Improve documentation

Aggregate improvements and SQL compatibility (dask-contrib#134)

* A lot of refactoring the the groupby. Mainly to include both distinct and null-grouping

* Test for non-dask aggregations

* All NaN data needs to go into the same partition (otherwise we can not sort)

* Fix compatibility with SQL on null-joins

* Distinct is not needed, as it is optimized away from Calcite

* Implement is not distinct

* Describe new limitations and remove old ones

* Added compatibility test from fugue

* Added a test for sorting with multiple partitions and NaNs

* Stylefix

0.3.0

Features:
* Allow for an sqlalchemy and a hive cursor input (dask-contrib#90)
* Allow to register the same function with multiple parameter combinations (dask-contrib#93)
* Additional datetime functions (dask-contrib#91)
* Server and CMD CLI script (dask-contrib#94)
* Split the SQL documentation in subpages and add a lot more documentation (dask-contrib#107)
* DROP TABLE and IF NOT EXISTS/REPLACE (dask-contrib#98)
* SQL Machine Learning Syntax (dask-contrib#108)
* ANALYZE TABLE (dask-contrib#105)
* Random sample operators (dask-contrib#115)
* Read from Intake Catalogs (dask-contrib#113)
* Adding fugue integration and tests (dask-contrib#116) and fsql (dask-contrib#118)

Bugfixes:
* Keep casing also with unquoted identifiers. Fixes dask-contrib#84. (dask-contrib#88)
* Scalar where clauses (dask-contrib#89)
* Check for the correct java path on Windows (dask-contrib#86)
* Remove # pragma once where it is not needed anymore  (dask-contrib#92)
* Refactor the hive input handling (dask-contrib#95)
* Limit pandas version (dask-contrib#100)
* Handle the case of the java version is undefined correctly (dask-contrib#101)
* Add datetime[ns, UTC] as understood type (dask-contrib#103)
* Make sure to treat integers as integers (dask-contrib#109)
* On ORDER BY queries, show the column names of the SELECT query (dask-contrib#110)
* Always refer to a function with the name given by the user (dask-contrib#111)
* Do not fail on empty SQL commands (dask-contrib#114)
* Fix the random sample test (dask-contrib#117)

Increase shown test coverage

Increase speed and parallelism of the limit algorithm and implement d…

…escending sorting (dask-contrib#75)

* Increase speed and parallelism of the limit algorithm

* Fixed docs

* Implement descending sorting. Fixes dask-contrib#10

* Replace two delayed usages. Thanks @mrocklin

* Remoe the reference to the function - to make it usable without dask-sql installation on the workers

Docker images with latest tag (dask-contrib#73)

* Deploy the latest tag using a newer version of the docker action

* Use the correct tag name in all examples

* Temporarily also build in PRs for testing

Fixed dependencies

Fix setup.py to be used without sphinx dependency

Implement a SQL server speaking the presto protocol (dask-contrib#56)

* Implement a SQL server speaking the presto protocol

This server can be used to run dask-sql in a standalone application.
It is e.g. possible to run it in a dask-cluster
and answer SQL queries from external.

* Stylefix

* Fixes to docker build setup

Implement a SQL server speaking the presto protocol (dask-contrib#56)

* Implement a SQL server speaking the presto protocol

This server can be used to run dask-sql in a standalone application.
It is e.g. possible to run it in a dask-cluster
and answer SQL queries from external.

* Stylefix

* Fixes to docker build setup