This chart provides as an overview of how content is added to the IR. It highlights key differences in the various methods. Links to more detail “how-tos” are provided below.
|
Submission Form
|
Metadata Import
|
Programming Script
|
SAF Method
|
Sword
|
Type
|
Manual
|
Partial Batch (metadata only)
|
Full Batch (metadata plus files)
|
Full Batch (metadata plus files) |
Auto Import
|
Ingest target
|
Per collection
Individual record
|
Across multiple collections
Max records: 3,000
|
Per collection
No max
typically used when have special conditions for ingest
|
Per collection
> 400 MB via GUI or used this method to prepare larger ingest requests for programmer to run via command line.
|
Specified collections only
|
Permissions
|
Collection submitter1
Community/Collection admin
|
Curators (Amanda, Shannon, Monica).
System Admin.
Process through Admin GUI
|
Programmers (Ying or Sid).
System Admin.
Process through Command line
|
DSpace Admin
Process via GUI
|
|
Typical Use Cases
|
Faculty Publications and special projects
|
- Ingest new items by library staff for single or small sets (e.g. UA, Digitized T/Ds, Faculty Publications)
|
|
New item ingests only
Useful method for both access and masters
|
possible future options:
-
Shared Shelf
-
Research Data
|
Metadata Input format
|
Online form2
|
CSV
|
Excel
|
XML
compile SAF via overnight perl script on project server
|
Sword / “Push” API
|
Requirements
|
-
Title
-
File (bitstream)
-
Non-exclusive license
- additional fields may be required per collection.
|
-
First two columns must be id and collection
-
Include lang. tag in headers
-
Use double pipe (||) to separate multiple values
-
Recommended but not required: dc.identifier.digital (object id)
-
Files must be added manually (on the same day as ingest)
|
-
Do Not include lang. tag in headers
-
Use semi colon to separate multiple values
-
Include dc.identifier.digital (object id) which must match filename prefix3
-
filename must be unique and no id can be a prefix of another filename4
-
Files stored either under one subfolder “contents” or in subfolder by object id (if more than one file is associated with object)
-
Provide collection handle in ingest request and give organization / location of files
|
-
Do Not include lang. tag in headers
-
Use semi colon to separate multiple values
-
Include dc.identifier.digital (object id) which must match filename prefix3
-
filename must be unique and no id can be a prefix of another filename4
-
Files stored under one subfolder “contents”
|
|
Optional features
|
|
|
bitstream descriptions
|
bitstream descriptions |
|
additional info or guidelines
Deposit Checklist
|
|
Metadata batch process
|
|
DSpace UI Batch Ingest (SAF) guidelines (google doc restricted to Rice) |
|
Notes
1 Collection submitter may only upload files at time of submission. Further edits (replacing, adding or deleting) files requires system admin permission.
2 The default submission form for the IR is designed for deposits of articles or other types of faculty publications. For example it includes fields for embargo periods, dc.types related to faculty papers and articles and contains fields for issue identification. Therefore using this default form for non-faculty papers may require editing the record to add fields recommended in other types of collections.
3 The filename prefix is defined as the alphanumeric string before any underscore symbol, suffix (e.g. page number) and file extension
4 The filename must be unique and no object id can be a prefix of another filename
Comments (0)
You don't have permission to comment on this page.