Move rows into and out of SeaTable big data storage

SeaTable caps the number of rows a base holds in its ordinary grid; archiving moves rows into "big data" storage, which is much larger and still queryable with seatable_query(), at the cost of no longer being editable in the web interface. seatable_archive_rows() moves rows out of the grid, seatable_unarchive_rows() brings them back.

Usage

seatable_archive_rows(
  table,
  view_name = NULL,
  view_id = NULL,
  base = NULL,
  con = default_connection()
)

seatable_unarchive_rows(
  table,
  row_ids,
  base = NULL,
  con = default_connection()
)

Arguments

table

Name of the table.

view_name

Name of the view whose rows should be archived. Supply exactly one of view_name or view_id.

view_id

Id of the view whose rows should be archived.

base

Optional base name or Base object; discovered from table when NULL.

con

A seatable_connection.

row_ids

Character vector of row ids (the _id column) to unarchive.

Value

The SeaTable response, invisibly.

Details

The two directions select rows differently, because the API does. Archiving takes a view: make a view holding exactly the rows you want archived (a filtered view is the usual way) and name it. Earlier versions of the endpoint accepted a SQL WHERE clause; that is gone, and a view is now the only way to choose rows. Unarchiving instead takes explicit row_ids, which you can get by querying the archived rows with seatable_query("SELECT _id FROM ...").

Archiving is not easily undone in bulk — you need the row ids to bring rows back — so check the view holds what you expect before calling.

Examples

if (FALSE) { # \dontrun{
# Archive everything in the (filtered) view called "to archive"
seatable_archive_rows("neurons", view_name = "to archive", con = con)

# ...and bring two rows back
seatable_unarchive_rows("neurons",
  row_ids = c("FoDxhChYQSycLm88JZ11RA", "Yb3kQ0rRQ0aXk3lWk0mWLQ"),
  con = con)
} # }