, token);
token->loc = lexer->pos;
token->type = TOKEN_INT_LITERAL;
usize output;
if (lex_parse_number(lexer->stream, &lexer->pos, &output) == true) {
token->value.n = output;
} else {
LEX_ERROR("Unexpected number literal");
token->type = TOKEN_UNKNOWN;
}
goto END;
/* clang-format off */
case 'a': case 'b': case 'c': case 'd': case 'e': case 'f': case 'g':
@@ -687,25 +431,9 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
case 'O': case 'P': case 'Q': case 'R': case 'S': case 'T': case 'U':
case 'V': case 'W': case 'X': case 'Y': case 'Z': case '_':
/* clang-format on */
// TOKEN_IDENT
// TODO
// if ((ch == 'L' && ch == '\'') || (ch == 'L' && ch == '"')) {
// LEX_ERROR("unsupport wide-character char literal by `L` format");
// }
cstring_t str = cstring_new();
cstring_push(&str, stream_next_char(stream));
lexer_next_pos(lexer);
while (1) {
ch = stream_peek_char(stream);
if ((ch >= 'a' && ch <= 'z') || (ch >= 'A' && ch <= 'Z') ||
(ch == '_') || (ch >= '0' && ch <= '9')) {
stream_next_char(stream);
lexer_next_pos(lexer);
cstring_push(&str, ch);
continue;
}
break;
}
cbool ret = lex_parse_identifier(lexer->stream, &lexer->pos, &str);
Assert(ret == true);
int res = keyword_cmp(cstring_as_cstr(&str), cstring_len(&str));
if (res == -1) {
@@ -724,14 +452,14 @@ void lexer_get_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
}
goto once_char;
triple_char:
stream_next_char(stream);
lexer_next_pos(lexer);
core_stream_next_char(stream);
core_pos_next(&lexer->pos);
double_char:
stream_next_char(stream);
lexer_next_pos(lexer);
core_stream_next_char(stream);
core_pos_next(&lexer->pos);
once_char:
stream_next_char(stream);
lexer_next_pos(lexer);
core_stream_next_char(stream);
core_pos_next(&lexer->pos);
token->type = type;
END:
LEX_DEBUG("get token `%s` in %s:%d:%d", get_tok_name(token->type),
@@ -746,6 +474,7 @@ void lexer_get_valid_token(smcc_lexer_t *lexer, lexer_tok_t *token) {
type = get_tok_subtype(token->type);
AssertFmt(type != TK_BASIC_INVALID, "Invalid token: `%s` at %s:%d:%d",
get_tok_name(token->type), token->loc.name, token->loc.line,
token->loc.column);
token->loc.col);
Assert(type != TK_BASIC_INVALID);
} while (type == TK_BASIC_EMPTYSPACE || type == TK_BASIC_COMMENT);
}