-
Notifications
You must be signed in to change notification settings - Fork 310
Description
Description:
In Data Loader v64.0.0 (and later), there is a critical UI/UX regression where the "Auto-Match" logic incorrectly overwrites the mapping loaded from an .sdl file. This behavior makes it impossible to intentionally leave specific fields unmapped if the CSV header matches the Salesforce API name.
Steps to Reproduce:
①Prepare a CSV file where a header matches a Salesforce API name (e.g., RecordTypeId).
②Launch Data Loader v64 and go to the Mapping step.
③Click "Choose an Existing Map" and load an .sdl file that does not map the RecordTypeId field (it should be left blank).
④Observe that the mapping is correctly loaded (the field is unmapped).
⑤Click the "Create or Edit Map" button to adjust other fields.
Actual Result:
The moment the "Edit" window opens, Data Loader automatically runs its Auto-Match logic and fills in the RecordTypeId mapping, overwriting the blank setting from the .sdl file. The user must manually remove this mapping every time they open the window.
Expected Result:
The mapping loaded from the .sdl file must be preserved. Auto-Match should only run when the user explicitly clicks the "Auto-Match Fields to Columns" button, or at least it should not overwrite existing configurations (including intentional blanks) provided by a mapping file.
Comparison with Previous Versions:
In version 60.0.0, the loaded mapping was respected, and opening the edit window did not result in forced auto-mapping of unmapped fields.
Environment:
Data Loader Version: 64.1.0
OS: Windows