datapack (version 1.0.0)

DataObject-class: DataObject wraps raw data with system-level metadata

Description

DataObject is a wrapper class that associates raw data or a data file with system-level metadata describing the data. The system metadata includes attributes such as the object's identifier, type, size, checksum, owner, version relationship to other objects, access rules, and other critical metadata. The SystemMetadata is compliant with the DataONE federated repository network's definition of SystemMetadata, and is encapsulated as a separate object of type SystemMetadata that can be manipulated as needed. Additional science-level and domain-specific metadata is out-of-scope for SystemMetadata, which is intended only for critical metadata for managing objects in a repository system.

Methods

• initialize

{: Initialize a DataObject} getData{: Get the data content of a specified data object} getIdentifier{: Get the Identifier of the DataObject} getFormatId{: Get the FormatId of the DataObject} setPublicAccess{: Add a Rule to the AccessPolicy to make the object publicly readable.} addAccessRule{: Add a Rule to the AccessPolicy} canRead{: Test whether the provided subject can read an object.}

Details

A DataObject can be constructed by passing the data and SystemMetadata to the new() method, or by passing an identifier, data, format, user, and DataONE node identifier, in which case a SystemMetadata instance will be generated with these fields and others that are calculated (such as size and checksum).

Data are associated with the DataObject either by passing it as a 'raw' value to the 'dataobj' parameter in the constructor, which is then stored in memory, or by passing a fully qualified file path to the data in the 'filename' parameter, which is then stored on disk. One of dataobj or filename is required. Use the 'filename' approach when data are too large to be managed effectively in memory. Callers can access the 'filename' slot to get direct access to the file, or can call 'getData()' to retrieve the contents of the data or file as a raw value (but this will read all of the data into memory).

See Also

datapack

Examples

data <- charToRaw("1,2,3\n4,5,6\n")
do <- new("DataObject", "id1", dataobj=data, "text/csv", 
  "uid=jones,DC=example,DC=com", "urn:node:KNB")
getIdentifier(do)
getFormatId(do)
getData(do)
canRead(do, "uid=anybody,DC=example,DC=com")
do <- setPublicAccess(do)
canRead(do, "public")
canRead(do, "uid=anybody,DC=example,DC=com")
# Also can create using a file for storage, rather than memory
tf <- tempfile()
con <- file(tf, "wb")
writeBin(data, con)
close(con)
do <- new("DataObject", "id1", format="text/csv", user="uid=jones,DC=example,DC=com", 
  mnNodeId="urn:node:KNB", filename=tf)