Jump to: navigation, search

Orion/Server API/Transfer API

Under Construction: This API is not yet implemented

The transfer API is used for batch import and export of directory trees to/from an Orion workspace.

Import

Starting a file upload

Overview
An upload is initiated via a POST request. The request URL indicates where the file should be located once the upload is complete. The request must indicate the total size of the file the server should expect.
HTTP Method
POST
Example Request
POST /xfer/MyProj/SomeFolder
Slug: MyTitle
X-Xfer-Content-Length: 901024

  
Example Response
HTTP/1.1 200 OK
Location: /xfer/import/23A43


Detailed Explanation
The Location header in the response indicates the URL where chunks of the file should be uploaded. Once the server has received all chunks, the import into the workspace is completed. An upload that is not completed in a timely fashion from the initial POST is discarded by the server.


After initiating the transfer, uploads are performed as many times as required using PUT actions:


Overview
Uploads a chunk of a file to the server.
HTTP Method
PUT
Example Request
PUT /xfer/import/23A43
Content-Length: 32768
Content-Type: application/zip
Content-Range: bytes 0-32767/901024

  
Example Response
HTTP/1.1 308 Resume Incomplete
Range: bytes 0-30000


Detailed Explanation
A response of 308 indicates that the server has not yet received all required bytes, and the Range response header indicates the bytes known to the server so far. When the server has received all bytes, a 201 Created response is returned, with a Location header specifying the location of the newly added file or directory.



  • POST request, returns a URI for resumable upload (uploading in chunks will allow for progress reporting). Request indicates whether we are uploading a file that should be stored directly in the workspace, or whether it is an archive that will need to be extracted into the workspace on the server.
  • POST Response Location indicates upload URL
  • Series of PUT requests to upload URL, using Content-Range header
  • Responses indicate what total range server already knows
  • Once server has all the bytes the file is stored or extracted into the workspace

Export

  • GET request /xfer/export/<filename.zip>
  • Request body contains:
    • Array of files/directories to export - depth infinity is assumed, children should be listed independently if depth infinity is not desired
    • Content type of response (zip, tar, etc)