Bulk API 2.0: BOM in CSV Files Causes Partial Failure Mapping Issues
0 reputation · 19 Jul 2020, 22:34 UTC
Problem Context
When ingesting CSV data with Salesforce Bulk API 2.0, a file that begins with a UTF‑8 Byte Order Mark (BOM) can trigger partial failures. The API reports these failures, but the returned error log does not include the record number or line offset, making it difficult to pinpoint the problematic rows.
Goal and Constraints
The objective is to measure the latency and ambiguity introduced by this behavior before applying any optimization. Constraints include the need to work with API version 58.0 or later, the lack of formal documentation on BOM handling in Bulk API 2.0, and the requirement to avoid impacting production data loads.
Unresolved Questions
1. What API calls or log entries can be leveraged to correlate error rows with CSV lines when a BOM is present?
2. Is there a configuration setting or recommended pre‑processing step that ensures record numbers are preserved in the error log?
3. How does the current Bulk API 2.0 implementation treat BOM characters during parsing, and can this behavior be overridden or documented in a future release?
salesforce