Bases Data Studio

Import local CSV and JSON data into Obsidian notes and explore the results in Bases. The plugin runs locally and does not send data to a server.

This is a development preview. Requires Obsidian 1.15.0 or later. Test in a backup vault before importing important data.

Import data

Run Bases Data Studio: Import CSV or JSON, choose a CSV with headers or a JSON array of objects, select a destination folder and a stable unique key, map fields to properties, then review the preview before importing. The preview shows new records, possible updates, skipped records and conflicts. Expand an update to see old and new values, then check only the properties you want to change. No existing property is updated unless selected. The note body and unselected properties are preserved. If a selected property or record identity changes after preview, that note is not updated; preview again.

After an import, the last folder, key field and field mappings are saved locally and restored for matching fields next time. Review them before each import, especially when switching datasets. Source data and field values are not stored in these settings. Updates are not an automatic synchronization and there is no built-in undo; back up your vault before updating important notes. The log records counts and errors, not the old values.

Use a separate destination folder for each dataset and keep using the same folder and key field when reimporting. Each imported note records its key and source folder in bases_data_studio_key and bases_data_studio_source; leave these properties intact so the plugin can recognize a note even if it is moved or renamed. It scans Markdown notes in the vault when preparing and confirming an import. Changing the original source folder for a dataset changes its identity. Duplicate or empty input keys stop the preview; occupied paths and ambiguous existing records are reported as conflicts and skipped.

The plugin creates Imported records.base and an import log in the destination folder. Existing views are not overwritten. In particular, the old Imported data.base from 0.1.0 is kept; open the new view to avoid including logs and the base itself. If interrupted, repeat the import, then inspect the log and notices for errors.

CSV values are converted to numbers or booleans conservatively; selected keys and leading-zero identifiers remain text. CSV column names must be nonempty and unique. JSON values retain their types; supported values are strings, numbers, booleans, null, and lists of strings. Nested objects, other lists and duplicate JSON keys are rejected. Quote high-precision JSON numbers to preserve them as text.

Upgrading from 0.1.0

The old release converted and trimmed keys, so their original values cannot always be recovered. When a legacy note might match an input record, the preview marks it as a conflict and skips it. Check the original data and note manually. Once you have identified the correct record, set its bases_data_studio_key property to the exact original key as text and add bases_data_studio_source with the original destination folder path. Then preview again. Keep both properties when moving or renaming the note. Take a backup before making these changes.

Development

Based on the official Obsidian sample plugin. Run npm install, npm test, npm run lint, and npm run build. Copy main.js, manifest.json, and styles.css into .obsidian/plugins/bases-data-studio/, then reload and enable the plugin.

Roadmap

Future work includes richer conflict resolution, reversible updates, synchronization, and reporting.

Release

Create a GitHub release with a tag matching the manifest version and attach main.js, manifest.json, and styles.css. Submit the repository through the Obsidian Community Directory after reviewing its requirements.