Support join_by()
#1074
Support join_by()
#1074
Conversation
R/join-cols-compat.R
Outdated
| check_dots_empty0(...) | ||
|
|
||
| if (!is.character(vars)) { | ||
| message <- glue("Join columns in `{input}` must be character vectors.") |
There was a problem hiding this comment.
When does this error arise? Is it an internal error?
There was a problem hiding this comment.
Probably, it is an internal error. I basically only copied over the file from dplyr https://github.com/tidyverse/dplyr/blob/main/R/join-cols.R.
| return(x) | ||
| check_join_by_supported <- function(by, call = caller_env()) { | ||
| if (any(by$filter != "none")) { | ||
| cli_abort("Rolling joins aren't supported on database backends.", call = call) |
There was a problem hiding this comment.
It seems likely we'll need to relax this on a per backend basis at some point, but I don't think we need to do it prospectively.
There was a problem hiding this comment.
I don't think there are many backends that support ASOF style joins though. None of the common ones like Postgres, SQLite, MS SQL Server, etc
There was a problem hiding this comment.
But at least there are usually some workarounds to get the same results. In case I need a rolling join in the database I might just implement the workaround in dbplyr 😄
| @@ -284,7 +284,8 @@ sql_join_tbls <- function(con, by, na_matches = "never") { | |||
| sql_expr_matches(sql(lhs[[i]]), sql(rhs[[i]]), con = con) | |||
| }) | |||
| } else { | |||
| compare <- paste0(lhs, " = ", rhs) | |||
| by$condition[by$condition == "=="] <- "=" | |||
There was a problem hiding this comment.
This is the only real change to the SQL generation right? Pretty much everything else is improved input validation and new handling of output columns?
Support for the awesome new
join_by()clause 😄