Restore a base from a dump — air

Recreates a base from output of air_dump(). When attachments is "file", uploads attachment files from the dump directory after record creation. For CSV dumps, automatically detects and parses the flattened format.

Usage

air_restore(
  dump,
  base_name = NULL,
  workspace_id = NULL,
  attachments = c("file", "meta"),
  attachment_dir = NULL,
  restore_linked_fields = TRUE,
  .token = NULL
)

Arguments

dump: Either a list (from air_dump(format = "list")) or a path to a dump directory (from air_dump(format = "json") or air_dump(format = "csv")).
base_name: Name for the new base. If NULL, uses a generated name.
workspace_id: Workspace ID to create the base in.
attachments: How to handle attachment fields: "meta" (default) keeps only metadata (filename, url, size, type); "file" downloads to attachment_dir; "blob" downloads as in-memory raw vectors.
attachment_dir: Directory for downloading attachments (required when attachments = "file"). Files are saved as {attachment_dir}/{record_id}/{filename}.
restore_linked_fields: If TRUE (the default), after all records are created, linked-record fields (multipleRecordLinks) and their dependent computed fields (rollup, lookup, count) are recreated with remapped table/field IDs, and the link cell values (record-to-record connections) are repopulated by remapping old record IDs to the newly created record IDs. Set to FALSE to skip this step (faster, but links will be empty).
.token: Personal access token (resolved via air_token() if NULL).

Value

The new base ID (invisibly).

Linked-record fields

Linked-record fields (multipleRecordLinks) and the computed fields that depend on them (rollup, lookup, count) are skipped during the initial field-creation pass because their options reference base-specific IDs. When restore_linked_fields = TRUE (the default), after all records are inserted a two-step pass runs:

Field definitions: multipleRecordLinks fields are recreated with linkedTableId remapped to the new base's table IDs.
Cell values: the link columns in the dump contain old record IDs. These are remapped to the new record IDs (matched by insertion order) and written back via air_upsert().

Examples

if (FALSE) { # \dontrun{
# Restore from a directory dump
air_restore("backup/", workspace_id = "wspXXXXXX")

# Restore from a CSV dump
air_restore("backup/", workspace_id = "wspXXXXXX", format = "csv")
} # }