ts-ids-es-json-generator

Introduction

This simple utility generates elasticsearch.json using schema.json.

How to Use

To run the generator, simply provide the program with the location of the IDS root directory:

$ pipenv run python -m ids_es_json_generator ~/path/to/ids/folder

elasticsearch.json may be manually edited after automatic generation (e.g. to update non-searchable fields).

Output

The script saves elasticsearch.json to the same directory you provided.

Assumptions

The generator operates under the following assumptions:

• It's the IDS designer's responsibility to make sure the schema and elasticsearch.json are correct. The generator will not validate the IDS for you.

Restrictions

• JSON pointers are dereferenced to create the mapping, so schemas with circular references or missing references will fail.

Tests

Run the tests using the following command:

$ pipenv run python -m pytest

How it works

Anywhere an array of objects appears in the schema, a nested field type will be created in the output elasticsearch.json, including arrays of objects contained anywhere within other arrays of objects. The steps are:

1. The schema is loaded into a dict with all of its JSON pointers dereferenced using jsonref, so that nested fields involving definitions are included in the elasticsearch mapping.
2. find_array_object will find all arrays of objects and create a dict representing their locations in the schema (see test case test_array_simple_nested_case in __tests__/unit/test_find_array_object.py).
3. generate_es_mapping uses this dict to create the elasticsearch mapping (see test case test_nested_keys_case in __tests__/unit/test_generate_es_mapping.py).

See the example files and test cases for more example usage.