...
Islandora Workbench uses YAML files to configure its operations. These files are documented in detail. Here is an example config file, including a link to a sample CSV (file uploaded to wiki, update csv link below after published):
...
task:
...
create
...
host:
...
"https://islandora.traefik.me/"
...
username:
...
xxxx
...
password:
...
xxxx
...
media_type:
...
file
...
input_csv:
...
...
PID
...
csv_field_templates:
...
-
...
field_rights:
...
"http://rightsstatements.org/vocab/CNE/1.0/"
...
-
...
field_member_of:
...
103
...
-
...
field_model:
...
13
...
-
...
field_resource_type:
...
25
...
-
...
field_display_hints:
...
21
...
default_file_mimetype:
...
'image/tiff'
...
default_file_extension:
...
".tif"
...
use_node_title_for_media:
...
1
...
allow_adding_terms:
...
true
NOTE: This CSV associated with this config file uses file URLs. To use a file directory, the input_dir configuration option may be used. More information is available in the Workbench documentation.
...
Resource Type
Default fields are:
Term Name | External URI |
Collection | |
Dataset | |
Image | |
Interactive Resource | |
Moving Image | |
Physical Object | |
Service | |
Software | |
Sound | |
Still Image | |
Text |
New terms can be created on the fly (during ingest). If your metadata uses terms that already exist by default, you can reference the term name, its ID, or its URI (if it has one) in the CSV.
...
System Model
Available terms are:
Term Name | External URI |
Audio | |
Binary | |
Collection | |
Compound Object | |
Digital Document | |
Image | |
Newspaper | |
Page | |
Paged Content | |
Publication Issue | |
Video |
NOTE: The Born-Digital i8 theme requires specific combinations of Resource Type and Model terms in order for compound objects, collections, and paged objects to display correctly. Please refer to Appendix B: Born-Digital i8 Theme Object View Configurations.
...
The machine names for each field are what need to be used as the column headers in the CSV. To find the machine name for a field, go to Structure > Content Type > Repository Item > Manage Fields.
Label | Machine Name or Workbench-required name (use for column header in CSV) | Field Type and Notes |
Title | title | text field (see more information above, under CSV Required Fields) |
Alternative Title | field_alternative_title | text field |
Identifier | field_identifier | text field; multi-value |
Resource Type | field_resource_type | taxonomy reference (see more information above, under CSV Required Fields) |
Genre | field_genre | taxonomy reference; multi-value |
Linked Agent | field_linked_agent | typed relation field; multi-value - see more information about how to set up this field below, under Linked Agent. |
Date Created | field_edtf_date_created | EDTF field; multi-value; must be in EDTF format - see more information about how to set up this field below, under EDTF Formats. |
Date Issued | field_edtf_date_issued | EDTF field; multi-value; must be in EDTF format - see more information about how to set up this field below, under EDTF Formats. |
Date | field_edtf_date | EDTF field; multi-value; must be in EDTF format - see more information about how to set up this field below, under EDTF Formats. |
Edition | field_edition | text field; multi-value |
Place Published | field_place_published | text field; multi-value |
Language | field_language | taxonomy reference field; multi-value |
Description | field_description_long | text-formatted-long, can support line breaks If your metadata has line breaks, as long as they are included in the cell in the spreadsheet, saving as a CSV should wrap the content of the cell in quotes and the line break will be preserved. |
Table of Contents | field_table_of_contents | text-formatted-long, can support line breaks |
Physical Form | field_physical_form | taxonomy reference; multi-value |
Extent | field_extent | text field; multi-value |
Rights | field_rights | text field; multi-value; if you want this to be an external link, field needs to be changed to a Link type field. |
Subject | field_subject | taxonomy reference (from corporate body, family, geographic location, person, or subject vocabularies); multi-value; data in CSV must include a namespace before the term (i.e., person: , corporate_body: , geo_location: , family: , or subject: ). |
Geographic Subject | field_geographic_subject | taxonomy reference (from Geographic Location vocabulary); multi-value |
Coordinates | field_coordinates | geolocation fields (latitude and longitude); multi-value - see more information about how to set up this field below, under Coordinates |
Coordinates (Text) | field_coordinates_text | text field; multi-value |
Temporal Subject | field_temporal_subject | taxonomy reference (from Temporal vocabulary); multi-value |
Subjects (name) | field_subjects_name | taxonomy reference (from person, family, or corporate body); multi-value; data in CSV must include a namespace before the term (i.e., person: , family: , or corporate_body) |
Dewey Classification | field_dewey_classification | text field; multi-value |
Library of Congress Classification | field_llc_classification | text field; multi-value |
Classification (Text) | field_classification | text field; multi-value |
Local Identifier | field_local_identifier | text field; multi-value |
ISBN | field_isbn | text field; multi-value |
OCLC Number | field_oclc_number | text field; multi-value |
Note | field_note | text-formatted-long, can support line breaks; multi-value |
System Model | field_model | taxonomy reference (from Islandora Models vocabulary); (see more information above, under CSV Required Fields) |
Member of | field_member_of | node reference; multi-value; points to the containing collection or parent node - see more information below under Member Of |
Language | n/a | (dropdown; defaults to English - ignore this for the purposes of the CSV) |
Access Control | field_access_terms | taxonomy reference (from Islandora Access vocabulary) |
Display hints | field_display_hints | taxonomy reference (from Islandora Display vocabulary) - see more information below, under Display Hints |
PID | field_pid | text field |
Weight | field_weight | number (integer); indicates the order of a resource in a collection of resources (used for compound objects and paged content) |
Adding New Fields
Follow these steps to add new fields as required:
...
All dates must follow EDTF formatting rules. Here are some examples:
EDTF Input | Front-End Output |
1933? | 1933 (year uncertain) |
1945~ | 1945 (year approximate) |
2016-04-12 | 2016-04-12 |
1860/1880? | 1860 to 1880 (year uncertain) |
1870/1880 | 1870 to 1880 |
Links
A Link field type stores URLs and link text in separate data elements.
...
These terms, from the Islandora Display vocabulary, will define which viewer is used on an object page.
Term Name | Used For |
Open Seadragon | Large Image, Page |
PDFjs |
Configure the CSV
Your CSV will include only columns that are a) required and b) have data in them. You should not include columns in your CSV that do not have data. This will cause an error during Workbench’s configuration check process.
If the collection(s) that will contain the objects have NOT yet been created in Drupal, you can include rows for the collection(s) in your CSV. Each object will also have a row, and will reference the id of the collection it belongs to. For example:
title | id | parent_id | field_member_of |
Test Collection | 55 | ||
Easthampton Town Hall | 1 | 55 | |
Nehemiah Strong House | 2 | 55 | |
Amherst College, Lawrence Observatory | 3 | 55 |
If the collection(s) that will contain the objects already exist in Drupal, you can use the field_member_of column to reference the node ID of the collection(s). (In this case, you will not use the parent_id column except for compound objects and paged content; see Configuring Complex Objects in the CSV for more information.)
...
Assuming, for example, your top-level collection had a node ID of 100, your CSV would look like this (this is only showing a portion of the columns):
title | id | parent_id | field_member_of |
Easthampton Town Hall | 1 | 100 | |
Nehemiah Strong House | 2 | 100 | |
Amherst College, Lawrence Observatory | 3 | 100 |
Anchor #ConfigComplex #ConfigComplex
| #ConfigComplex | |
| #ConfigComplex |
...
Assuming your Newspaper Parent node has not yet been created, and you’d like it to be contained within another top-level collection whose node has also not yet been created, the table below shows how the three newspaper parts (and the top-level collection) would be laid out in your CSV:
title | id | parent_id | field_weight |
Sample Collection | 1 | ||
Connecticut Western News (Newspaper) | 2 | 1 | |
Connecticut Western News Vol. 1 No. 7 | 3 | 2 | |
Connecticut Western News Vol. 1 No. 7 - page 1 | 4 | 3 | 1 |
Connecticut Western News Vol. 1 No. 7 - page 2 | 5 | 3 | 2 |
Connecticut Western News Vol. 1 No. 7 - page 3 | 6 | 3 | 3 |
Connecticut Western News Vol. 1 No. 7 - page 4 | 7 | 3 | 4 |
The Newspaper Parent is a child of the containing collection (or it may have no parent). The Newspaper Issue is a child of the Newspaper Parent, and each Newspaper Page is a child of the Newspaper Issue.
...
If, however, your Newspaper Parent is already present on your site, you can use field_member_of to identify the parent using its node ID. In that case, the Newspaper Issue row would have nothing in the parent_id column, but would include the Newspaper Parent’s node ID in the field_member_of column. The Newspaper Page rows would have the Newspaper Issue’s id in the parent_id column. These rows of your CSV would look like this (assuming the Newspaper Parent’s node ID were “100”):
title | id | parent_id | field_weight | field_member_of |
Connecticut Western News Vol. 1 No. 7 | 1 | 100 | ||
Connecticut Western News Vol. 1 No. 7 - page 1 | 2 | 1 | 1 | |
Connecticut Western News Vol. 1 No. 7 - page 2 | 3 | 1 | 2 | |
Connecticut Western News Vol. 1 No. 7 - page 3 | 4 | 1 | 3 | |
Connecticut Western News Vol. 1 No. 7 - page 4 | 5 | 1 | 4 |
Books
Configuring Books in your CSV is very similar to configuring Newspapers. The parent Book object will have child Page objects.
...
Assuming your Book Parent node has not yet been created, and you’d like it to be contained within a collection whose node has also not yet been created, the table below shows how the two Book parts (and the collection) would be laid out in your CSV:
title | id | parent_id | field_weight |
Sample Collection | 1 | ||
On the Tides at Malta (Book) | 2 | 1 | |
On the Tides at Malta - page 1 | 3 | 2 | 1 |
On the Tides at Malta - page 2 | 4 | 2 | 2 |
On the Tides at Malta - page 3 | 5 | 2 | 3 |
On the Tides at Malta - page 4 | 6 | 2 | 4 |
On the Tides at Malta - page 5 | 7 | 2 | 5 |
The Book Pages must each have a field_weight assigned.
...
If, however, the collection into which you are ingesting the Book is already present on your site, you can use field_member_of to identify the parent using its node ID. In that case, the Book Parent would not have anything in the parent_id column, but would include the collection node ID in the field_member_of column. The Book Page rows would have the Book Parent’s id in the parent_id column. These rows of your CSV would look like this (assuming the collection node ID were “100”):
title | id | parent_id | field_weight | field_member_of |
On the Tides at Malta (Book) | 1 | 100 | ||
On the Tides at Malta - page 1 | 2 | 1 | 1 | |
On the Tides at Malta - page 2 | 3 | 1 | 2 | |
On the Tides at Malta - page 3 | 4 | 1 | 3 | |
On the Tides at Malta - page 4 | 5 | 1 | 4 | |
On the Tides at Malta - page 5 | 6 | 1 | 5 |
Compound Objects
Configuring Compound Objects in your CSV is very similar to Newspapers and Books. The Compound Object will have a containing “parent object” as well as one or more children. The containing parent object will not be visible on the front-end, aside from its metadata (it cannot have any media of its own; that would never be seen).
...
Assuming you’d like the Compound Object to be contained within a collection whose node has also not yet been created, the table below shows how the two Compound Object parts (and the collection) would be laid out in your CSV:
title | id | parent_id | field_weight |
Sample Collection | 1 | ||
Historic Western Mass (Compound Object) | 2 | 1 | |
Amherst 1886 (Child Object 1) | 3 | 2 | 1 |
Adams 1882 (Child Object 2) | 4 | 2 | 2 |
The child objects must each have a field_weight assigned.
...
If, however, the collection into which you are ingesting the Compound Object is already present on your site, you can use field_member_of to identify the parent using its node ID. In that case, the Compound Object parent would not have anything in the parent_id column, but would include the collection node ID in the field_member_of column. The child object rows would have the Compound Object parent’s id in the parent_id column. These rows of your CSV would look like this (assuming the collection node ID were “100”):
title | id | parent_id | field_weight | field_member_of |
Historic Western Mass (Compound Object) | 1 | 100 | ||
Amherst 1886 (Child Object 1) | 2 | 1 | 1 | |
Adams 1882 (Child Object 2) | 3 | 1 | 2 |
Check, Then Run
You should always check your configuration and spreadsheet are valid before running the ingest. Fortunately, Islandora Workbench makes this easy with the --check command:
...
Appendix B: Born-Digital i8 Theme Object View Configurations
Object Type | Resource Type term | System Model term |
Collection | Collection | Collection |
Audio | Sound | Audio |
Basic Image (jpg) | Still Image | Image |
Binary | [any] | Binary |
Book (parent of pages) | Collection | Paged Content |
Compound Object | Collection | Compound Object |
Large Image (tiff) | Still Image | Image |
Newspaper (parent of issues) | Collection | Newspaper |
Newspaper Issue (parent of pages) | Collection | Publication Issue |
Newspaper Page | Text | Page |
Page | Text | Page |
[any] | Digital Document | |
Video | Moving Image | Video |