Zotero-to-DSpace

A tool to migrate Zotero metadata to DSpace. It converts Zotero-CSV to DSpace-CSV format. Usage example: It is used to migrate data from Zotero to UCY's institutional repository (https://gnosis.library.ucy.ac.cy/)

License: CC BY-NC-SA 4.0 (https://creativecommons.org/licenses/by-nc-sa/4.0/)

Using:

• langid 1.1.6
• numpy 1.16.2
• python-stdnum 1.11
• PyYAML 5.1
• tqdm 4.31.1

Getting Started

These instructions will install requirements and allow you to execute the code.

Prerequisites

(Optional) Create a virtual environment to install

Configuring your project

• Install requirements

pip install -r requirements.txt

• To view usage instructions use the command ./convert.py -h:

$ ./convert.py -h
usage: convert.py [-h] [-i INPUT_FILE] [-o OUTPUT_FILE] [-hdl HANDLE] [-ao]

This is a script to convert Zotero CSV export to DSpace CSV format.

optional arguments:
  -h, --help            show this help message and exit
  -i INPUT_FILE, --input-file INPUT_FILE
                        This is the filename of the input csv file from
                        Zotero. If none is specified, input.csv is used.
  -o OUTPUT_FILE, --output-file OUTPUT_FILE
                        This is the filename of the generated csv file (to be
                        imported to DSpace). If none is specified, output.csv
                        is used.
  -hdl HANDLE, --handle HANDLE
                        This is the collection handle (target collection) If
                        none is specified, 7/1234 is used.
  -ao, --add-orcids     Add ORCiDs from file. It loads ORCiDs from
                        ./settings/orcid.csv. This CSV file includes 3 columns
                        (surname, name, ORCiD).

• Place a Zotero CSV file in the same directory as convert.py

• Convert to DSpace CSV format using the command:

./convert [-i INPUT_FILE] [-o OUTPUT_FILE] [-hdl HANDLE] [-ao]

Note 1: If no input file is declared with -i, input.csv is assumed.

Note 2: If no output file is declared with -o, output.csv is assumed.

Note 3: If no handle is declared with -hdl, 7/1234 is assumed.

Note 4: An additional custom metadata value can be added by using the -ao option. This is optional and a custom implementation.

Configuration

• Set the equivalent of Zotero language in ISO format in languages_zotero.csv. This is used to generate dc.language.iso
• Set the header of the generate DSpace CSV file using values in dspace_csv_header.yml. The file comprises of 3 sections:
  1. initial_fieldnames are header values that always exist. Note that dc.type.uhtype is a custom field used at gnosis.ucy.ac.cy, to store item type. In your case you could copy values in dc.type and delete this column.
  2. fieldnames_with_language are fields that have a language value, for example dc.title[el] or dc.title[en].
  3. fieldnames_with_no_language are fields (headers) without language distinction, for example dc.source.uri[].
• Set metadata mapping in metadata_mapping.yml. In it are 2 main groups of settings:
  1. metadata_with_language. These are fields that should have their language detected. For example if a title in in english, it should be set in dc.title[en] column. If a title in in greek, it should be set in dc.title[el] column.
  2. metadata_without_language. These are fields that should not have their language detected. For example dc.source.uri[] column.
  • On the left you have the DC field.
  • On the right you have the column number of Zotero's CSV file. CAUTION: numbering starts at zero (0)!
• Set type translation in types.yml. This configuration file allows to translate a Zotero type to a "friendlier" value in case you need to index it and present it to the end-user.

Note: Any metadata field set in metadata_mapping.yml must also exist in dspace_csv_header.yml