rockset 0.4.0a20180709191949

Rockset Python Client and "rock" CLI

This package includes the Rockset Python SDK that includes the following:

• Native Python API to connect to Rockset, manage data and execute queries. This API includes Rockset’s query builder that allows for programmatic construction of advanced queries using native Python expressions.

• rock command line utility to interact, manage and operate Rockset from a shell command line. rock CLI includes a SQL REPL command-line that can be entered using rock sql.

• SQL connector for Python, which conforms to the Python DB API 2.0 specification. Refer to https://www.python.org/dev/peps/pep-0249 for more details.

• SQLAlchemy dialect to bridge Rockset and SQLAlchemy applications. SQLAlchemy users can connect to Rockset using ‘rockset://<api_key>:@<api_server>’

Rockset documentation is available at: http://docs.rockset.com

Release Notes

• 0.4.0-alpha.20180515+ecbe2a0

  • Moved to new /org/{org} REST API

  • Add SQL support that is accessible from rock sql

  • Field _id defines the unique identified for a document (:id deprecated)

  • Field _meta captures all other system generated fields in documents (:collection etc deprecated)

• 0.3.0-alpha.20180314+410b55e

  • Add py query builder support for <, <=, > and >= operators for string values

• v0.3.0-alpha.20180206+168be7d

  • Add F.collect() support to accumulate entire parent documents as part of a bucket aggregation

• v0.3.0-alpha.20171206+7debcb7

  • Add collect aggregate function to support bucket aggregations

  • Add support for F.count() to do equivalent of count(*)

• v0.3.0-alpha.20171121+d8c7961

  • Sensitive fields can now be masked or anonymized, when collections are created from AWS S3. See rock help create collection for more details.

  • AWS S3 bucket path prefixes can be provided to create a collection from only a subset of data in an AWS S3 bucket.

  • Support Parquet file format for loading data into collections.

• v0.3.0-alpha.20170919+49e62f5

  • Simpler REST API 0.3

  • Add fence API and write APIs returning commit markers

  • Add View creation progress indicator to describe view API

  • More user-friendly error messages and Python exception strings

  • Enforce and document various system-wide limits

• v0.2.2-alpha.20170824+dff8385

  • Support for next version of api_key and got rid of api_secrets

  • Added support for adding multiple documents in single API call

  • Added rockset.source to better support managing S3 sources while creating collections and collection sources while creating views.

  • Better client support for views with multiple analyzer mappings

  • py/client supports async_requests to have multiple concurrent queries pending

  • Support for sampling queries and queries on nested documents

  • Backend: better query optimization support

• v0.2.1-alpha.20170707+88ab952

  • Added support for views

• v0.2.1-alpha.20170629+4492c9a

  • py/client now respects HTTP_PROXY and HTTPS_PROXY ENV variables

• v0.2.1-alpha.20170629+7b050f5

  • Fix bug that made limit queries not composable

• v0.2.1-alpha.20170628+303f256

  • Added cursors with automatic pagination support

  • Added support for limit & skip queries

• v0.2.1-alpha.20170624+e6aee25

  • Added support for lookup operator for JOINs

  • Auth token auto refresh will support long lived Client() objects

• v0.2.1-alpha.20170615+f3df06e

  • Parallel aggregations for faster performance

  • Aggregation support for nested fields

  • Aggregation support over large data sets

  • Rewrote backend service discovery to make it elastic

  • User friendly messages for all py/client exceptions

• v0.1.1-alpha

  • First alpha release