CSV Column Analyzer

In short The failures that break a CSV import are rarely in the data you are looking at. They are ragged rows, where one line has more or fewer fields than the header, and they usually come from an unescaped comma inside a value or a line break inside a quoted field. This profiles every column — type, blank count, distinct values, ranges — and flags those rows specifically, because a single ragged row is enough to shift every subsequent column and turn a clean import into silent nonsense. The column profile catches the other class of problem: a column that is 98% numbers and 2% text, a date column with three different formats in it, or an identifier column with unexpected duplicates. Those pass an import cleanly and produce wrong answers afterwards, which is worse. Type detection is a heuristic — a column becomes a date when enough of its values parse as one under recognisable formats — so a column of ambiguous values like 01/02/2026 is worth checking by eye, since the same string is January and February in different conventions. The file never leaves your browser, and it is built for inspecting an export rather than processing a very large dataset.