Best Practices for Creating a JSON Schema

Below, you will find some of the good practices we should follow to create a JSON schema. We are following this guideline on ls front-end/back-end decoupling project.

We use the JSON schema to define the structure of the data and the validation rules for JSON data.

• Provide clear titles and descriptions for your schema and its properties to make it easier to understand.
• Specify which properties are required to ensure that the JSON data includes all necessary information.
• Define the correct data types for each property (e.g., string, number, object, array, boolean, null).
• Use constraints like minLength, maxLength, minimum, maximum, pattern, etc., to enforce rules on the data.
• Use the enum keyword to specify a set of allowed values for a property.
• Clearly define the structure of nested objects and arrays, including their properties and constraints.
• Use $ref to reference reusable schema definitions to avoid duplication and maintain consistency.
• Include a version number in your schema to manage changes and ensure compatibility.
• Use tools to validate your schema against sample data to ensure it works as expected.
• Provide documentation or examples to help users understand how to use the schema.

Detailed Solr fields

Solr query to get fields that does not exit -field:[* TO *] ==> -author:["" TO *]

Miss or empty fields in Solr index:

• author
• bothPublishDate
• date
• enumPublishDate

Required fields in Solr index:

• title_display
• id
• record_no
• rights

Calculated fields in ls interface:

• bookID => may be possible for an item not to have a book ID - if it is not a Google-scanned item and has no OCLC #, ISBN, or LCCN? But it is probably not currently used in the front end at least for full-text (although there could be some connection to the link in pageturner to the item in google books, or the find in a library link, or something?)
• "useDate" => if the bothPublishDate is empty and date is empty, then useDate does not exist
• date => it will be empty if date is not included in the solr index
• enumDate => it will be empty if enumPublishDate is not included in the solr index
• bothDate => it will be empty if bothPublishDate is not included in the solr index

==> Many of the item fields are not currently used in the front end and could be omitted (bookID, EnumDate, BothDate, ABLabel, result_number, relevance, PtSearchHref, PtHref)

Python libraries for JSON schema

• https://python-jsonschema.readthedocs.io/en/latest/referencing/
• https://builtin.com/software-engineering-perspectives/python-json-schema