Make a table description directly.

Build minimal structures (table name and column names) needed to represent data from a remote table.

mk_td(
  table_name,
  columns,
  ...,
  qualifiers = NULL,
  q_table_name = NULL,
  head_sample = NULL,
  limit_was = NULL
)

table_source(
  table_name,
  columns,
  ...,
  qualifiers = NULL,
  q_table_name = NULL,
  head_sample = NULL,
  limit_was = NULL
)

Arguments

table_name	character, name of table
columns	character, column names of table (non-empty and unique values).
...	not used, force later argument to bind by name
qualifiers	optional named ordered vector of strings carrying additional db hierarchy terms, such as schema.
q_table_name	optional character, qualified table name, note: has to be re-generated for different DB connections.
head_sample	optional, head_sample of table as an example
limit_was	optional, row limit used to produce head_sample.

Value

a relop representation of the data

Details

Generate a query that returns contents of a table, we could try to eliminate this (replace the query with the table name), but there are features one can work with with the query in place and SQL optimizers likely make this zero-cost anyway.

Functions

• table_source: old name for mk_td

See also

db_td, local_td

db_td, local_td, rq_copy_to, materialize, execute, to_sql

Examples

if (requireNamespace("DBI", quietly = TRUE) && requireNamespace("RSQLite", quietly = TRUE)) {
  my_db <- DBI::dbConnect(RSQLite::SQLite(), ":memory:")
  rq_copy_to(my_db,
             'd',
             data.frame(AUC = 0.6, R2 = 0.2),
             overwrite = TRUE,
             temporary = TRUE)
  d <- mk_td('d',
                    columns = c("AUC", "R2"))
  print(d)
  sql <- to_sql(d, my_db)
  cat(sql)
  print(DBI::dbGetQuery(my_db, sql))
  DBI::dbDisconnect(my_db)
}
#> [1] "mk_td(\"d\", c( \"AUC\", \"R2\"))"
#> SELECT
#>  `AUC`,
#>  `R2`
#> FROM
#>  `d`
#>   AUC  R2
#> 1 0.6 0.2