Skip to content
| Marketplace
Sign in
Visual Studio Code>Data Science>GridLens - CSV Editor & XLSX ViewerNew to Visual Studio Code? Get it now.
GridLens - CSV Editor & XLSX Viewer

GridLens - CSV Editor & XLSX Viewer

Dilo Tools

| (0) | Free
Offline keyed CSV comparison with duplicate-key checks, CSV/TSV editing, column summaries and read-only XLSX viewing.
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Copied to clipboard
More Info

GridLens — keyed CSV comparison without rewriting your files

Compare CSV/TSV snapshots by a unique key, edit delimited files as a grid, and browse XLSX values inside VS Code. Comparison uses literal strings: 001 is not 1, and 02.00 is not 2.00.

Why use it?

  • Find changed, added and removed records even when rows are reordered.
  • Duplicate or blank keys block matching instead of creating ambiguous results.
  • Match columns by exact header name when header mode is enabled; column order may differ.
  • Inspect column statistics and save a JSON report, without rewriting either comparison source.
  • All current CSV editing, comparison, and summary features are free. There is no active paid checkout.

GridLens CSV grid using synthetic data

Keyed comparison showing literal amount changes and added/removed keys

These images show the same interface in a synthetic local preview. File selection and report saving are validated separately inside VS Code; no customer spreadsheets are shown.

Getting started

Right-click a CSV, TSV or XLSX file → GridLens: Open as Grid. Alternatively, use Reopen Editor With… from the editor tab. GridLens does not replace your default editor automatically.

Compare two snapshots

  1. Open samples/before.csv with GridLens and enable First row is header.
  2. Click Compare CSV, choose samples/after.csv, and choose id in both key-column prompts.
  3. Result: two changed records, one added, one removed. Row reordering alone is ignored, but 02.00 versus 2.00 is reported because values are not coerced.
  4. Try samples/duplicate-keys.csv: matching is blocked and the duplicate source rows are shown.

The current file uses its in-editor text; the second file is read from its saved UTF-8 disk snapshot. Comparison and summaries include the full source data, not just the sorted/filtered display. Tables must be rectangular. In header mode, schemas need the same exact, nonblank, unique names. Neither file is merged or automatically repaired.

  • CSV/TSV: double-click a cell or press Enter to edit. Enter applies; Shift+Enter inserts a newline. Save and undo use VS Code's document system.
  • Add a blank row or delete a selected source row.
  • Filter rows, sort column views, and optionally treat the first row as a header. These operations change the display, not the underlying order in the file.
  • XLSX: choose a worksheet from the selector. Values are read-only; no workbook data is written.
  • Both rows and columns render only the visible slice.
  • Select a cell, then click Column summary to inspect that column. Numeric aggregates use finite decimal/scientific values; identifiers are not changed.

Limits and privacy

No telemetry, remote scripts, analytics or spreadsheet uploads. Current free features make no network calls. Paid activation is disabled in this release; future activation will contact Lemon Squeezy with the license key and instance ID only, never spreadsheet content. Files stay in your local/remote VS Code workspace; if using VS Code Remote, parsing runs on that workspace's extension host.

Comparison JSON contains exact counts and up to 1,000 sampled changes; the dialog shows up to 100 of those. Saved reports can contain sensitive cell values: choose an appropriate local destination before sharing them. Row indices in JSON are zero-based, while the dialog displays one-based row numbers.

CSV limits: 50 MiB of decoded text, 100,001 rows, 512 columns, one million cells. Malformed quoting is rejected rather than silently changed. Existing cell spelling and quotes are retained; VS Code itself may normalize mixed line endings or encoding when saving.

XLSX limits: 10 MiB compressed, 30 MiB unpacked, 10 MiB per ZIP entry, 100 sheets and one million cells. Some valid complex workbooks may exceed these safety limits. Encrypted, ZIP64 and split archives are unsupported. Charts, formatting and merge layout are not rendered. Dates use ISO text. Formulas are not calculated; cached values are displayed, or the formula when a cached result is absent. Close and reopen after external workbook changes.

v0.2 is free. No Pro product or checkout is available yet. Prepared paid-export and license foundations are not enabled. XLSX editing, calculations, automatic merging, and full comparison exports beyond the sample limit are not included.

Support

Contact dilojbusiness@gmail.com. Do not attach sensitive spreadsheets; provide a small anonymized example.

Develop and test

Node.js 22 is used locally. Install with npm ci, run npm run check and npm test. scripts/fidelity-benchmark.cjs measures synthetic GridLens-only parser round trips, not competitor performance or customer demand. tests/host/index.cjs exercises real VS Code editing and report export in an isolated profile. Core source is MIT; bundled upstream notices are in THIRD_PARTY_NOTICES.txt.

  • Contact us
  • Jobs
  • Privacy
  • Manage cookies
  • Terms of use
  • Trademarks
  • Your Privacy Choices
  • Consumer Health Privacy
© 2026 Microsoft