You may not always know the type of table you need to read. For example, if a given table is a Delta table you may need to read it differently than if it were a Parquet table.

This article explains how you can use Python code in a Databricks notebook to programmatically determine if a table is a Delta table or not.

Instructions

1. Attach your notebook to an all-purpose cluster.
2. Copy the example code to your notebook.
3. Replace the following values in the example code:
   • <table-name-to-check> - The name of the table you want to read
4. Run the cell.

If the table is a Delta table, the example code returns Yes, it is a Delta table.

If the table is not a Delta table, the example code returns No, it is not a Delta table.

You can use this example code as a basis to build an automatic check into your notebook code.

Example code

%python

def delta_check(TableName: str) -> bool:
  desc_table = spark.sql(f"describe formatted {TableName}").collect()
  location = [i[1] for i in desc_table if i[0] == 'Location'][0]
  try:
    dir_check = dbutils.fs.ls(f"{location}/_delta_log")
    is_delta = True
  except Exception as e:
    is_delta = False
  return is_delta

res = delta_check("<table-name-to-check>")

if (res="True")
 print("Yes, it is a Delta table.")
else
 print("No, it is not a Delta table.")