Discover and read tables through a Fabric SQL endpoint — fabric_sql

These helpers provide a target-independent metadata and table-read layer for Fabric SQL endpoints. They accept Lakehouse, Warehouse, Warehouse snapshot, and SQL Database objects, or the same direct server inputs as fabric_sql_query(). Discovery uses SQL catalog metadata views and is limited by the caller's SQL metadata permissions.

Usage

fabric_sql_tables(
  server,
  schema = NULL,
  detail = TRUE,
  database = NULL,
  target_type = c("auto", "lakehouse", "warehouse", "sql_database",
    "sql_analytics_endpoint"),
  backend = c("odbc", "adbc"),
  token = NULL,
  ...
)

fabric_sql_views(
  server,
  schema = NULL,
  detail = TRUE,
  database = NULL,
  target_type = c("auto", "lakehouse", "warehouse", "sql_database",
    "sql_analytics_endpoint"),
  backend = c("odbc", "adbc"),
  token = NULL,
  ...
)

fabric_sql_read_table(
  server,
  table,
  schema = NULL,
  columns = NULL,
  limit = NULL,
  result = c("tibble", "arrow_stream"),
  database = NULL,
  target_type = c("auto", "lakehouse", "warehouse", "sql_database",
    "sql_analytics_endpoint"),
  backend = c("odbc", "adbc"),
  token = NULL,
  ...
)

Arguments

server: Fabric SQL endpoint, portal connection string, or discovered SQL-capable item object.
schema: Optional schema filter. fabric_sql_read_table() defaults to the schema in a discovered table row, otherwise "dbo".
detail: Whether table or view discovery should retrieve column metadata.
database: Optional catalog/database. An explicit value overrides one discovered from server.
target_type: Kind of Fabric SQL target. Keep "auto" unless a custom hostname prevents automatic identification.
backend: SQL driver backend, either "odbc" or "adbc".
token: Optional SQL access token or audience-aware token-provider function.
...: Additional authentication, driver, endpoint, timeout, verbosity, and retry options passed to fabric_sql_query(). Query text, parameters, read-only status, and idempotency are controlled by these helpers and cannot be supplied here.
table: Table or view name, or a one-row record containing name and optionally schema.
columns: Optional unique column names to project.
limit: Optional non-negative maximum number of rows to return.
result: Result representation for fabric_sql_read_table(); either a tibble or a single-use Arrow stream.

Value

fabric_sql_tables() and fabric_sql_views() return a tibble with object name, schema, full_name, type, optional view definition, list-column columns, and the unmodified discovery row in raw. fabric_sql_read_table() returns a tibble or nanoarrow_array_stream.

References

System information schema views

SQL module definitions

Fabric SQL analytics endpoints

Examples

if (FALSE) { # \dontrun{
workspace <- fabric_workspaces()[[1L]]
warehouse <- fabric_warehouses(workspace)[[1L]]

tables <- fabric_sql_tables(warehouse, schema = "dbo")
rows <- fabric_sql_read_table(warehouse, tables[1L, ], limit = 1000)
views <- fabric_sql_views(warehouse)
} # }