Naming conventions

Fivetran provides the options to customize your connection's schema name when you first set up a connection. When you use Fivetran's standard naming, the schema name becomes the connection name and must be unique for each connection. However, this rule does not apply if you enable Source naming.

We use one of the following naming conventions to name your schemas, tables, and columns in the destination:

• Fivetran naming
• Source naming

Fivetran naming

You can use this naming convention to standardize the schemas, tables, and columns names when they are replicated from the source to your destination. Depending on the connector and object type, we use one of the following naming rule sets:

• Table and column naming rule set
• Schema naming rule set

Destination schema vs destination schema prefix

Database connectors can replicate multiple schemas to a destination. When setting up a database connection, you must specify a destination schema prefix. The schema prefix that you create is appended to the names of the source schemas.

For example, let's say an Azure SQL database connection has two schemas, schema1 and schema2. If you specify the destination schema prefix as azure_sql_db, the replicated schemas in your destination would be azure_sql_db_schema1 and azure_sql_db_schema2.

For non-database connectors you specify the destination schema name. These types of connectors replicate a single schema to your destination. We load every table you selected from the source into this schema. For our webhooks and file connectors, you choose the table name in addition to choosing the schema name. For all connectors except webhooks and file connectors, Fivetran auto-generates the table and column names from the names of corresponding objects.

Table and column naming rule set

Fivetran has two different table and column naming rule sets:

• Naming rule set for non-database connectors
• Naming rule set for database connectors

Table and column naming rule set for non-database connectors

For non-database connector types, when we name tables and columns in the destination, we apply the following rules in the order in which they are listed:

• Transliterate non-ASCII characters. We support transliterations for the following character sets:
  • Han
  • Katakana
  • Arabic
  • Cyrillic
• Remove all accent marks from characters
• Convert any character that is not a letter, digit, or underscore to underscore. A space is considered a character.
• Divide alphanumeric strings containing letters and numbers into multiple words and numbers joined with underscores. A change from lower-case letter to an upper-case letter and alternation of numbers and letters are recognized as separators. However, two or more adjacent capital letters are considered a part of one word. For example, "anotherName" becomes "another_name", but "ANOTHERName" becomes "anothername". "MyName123" becomes "my_name_123".
• Convert multiple underscores to a single underscore
• Convert all upper-case letters to lower-case
• Prepend names that start with a number with an underscore
• For table names only, we remove any underscore prefixes if they are followed by a letter

Table and column naming rule set for database connectors

For database connectors, we do minimal name-changing between the source and destination. There are two exceptions to this rule:

1. Schema names that are prepended by the connection name.
2. Schema, table, or column names that violate Fivetran’s naming conventions.

In these cases, we apply the following rules in the order in which they are listed:

• Non-ASCII characters (such as Chinese characters) are replaced with a transliteration.
• Remove all accent marks from characters
• Convert any character that is not a letter, digit, or underscore to underscore. A space is considered a character.
• Convert all upper-case letters to lower-case
• Delete any characters that exceed our 127 character limit.

Schema naming rule set

When we name schemas in the destination, we apply the following rules in the order in which they are listed:

• Transliterate non-ASCII characters. We support transliterations for the following character sets:
  • Han
  • Katakana
  • Arabic
  • Cyrillic
• Remove all accent marks from characters
• Convert any character that is not a letter, digit, or underscore to underscore. A space is considered a character.
• Convert all upper-case letters to lower-case

Source Naming

Source naming does not follow Fivetran’s naming conventions. When enabled, Fivetran bypasses the renaming rules and preserves the original or source names of the schemas, tables, and columns with UTF-8 values without any modifications, and replicates them to your destination.

Some destinations may have reserved keywords or restrictions that prevent certain names from being used. In such instances, Fivetran displays a warning message and excludes the schemas, tables, or columns from the sync.

Preserving source names is incompatible with QuickStart transformations. Therefore, if you enable Source naming, QuickStart transformations is automatically disabled for your connection.

Supported connectors and destinations

Fivetran supports Source naming for a subset of its connectors and destinations. However, this feature is not supported in data lakes.

Supported connectors

Fivetran supports Source naming for the following connectors:

Supported destinations

Fivetran supports Source naming for the following destinations:

• Azure Synapse
• BigQuery
• Databricks
• MySQL
• PostgreSQL
• Redshift
• Snowflake
• SQL Server