Settings you can trust next week
The settings file stores the engine version, every dialect option, and the evidence: whether it was confirmed on the full file or only a sample, record counts, and the delimiter and encoding confidence with alternatives. Import it later to parse the next delivery the same way. Imports are validated strictly; unknown options are refused.
Code examples are not promises
Papa Parse, Python’s csv module and pandas each make different choices. pandas renames duplicate headers and its comment option cuts lines in the middle; Papa Parse picks one newline style per file; Python’s csv accepts every line ending. Each example names its library version and lists the gaps that matter for your settings. Test them with your own fixtures before relying on identical counts.
The example
A nightly tab-separated feed with a generator banner, comment lines and a blank line. Set “Skip records” to 1 and the comment marker to “#”, and the generated code follows.
Questions
Does ParseCSV run the code?
No. Code is generated text for you to copy into your own environment. Nothing is executed on this site.
Is there an API I can call?
No. ParseCSV runs entirely in your browser and has no server endpoint. Use the settings file and generated code in your own pipeline.