You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I think there are two ways to solve the problem, the first one is to separate the struct in Tokenizer, like in
we could try to match "struct" here and if it matches, do the same logic as the non-quote "match".
The second one is to do a rewrite in Datafusion, but it also includes the parse of "t1."struct(xxxxx)", but we have to deal with I think the best way to do it is in SQL-parser.
However, the tricky part is how to determine the pattern of parse quoted "struct(xxx)", since the tokenizer's next() method does not have a state to save the previous token(my current thought is when the last token is Period and the dialect is a general or big query, we do the logics above). I don't know whether it is the best way. Could you please give me a hint about it? @alamb
related to apache/datafusion#9891
The text was updated successfully, but these errors were encountered: