Skip to main content
Docs
Feature Requests Onboarding Trust Center Status Page

Community

Destination Improvement: Japanese tables and columns are converted to Chinese reading alphabetical notation.

Planned
悟 相樂 User

This is a request for improvement of a problem in which tables and columns registered in Japanese in a data source are converted to alphabetical Chinese readings and the tables and columns are created.

We would like to have one of the following priorities addressed.

First:

If Destination's DWH supports Japanese column names, then we would like to synchronize with the Japanese names. Example: "顧客ID" ➟ Load as "顧客ID"

Second:

Register column names with English translations of Japanese names. Example: "顧客ID" ➟ loaded as "customer_id"

Third choice:

Register the column name in romaji reading of Japanese. Example: "顧客ID" ➟ loaded as "kokyaku_id"

Please sign in to leave a comment.

Comments

4 comments

Date Votes
  • Official comment
    Egidio Terra Fivetranner

    Hi 悟 相樂,

    This feature is now support by over 90 connectors and many destinations.
    You can find the details here: https://fivetran.com/docs/core-concepts#fivetrannaming

    Best, 

  • Egidio Terra Fivetranner
    • Edited

    Hi 悟 相樂,

    We’re in the process of changing how destination names are created. You’ll soon have the option to retain the input source name and propagate it to the destination. Let us know if you have any further questions.

    Best,

     

  • 悟 相樂 User

    That's fantastic news! I can't wait for the new feature.

  • 悟 相樂 User

    Hi Egidio,

    Thank you for the update! I tested the Source Naming feature with a Google Sheets connector and Snowflake destination, and I'm happy to confirm that Japanese column names are now being loaded correctly without conversion.

    However, I noticed some behaviors that might be worth considering for future improvements:

    1. Double-quote wrapping for all identifiers

    When using Source Naming, all schema names, table names, and column names are wrapped in double quotes (") in the DDL statements. This means that even if I enter schema and table names in lowercase during connector setup, I need to use double quotes and match the exact case when querying in Snowflake.

    For example, if I set the schema name as my_schema, I have to query it as "my_schema" rather than just MY_SCHEMA.

    2. Fivetran system columns are created in lowercase

    I also noticed that Fivetran's system columns such as _fivetran_synced and _row are being created in lowercase with double quotes. This is inconsistent with the standard Fivetran Naming behavior, where these columns would be created as _FIVETRAN_SYNCED and _ROW in Snowflake.

    This inconsistency can cause confusion when writing queries, especially for teams that use both Source Naming and Fivetran Naming connectors in the same destination.

    Suggestions:

    • It would be helpful if double-quote wrapping could be applied only to columns that contain non-ASCII characters, while keeping ASCII-only schema/table/column names without quotes.
    • Fivetran system columns could follow the standard uppercase convention regardless of the naming option selected.

    Thank you again for implementing this feature!

    Best regards