CHouseUI
Using CHouse UIExplorer

Upload & preview

Load CSV, TSV and JSON files into existing tables and sample data with pagination.

In the app
Explorer › Import data/explorer

Load files into existing tables and inspect rows without writing SQL.

File upload

Upload CSV, TSV or JSON files into an existing table (requires table:insert):

  1. Open the table in the Explorer and choose Upload data.
  2. Select the file — .csv, .tsv or .json.
  3. Map/confirm the target columns (header handling follows the format: headers on first row for CSV/TSV, object keys for JSON).
  4. Start the upload — rows are inserted via the server proxy in batches.

Behavior notes:

  • Uploads append to the table; they don't truncate
  • Format errors surface per batch with row context — fix the file and retry rather than fighting partial state
  • Very large files: prefer ClickHouse-native bulk loading (clickhouse-client) and keep the UI upload for team-sized datasets
  • All inserts pass data access rules — the target table must be allowed

Data preview

Every table opens with a paginated sample:

  • Bounded reads (server caps rows per page — configurable via preferences' max-rows default)
  • Sortable columns and quick filters on the sampled page
  • Column types shown alongside values — catches Nullable/stringly-typed issues early

The preview is a SELECT … LIMIT through the security pipeline; table:select gates it.

From preview to action

Next stepHow
Run a real queryOpen in SQL editor
Export the sampleExports — CSV/JSON/TSV
Check schema hygieneSchema advisor
Diagnose part healthParts

Tip: If an upload lands with unexpected types (numbers parsed as strings), check the source file encoding/quoting — ClickHouse types are inferred from the table schema, not the file.

Edit this page on GitHub ↗