Data · THE NO-PANIC PLAN
Validate a CSV before importing it
For a person preparing a comma-separated or other delimiter-separated file for a known CRM, spreadsheet, database, marketplace or other importer. CSV is interpreted differently by different applications: first obtain the destination's current import specification and treat that as the contract. RFC 4180 is an informational common-format reference, not a universal guarantee that every importer accepts the same encoding, separator, header, quoting or value rules. This workflow keeps the source unchanged, uses Nirmion tools only for their stated supporting tasks, and requires a controlled test import before loading production data. Check tool data-handling behavior and your organization's policy before processing confidential or personal rows; use synthetic samples or an approved local process if needed.
MISSION Check a CSV's structure, fields, data values and importer settings against the destination system before a production import.
Check the CSV before importTHE REAL-WORLD BIT
What happens outside this browser tab?
Get the receiving system's import contract, inspect the file's dialect and structure, map source headers to destination fields, check values and duplicates, then test in a non-production destination and reconcile the result before production use.
YOUR CHECKLIST, WITH FEWER DRAMATIC SIGHES
One step at a time.
Follow the order below. If a step names a Nirmion tool, its link is right there with it.
- 01
Get the import specification and preserve a baseline copy
Name the exact system, table/list and import method that will receive the file, then obtain its current documentation or administrator-approved mapping. Record the required header names/order, mandatory and optional fields, data types and formats, maximum lengths, null/blank handling, key/unique rules, accepted character encoding, delimiter, quote/escape behavior, header-row setting, file-size limits and duplicate/error behavior. Save an unchanged source file and record its name, date, size and expected record count; do not clean the only copy. If the receiving application has no written contract, ask its owner for a sample import and acceptance rules before changing data. BigQuery's official CSV loader documents configurable encoding, delimiters and parsing options and product-specific date requirements, illustrating why CSV settings are destination-specific; those BigQuery rules must not be applied to another system. (Sources 1, 2)
- 02
Check the file dialect, header and record shape
Inspect the raw text with a tool that respects CSV quoting instead of counting commas in a text editor: a comma inside a quoted field is data, and quoted fields may contain commas, quote characters or line breaks. Confirm the observed separator, quote and escape convention, line ending, encoding and whether the first row is a header. Check that header names are present, unambiguous and match the agreed mapping, and that every parsed record has the expected number of fields, including the final row. RFC 4180 describes a common comma-separated format and says records should have the same field count, but it is informational and explicitly acknowledges implementation differences. Nirmion CSV Field Mapping Builder (tool 922) can help map source column names to the documented destination fields; review every mapping and do not treat that mapping output as a parser or schema validator. (Sources 1, 2, 3)
- 03
Match the parser settings without altering values accidentally
Compare the file's actual encoding, delimiter, quote character, header presence and newline handling with the target import form. If you intentionally need to change a supported delimiter, CSV Delimiter Converter (Nirmion tool 143) can convert between its listed separators; verify the converted file still parses the same number of fields and records and preserves quoted content. A delimiter converter does not infer the destination schema, repair malformed values or prove that an import will succeed. Do not open and re-save the only copy in a spreadsheet program without checking its import behavior: Microsoft's Excel guidance notes that default column settings can reinterpret dates and numbers, including stripping leading zeros. When inspecting in Excel, import through its text/CSV flow and set sensitive identifiers or code-like columns to text where appropriate; then compare representative values back to the source. Preserve the original and record every intentional conversion. (Sources 1, 2, 4)
- 04
Validate mapped values, required fields and duplicate keys
For each destination field, check required-versus-optional status, permitted types and formats, blank/null interpretation, date and time conventions, numeric precision, text length, allowed codes and any documented range or cross-field rule. Keep identifiers such as postal codes, SKU codes and account-like strings as text if the destination expects them; do not normalize or fill missing values without an approved business rule. Check unique keys and repeated rows against the target's actual duplicate policy. Nirmion CSV Duplicate Finder (tool 142) can identify duplicate records by selected columns and preserve source row numbers for review, but it cannot decide whether repeated records are valid updates or errors and may not detect near-duplicates. Use representative examples for empty values, quotes, embedded line breaks, non-ASCII characters, leading zeroes, long text, invalid dates and duplicate keys; reconcile any fixes to the source-of-truth owner and retain a change log. W3C's tabular-data model treats metadata such as column datatypes and constraints as part of interpreting and validating a table, supporting validation against an explicit schema instead of visual appearance alone. (Sources 2, 3, 4)
- 05
Run a test import and reconcile before production
Load a small but representative copy into the destination's sandbox, staging area or test list using the exact planned import options. Save the importer settings and error report; confirm the detected header, parsed row/column totals, accepted/rejected counts, field mapping, date/number interpretation, encoding and duplicate behavior. Compare key fields from several records—including edge cases—to the unchanged source and investigate every skipped row or coercion. If the tool offers only a preview, do not mistake it for a successful commit; verify the destination state after a test load. Agree on a rollback or restore path and an import window with the system owner, then use the approved final file and settings in production. Reconcile total rows and key totals after loading, log the final file version and result, and keep the source and validated artifact in their approved retention locations. BigQuery's documentation warns that autodetection and parsing depend on its load settings, while Excel can apply default conversions; confirm the behavior in the actual target instead of assuming a generic CSV check guarantees safe import. (Sources 1, 2, 4)
THE HELPER CREW
Tools for the fiddly bits.
These are the currently published Nirmion tools matched to this guide. Open a tool page for its accepted inputs and limits.
RECEIPTS, PLEASE
Sources & review notes
Each source is linked to the steps it supports. Open it to check its scope and current guidance.
Source checked 2026-10-06