Import

Import is available for admins and managers and offers functionality to import data from CSV files. Currently lists can be imported containing:

  • Name

  • Alias

  • Description

  • Dates

  • GIS data

  • Types

  • Value types

  • References

  • Reference systems

  • Administrative unit

  • Historical place

  • origin_id

  • Place hierarchy

Preparations

Automatic imports can be dangerous for data integrity and it may be very time consuming to reverse them so we strongly advise:

  • SQL backups before import, an existing backup not older than a day is enforced

  • Use the preview (enabled by default) and check if the data looks alright

The import operation is encapsulated in a transaction, meaning if there is an error in the script, nothing will be imported.

The file:

  • Take a look at the example.csv or example_place_hierarchy.csv

  • Make sure the extension is spelled correctly in lower case e.g. my_data.csv

  • In the first row should be the header names

  • Each following row is one data set. Values are separated by commas

  • Text can be enclosed in double quotes (”), especially if the contain commas

Projects

To retrace imported entities they have to be associated with a project. If there is none (or not the right one) available, you’ll have to create a new one. The name and description can be updated later.

Import class

Only one class can be imported at a time, so you have to choose one of the available classes.

Possible import fields

The header can contain the following titles. Columns with other titles won’t get imported but shown as an error message.

  • name - required, an error will be displayed if the header is missing. A warning will be displayed, if names in data rows are missing and these won’t get imported.

  • alias - only available for person, group and place, see Alias

  • description - a description can be provided

  • origin_id - It has to be unique per project so if you have multiple like a person and place with id = 1 you can prefix them in the document e.g. person_1, place_1 before importing them

  • begin_from - used for dates, see Dates

  • begin_to - used for dates, see Dates

  • end_from - used for dates, see Dates

  • end_to - used for dates, see Dates

  • type_ids - used to link to types, see Types

  • value_types - used to link to a value type, see Value types

  • references - used to link existing references, see References

  • wkt - only available for places and artifacts, see WKT coordinates

  • reference_system_* - used to link existing external reference systems, see External reference systems

  • administrative_unit - only available for places, id of existing administrative unit

  • historical_place - only available for places, id of existing historical place

  • parent_id - only available for place, id of a super unit in a place hierarchy, see Place hierarchy

  • openatlas_parent_id - only available for place, id of a super unit existing in OpenAtlas, see Place hierarchy

  • openatlas_class - only available for place and only used with parent_id, see Place hierarchy

Alias

Alias can be entered as string. Multiple aliases can be separated with semicolon (;). If an Alias contains a comma (,) please surround the whole filed with double quotes(”)

Dates

Dates can be entered in the format YYYY-MM-DD in the fields begin_from and end_from. You can also use time spans in combinations with the fields begin_to and end_to, see: Date

  • If the date format is incorrect, they will be displayed in red and won’t be imported

  • Missing values for time spans will be discarded silently, e.g. a valid value in begin_to, but an empty value in begin_from

  • There are no advanced checks between dates, e.g. end dates can be before begin dates. You should check them after the import at Data integrity checks

Types

It is possible to link entities to Type at the import which can be very useful e.g. if you have a custom type Case studies to link them all in one go.

  • Type ids can be entered at the column type_ids

  • You can enter multiple separated with a space

  • The id of a Type can be looked up at the detail view of a Type

Value types

It is possible to link entities to value Type at the import.

  • Value Type can be entered at the column value_types

  • Type id and value are separated with a semicolon (;), e.g. 1234;-13.65

  • Value Type need always a value

  • You can enter multiple separated with a space

  • The id of a value Type can be looked up at the detail view of a value Type

References

It is possible to link existing Reference to imported entities.

  • Reference ID and pages are separated with a semicolon (;), e.g. 1234;56-78

  • To link Reference without page number, just add the ID without semicolon (;)

  • You can enter multiple Reference separated with a space, e.g. 1234;56-78 5678

  • The ID of a Reference can be looked up at the detail view of the entity

WKT coordinates

For places and artifact (multi)point, (multi)polygon, (multi)linestring coordinates or GeometricCollection can be imported. Keep in mind to use the WGS84 geodetic system (EPSG 4326). Coordinates will be imported as WKT. It is only possible to import one geometry for each entry. Since the WKT format uses commas (,), surround the coordinates with double quotes (”)

Example:

“LINESTRING (12.458533781141528 41.922205268362234, 12.53062334955289 41.917606998887024, 12.52169797441624 41.888476931243254)”

External reference systems

It is possible to link the imported entity to an existing Reference System. In this case, the header has to be named reference_system_* with the name of the external Reference System appended, e.g. reference_system_wikidata. If spaces occur in the name, please substitute them with underscore (_), e.g. reference_system_getty_aat.

The entry consist of two values, separated by a semicolon (;). The first value is the identifier, e.g. Q54123, the second value is the match type (close_match or exact_match)

Example:

Q54123;close_match

Place hierarchy

The parent_id or openatlas_parent_id is used to generate a Place hierarchy together with Feature, Stratigraphic unit, Artifact, and Human remains. The parent_id has to be an origin_id of a row in the current import file. The openatlas_parent_id has to be an existing OpenAtlas entity ID with the correct class. To declare, which entry has a specific class, the openatlas_class column is used. Here the following classes can be entered: Place, Feature, Stratigraphic unit, Artifact, and Human remains. This is case-insensitive. For an example, please see: example_place_hierarchy.csv

For questions about the correct place hierarchy structure, please see the archaeological sub units at the OpenAtlas Model.

Import options

  • File - select the file you’ll want to import

  • Preview - if this option is selected, nothing will be imported and you see a preview

  • Check for duplicates - if selected, the chosen class e.g. person will be searched for already existing names. The search is not case-sensitive e.g. “King Arthur” would be found even it is spelled “KiNg ArThUr”. If duplicates are found a warning is printed but this doesn’t stop the import so check it before with the preview.

After the import

When the import went through, you’ll see a summary which data was imported (like the preview). Also, you can browse the projects to see which imported entities are associated with them. If you enabled the advanced layout you can also see in the detail view of an entity from which project it was imported, which user did the import and the origin_id value.

Although the script makes a lot of validation checks it’s always a good idea to run Data integrity checks after each import.