Skip to content
| Marketplace
Sign in
Visual Studio Code>Data Science>Wide CSVNew to Visual Studio Code? Get it now.
Wide CSV

Wide CSV

Ronica

|
37 installs
| (0) | Free
Read and edit wide CSV/TSV files without scrolling sideways: key columns in the table, long text and JSON in row details. Filter, sort, edit cells, live reload.
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Copied to clipboard
More Info

Wide CSV

Read and edit wide CSV and TSV files in VS Code, Cursor, and Windsurf without scrolling sideways. The columns you scan stay in the table; expand any row to read everything else — long notes, descriptions, JSON — in full.

Filtering a column to two values from its header funnel, then expanding a row whose description is rendered as Markdown

Open a file

  • Right-click a .csv, .tsv or .tab file and choose Open in CSV Viewer, or
  • click the table icon in the editor title bar while the file is open as text.

The first time you open one of these files as text, CSV Viewer asks once whether to view it as a table. Choose Always for CSV Files to make the viewer the default (you can also do this yourself through Open With… → Configure default editor). Double-click a cell (or press F2 on a row) to edit it, and Ctrl/Cmd+S saves. Open as Text in the toolbar or the editor title bar takes you to the text editor for anything beyond cell values; it asks you to save or revert first if you have unsaved edits.

What it does

  • Table + row details. Short columns go in the table; long text, multi-line and JSON columns start in each row's details. Change the split any time with Columns.
  • Reads long values properly. Row details wrap prose, pretty-print JSON, clamp long values with More, and have a copy button per field. Columns that hold Markdown (headings, lists, tables, code, links) are shown rendered.
  • Find rows fast. Search all columns, or add filter rules ("Keep rows where status equals open"). Click the funnel in a column header to pick several values from that column at once, like a spreadsheet. Right-click any value to filter by it or copy it.
  • Sort by clicking a header; Shift+click adds a second key.
  • Handles messy exports. Detects comma, semicolon, tab or pipe; warns and offers a one-click fix when quotes are malformed; reloads when the file changes on disk.
  • Private, and careful with your file. Edits change only the cells you touch; everything else is written back byte for byte (quoting, line endings, blank lines, the final newline). No telemetry, no network, works in untrusted and virtual workspaces.

Reading wide files

When a file is first opened, CSV Viewer looks at the first 200 rows and puts up to 8 short columns in the table, in file order. Columns that hold long text, line breaks or JSON start in the row details instead. The toolbar shows the split (Columns 8/12), and the choice you make there is remembered per file.

Click a row's arrow (or the row) to expand it. The details show the columns that aren't in the table, followed by any table column whose value was cut off. In the details:

  • JSON objects and arrays are formatted, with a Raw / Formatted switch.
  • Markdown is rendered when a column looks like it (headings, fenced code, links, tables, or a mix of lists, bold, quotes and inline code in at least 1 in 10 sampled values). Click Raw to see the source; click Markdown on any long, multi-line or Markdown-looking value in another column to render it. The choice applies to the whole column and is remembered per file. Raw HTML in a cell is shown as text, images are never loaded (they appear as a link), and only http, https and mailto links are clickable. Table cells always show the raw text.
  • Long values show six lines with More / Less; very large values (over 10,000 characters) add Show all.
  • Each field has a copy button, and empty fields show a dash.

Right-click a cell or field for Copy Value, Copy Row as CSV and Copy Row as JSON.

Filtering and sorting

The search box matches across every column, including ones in the row details. Filters adds rules that read as sentences: Keep or Hide rows where a column (or any column) contains, equals, starts with, ends with or matches a regex, is empty, or compares as a number (>, <, >=, <=). Rows must match all rules. A rule that can't apply — its column is gone, or it has no value yet — is skipped with a note rather than hiding everything, and a regex that takes more than 2 seconds is skipped instead of freezing the view. The button shows how many rules are active (Filters • 2).

To keep or hide several exact values of one column, click the funnel next to its name in the header (or right-click a cell and choose Filter ‹column› by Values…). A list shows each distinct value with how many rows have it, with a search box and Select all / Clear for the values you've searched to; nothing changes until you press OK. The funnel turns blue while that column is filtered. The result is an is any of rule in Filters, where you can switch it to Hide or edit its values. The list always counts the whole file, not just the rows currently showing.

Click a column header to sort ascending, then descending, then not at all; Shift+click adds another key. Sort lists every sort key, lets you flip or remove each one, and can sort by columns that are in the row details. Numbers sort as numbers, text sorts alphabetically, and empty cells always go last.

File format

The gear button opens File format:

  • Separator: Auto (shows what was detected), Comma, Semicolon, Tab, Pipe, or a custom separator of up to 5 characters.
  • First row is header: turn off for files without a header row.
  • Quoted fields: on by default. If a stray " makes rows run together, a banner offers Read Quotes as Plain Text, which turns this off.

These choices are remembered per file. .tsv and .tab files default to Tab. They are locked while there are unsaved edits; save or revert first. The first edit to a file set to Auto fixes the separator that was detected, so the file reopens the way you edited it.

Editing

Double-click a cell, press F2 on a focused row (then Tab / Shift+Tab to move across the row), or right-click and choose Edit Value. Enter commits, Esc cancels, and clicking elsewhere commits. Multi-line values, JSON, Markdown and anything over 500 characters edit in the row details instead, in a text box where Enter adds a line and Ctrl/Cmd+Enter applies; every field in the details also has an edit button.

Edited cells get a blue mark until you save, and Revert Cell in the right-click menu puts the original back. Ctrl/Cmd+S saves, and Ctrl/Cmd+Z / Ctrl/Cmd+Shift+Z undo and redo, as in any editor. The tab shows the usual unsaved dot, Revert File discards your edits, Save As works, and unsaved edits survive a restart (hot exit). If the file changes on disk while you have unsaved edits, the viewer keeps your edits and asks before overwriting.

A save rewrites only the rows you edited, and quotes only the fields that need it, so version control sees a minimal diff.

What can't be edited: header names, adding or removing rows or columns, files that aren't valid UTF-8 (they open read-only), and values over 1 MB. Use Open as Text for those. Turn editing off entirely with csvViewer.editing.

Keyboard

  • Tab moves through the toolbar, the column headers and into the table.
  • In the table: ↑ / ↓ move between rows, → / ← (or Enter) expand and collapse, Home / End jump to the first and last row, Shift+F10 opens the row menu, F2 edits the row's first cell.
  • While editing a cell: Enter commits, Esc cancels, Tab / Shift+Tab move to the next or previous cell. In the row details, Ctrl/Cmd+Enter applies.
  • Ctrl/Cmd+Z and Ctrl/Cmd+Shift+Z undo and redo edits.
  • Cmd/Ctrl+F focuses the search box. Alt+← / Alt+→ change page. Esc closes popovers and menus.

Column headers, sort state, expanded rows and row counts are exposed to screen readers.

Limits

  • Only the values of existing cells can be edited: no adding, removing or moving rows or columns, no renaming headers, and no files that aren't valid UTF-8.
  • It shows what is saved on disk plus its own unsaved edits. An unsaved edit in a text editor appears after you save, and the viewer's unsaved edits are not visible to a text editor until you save.
  • Files over 50 MB show a "may be slow" warning; files over 512 MB are not opened.
  • Numbers written with a decimal comma (12,50) or a thousands separator (1,000) are treated as text when sorting and filtering.

Privacy

No telemetry and no network requests. Files are read through the editor's own file API; nothing is written unless you save and no processes are started.

Settings

Setting Default Description
csvViewer.defaultTableColumns 8 The most columns shown in the table when a file is first opened. Long-text and JSON columns start in the row details regardless.
csvViewer.editing true Allow changing cell values in the table and row details. Nothing is written until you save.
csvViewer.suggestOnOpen true Ask once whether to view a CSV/TSV/TAB file as a table when it is first opened as text.

Works well with

  • Rainbow CSV colors the columns of a CSV in the text editor.
  • Edit CSV opens an editable grid.

Wide CSV edits cell values; Edit CSV is the tool for adding rows and columns.

Feedback

See the changelog for release notes, and open an issue for bugs or requests. To build or test the extension yourself, see CONTRIBUTING.md.

Credits and license

MIT licensed. Parsing by Papa Parse (MIT). Icons from Codicons by Microsoft (CC BY 4.0).

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