Download or publish CRANT synapse snapshots

crant_download_snapshot() downloads the newest published synapse snapshots into your snapshot folder; select one with crant_use_snapshot() to use it. The first download fetches the static data and a full snapshot (about 1.8 GB); later ones usually just fetch a few small checkpoints.

crant_publish_snapshot() is for whoever maintains the snapshots: it makes the newest snapshot in root, and the earlier checkpoints it is built on, available for download from a folder served over https.

Usage

crant_download_snapshot(url = NULL, root = crant_snapshot_root(create = TRUE))

crant_publish_snapshot(dest, root = crant_snapshot_root(), snapshot = "latest")

Arguments

url

The address of the published snapshots (the folder or its manifest.json). The default (NULL) uses the crantr.snapshot_url option if set, and otherwise the standard address, which needs access to the CRANT CAVE datastack to work out.

root

The local snapshot folder. Defaults to crant_snapshot_root().

dest

The folder to publish into.

snapshot

The snapshot to publish; "latest" (the default) for the newest one in root.

Value

crant_download_snapshot(): the tags downloaded, invisibly. crant_publish_snapshot(): the manifest, invisibly.

Details

Snapshots are published with a manifest.json that lists every file with its size and md5. A download only fetches snapshots that are not already in root (snapshots never change once made) and checks each file's md5, so an interrupted download is never mistaken for a snapshot; just run it again. Downloads use the curl command line tool and resume partial files.

Publishing hard links the files into dest when it is on the same filesystem as root (otherwise they are copied), so it takes no extra space. manifest.json is written last.

See also

crant_use_snapshot(), crant_update_snapshot()

Examples

if (FALSE) { # \dontrun{
crant_download_snapshot()
} # }