Large CSV Viewer

Open CSV and TSV files of hundreds of MB or several GB in VS Code. The first rows are on screen in a fraction of a second, while the rest of the file is indexed in the background.

Measured on a 496 MB CSV with 8.7 million rows (quoted fields, line breaks inside fields, UTF-8), VS Code 1.142, same machine and file for every extension:

• First rows on screen in about 0.2 s (median of 5 runs, from the open command to the rows being visible).
• Whole file indexed in about 1 s. You can scroll while that runs; the row count fills in as it goes.
• Jump to any row, including the last one: a few milliseconds.
• Filter, search, sort and column stats across all 8.7 million rows run in a background thread, so VS Code stays responsive. Filter results start to appear after about 0.2 s and grow while the scan continues.
• Extra memory for the index: about 20-35 MB.

Features

• Read-only. The file is never changed.
• Detects the delimiter (, ; tab |) and the encoding (UTF-8 with or without BOM, UTF-16 LE/BE with or without BOM).
• Handles quoted fields, escaped quotes (""), and line breaks inside fields correctly (RFC 4180).
• Click a header to sort ascending, then descending, then back to file order. Numeric columns sort as numbers.
• Filter one column or all columns:
  • text contains (case-insensitive), !text does not contain
  • =text equals, !=text not equal
  • >10, >=10, <5, <=5 numeric comparisons
  • /regex/i regular expression
• Search with Find next (same syntax), and jump to row.
• Column stats: count, empty, min, max, and distinct count (exact up to 10,000 values, then a HyperLogLog estimate within about 1%).
• Cancel button for long operations.
• No telemetry and no network access. No dependencies at runtime. The package is about 20 KB.

How to open a file

Right-click a .csv, .tsv, .tab or .psv file in the Explorer and choose Open in Large CSV Viewer. You can also use the button in the editor title bar, run the command Large CSV: Open in Large CSV Viewer, or use Reopen Editor With... > Large CSV Viewer. To make it the default for CSV files, add this to your settings: "workbench.editorAssociations": { "*.csv": "digivibetech.largeCsv" }

Settings

• largeCsv.delimiter: auto (the default), ,, ;, tab or |.

Limits

• The first row is always treated as the header.
• Filter, search, sort and stats wait until the index is complete (about 1 s per 500 MB on a fast disk), then scan the file at roughly 60 MB/s.
• Sort keeps a per-row key in memory: about 8 bytes per row for numbers, plus the distinct values for text. For 9 million rows that is about 300-400 MB of RAM while sorting.
• Text sort uses the full value when a column has up to 3 million distinct values. Above that, it sorts on the first 32 characters.
• Local files only (no remote or virtual file systems).

Privacy

The extension reads only the file you open. It sends nothing and makes no network requests. See https://digivibetech.com/privacy/large-csv-viewer/

Support

https://digivibetech.com/support/ · DigiVibeTech, Amersfoort, the Netherlands

Price

Free.