Most dms do not provide support for all fields, therefore nearly all fields are optional. If your dms does not provide support for eg. a "project" or a "directory" field - simply omit them in the ´meta.json`. During import you can ignore those fields not supported by your dms.
This is highly dependent on the dms. It contains of the following simple generics steps:
- Iterate over all elements in the gunzipped/untared export-archive
- Untar all encountered containers
- Import the found document
You might use the reference implementation dms-exchange-java for this task.
That's up to the dms provider. Some suggestions/examples how it could work:
- A simple download-link
- An option to issue an export, which creates a background-job that processes the documents (large amounts of documents might take a while), and notifies the user when the download is ready.
To my knowledge I found the following related formats (give me a hint if you know others):
- docTag - An initiative by mostly billing-related services that has been started and ceased in 2012. Provides some JSON-Schemas that are invalid (eg. 'date' is no valid format) and code-examples.
- ODMA (Open Document Management Alliance) - The ODMA defined APIs for desktop applications accessing dms server. As far I could see it is DLL based and regarding to Wikipedia superseded by other ways of accessing the data inside an dms, namly WebDAV. The website looks out-dated and seems not be maintained anymore.