Find incomplete CSV records
Report every record whose cell count doesn't match the header: too few cells (a truncated record) or too many (an unescaped comma). Read-only: nothing is changed. Quote-aware, so a record with a quoted field containing a literal newline still counts as one record with the right cell count. Runs entirely on your device.
Runs on your device. The file is never uploaded.
Find incomplete CSV records reports every row whose cell count differs from the header's, with the record number, its count and the expected one. Records are numbered from 1 among data rows only, since a quoted newline lets one record span several lines. Nothing is written and no file comes back, so a suspect export is safe to inspect.
Questions
Does it change my file?
No. This tool is read only. It reports which records do not match the expected cell count and produces no output file at all, so you can inspect a suspect export without risking it. Fix the rows yourself, or run a tool that does rewrite, once you know what is wrong.
How does it decide what is incomplete?
The first row sets the expected cell count. Every data row with a different count is reported, whether it has too few cells, meaning a truncated record, or too many, usually an unescaped comma inside an unquoted field. Turn hasHeader off and the first row still defines the expected count but is also checked as data.
Why does it say record instead of line?
Because a record can legitimately span several physical lines when a quoted field contains a newline, which makes a line number either meaningless or wrong. Records are numbered from 1, counting data rows only, so the number stays stable no matter how the file is laid out.
What does the output look like?
A summary line naming your file, how many records failed out of how many, and the column count from the header, then one line per bad record with its number, its cell count and the expected count. If nothing is wrong you get a single line saying all records have the right number of cells.
Why did it say my file is empty?
Because parsing produced no records at all. A wholly blank file, or one containing only blank lines, has no header to take a cell count from, so there is nothing to check against and the tool refuses rather than reporting a clean bill of health.
Is my data uploaded to be checked?
No. The file is parsed in your browser and the report is built there. Nothing is sent to a server and there is no account. The whole file is held in memory while it is parsed, which is the practical size limit for a large export.