Twig Recipe Cards for Common Use Cases
The Twig Recipe Cards below reference common Metadata transformation, display, or other use cases/needs you may have in your own Archipelago repository.
Getting Started Working with Twig in Archipelago
We recommend reading through our main Metadata Display Preview and Twigs in Archipelago documentation overview guides, and also our Working with Twig primer before diving into applying any of these recipes in your own Archipelago.
AMI Ingest Template Adaptations -- Common Use Cases and Twig Recipe Cards:
Use Case #1: I used AMI LoD Reconciliation to reconciliate the values in my AMI Set Source CSV mods_subject_topic column against both LCSH and Wikidata. I would like to map the reconciliated values into the Archipelago default subject_loc and subject_wikidata JSON keys.
Twig Recipe Card for Use Case #1:
{#- LCSH -#}
{% if data.mods_subject_topic|length > 0 %}
"subject_loc": {{ data_lod.mods_subject_topic.loc_subjects_thing|json_encode|raw }},
{% endif %}
{#- Wikidata -#}
{% set subject_wikidata = [] %}
{% for source, reconciliated in data_lod %}
{% if (('subject' in source) or ('genre' in source)) and reconciliated.wikidata_subjects_thing and reconciliated.wikidata_subjects_thing|length > 0 %}
{% set subject_wikidata = subject_wikidata|merge(reconciliated.wikidata_subjects_thing) %}
{% endif %}
{% endfor %}
Use Case #2: I have both columns containing a mods_subject_authority_lcsh_topic (labels) and corresponding mods_subject_authority_lcsh_valueuri (URIs) data in my AMI Set Source Data CSV that I would like to pair and map into the Archipelago default subject_loc JSON key.
Twig Recipe Card for Use Case #2:
{%- if data['mods_subject_authority_lcsh_topic'] is defined and not empty -%}
{% set subjects = data["mods_subject_authority_lcsh_topic"] is iterable ? data["mods_subject_authority_lcsh_topic"] : data["mods_subject_authority_lcsh_topic"]|split('|@|') %}
{% set subject_uris = data["mods_subject_authority_lcsh_valueuri"] is defined ? data["mods_subject_authority_lcsh_valueuri"] : '' %}
{% set subject_uris_list = subject_uris is iterable ? subject_uris : subject_uris|split('|@|') %}
"subject_loc": [
{% for subject in subjects %}
{
"uri": {{ subject_uris_list[loop.index0]|default('')|json_encode|raw }},
"label": {{ subject|json_encode|raw }}
}
{{ not loop.last ? ',' : '' }}
{% endfor %}
],
{%- endif -%}
Use case #3: I have dc.creator and dc.contributor columns in my AMI Set Source Data CSV with simple JSON-encoded values (e.g. source column cells contain ["Name 1, Name 2"]) that I would like to map to the Archipelago default creator_lod JSON key.
Twig Recipe Card for Use Case #3:
{% if data['dc.creator']|length > 0 or data['dc.contributor']|length > 0 %}
{% set total_creators = (data["dc.creator"]|length) + (data["dc.contributor"]|length) %}
{% set current_creator = 0 %}
"creator_lod": [
{% for creator in data["dc.creator"] %}
{% set current_creator = current_creator + 1 %}
{% set creator_source = data["dc.creator"][loop.index0] %}
{
"name_uri": null,
"agent_type": null,
"name_label": {{ creator|json_encode|raw }},
"role_label": "Creator",
"role_uri": "http://id.loc.gov/vocabulary/relators/cre"
}
{{ current_creator != total_creators ? ',' : '' }}
{% endfor %}
{% for creator in data["dc.contributor"] %}
{% set current_creator = current_creator + 1 %}
{% set creator_source = data["dc.contributor"][loop.index0] %}
{
"name_uri": null,
"agent_type": null,
"name_label": {{ creator|json_encode|raw }},
"role_label": "Contributor",
"role_uri": "http://id.loc.gov/vocabulary/relators/ctb"
}
{{ current_creator != total_creators ? ',' : '' }}
{% endfor %}
],
{% endif %}
creator_lod JSON key.
Twig Recipe Card for Use Case #4:
Click to view the full Recipe Card
{#- START Names from LoD and MODS CSV with/without URIS. -#}
{# Updated August 26th 2022, by Diego Pino. New checks/logic for mods_name_type_role_composed_or_more_namepart
- Check first IF for a given namepart there is already reconciliaton.
- IF not i check if there is a matching valueuri,
- If not leave the URL empty and use the value in the namepart (label) only?
- Only check/use mods_name_corporate/personal_namepart field IF there are no other fields
- That specify Roles. Since normally in ISLANDORA that field (no role) is a Catch all names one
- And in that case USE creator as the default ROLE
#}
{%~ set creator_lod = [] -%}
{# Used to keep track of parts after the type (corporate, etc) that are no roles
but authority properties. Add more if you find them #}
{% set roles_that_are_no_roles = ['authority_naf','authority_marcrelator',''] %}
{# Used to keep track of the ones that are reconciled already #}
{%- set name_has_creator_lod = [] -%}
{%- for key,value in data_lod -%}
{%- if key starts with 'mods_name_' and key ends with '_namepart' -%}
{# If there is mods_name_SOMETHING_namepart in data_lod we keep track so we
do not try afterwards to use that Sources KEY from the CSV.
#}
{%- set name_has_creator_lod = name_has_creator_lod|merge([key]) -%}
{# Now we remove 'mods_name_' and '_namepart' #}
{%- set name_type_and_role = key|replace({'mods_name_':'', '_namepart':''}) -%}
{# We will only target personal or corporate. If any of those are missing we skip? #}
{% set name_type = null %}
{%- if name_type_and_role starts with 'personal_' -%}
{% set name_type = 'personal' %}
{%- elseif name_type_and_role starts with 'corporate_' -%}
{%- set name_type = 'corporate' -%}
{%- endif -%}
{%- if name_type is not empty -%}
{#- Now we remove 'type', e.g 'corporate_' -#}
{%- set name_role = name_type_and_role|replace({(name_type ~ '_'):''}) -%}
{# in case the name_role contains one of roles_that_are_no_roles, e.g
something like `creator_authority_marcrelator` we remove that #}
{% for role_that_is_no_role in roles_that_are_no_roles %}
{%- set name_role = name_role|replace({(role_that_is_no_role):''}) -%}
{% endfor %}
{# After removing all what can not be a role if we end with an empty #}
{% if name_role|trim|length == 0 %}
{%- set name_role = "creator" %}
{% else %}
{%- set name_role = name_role|replace({'\\/':'//' , '_':' '})|trim -%}
{% endif %}
{#- we iterate over all possible vocabularies and fetch the reconciliated names from them (if any) -#}
{%- for approach, names in value -%}
{#- if there are actually name pairs (name and uri) that were reconciliated we use them -#}
{%- if names|length > 0 -%}
{#- we call the ami_lod_reconcile twig extension with the role label using the LoC Relators endpoint in english and get 1 result -#}
{%- set role_uri = ami_lod_reconcile(name_role|lower|capitalize,'loc;relators;thing','en',1) -%}
{#- for each found name pair in a list of possible LoD reconciliated elements we generate the final structure that goes into "creator_lod" json key -#}
{%- for name in names -%}
{%- set creator_lod = creator_lod|merge([{'role_label': name_role|lower|capitalize, 'role_uri': role_uri[0].uri, "agent_type": name_type, "name_label": name.label, "name_uri": name.uri}]) -%}
{%- endfor -%}
{%- endif -%}
{%- endfor -%}
{% endif -%}
{%- endif -%}
{%- endfor -%}
{# Now go for the RAW CSV data for names #}
{%- for key,value in data -%}
{# here we skip values previoulsy fetched from LoD and stored in name_has_creator_lod #}
{%- if key not in name_has_creator_lod and key starts with 'mods_name_' and key ends with '_namepart' -%}
{# If there is mods_name_SOMETHING_namepart in data_lod we keep track so we
do not try afterwards to use that Sources KEY from the CSV.
#}
{%- set name_has_creator_lod = name_has_creator_lod|merge([key]) -%}
{# Now we remove 'mods_name_' and '_namepart' #}
{%- set name_type_and_role = key|replace({'mods_name_':'', '_namepart':''}) -%}
{# We will only target personal or corporate. If any of those are missing we skip? #}
{%- set name_type = null -%}
{%- if name_type_and_role starts with 'personal_' -%}
{%- set name_type = 'personal' -%}
{%- elseif name_type_and_role starts with 'corporate_' -%}
{%- set name_type = 'corporate' -%}
{%- endif -%}
{% if name_type is not empty %}
{# Now we remove 'type', e.g 'corporate_' #}
{%- set name_role = name_type_and_role|replace({(name_type ~ '_'):''}) -%}
{# in case the name_role contains one of roles_that_are_no_roles, e.g
something like `creator_authority_marcrelator` we remove that #}
{% for role_that_is_no_role in roles_that_are_no_roles %}
{%- set name_role = name_role|replace({(role_that_is_no_role):''}) -%}
{% endfor %}
{# After removing all what can not be a role if we end with an empty #}
{% if name_role|trim|length == 0 %}
{%- set name_role = "creator" %}
{% else %}
{%- set name_role = name_role|replace({'\\/':'//' , '_':' '})|trim -%}
{% endif %}
{# Now we check if there is a corresponding _valueuri for this #}
{% set name_uris = [] %}
{%- if data[('mods_name_' ~ name_type_and_role ~ '_valueuri')] is not empty
and data[('mods_name_' ~ name_type_and_role ~ '_valueuri')] != '' -%}
{%- set name_uris = data[('mods_name_' ~ name_type_and_role ~ '_valueuri')]|split('|@|') -%}
{%- endif -%}
{%- set role_uri = ami_lod_reconcile(name_role|lower|capitalize,'loc;relators;thing','en',1) -%}
{#- we split and iterate over the value of the mods_name key -#}
{# NOTE. THIS IS TARGETING Anything after the year 1000, or 2000 #}
{%- for index,name in value|replace({'|@|1':', 1', '|@|2':', 2', '|@|-':', -'})|split('|@|') -%}
{%- if name is not empty and name|trim != '' -%}
{%- set name_uri = null -%}
{# Here we can check if one of the names IS not a name (e.g a year? #}
{#- we call the ami_lod_reconcile twig extension with the role label using the LoC Relators endpoint in english and get 1 result -#}
{%- if name_uris[index] is defined and name_uris[index] is not empty -%}
{%- set name_uri = name_uris[index] -%}
{%- endif -%}
{%- set creator_lod = creator_lod|merge([{'role_label': name_role|lower|capitalize, 'role_uri': role_uri[0].uri, "agent_type": name_type, "name_label": name, "name_uri": name_uri}]) -%}
{%- endif -%}
{%- endfor -%}
{%- endif -%}
{%- endif -%}
{%- endfor ~%}
{# Use reduce filter + other logic for depulicating #}
{% set creator_lod = creator_lod|reduce((unique, item) => item in unique ? unique : unique|merge([item]), []) %}
"creator_lod": {{ creator_lod|json_encode|raw -}},
{#- END Names from LoD and MODS CSV with/without URIS. -#}
Use Case #5: I have geographic location information that I would like to reconciliate against Nominatim and map into the default Archipelago 'geographic_location' key. I have AMI Source Data CSVs which contain values/labels and some which contain coordinates.
Twig Recipe Card for Use Case #5 with variation notes:
{#- <-- Geographic Info and terms:
Includes options for geographic info for:
- Nominatim lookup by value/label
- Nominatim lookup by coordinates
-#}
{#- use value for Nominatim search -#}
{% if data.mods_subject_geographic|length > 0 %}
{% set nominatim_from_label = ami_lod_reconcile(data.mods_subject_geographic,'nominatim;thing;search','en') -%}
"geographic_location": {{ nominatim_from_label|json_encode|raw }},
{% endif %}
{#- use coordinates for Nominatim search, if provided -#}
{% if data.mods_subject_cartographics_coordinates|length > 0 %}
{% set nominatim_from_coordinates = ami_lod_reconcile(data.mods_subject_cartographics_coordinates,'nominatim;thing;reverse','en') -%}
"geographic_location": {{ nominatim_from_coordinates|json_encode|raw }},
{% endif %}
{#- Geographic Info and terms --> #}
Use Case #6: I have date values in a dc.date column that contain instances of 'circa' or 'Circa' where I would like to replace with the EDTF-friendly '~' instead and map to the Archipelago default 'date_created_edtf' JSON key.
Twig Recipe Card for Use Case #6:
{% if data['dc.date'] is defined %}
{% set datecleaned = data['dc.date']|replace({"circa ":"~", "Circa ":"~"}) %}
"date_created_edtf": {
"date_to": "",
"date_free": {{ datecleaned|json_encode|raw }},
"date_from": "",
"date_type": "date_free"
},
{% endif %}
More recipe cards will be added over time. Please see our Archipelago Contribution Guide to learn about contributing your own recipe card or other documentation.
Thank you for reading! Please contact us on our Archipelago Commons Google Group with any questions or feedback.
Return to the Archipelago Documentation main page.
Last update:
December 7, 2022
Created: December 7, 2022
Created: December 7, 2022