Page History
Table of Contents | ||||||
---|---|---|---|---|---|---|
|
Item Importer and Exporter
DSpace has a set of command line tools for importing and exporting items in batches, using the DSpace Simple Archive Format (SAF). Apart from the offered functionality, these tools serve as an example for users who aim to implement their own item importer.
DSpace Simple Archive Format
The basic concept behind the DSpace's Simple Archive Format (SAF) is to create an archive, which is a directory containing one subdirectory per item. Each item directory contains a file for the item's descriptive metadata, and the files that make up the item.
Code Block |
---|
archive_directory/ item_000/ dublin_core.xml -- qualified Dublin Core metadata for metadata fields belonging to the 'dc' schema. metadata_[prefix].xml -- metadata in another schema. The prefix is the name of the schema as registered with the metadata registry. contents -- text file containing one line per filename. collections -- (Optional) text file that contains the handles of the collections the item will belong to. Each handle in a row. -- Collection in first line will be the owning collection. handle -- contains the handle assigned/to be assigned to this resource relationships -- (Optional) If importing Entities, you can specify one or more relationships to create on import file_1.doc -- files to be added as bitstreams to the item. file_2.pdf item_001/ dublin_core.xml contents file_1.png ... |
dublin_core.xml or metadata_[prefix].xml
The dublin_core.xml
or metadata_[prefix].xml
file has the following format, where each metadata element has its own entry within a <dcvalue>
tagset. There are currently three tag attributes available in the <dcvalue>
tagset:
element
- the Dublin Core elementqualifier
- the element's qualifierlanguage
- (optional) ISO language code for elementCode Block <dublin_core> <dcvalue element="title" qualifier="none">A Tale of Two Cities</dcvalue> <dcvalue element="date" qualifier="issued">1990</dcvalue> <dcvalue element="title" qualifier="alternative" language="fr">J'aime les Printemps</dcvalue> </dublin_core>
(Note the optional language tag attribute which notifies the system that the optional title is in French.)
When providing urls as values for fields that contain the ampersand (&) symbol, the ampersands in these urls have to be encoded as &
Every metadata field used, must be registered via the metadata registry of the DSpace instance first. See Metadata and Bitstream Format Registries.
Note | ||
---|---|---|
| ||
It is recommended to minimally provide "dc.title" and, where applicable, "dc.date.issued". Obviously you can (and should) provide much more detailed metadata about the Item. For more information see: Metadata Recommendations. |
contents file
The contents
file is a plain text document that simply enumerates, one file per line, the bitstream file names. See the following example:
...
Please notice that the license
is optional, and if you wish to have one included, you can place the file in the .../item_001/ directory, for example.
The bitstream name may optionally be followed by any of the following:
\tbundle:BUNDLENAME
\tpermissions:PERMISSIONS
\tdescription:DESCRIPTION
\tprimary:true
Where '\t' is the tab character.Tab character (not a literal "\t" string)
'BUNDLENAME' is the name of the bundle to which the bitstream should be added. Without specifying the bundle, items will go into the default bundle, ORIGINAL.
'PERMISSIONS' is text with the following format: -[r|w] 'group name'
- -r = read permissions for this group
- -w = write permissions for this group
'DESCRIPTION' is text of the files description.
...
'IIIFHEIGHT' is the image height that will be used for the IIIF canvas.
relationships file
Note | ||
---|---|---|
| ||
This feature was added in 7.1. Currently the 'relationships' file is only supported on import ('add' mode) of an SAF package. See note at bottom of this section about using the "metadata_relation.xml" if you wish to export & update relationships. |
...
Note | |||||
---|---|---|---|---|---|
| |||||
If you already know the UUID of an existing Entity that you want to relate to, you can also create/update the "metadata_relation.xml" file to add/update the relationship, similar to:
The "relationships" file is primarily for creating relationships between Entities in the same import batch. Of course, you can also choose to use the "relationships" file to create new relationships to existing Entities instead of creating/updating the "metadata_relation.xml" file. The main advantage of the "metadata_relation.xml" file is that it is used both on export and import, while the "relationships" file is only used on import at this time. |
Configuring metadata_[prefix].xml
for a Different Schema
It is possible to use other Schema such as EAD, VRA Core, etc. Make sure you have defined the new schema in the DSpace Metadata Schema Registry.
- Create a separate file for the other schema named
metadata_[prefix].xml
, where the[prefix]
is replaced with the schema's prefix. - Inside the xml file use the same Dublin Core syntax, but on the
<dublin_core>
element include the attributeschema=[prefix]
. Here is an example for ETD metadata, which would be in the file
metadata_etd.xml
:Code Block <dublin_core schema="etd"> <dcvalue element="degree" qualifier="department">Computer Science</dcvalue> <dcvalue element="degree" qualifier="level">Masters</dcvalue> <dcvalue element="degree" qualifier="grantor">Michigan Institute of Technology</dcvalue> </dublin_core>
Importing Items
Before running the item importer over items previously exported from a DSpace instance, please first refer to Transferring Items Between DSpace Instances.
Command used: |
|
Java class: |
|
Arguments short and (long) forms: | Description |
| Add items to DSpace ‡ |
| Replace items listed in mapfile ‡ |
| Delete items listed in mapfile ‡ |
| Source of the items (directory) |
| Destination Collection by its Handle or database ID |
| Where the mapfile for items can be found (name and directory) |
| Email of eperson doing the importing |
| Send submission through collection's workflow |
| Kicks off the email alerting of the item(s) has(have) been imported |
| Test run‚ do not actually import items |
| Apply the collection template |
| Resume a failed import (Used on Add only) |
| Command help |
| Name of zipfile |
...
The item importer is able to batch import unlimited numbers of items for a particular collection using a very simple CLI command and 'arguments'.
Adding Items to a Collection from a directory
To add items to a collection, you gather the following information:
...
The above command would cycle through the archive directory's items, import them, and then generate a map file which stores the mapping of item directories to item handles. SAVE THIS MAP FILE. You can use it for replacing or deleting (unimporting) the mapped items.
Testing. You can add --testvalidate
(or -tv
) to the command to simulate the entire import process without actually doing the import. This is extremely useful for verifying your import files before doing the actual import.
Adding Items to a Collection from a zipfile
To add items to a collection, you gather the following information:
...
The above command would unpack the zipfile, cycle through the archive directory's items, import them, and then generate a map file which stores the mapping of item directories to item handles. SAVE THIS MAP FILE. You can use it for replacing or deleting (unimporting) the mapped items.
Testing. You can add --testvalidate
(or -tv
) to the command to simulate the entire import process without actually doing the import. This is extremely useful for verifying your import files before doing the actual import.
Replacing Items in a Collection
Replacing existing items is relatively easy. Remember that mapfile you saved above? Now you will use it. The command (in short form):
...
Code Block |
---|
[dspace]/bin/dspace import -r -e joe@user.com -c collectionID -s zipfile_dir -z filename.zip -m mapfile |
Deleting or Unimporting Items in a Collection
You are able to unimport or delete items provided you have the mapfile. Remember that mapfile you saved above? The command is (in short form):
...
Code Block |
---|
[dspace]/bin/dspace import --eperson=joe@user.com --delete --mapfile mapfile |
Other Options
- Workflow. The importer usually bypasses any workflow assigned to a collection. But add the
--workflow
(-w
) argument will route the imported items through the workflow system.
...
- Resume. If, during importing, you have an error and the import is aborted, you can use the
--resume
(-R
) flag to resume the import where you left off after you fix the error. Specifying the owning collection on a per-item basis from the command line administration tool
If you omit the -c flag, which is otherwise mandatory, the ItemImporter searches for a file named "collections" in each item directory. This file should contain a list of collections, one per line, specified either by their handle, or by their internal db id. The ItemImporter then will put the item in each of the specified collections. The owning collection is the collection specified in the first line of the collections file.
If both the -c flag is specified and the collections file exists in the item directory, the ItemImporter will ignore the collections file and will put the item in the collection specified on the command line.
Since the collections file can differ between item directories, this gives you more fine-grained control of the process of batch adding items to collections.
UI Batch Import
Info |
---|
Available in DSpace 7.4 and above. |
...
Note |
---|
It is also possible to start an "import" directly from the "Processes" menu. This allows you to specify additional options/flags which are normally only available to the command-line "import" tool (see documentation above). |
Exporting Items
The item exporter can export a single item or a collection of items, and creates a DSpace simple archive in the aforementioned format for each exported item. The items are exported in a sequential order in which they are retrieved from the database. As a consequence, the sequence numbers of the item subdirectories (item_000, item_001) are not related to DSpace handle or item IDs.
...
Using the -x argument will do the standard export except for the bitstreams which will not be exported. If you have full SAF without bitstreams and you have the bitstreams archive (which might have been imported into DSpace earlier) somewhere near, you could symlink original archive files into SAF directories and have an exported collection which almost doesn't occupy any space but otherwise is identical to the exported collection (i.e. could be imported into DSpace). In case of huge collections -x mode might be substantially faster than full export.
UI Batch
...
Export
Info |
---|
Available in DSpace 7.4 and above. |
Batch export can also take place via the Administrator's UI. The default file size upload limit is 512MB, and is being configured in the spring boot application.properties file.
The steps to follow are:
- Login as an Administrator.
- In the side menu, select "Import" → "Batch Export (ZIP)"
- Select or search for the Collection to export from:
- Clicking "Export" will start the Batch Export. This creates a new "Process" which begins export process. Depending on the size of the export, this process may take some time to complete. You can refresh to page to see the current status, or go back to the list of processes ("Processes" menu in sidebar) to check on its status. Once the process is COMPLETED, you will see a log of the results and an exported ZIP file which you can download for the results.
- All prior exports will be listed in the "Processes" menu, until their corresponding process entry is deleted. Once you are satisfied with the export and have downloaded the ZIP, you may wish to delete that process entry in order to free up storage space (as your exported ZIP will be retained in DSpace until the process is deleted). A "process-cleaner" script can also be started from the "Processes" page which can be used to bulk delete old processes.
...