RQLAlchemy

Resource Query Language extension for SQLAlchemy

Overview

Resource Query Language (RQL) is a query language designed for use in URIs, with object-style data structures.

rqlalchemy is an RQL extension for SQLAlchemy. It easily allows exposing SQLAlchemy tables or models as an HTTP API endpoint and performing complex queries using only querystring parameters.

Installing

pip install rqlalchemy

Usage

RQL queries can be supported by an application using SQLAlchemy by adding the rqlalchemy.RQLQueryMixIn class as a mix-in class to your base Query class:

from sqlalchemy.ext.declarative import declarative_base
from sqlalchemy.orm import Query as BaseQuery

from rqlalchemy import RQLQueryMixIn

# create the declarative base
Base = declarative_base()

# create the custom query class
class RQLQuery(BaseQuery, RQLQueryMixIn):
    _rql_default_limit = 10
    _rql_max_limit = 100

# assign the custom query class to the declarative base
Base.query_class = RQLQuery

With that in place, you can perform RQL queries by passing the querystring to the query rql() method. For example, if you have a Flask HTTP API with an users collection endpoint querying your User model:

from urllib.parse import unquote

from flask import request

@app.route('/users')
def get_users_collection():
    qs = unquote(request.query_string.decode(request.charset))
    query = session.query(User).rql(qs)
    users = query.rql_all()

    return render_response(users)

Aggregates

As with the base SQLAlchemy Query class, you can retrieve results with the all() method, or by iterating over the query, however, if you want to support RQL expressions with aggregate functions or querying functions that result in a subset of columns, you must retrieve the results with rql_all().

Pagination

RQLAlchemy offers limit/offset pagination with the rql_paginate() method, which returns the requested page, the RQL expressions for previous and next pages if available, and the total number of items.

from urllib.parse import unquote

from flask import request

@app.route('/users')
def get_users_collection():
    qs = unquote(request.query_string.decode(request.charset))
    query = session.query(User).rql(qs)
    page, previous_page, next_page, total = query.rql_paginate()

    response = {"data": page,
                "total": total,
               }

    if previous_page:
        response["previous"] = '/users?' + previous_page

    if next_page:
        response["next"] = '/users?' + next_page

    return render_response(response)

Keep in mind that pagination requires a limit, either a _rql_default_limit value, a querystring limit(x), or the limit parameter to the rql() method. Calling rql_paginate() without a limit will raise RQLQueryError.

Reference Table

RQL	SQLAlchemy	Obs.
QUERYING
select(a,b,c,...)	.query(Model.a, Model.b, Model.c,...)
values(a)	[o.a for o in query.from_self(a)]
limit(count,start?)	.limit(count).offset(start)
sort(attr1)	.order_by(attr)
sort(-attr1)	.order_by(attr.desc())
distinct()	.distinct()
first()	.limit(1)
one()	[query.one()]
FILTERING
eq(attr,value)	.filter(Model.attr == value)
ne(attr,value)	.filter(Model.attr != value)
lt(attr,value)	.filter(Model.attr < value)
le(attr,value)	.filter(Model.attr <= value)
gt(attr,value)	.filter(Model.attr > value)
ge(attr,value)	.filter(Model.attr >= value)
in(attr,value)	.filter(Model.attr.in_(value)
out(attr,value)	.filter(not_(Model.attr.in_(value)))
contains(attr,value)	.filter(Model.contains(value))	Produces a LIKE expression when filtering against a string, or an IN expression when filtering against an iterable relationship
excludes(attr,value)	.filter(not_(Model.contains(value)))	See above.
and(expr1,expr2,...)	.filter(and_(expr1, expr2, ...))
or(expr1,expr2,...)	.filter(or_(expr1, expr2, ...))
AGGREGATING		All aggregation functions return scalar results.
aggregate(a,b(c),...)	.query(Model.a, func.b(Model.c)).group_by(Model.a)
sum(attr)	.query(func.sum(Model.attr))
mean(attr)	.query(func.avg(Model.attr))
max(attr)	.query(func.max(Model.attr))
min(attr)	.query(func.min(Model.attr))
count()	.query(func.count())