Run a SQL query against a SeaTable server

Usage

seatable_query(
  sql,
  con = default_connection(),
  limit = 100000L,
  base = NULL,
  python = FALSE,
  convert = TRUE,
  paginate = TRUE,
  chunksize = NULL,
  collapse_lists = TRUE,
  retries = 3L,
  progress = interactive()
)

Arguments

sql: A single SQL SELECT statement. The table name is parsed from the FROM clause.
con: A seatable_connection.
limit: Maximum rows to return. The default fetches all rows, paginating as needed.
base: Optional base name or Base object; discovered from the table name when NULL.
python: If TRUE, return the raw pandas DataFrame instead of an R data.frame.
convert: Passed to the SeaTable Python query call.
paginate: Whether to page through large results automatically (default TRUE). SeaTable's SQL endpoint silently caps a single call at a server-specific maximum; pagination returns the full result.
chunksize: Advanced: force LIMIT/OFFSET paging in fixed windows of this size. The default NULL auto-detects the server's per-call cap.
collapse_lists: Whether to collapse multiple-select (and other list) columns into simple character vectors. The default TRUE comma-joins multi-valued cells; FALSE keeps them as list-columns. A string is used as the separator instead of ",".
retries: Number of times to retry a page that fails because the server is rate-limiting us (HTTP 429), with exponential backoff between attempts. Long paginated reads are the usual way to hit a quota, and losing the whole read to one throttled page is expensive. Set to 0 to fail fast.
progress: Whether to report each page of a paginated read as it arrives. Defaults to TRUE in an interactive session. Reading a large table takes minutes, and silence is hard to distinguish from a hang.

Value

An R data.frame (or a pandas DataFrame when python = TRUE). If a paginated read stops early because a page kept failing, the rows read so far are returned along with a warning saying where it stopped, rather than throwing away a read that may have taken minutes.

Examples

if (FALSE) { # \dontrun{
con <- seatable_connection("https://cloud.seatable.io/", "BANCTABLE_TOKEN",
                           workspace_id = "57832")
seatable_query("SELECT * FROM banc_meta", con = con)
} # }