1. Run the six-contact example
A CSV can open successfully and still contain an extra column, a duplicated ID or a date with two possible meanings. Start with fictional contacts so you can learn what the checker reports before using your own source.
- Open the CSV import checker and choose Try a sample. It runs the example and selects ContactID for the unique ID check.
- Confirm First record contains headings is checked. Leave Date-like values at Flag ambiguous dates.
- Expect 6 data records, 4 expected columns, 3 to fix and 5 to review.
The downloadable contacts-to-review.csv contains the same example. If you load that file instead, run Check CSV once, select ContactID under Unique ID column, then check again. Without an ID selection, duplicate and blank IDs are not checked.
ContactID,Name,PostalCode,StartDate
C001,Alex Morgan,02108,2026-10-01
C002,"Rivera, Sam",00123,04/05/2026
C002,Lee Chen,10001,2026-02-30
,Mina Patel,00901,2026-10-08
C005,Ana Silva,10115,2026-10-09,extra cell
C006,"Jo ""JJ"" Park",75001,2026-10-10Dates in this exercise are fixed sample values.
2. Read the findings without changing good data
| Finding | Where in this example | What to decide |
|---|---|---|
| Exact duplicate ID | Record 4: Lee Chen has C002, already used by Sam | Confirm whether this is a second person or a repeated record. Do not invent a replacement key in real data. |
| Empty ID | Record 5: Mina Patel | Get the correct ID from the source system. |
| Column count | Record 6: Ana Silva has a fifth value | Determine whether the extra value is accidental or belongs to a missing column. |
| Leading zeros | Postal codes 02108, 00123 and 00901 | Keep the zeros. Plan a Text destination field. |
| Ambiguous date | Sam: 04/05/2026 | Ask whether it means 4 May or April 5. |
| Invalid calendar date | Lee: 2026-02-30 | Check the original date; February has no day 30. |
Record numbers include the header, so the first contact is record 2. A finding also shows its starting line in the source. For quoted values containing line breaks, record and line numbers can differ.
To fix identifies structural and selected-ID problems. To review asks for a decision; it is not a command to change every flagged value. The checker does not know your business rules. Duplicate checks use exact, case-sensitive text, with separate warnings for case or outer-space similarities. They do not simulate FileMaker’s matching behavior.
3. Correct a copy using confirmed source values
For this fictional exercise, assume the source confirms Lee’s ID is C003, Mina’s ID is C004, Sam started on 4 May 2026, Lee started on 28 February 2026, and Ana’s extra cell was accidental. Those are exercise facts, not corrections the tool can infer.
- Make those five changes in a copy of the source. Keep the three postal codes exactly as written.
- Keep the comma inside Sam’s name within double quotes. Keep the doubled quotes around JJ; they represent literal quotation marks.
- Export the copy as UTF-8 CSV. Reopen it in the checker and verify the preview.
ContactID,Name,PostalCode,StartDate
C001,Alex Morgan,02108,2026-10-01
C002,"Rivera, Sam",00123,2026-05-04
C003,Lee Chen,10001,2026-02-28
C004,Mina Patel,00901,2026-10-08
C005,Ana Silva,10115,2026-10-09
C006,"Jo ""JJ"" Park",75001,2026-10-10You can use contacts-reviewed.csv to compare your result. The checker does not edit your input or create a repaired CSV. Download report · TXT saves a findings report, not an import file.
4. Recheck and keep the intentional warnings
Run Check CSV once to load the headings. Select ContactID if needed, set Date-like values to Expect YYYY-MM-DD, then check again.
The reviewed file should show 6 records, 4 columns, 0 to fix and 3 to review. All three review findings are the postal codes with leading zeros. That is the expected result: their values are useful and should stay intact.
A date-order setting only changes what the checker looks for. It does not convert the CSV or configure FileMaker. If a spreadsheet has already turned 00123 into 123, the checker cannot reconstruct the missing zeros. Return to the original export.
5. Test the import separately
The checker has inspected one text file. It has not checked your FileMaker fields, existing records, relationships, validation or auto-enter rules.
- Use an empty practice table in a copy of your FileMaker file. For this exercise, preserve ContactID and PostalCode in Text fields.
- Choose the reviewed CSV, use Add new records, designate the header as field names and map the four columns explicitly.
- Review auto-enter behavior before importing. Check whether StartDate is interpreted as intended; a temporary Text field is useful when you need to inspect dates before converting them.
- Confirm six added contacts, distinct IDs, the three postal codes, Sam’s comma, Jo’s quotation marks and both corrected dates.
Claris documents the separate choices for field mapping, header rows, import actions and auto-enter options in Importing data into an existing file. Follow our contact import walkthrough for the wider process. If you are updating existing records, use the matching-record update guide instead of treating this addition exercise as an update recipe.
Use the same routine on your own CSV
Keep an untouched original, confirm the delimiter and headings, choose an ID only when that column should be unique, and rerun after each correction. The preview shows the first 20 data records, while the analysis checks the full accepted input. Do not treat the preview as the whole file.
The tool accepts UTF-8 files up to 5 MiB, with up to 50,000 data records, 100 columns and 1,000,000 cells. It retains the first 1,000 findings and reports overall totals. After fixing early problems, recheck to reveal later findings.
CSV contents stay in the browser tab and are not uploaded or saved in browser storage. Downloaded reports contain value excerpts, so review them before sharing. Clear removes the tab’s input and results; it does not delete downloaded files.