- Why does Excel break Thai text in CSV files, and how does this tool avoid it?
- Excel assumes a legacy encoding unless the file starts with a UTF-8 byte-order mark. CSV and TSV files downloaded from this tool start with that BOM, so Thai and other non-Latin text opens correctly with a double-click.
- Can it handle commas and line breaks inside a field?
- Yes. The parser follows RFC 4180: a field wrapped in quotes may contain the delimiter, doubled quotes ("") and even line breaks, so a cell like an address spanning two lines converts to one JSON value with the newline preserved.
- Should I convert to JSON objects or arrays?
- Objects (one {column: value} per row) are what most APIs and scripts expect, and they need a header row. Arrays keep every row as a plain list, which suits spreadsheets without headers or when column order matters. NDJSON writes one object per line — the format logging pipelines and jq handle best.
- How do the SQL inserts escape values?
- Column and table names are double-quoted with internal quotes doubled, and string values are single-quoted with ' doubled — the standard SQL escape. With Types on, numeric fields are written unquoted, true/false become TRUE/FALSE and empty fields become NULL.
- How does JSON to CSV handle nested objects?
- One level is flattened: the columns are the union of every object's top-level keys, in first-seen order, and a nested object or array is written as its JSON text in that cell. Rows missing a key get an empty cell.