Data tables

Import spreadsheet data

Preview and import CSV, TSV, or supported XLSX data into a new Kitable or an existing table with reviewed field types.

Updated

By Kition Docs

Verified against Kition 0.1.41 (d3e1b931) on 2026-08-29

Supported entry points and formats

  • Workspace Create menu → Import spreadsheet — create a new .kitable from CSV, TSV, or XLSX
  • Open table item menu → Import data — import into the active table
  • .csv and .tsv — supported by the current runtime and by the compatibility path for an existing table
  • .xlsx — requires a runtime that advertises XLSX import support
  • Legacy .xls — not supported; save it as XLSX, CSV, or TSV first

Review the preview

The import dialog shows the detected format, row count, field count, selected worksheet, sample values, and warnings. For a workbook with multiple visible sheets, choose the sheet to import.

Review each inferred field and change its type when needed. Import overrides are limited to text, long_text, number, date, datetime, single_select, multi_select, checkbox, and url.

Choose how an existing table is written

Although the runtime contract includes an upsert mode, the 0.1.41 import dialog does not expose an Update by key or unique-key picker. Do not depend on an upsert workflow from the public UI.

Creating a new Kitable uses append mode against the new target. The dialog submits automatic schema handling with the field overrides you reviewed.

  • Append — keep existing records and add imported records
  • Replace — replace through the runtime import job

How the compatibility CSV path behaves

When the updated runtime import capability is unavailable, an existing table can still use the client compatibility path for CSV or TSV. It detects comma, tab, or semicolon delimiters; preserves quoted commas, escaped quotes, and quoted newlines; and normalizes blank or duplicate headers.

A pristine blank table can have its default schema and empty rows replaced by the inferred CSV schema. For an established table, the compatibility importer keeps existing fields, adds missing fields by normalized title, and imports the normalized CSV rows.

Inference is a starting point

The compatibility analyzer can infer checkbox, number, datetime, date, URL, long-text, and single-select fields; otherwise it uses text. Its Boolean tokens are true/false and yes/no, not the old documented 1/0 rule.

Use a header row, prefer UTF-8, normalize mixed date formats, and inspect columns with identifiers, leading zeros, large numbers, or sparse categorical values before importing.

Progress, cancellation, and recovery

Runtime imports report queued, parsing, validation, schema-writing, record-writing, and finalization progress and can be canceled from the dialog. Completion reports created, updated, skipped, and field counts.

Do not rely on old fixed timing estimates or row-count limits. For important data, keep the source file, import into a test table first, and verify record and field counts before deleting the original.

Implementation sources

Related pages

Ready when you are.

Kition is a desktop AI workspace for connected documents, structured tables, editable designs, whiteboards, visual workflows, and reviewable AI agents.