Aside from providing a great way to organise your data set and add metadata to it, CloudStor Collections does a number of things that make the Collections different from normal zip files.
- Checksums . When a collection is packaged, CloudStor Collections creates a file that contains a unique code (known as a checksum) for every file in the collection. Importantly, any changes to a file will result in a different unique number being created when the Collection is packed. By comparing the checksum values in the manifest to the contents of a file, it is possible to tell whether the contents of the collection have changed after it was packed. This is very important for ensuring that once a collection is created, the contents cannot be unknowingly altered.
- Human readability . Each collection contains a ‘readme.html’ file. This file contains the metadata associated with the collection, as well as a complete list of files and folders. This file is laid out in a way that is easy for another person to read.
- Machine readability . Each collection also contains a ‘manifest.json’ file. Like the readme.html, this file contains details of the metadata and data in the collection, only in this case it is designed to be read by another program. This means that other programs can work out what each collection contains without sifting through all the contents of the collection.