Hey @auxten! First of all congratulations on launching this project!! I'm actually excited to crate the first Issue!! I was thinking if I could help out in implementing the DDL. I've myself implemented a simple DB in Python and although it did not support DDL, it was possible to create custom tables(all columns are strings) and the start. You can check it out here NaiveDB. I'm a newbie to the world of Golang, and therefore would like some guidance from you on how i would go about helping you out.

感谢你开源的项目啊，对于学习数据库很有帮助。 请教一个问题：我看测试用例 SELECT id, username, FROM table WHERE id > 5 AND id < 7 LIMIT 3 查询的时候，过程是，先遍历所有的row，再在内存中过滤，最后返回满足要求的row。 我理解的是应该会先根据where条件在B+树种过滤好再返回吧，而不是在内存中，而且也只会返回 id, username 其他的字段不会返回（这个字段过滤是在内存中吗？还是说在存储层就已经过滤了？）。还是我哪里理解错了？谢谢。

Translation of the question:

Thank you for your open source project, it is very helpful for learning database theory. Ask a question: I see the test case SELECT id, username, FROM table WHERE id> 5 AND id <7 LIMIT 3 When querying, the process is to first traverse all rows, then filter in memory, and finally return the row that meets the requirements. . What I understand is that it should be filtered in the B-Tree based on where conditions before returning, not in memory, and only id will be returned, and other fields of username will not be returned (is this field filtered in memory? Or Say it has been filtered at the storage layer?). Or am I getting it wrong? Thank you.

when trying example i get error.

this is due to firefox being a json browser and the examples are returning plain/text but saying they are returning json.

writer.Header().Set("Content-Type", "application/json")
q := request.URL.Query()
query := q.Get("q")
if query != "" {
	var (
		ast *parser.SelectTree
	)

	p := &parser.Parser{}
	if p.GetSQLType(query) != parser.SELECT {
		_, _ = fmt.Fprintf(writer, "not a SELECT statement")
		return
	}

for all the error printing lines, this will fix;

example; http.Error(writer, "not a SELECT statement", http.StatusBadRequest) from _, _ = fmt.Fprintf(writer, "not a SELECT statement")

but the data written will also have to be converted to json.