Fast, safe, token-based formatter for very large XML, JSON, and HTML files in VS Code & Cursor.
A formatter built to handle huge documents without blocking the editor — with structural safety checks, minimal edits, worker-thread offloading and performance visibility.
Unique features
Worker-thread formatting for large files
Offloads formatting to a Node worker when a file exceeds a configurable byte threshold (per language) so VS Code stays responsive.
Structural validation with automatic safe fallback
Re-tokenizes formatted output and compares a structural signature; if the structure changes, the extension preserves the original content to avoid corrupting XML, JSON, or HTML.
Token-aware tokenizer & formatter (XML, JSON, and HTML)
XML: declarations, DOCTYPE, processing instructions, comments, CDATA, quoted attributes. JSON: objects, arrays, strings, numbers, literals. HTML: void elements, comments, DOCTYPE, and raw-text elements (script, style, textarea, title, …) — more robust than regex-based formatters.
Minimal, offset-based edits
Computes the smallest replacement range and applies only that edit to reduce undo churn and preserve editor state (selections, cursors).
Performance visibility
Optional info popups show duration, worker/fallback status, size/threshold, edit count and token count.
Diagnostics and graceful error handling
Emits warnings for unterminated or invalid tokens; worker failures fall back to a safe main-thread pass.
|
Normal formatters |
Large File Formatter |
| Where it runs |
Main thread (same as the editor) |
Worker thread for large files |
| On huge files |
Editor can freeze or time out |
Stays responsive; work runs off the main thread |
| Speed on large files |
Often slow or unusable (MB-sized files) |
Much lower formatting time for the same file size |
Example (XML, ~20 MB file):
File Size: 20.66 MB
Tokens: 3,200,003
Formatting Time: 5133 ms
Worker Thread: enabled
Memory Mode: large-file
That’s a format time that most built-in or main-thread formatters cannot match on files this size — and the editor stays usable while it runs. Try the same file with a normal formatter and compare.
Commands
Large File Formatter: Format Current Document — format the current XML, JSON, or HTML document (also bound to the standard Format Document action).
Settings
large-file-formatter.insertFinalNewline (boolean, default: true)
Append a trailing newline when formatting XML, JSON, or HTML.
large-file-formatter.workerThresholdBytes (number, default: 131072)
Minimum XML document size (bytes) before formatting is offloaded to a worker.
large-file-formatter.jsonWorkerThresholdBytes (number, default: 131072)
Minimum JSON document size (bytes) before formatting is offloaded to a worker.
large-file-formatter.htmlWorkerThresholdBytes (number, default: 131072)
Minimum HTML document size (bytes) before formatting is offloaded to a worker.
large-file-formatter.showFormatTiming (boolean, default: true)
Show a popup with formatting duration after each format.
large-file-formatter.showFormatDetails (boolean, default: true)
Include worker/fallback/size/edits/token details in the timing popup.
- Designed for very large XML, JSON, and HTML files where main-thread formatters can freeze the editor.
- Prioritizes safety — structural validation plus automatic fallback prevents accidental corruption.
- Minimizes editor churn through minimal-range edits and provides visibility into performance and decisions.
Quick start
- Install and enable the extension in VS Code.
- Open an XML, JSON, or HTML file and run “Format Document” or use the extension commands.
- Tune behavior under
Preferences → Settings → Large File Formatter.
Architecture (simplified)
flowchart TB
Editor --> XmlProv[XML Provider]
Editor --> JsonProv[JSON Provider]
Editor --> HtmlProv[HTML Provider]
XmlProv --> Client[FormatWorkerClient]
JsonProv --> Client
HtmlProv --> Client
Client -->|language + options| Worker[Worker]
Worker --> Dispatch{Dispatch}
Dispatch -->|xml| XmlPipeline[XML pipeline]
Dispatch -->|json| JsonPipeline[JSON pipeline]
Dispatch -->|html| HtmlPipeline[HTML pipeline]
XmlPipeline --> Edits[Minimal edits]
JsonPipeline --> Edits
HtmlPipeline --> Edits
Edits --> Provider
Known constraints
- XML/HTML: avoids aggressive reflow in mixed-content nodes to preserve semantics.
- HTML: void elements (
br, img, …) are treated as self-closing for indentation; script/style/textarea/title content is preserved as raw text.
- For malformed XML, JSON, or HTML, the formatter may fall back to the original input to guarantee safety.
Contributing
GitIngest is open source and we welcome contributions.
- 🌟 Star the repository – Help others discover Large File Formatter
- 🐛 Report bugs – Open an issue
- 💡 Suggest features – Share ideas via GitHub Discussions or issues
- 🔧 Contribute code – See CONTRIBUTING.md for development setup, code style, and how to submit pull requests
📜 License
This project is licensed under the MIT License. See the LICENSE file for complete details.
| |
