Resolves one versioned object in cloud storage. When version = "latest",
returns the most recently updated object matching the prefix and extension.
Usage
cloud_object_name(
prefix,
version = "latest",
extension = "",
provider,
exact_match = FALSE,
options
)Arguments
- prefix
A character string specifying the file prefix.
- version
A character string specifying the version. Default is "latest".
- extension
A character string specifying the file extension.
- provider
A character string specifying the cloud storage provider.
- exact_match
A logical value indicating whether to match prefix exactly.
- options
A named list of cloud storage provider options.
Value
A character string with the full object name, or character(0) if
no matching objects are found.
Resolves one object, never enumerates a bucket
This function is not a listing helper. It always returns a single name —
selected_rows$name[1] — even though it groups internally by base_name and
ext. Point it at a prefix that matches many distinct base names (for
example one object per GPS trip, where tens of thousands share a prefix) and
it silently hands back an arbitrary one of them rather than erroring. It is
built for the versioned-snapshot pattern <prefix>__<version>__.<ext>, where
the prefix identifies exactly one logical dataset and the only question is
which version of it to take.
To enumerate objects, use cloud_object_names(), which returns every match.
The single-name return type is depended upon by downstream pipelines and will not change.
See also
cloud_object_names() for enumeration.