Random Master Data Generator for Pure

This tool creates random data for Pure in valid XML. Variable size datasets containing orgs and persons can be generated.

Getting Started

1. Install Python 3 and pip if you have not already.
2. See requirements.txt for required Python modules. To install these modules, use pip install -r requirements.txt.
3. Download code above and put all files into the same working directory.
4. Download commons.xsd, person.xsd and organisation.xsd and place into the working directory. (Get these from the Pure Administrator)
5. Run generator.py while passing the desired parameters (use --help to get an overview of possible parameters).
6. Output orgs + person XML files.
7. Upload to Pure.

Parameters

Parameter	Defaults	Description
--help	Display help.
--simple/--complex	--complex	Create simple or complex XML output.
--photos / --no-photos	--no-photos	Whether to create photos for persons. Note: Disabled with --simple.
--persons_out	persons.xml	Persons output file.
--orgs_out	orgs.xml	Orgs output file.
--i_orgs		Existing org structure file.
--persons	1	Number of persons to create.
--orgs	1	Maximum number of children to create at each level of the hierarchy (hierarchy always has tree-height = 2).
--submission	en-GB	Locale for submission. E.g. 'en-GB'.
--locale	en	Locale for data generation. E.g. 'en'.
--validate	false	Validate XML output. Requires the XSDs in working directory.
--excel/--no-excel	--no-excel	Output to Excel file.

Usage Notes

• Org types generated are: university, faculty and department. If other types are required, use the --i_orgs to input an existing org structure.
• If modifying the XSL files it is recommended to use --validate to ensure XML output conforms to the schema definition.
• Qualifications.json contains a pre-populated set of academic qualifications used by the script. This can be customized by fetching the classification scheme from the Pure API.
• Usage of classifications has been kept to a minimum, so it should not be necessary to customize a lot of classification schemes before uploading the XML files.
• Note: The Excel format does not correspond to the Pure masterlist spec, however support for this may be added in the future.
• Warning: Enabling photos will slow down the script, as Mimesis always downloads the file (even though we don't actually need it).

Examples

Create Chinese orgs and 50 persons (some with photos) in the corresponding submission locale:

python3 generator.py --persons 50 --locale zh --submission zh-CH --photos

Create a set of persons with Mexican locale based on an existing org hierarchy (filename existing_orgs.xml) and with US submission language:

python3 generator.py --locale es-mx --submission en-US --i_orgs existing_orgs.xml

Acknowledgments

This script uses https://github.com/lk-geimfari/mimesis for data generation.