Overview

Pbi_pyadomd is a fork of the original pyadomd library, designed to provide a typed Python interface for communicating with SQL Server Analysis Services (SSAS) instances. This library allows developers to execute DAX and XML queries against SSAS in a more structured and type-safe manner. It also includes the ability to stream results from long-running queries, such as those used in SSAS Trace subscriptions.

Installation

python -m pip install pbi_pyadomd

Examples

Reading Data from an SSAS Table

Note: in this version of pyadomd, executing a DAX returns a reader object rather than a cursor object. This is intended to provide a more consistent interface with the underlying C# Adomd library.

from pbi_pyadomd import Connection

conn_str = "Provider=MSOLAP;Data Source=localhost:51358;Initial Catalog=<db_UUID>;"
with Connection(conn_str) as conn:
    reader = conn.execute_dax("EVALUATE <table_name>")
    for row in reader.fetch_many():
        print(row)

Streaming Results

Note: when reading a query without a definite end, such as a SSAS Trace subscription, you can use the fetch_stream method to continuously read results as they become available. If you use the standard fetch_many method, it will block until the query completes, which may not occur.

Both the Reader and Connection with statements automatically close their underlying resources when they go out of scope, so you do not need to explicitly close them.

from pbi_pyadomd import Connection

conn_str = "Provider=MSOLAP;Data Source=localhost:51358;Initial Catalog=<db_UUID>;"
xml_command = """
<Subscribe xmlns="http://schemas.microsoft.com/analysisservices/2003/engine">
<Object>
    <TraceID>{{trace_name}}</TraceID>
</Object>
<SubscriptionId xmlns="http://schemas.microsoft.com/analysisservices/2018/engine/800">{{subscription_name}}</SubscriptionId>
</Subscribe>
"""
with Connection(conn_str) as conn:
    with conn.execute_xml(xml_command) as reader:
        for row in reader.fetch_stream():
            print(row)

            # this break clause is a non-functional example to illustrate how you might stop reading
            if row.val == "EndOfStream":
                break

Executing Non Queries

There are certain operations that do not return a result set, such as creating or altering database objects. You can execute these using the execute_non_query method.

from pbi_pyadomd import Connection

conn_str = "Provider=MSOLAP;Data Source=localhost:51358;Initial Catalog=<db_UUID>;"
with Connection(conn_str) as conn:
    result = conn.execute_non_query("CREATE TABLE <table_name> (id INT, name NVARCHAR(100))")
    print(result)  # prints "None"