Table and column naming

When writing table and column names to data warehouses, databases, and cloud storage using Bulk Syncs, Polytomic emulates a naming convention that is common when specifying database schemas: the only valid characters are alphanumeric and underscore characters with consistent casing (lowercase except for Snowflake which employs uppercase as a convention).

The source name is thus converted to match the above rule:

• The source name is lowercased (except when writing to Snowflake where it's uppercased to follow Snowflake convention).
• Separate words in the source name are concatenated with an underscore. For example: LoginID, loginID, Login ID all become login_id in your destination. However, loginid remains unchanged because there are no indicators of there being multiple words.
• Multiple underscores are replaced with a single underscore. For example: custom_salesforce_value__c becomes custom_salesforce_value_c.
• All non-alphanumeric sequences are replaced with a single underscore. For example: From field becomes from_field.

Resolving name collisions

What if two names normalize to the same one? For example, per the above rules, login_id and login__id should both normalize to login_id. In this case, to avoid name collisions Polytomic will append underscores to the colliding name until it becomes unique. Thus in this example, login__id will become login_id_.