MrlDB by Rémi "Mr e-RL" LANGDORPH

Copyright (c) 2019 Rémi LANGDORPH - mrerl@warlegend.net (under MIT license)

mrldb is a powerfull database handler for python

Links:

pypi project

github repo

Install with pypi:

pip install mrldb

This package supports the followings database systems: mariadb, mysql, cassandra, sqlite3

DOCUMENTATION

• MrlDBCluster(): a cluster of databases, the last created MrlDBCluster can be accessible via mdbcl

  • MrlDBCluster.get(name): return the MrlDB from the alias or the name

  • MrlDBCluster.add(name, db, aliases=[]): add a MrlDB object with the name (overwrite if an object has the same name), and link the aliases

  • MrlDBCluster.addalias(name, aliases): link aliases to the name

  • MrlDBCluster.get_cluster_infos(): return a dict with all the config of the connections {"name": {config...}}

  • MrlDBCluster [name]: return the db from the alias or the name

  • for name, connection in MrlDBCluster: iterate over a list of tuple (dbname, connection)

  • MrlDBCluster.dbs: a dict with all the connections {"conn1": <MrlDB>, ...}

  • MrlDBCluster.aliases: a dict with all the aliases and the realname {"alias": "realname", ...}

from mrldb import MrlDBCluster, mdbcl
mycluster=MrlDBCluster()

• mdbcl: return the last created MrlDBCluster

print(mdbcl)

• mdbstr: an object to generate sql commands(str), (the same as the one included in the MrlDB classes)
  • mdbstr.insert(table, data): return str command to insert a new record
    • table is a str of the table where the insert must be executed
    • data is dict with the columns and the values to insert {"col1": "value1", "col2", 3.14}
  • mdbstr.update(table, data, conds=None): return str command to update an existing(s) record(s)
    • table is a str of the table where the update must be executed
    • data is dict with the columns and the values to update {"col1": "value1", "col2", 3.14}
    • conds are a str object as "col1='test' and col2=5" or a NoneType object if you want to update all the records of your table
  • mdbstr.select(table, columns, conds=None): return str command to get record(s) value(s)
    • table is a str of the table where the select is executed
    • data is list with the columns to get or a "*" to get all the columns
    • conds are a str object as "col1='test' and col2=5" or a NoneType object if you want to select all the records of your table

MrlDB classes

• MrlDB base class: (Base of MrlDBCassandra, MrlDBMsql, MrlDBSqlite)

  • MrlDB.insert(table, data): insert a new record
    • table is a str of the table where the insert must be executed
    • data is dict with the columns and the values to insert {"col1": "value1", "col2", 3.14}
  • MrlDB.update(table, data, conds=None): update an existing(s) record(s)
    • table is a str of the table where the update must be executed
    • data is dict with the columns and the values to update {"col1": "value1", "col2", 3.14}
    • conds are a str object as "col1='test' and col2=5" or a NoneType object if you want to update all the records of your table
  • MrlDB.select(table, columns, conds=None): get record(s) value(s)
    • table is a str of the table where the select is executed
    • data is list with the columns to get or a "*" to get all the columns
    • conds are a str object as "col1='test' and col2=5" or a NoneType object if you want to select all the records of your table
  • MrlDB.init(): create (or ignoreif they already exists) the tables from the db structure
  • MrlDB.cursor: a connection, you can use MrlDB.cursor.execute(command)
  • MrlDB.structure: a dict of the db structure or a NoneType object if not specified
  • MrlDB._config: a dict of the connexion config
  • MrlDB._getinfos(): return MrlDB._config

• MrlDBCassandra(cluster, db=None, structure=None, username=None, password=None): a cassandra cluster handler, require library cassandra-driver

  • you can use the database you want or don't use it
  • username and passsword are only used with PlainTextAuthProvider, if you've configured users and password for your db, else, we're connecting as anonymous

• MrlDBMsql(host, database=None, structure=None, user=None, password=None): a cassandra cluster handler, require library mysql

  • host is the ip adress of the host or a dns-resolvable name of the host
  • you can use the database you want
  • username and passsword are only used with PlainTextAuthProvider, if you've configured users and password for your db, else, we're connecting as anonymous

• MrlDBSqlite(file, structure=None, autocommit=0): a sqlite file handler, require base library sqlite3 (not recommanded)

  • the file is sqlite3 db file
  • autocommit is the time in seconds (can be a float) between each autocommit, disabled if set 0 (by default)

STRUCTURE argument

with structure, you can get the column names with the results in a dict for each records

structure is an argument for all the DB classes, it must be a None oject or a dictionnary:

MrlDBCassandra(... ,structure={"table0": {"col1": "integer unique", "col2": "text"}, "table2": {"name": "text"}}, ...)

Tutorial script:

from mrldb import MrlDBCassandra, MrlDBCluster, mdbcl, mdbstr

• we add a simple cassandra cluster as connection to the mrldb cluster, we set cc0 as an alias to the db

mdbcl.add("cassandracluster0", MrlDBCassandra("127.0.0.2"), aliases=["cc0"])

• we can add other aliases to this connection:

mdbcl.addalias("cassandracluster0", ["cc0_", "cassandra0", "testcluster"])

• you can connect to your host with password and username

mdbcl.add("cassandracluster1", MrlDBCassandra("127.0.0.3", username="admin", password="something"))

• you can specify the database too

mdbcl.add("cassandracluster2", MrlDBCassandra("127.0.0.4", database="mydbtest", aliases=["cc3"]))

• and the best, you can provide database structure to have advanced features

mdbcl.add("cassandracluster3", MrlDBCassandra("127.0.0.5", database="mydbtest",
structure={"table0": {"col1": "integer unique", "col2": "text"}, "table2": {"name": "text"}}), aliases=["cc3"])

the following examples are working for all the differents database systems

SELECT

• with the correct structure, the following command will give you for each records a dict with the col name and the value, (you can replace the columns by a "*"

mdbcl.get("cc3").select(table="table0", columns=["col1"], conds=None)

result= [{"col1": 0}, {"col1": 1}...]

• you can use this without structure, it just return the results without dictionnarys

mdbcl.get("cc2").select(table="table0", columns=["col1"], conds=None)

result= [(0, ), (1, )...]

• you can use conditions

mdbcl.get("cc3").select(table="table0", columns="*", conds="col2='test'")

result= [{"col1": 0, "col2": "test"}, ...]

• in sql (not in cql !), you can do sql subrequests, we are formatting them with mdbstr class

mdbcl.get("cc3").select(table="table0", columns="*", conds=f"""(col2='test') or (col1 not in ({mdbstr.select(table='table2', columns='*', conds="name='john' ")}))""")

will execute this command: "SELECT * FROM table0 WHERE (col2='test') or (in (SELECT * FROM table2 WHERE name='john' ))"

result= [{"col1": 0, "col2": "test"}, ...]

INSERT

• data is a dict with all the values to insert

mdbcl.get("cc3").insert(table="table0", data={"col1": 5, "col2": "ok"})

UPDATE

• use data as the insert command, you can specify conditions with conds

mdbcl.get("cc3").update(table="table0", data={"col1": 5, "col2": "ok"}, conds="col2='test'")

DB init

• will create the table (or ignore if exists) as the structure

mdbcl.get("cc3").init()

with structure={"table0": {"col1": "integer unique", "col2": "text"}, "table2": {"name": "text"}}

will execute the following commands:

['CREATE TABLE IF NOT EXISTS table0(col1 integer unique, col2 text)',
 'CREATE TABLE IF NOT EXISTS table2(name text)']