Detect the delimiter before you parse
CSV tool
Delimiter guessing is fine until it is not. European decimals, quoted commas, and TSV exports renamed to .csv all create files that look comma-separated until you read them.
Why this helps
Stabilize column boundaries
Correct delimiters align headers with values. Wrong delimiters create phantom columns that look almost plausible.
Reduce silent coercion
Spreadsheet tools sometimes “help” by importing with locale assumptions. Explicit structure reduces surprises.
Pair with quoting checks
Delimiters interact with quotes. Detection is one input; quoting is the other half of the grammar.
How it works
- Inspect the header and first rows for stable field counts in CSV Unwrap.
- Compare candidate delimiters against quoting patterns.
- Lock the parse rule and validate on more rows.
Why filenames mislead
A .csv from Germany may be semicolon-delimited. A .txt may be tab-delimited. Treat extensions as weak evidence.
Detection versus domain knowledge
Heuristics suggest likely delimiters; your understanding of the domain confirms them. Combine both.
Delimiter mistakes look like data problems
Numbers split across columns and duplicated headers are often parse errors, not business anomalies.
FAQ
Is tab always safer than comma?
Tabs reduce collisions with literal commas, but tabs can appear inside quoted fields too. There is no delimiter free of edge cases.
What if two delimiters look plausible?
Validate against header field counts and sample rows. Stable schemas are the tiebreaker.
How does CSV Unwrap help?
You can inspect aligned columns quickly and iterate on parse assumptions without leaving the inspection workflow.