[R-sig-DB] SQL escaping/quoting proposal

[Hadley Wickham]

> Regarding (b), my reasoning is that when working with Postgres, I would want
> to allow Postgres to do its normal lower casing of unquoted identifiers.
> That is, `thisField` is a valid identifier, but passed in quoted case will
> be preserved, which means always having to quote the identifier in the
> future. Passed in quoted, Postgres will force it to `thisfield`, and if a
> client requests `thisField` unquoted, the correct field will be returned.

It might be reasonable to opt in to this behaviour, but it's dangerous
by default because R is case-sensitive. What happens if you're using
dbWriteTable with this data.frame?

df <- data.frame(a = 1, A = 2)

Hadley

-- 
Chief Scientist, RStudio
http://had.co.nz/