KitForma field notes

Check XML conversion fidelity and supported limits

Know what table extraction preserves, what remains in the original, and why some documents are refused.

Syntax checks are only the first gate

This tool accepts bounded UTF-8 XML 1.0 documents with one root element. It checks markup and namespace structure, refuses DTD and entity declarations, and rejects processing instructions. It does not load an XSD schema or claim that a well-formed document follows your supplier’s business rules.

Malformed XML produces an error instead of a partially repaired success. If the source uses another character encoding, export a UTF-8 copy in the original application. Removing an encoding label without converting the bytes can corrupt text. Keep the original before making any changes.

Keep the original for lexical and mixed-content fidelity

XML parsers apply defined normalization to line endings and attribute whitespace. A table of parsed values is therefore not an exact reproduction of the source bytes. Comments, formatting whitespace and namespace-prefix choices also do not naturally belong in ordinary record columns.

Mixed content combines text and nested elements, as in a note with an emphasized word. The converter extracts direct text and child values, and flags the mixed node. Original ordering remains available in source.xml inside the CSV ZIP and in the JSON rawXml field. Nodes outside your chosen record subtree are retained there too.

Use the limits as practical selection boundaries

The browser workflow is intentionally bounded: 2 MiB of input, 20,000 elements, 32 nesting levels and 500 unique element paths. The mapped output is further limited to 10,000 rows, 32 tables, 256 data columns per table and 250,000 cells. Expanded text and output-size budgets can reject smaller inputs with unusually large schemas.

Choose a narrower record level or split the source in its exporting application when a limit is reached. This release does not stream huge files, validate schemas, infer joins between unrelated records or restore missing source data. Before using an export operationally, compare row counts, representative exact identifiers, child-parent links and known totals in your destination system.

The evidence behind the check

Sources & assumptions

Formats, algorithms and supported features are documented below. Check the stated scope and inspect your output before using it.

  • W3C: XML 1.0 ↗

    XML character, markup and text-normalization rules; this tool supports a bounded subset without DTDs.

  • W3C: Namespaces in XML ↗

    Namespace URI and local name identify elements and attributes independently of prefix spelling.

  • SheetJS: Cell objects ↗

    The workbook writes text cells and no executable formulas.

  • OWASP: CSV injection ↗

    Spreadsheet applications can interpret CSV text as formulas; protected CSV and literal XLSX have different handling.

Reviewed 2026-10-18. Read our methodology or report an issue.

Essential local storage remembers this choice for up to 180 days. Your choice applies to this subdomain. Global Privacy Control and Do Not Track override optional measurement.

Advertising has a separate choice. Use Advertising preferences in the footer to allow or reject the optional guide-page banner.