Skip to content

Opinionated high level library to query Elasticsearch

License

Notifications You must be signed in to change notification settings

pmourlanne/fiqs

Repository files navigation

fiqs

Build Status

fiqs is an opinionated high-level library whose goal is to help you write concise queries agains Elasticsearch and better consume the results. It is built on top of the awesome Elasticsearch DSL library.

fiqs exposes a flatten_result function which transforms an elasticsearch-dsl Result, or a dictionary, into the list of its nodes. fiqs also lets you create Model classes, a la Django, which automatically generates an Elasticsearch mapping. Finally fiqs exposes a FQuery objects which, leveraging your models, lets you write less verbose queries against Elasticsearch.

Compatibility

fiqs is compatible with Elasticsearch 6.X and works with Python3

Documentation

Documentation is available at https://fiqs.readthedocs.io/

Code example

You define a model, matching what is in your Elasticsearch cluster:

    from fiqs import models

    class Sale(models.Model):
        index = 'sale_data'
        doc_type = 'sale'

        id = fields.IntegerField()
        shop_id = fields.IntegerField()
        client_id = fields.KeywordField()

        timestamp = fields.DateField()
        price = fields.IntegerField()
        payment_type = fields.KeywordField(choices=['wire_transfer', 'cash', 'store_credit'])

You can then write clean queries:

    from elasticsearch_dsl import Search
    from fiqs.aggregations import Sum
    from fiqs.query import FQuery

    from .models import Sale

    search = Search(...)
    metric = FQuery(search).values(
        total_sales=Sum(Sale.price),
    ).group_by(
        Sale.shop_id,
        Sale.client_id,
    )
    result = metric.eval()

And let fiqs organise the results:

    print(result)
    # [
    #     {
    #         "shop_id": 1,
    #         "client_id": 1,
    #         "doc_count": 30,
    #         "total_sales": 12345.0,
    #     },
    #     {
    #         "shop_id": 2,
    #         "client_id": 1,
    #         "doc_count": 20,
    #         "total_sales": 23456.0,
    #     },
    #     {
    #         "shop_id": 3,
    #         "client_id": 1,
    #         "doc_count": 10,
    #         "total_sales": 34567.0,
    #     },
    #     [...]
    # ]

Contributing

The fiqs project is hosted on Github

To run the tests on your machine use this command: python setup.py test Some tests are used to generate results output from Elasticsearch. To run them you will need to run a docker container on your machine: docker run -d -p 8200:9200 -p 8300:9300 elasticsearch:6.x.x and then run pytest -k docker.

License

See attached LICENSE file.

About

Opinionated high level library to query Elasticsearch

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages