feat: match 表达式 (P1 #8 收官)

- lexer: TOK_MATCH, TOK_MATCH_ARROW, TOK_UNDERSCORE
- parser: parse_match_stmt() desugar → let+if-else链
- 零 sema/codegen 改动
- 4个集成测试: enum/int literal/wildcard match

P1 全部完成: type alias + enum + array + impl + match

src/lexer/lexer.c

@@ -63,7 +63,8 @@ static TokenKind check_keyword(const Token* tok) {
     KW("bool",   TOK_BOOL);   KW("str",    TOK_STR);
     KW("void",   TOK_VOID);
     KW("struct", TOK_STRUCT); KW("type", TOK_TYPE);
-    KW("enum", TOK_ENUM);  KW("impl", TOK_IMPL);
+    KW("enum", TOK_ENUM);  KW("impl", TOK_IMPL);  KW("match", TOK_MATCH);
+    KW("_", TOK_UNDERSCORE);
     KW("true",   TOK_TRUE);   KW("false",  TOK_FALSE);
     #undef KW
     return TOK_IDENT;
@@ -126,6 +127,7 @@ Token* lex(Arena* a, const char* source, const char* filename,
         else if (c == '/') { tokens[idx++] = make_token(&l, TOK_SLASH, l.pos, 1); advance(&l); }
         else if (c == '%') { tokens[idx++] = make_token(&l, TOK_PERCENT, l.pos, 1); advance(&l); }
         else if (c == '=' && peek_next(&l) == '=') { tokens[idx++] = make_token(&l, TOK_EQ_EQ, l.pos, 2); advance(&l); advance(&l); }
+        else if (c == '=' && peek_next(&l) == '>') { tokens[idx++] = make_token(&l, TOK_MATCH_ARROW, l.pos, 2); advance(&l); advance(&l); }
         else if (c == '=') { tokens[idx++] = make_token(&l, TOK_ASSIGN, l.pos, 1); advance(&l); }
         else if (c == '!' && peek_next(&l) == '=') { tokens[idx++] = make_token(&l, TOK_BANG_EQ, l.pos, 2); advance(&l); advance(&l); }
         else if (c == '!') { tokens[idx++] = make_token(&l, TOK_BANG, l.pos, 1); advance(&l); }