NORTHMELD PRACTICAL GUIDES
Standardize Excel dates while preserving leading-zero IDs
Keep identifiers as text and normalize only dates with an unambiguous interpretation. In this four-row sample, Northmeld converts two date strings to YYYY-MM-DD, leaves an already normalized date unchanged, and preserves 03/04/2026 for review. The XLSX export keeps 00128 as a text identifier.
Prepare and download
Download dates-input.xlsx. Code cells are stored as text, Amount cells as numbers and the dates as strings. Import the Deliveries worksheet. The example does not repair IDs that another application has already converted to numbers.
Reproduce the result
Inspect the original cell values
Confirm that the first identifier is 00128, including both leading zeros. Check the four date values in the input table below and keep an unchanged source copy.
Set the date format
Select Normalize safe dates and choose YYYY-MM-DD as the target format. Do not apply unrelated numeric conversions to Code. Review the proposed changes before applying the plan.
Review the unresolved date
2026/09/09 should become 2026-09-09; 10 Sep 2026 should become 2026-09-10. Leave 03/04/2026 unresolved until the source owner confirms whether it means March 4 or April 3.
Verify the exported workbook
Export XLSX and reopen it. Check that all four codes retain their leading zeros, amounts remain numeric and 03/04/2026 has not been guessed. The downloadable CSV has the same text values but cannot store cell types.
Actual sample results
- Input data rows
- 4
- Output data rows
- 4
- Duplicates removed
- 0
- All four records remain. Two dates are reformatted and one already normalized date stays the same.
- 03/04/2026 is preserved unchanged because day/month order cannot be resolved safely from the value alone.
- Re-importing the XLSX confirms that Code values are strings and Amount values are numbers, including zero. This verifies types as well as visible values.
| Code | Delivery date | Amount |
|---|---|---|
"00128" | "2026/09/09" | 1250 |
"00129" | "10 Sep 2026" | 75 |
"00130" | "2026-09-11" | 0 |
"00131" | "03/04/2026" | 42 |
| Code | Delivery date | Amount |
|---|---|---|
"00128" | "2026-09-09" | 1250 |
"00129" | "2026-09-10" | 75 |
"00130" | "2026-09-11" | 0 |
"00131" | "03/04/2026" | 42 |
Limits and human review
- If the source already stores 128 as a number, the lost leading zeros cannot be inferred reliably without an external ID-length rule.
- CSV does not carry spreadsheet cell types. Opening it directly in Excel may trigger automatic conversions; import identifier columns as text or use the XLSX export.
- This sample covers date strings, not all Excel serial dates, time zones, formulas or workbook formatting. Resolve ambiguous dates with source evidence before applying a correction.