Skip to content

Latest commit

 

History

History
254 lines (157 loc) · 8.26 KB

CHANGELOG.md

File metadata and controls

254 lines (157 loc) · 8.26 KB

Changelog

4.11.0 (2023-12-06)

  • add entry point mara.commands (for mara-cli support)

4.10.0 (2023-11-21)

  • add cli group 'mara-db'. Mark old cli commands as deprecated (#74)

4.9.2 (2023-02-21)

Bugfixes 🐛

  • fix typo in depreciation warn messages and add warn category
  • fix JSONL data import into PostgreSQL db (#73)

4.9.1 (2023-02-08)

  • hotfix 🐛 issue with postgres cursor context (#72)

4.9.0 (2023-02-07)

Breaking Changes

  • the implementation of the formats option (#56) required to do changes to the functions in mara_db.shell. In case you defined custom implementations, you will have to adjust them.

Feat 🎉

  • adding a functional API to get the DB-API 2.0 of a database (#71)
  • add option to specify multiple formats like Parquet, Avro and ORC (#56, #64)

Changes 🚀

  • refactor the internally used cursor context logic to a more generic one (#68, #71)
  • a default database mara is not anymore defined (#67).
  • add deprecation warning when using parameter timestamp in query_command. This parameter will be removed in version 5.0. See #44.

Bugfixes 🐛

  • fix sqlalchemy_url for SQLServerDB
  • fix UI error when requesting schema view for Azure Synapse databases (#48)

4.8.0 (2022-09-01)

Breaking Changes

  • when using BigQuery you need to rename the BigQueryDB db config parameter service_account_private_key_file to service_account_json_file_name (#45)
  • when using BigQuery with copy/read operations you need to specify parameter gcloud_gcs_bucket_name in the BigQueryDB db config (#45)

Feat 🎉

  • Add Databricks db support (#62)
  • Add Snowflake db support (#52/#61)
  • Add sqlalchemy support for BigQuery (#45/#50)
  • Add sqlalchemy support for SQL Server (#46)
  • Add option to use either MSSQL Tools (sqlcmd/bcp) or sqsh to connect to SQL Server (#57)
  • Add documentation readthedocs.io documentation (#59)

Changes 🚀

  • Extend BigQuery functionality (#45)
  • add extras per database engine (#50/#53). Postgres and Readshift is still included by default.
    Note: It is recommended to always specifcy the extras for the database you use.
  • use client-side rendering for graphviz fallback (#51)

Bugfixes 🐛

  • add all package files in wheel
  • a proper error is now thrown when the user tries to look at the schema of a BigQuery database

4.7.0 - 4.7.1 (2020-10-23)

  • Add BigQuery support
  • Fix copy from PostgreSQLDB to BigQueryDB when delimiter_char is not set (#36)
  • Add config default_echo_queries (#38)
  • Add support for SQLServerDB port (#37)
  • Fix exit command sequence when previous command has exit code not zero (#40)

4.6.0 - 4.6.1 (2020-07-03)

  • Escape double quotes in copy_from_sdtin_command for PostgreSQL (#33)
  • Add overview page to visualization

required changes

If you use quotes in tables names in Copy commands, check whether they still work.

4.5.0 - 4.5.1 (2020-04-27)

  • Don't escape dollar sign in queries for SqlServer
  • Support echo sql queries for SqlServer
  • Bugfix copy_to_stdout_command for SqlServerDB

required changes

If use SQL Server and have queries that contain the $ sign, then please escape that one manually.

4.4.1 - 4.4.3 (2020-02-13)

  • Show warning when graphviz is not installed
  • Set fetch-count 10000 for the copy_to_stdout_command for PostgreSQLDB to handle out of memory error.
  • Add schema visualization support for SQL Server
  • Set mssql severity level to 10 (#25)

4.4.0 (2019-11-28)

  • Implement copy-from-sdtin command for Redshift (via tmp file on configuratble s3 bucket)
  • Refactor database schema visualization so that multiple databases can be implemented
  • Implement database schema visualization for MySQL
  • Add function mysql.mysql_cursor_context for connecting to MySQL databases via https://github.com/PyMySQL/mysqlclient-python
  • Allow to pass a dbs.PostgreSQLDB instance to postgresql.postgres_cursor_context

4.3.0 - 4.3.1 (2019-07-04)

  • Add travis integration and PyPi upload

4.2.0

  • Add new parameters delimiter_char and csv_format to all copy command functions (allows for better quoting JSONS, arrays, strings with tabs)
  • Add warnings for unused parameters
  • Make code a bit more pep-8 compliant

required-changes

  • Adapt own implementations of copy_to_stdout_command, copy_from_stdin_command & copy_command (add the two new parameters).
  • Test whether everything still works (has been working reliably in three big projects for 4 weeks now)

4.1.0

  • Revert commit 422c332 (Fix pg to pg copy command for json data). It was causing too much trouble.

4.0.0 - 4.0.1 (2019-04-12)

  • Allow MARA_AUTOMIGRATE_SQLALCHEMY_MODELS to be a function (in order to improve import speed)
  • Change MARA_XXX variables to functions to delay importing of imports
  • Fix pg to pg copy command for json data
  • Move some imports into the functions that use them in order to improve loading speed
  • Remove dependency_links from setup.py to regain compatibility with recent pip versions

required changes

  • Update mara-app to >=2.0.0

3.2.0 - 3.2.3 (2019-04-11)

  • Add oracle db access
  • Add SSL standard parameters to PostgreSQL connection string
  • Add missing footer parameter to Oracle copy to stdout command
  • Change arguments for sqsh client to return non zero exitcode in error case
  • Add single quotes around PostgreSQL passwords to prevent bash errors when the password contains certain characters

3.1.0 - 3.1.2 (2018-08-30)

  • Make graphviz engine in schema visualization selectable
  • Implement Redshift DB
  • Show enums in schema drawing for constrained tables
  • Extend copy_to_stdout_command with "footer" argument for PostgreSQL DB

3.0.0 - 3.0.2 (2018-04-27)

  • Move sqlalchemy auto-migration from mara-app to mara-db
  • Remove config.mara_db_alias function
  • Move function sqlalchemy.postgres_cursor_context to module postgresql
  • Remove sqlalchemy/session_context context handler
  • Import graphviz only when needed
  • Update / improve documentation
  • Add port to sqlalchemy postgres connection string
  • Extend copy_to_stdout_command with "header" argument

required changes

  • Replace all occurrences of mara_db.config.mara_db_alias() with 'mara'
  • Replace mara_db.sqlalchemy.postgres_cursor_context with mara_db.postgresql.postgres_cursor_context
  • Change all usages of mara_db.sqlalchemy.session_context to psycopg2 using mara_db.postgresql.postgres_cursor_context

2.3.0 - 2.3.1 (2018-04-03)

  • Switch dependency links in setup.py from ssh to https
  • Add psycopg2 as dependency

2.2.0 (2018-02-28)

  • add web ui for visualizing database schemas (postgres only currently)
  • improve acl
  • Fix bug in schema drawing
  • Quote strings when copying from sqlite
  • NULL value handling when copying from sqlite

2.1.0 - 2.1.3 (2018-01-19)

  • add SQLite support
  • don't use sqlalchemy session in postgres_cursor_context because it creates to many setup queries on each instantiation
  • always append ';\n\go' to queries against SQL Server
  • remove default-character-set=utf8mb4 from My SQL queries

2.0.0 - 2.0.1 (2017-12-20)

  • change database configuration from sqalchemy urls to custom database specific classes
  • create sqlalchemy session contexts from configuration objects
  • add functions for creating shell commands for accessing databases
  • add documentation
  • bug fixes
  • various smaller improvements in mara_db/shell.py

required changes

This version is pretty much incompatible with previous versions. See README.md for new usage patterns.

1.1.0 (2017-12-04)

  • Replace config function databases with database_urls
  • Add functions for client command creation

required changes

  • Change database configurations from
from sqlalchemy import engine

def databases() -> {str: engine.Engine}:
     """The list of database connections to use, by alias"""
    return {'mara': engine.create_engine('postgresql+psycopg2://root@localhost/mara')}

to

import sqlalchemy.engine.url

def database_urls() -> {str: sqlalchemy.engine.url}:
     """The list of database connections to use, by alias"""
    return {'mara': sqlalchemy.engine.url.make_url('postgresql+psycopg2://root@localhost/mara')}

1.0.0 - 1.0.1 (2017-03-08)

  • Initial version
  • Minor bug fixes and code style issues