Elastic-dql
========
Advanced search language for Django and elasticsearch based on DjangoQL project.
DjangoQL is a library that provides search language that works with django orm. elastic-dql extends DjangoQL project to
generate elasticsearch queries.<br><br>
elastic-dql suppports logical operators and parenthesis.It also provides to apis to get index mappings and value
suggestions for keyword fields.
Contents
--------
* [`Installation`](#installation)
* [`Generating Elasticsearch Queries`](#generating-elasticsearch-queries)
* [`Custom SchemaFactory`](#custom-schemafactory)
* [`Mappings and Suggestions (auto-complete) api`](#mappings-and-suggestions-api)
* [`Features`](#features)
* [`TODO Tasks`](#todo-tasks)
* [`DjangoQL project`](#djangoql-project)
* [`Language reference`](#language-reference)
* [`License`](#license)
Installation
------------
```shell
$ pip install elastic-dql
```
Add ``'elastic_dql'`` to ``INSTALLED_APPS`` in your ``settings.py``:
```python
INSTALLED_APPS = [
...
'elastic_dql',
...
]
```
Add ``ELASTIC_DQL`` section in ``settings.py``:
```python
ELASTIC_DQL = {
"schema_factory": "elastic_dql.schema.SchemaFactory",
"default_schema": "elastic_dql.schema.ElasticDjangoQlSchema",
"default_index": None,
"accept_index_param": True, # if False default_index should be specified
"connection": {
"hosts": ["http://localhost"],
}
}
```
this values are default configs.if you have just one elasticsearch index you better set a default index otherwise you
should pass ``index`` parameter in ``mappings`` and ``suggestions`` apis.
Generating Elasticsearch Queries
--------------------------------
to create elasticsearch queries you should follow these lines:
```python
from elastic_dql.query import get_query
index_name = "your_elasticsearch_index"
query = 'name = "mohammad" and age = 10'
elastic_query = get_query(index_name, query)
```
Custom SchemaFactory
--------------------
SchemaFactory handles limits to access elasticsearch.by default elastic-dql uses ``elastic_dql.schema.SchemaFactory``
and allows to access to all indexes and fields.
To make some limits at first you should create custom SchemaFactory:
```python
from elastic_dql.schema import SchemaFactory
class CustomSchemaFactory(SchemaFactory):
include_indices = ('*',)
exclude_indices = ()
index_field_limits = {
"some-index": ["password_field","other_limited_field"]
}
```
after implementing ``CustomSchemaFactory`` add the class path to ``settings.py``:
```python
ELASTIC_DQL = {
"schema_factory": "path.to.CustomSchemaFactory",
...
}
```
> :warning: **you must either fill include_indices or exclude_indices not both**
Mappings and Suggestions api
--------------------------------------------
``Mappings api`` : provides field mappings of index (default_index or get from url parameters)
``Suggestions api`` : provides field value suggestion for ``Keyword`` fields.can use for auto-complete.
To use this apis you must add elastic_dql urls:
```python
from elastic_dql.urls import get_urls
urlpatterns = [
...
] + get_urls()
```
OR
```python
from django.urls import include
urlpatterns = [
...
include("elastic_dql.urls"),
...
]
```
```shell
$ curl localhost:8000/mappings?index=your_index
```
> :warning: **if use default_index, index parameter will be skipped**
```shell
$ curl localhost:8000/suggestions/some_keyword_field?index=your_index&search=values_must_contains_this
```
> :warning: **search is optional - if use default_index, index parameter will be skipped**
Features
--------
- index and field limiting with ``SchemaFactory`` Customization
- mappings and suggestions apis
TODO Tasks
----------
- add pagination to suggestions api
- make library compatible with elastic-dsl query generator
- add async for elasticsearch communications
- cache field mappings and invalidate it with user defined duration
- compatibility test with some elasticsearch (python lib) versions
- handle all elasticsearch fields now it supports (long,unsigned_long,text,keyword,float,int,date,boolean)
DjangoQL project
------------------
DjangoQL github page:
[DjangoQL](https://github.com/ivelum/djangoql)
Language reference
------------------
The query language is as same as djangoql query language
[Language refrence](https://github.com/ivelum/djangoql#language-reference>)
License
-------
MIT