| title | Node:FS |
|---|---|
| sidebarTitle | node:fs |
| description | Read and write files in Edge Scripts using the Node.js-compatible file system API |
Edge Scripting provides support for the Node.js file system API through the
node:fs and node:fs/promises modules. This allows you to read, write, and
manipulate files and directories within your Edge Scripts.
The file system API is based on Deno's sandboxed environment with Node.js
compatibility, supporting both the modern Promise-based API (node:fs/promises)
and the traditional callback-based API (node:fs).
We recommend using the Promise-based API for cleaner async/await syntax.
Each script has now access to a Virtual File System that allow to handle file based operations.
This file system lives on the Virtual Memory available of the worker.
Currently the file directory is composed of two folder:
/ # Root directory
├── home/
│ └── user/ # Your current working directory, R/W access
└── tmp/ # Temporary directory, R/W access
You can create directories where you want as permissions are not enforced yet.
All files are opened in w+.
In the Virtual File System, paths are limited to 4096 chars, for instance
foo/bar is 7 characters. It's also limited to 48 segments (a/b/c is 3
segments).
Capacity checks for creating, copying and moving are done before the operation. This can result in trying to copy a large file resulting in an out of memory error, even though there is space left to create new files or directories.
EdgeScripting supports both the modern Promise-based API and the traditional callback-based API.
import * as fs from "node:fs/promises";
// Use with async/await
const data = await fs.readFile("/tmp/file.txt", "utf-8");import * as fs from "node:fs";
// Use with callbacks
fs.readFile("/tmp/file.txt", "utf-8", (err, data) => {
if (err) {
console.error("Error:", err);
return;
}
console.log("Data:", data);
});EdgeScripting's file system implementation operates in a sandboxed environment with limitations. Understanding these constraints is essential for building reliable applications.
Always handle file system errors gracefully. The sandboxed environment may behave differently from traditional Node.js environments.Symbolic links and hard links are not supported in the EdgeScripting runtime yet.
The following operations will throw errors or behave unexpectedly: - `symlink()`, `symlinkSync()` - Creating symbolic links - `readlink()`, `readlinkSync()` - Reading symbolic links - `link()`, `linkSync()` - Creating hard links - `lstat()` may not distinguish between files and symlinks correctlyWorkaround: Use copyFile() instead of creating links, or restructure
your application to avoid link dependencies.
File modification and access times are not reliably available yet.
The following `Stats` object properties may return incorrect or placeholder values: - `atime` - Last access time - `mtime` - Last modification time - `ctime` - Last status change time - `atimeMs`, `mtimeMs`, `ctimeMs` - Millisecond timestamps - `birthtime`, `birthtimeMs` - File creation timeReliable Stats properties:
size- File size in bytesisFile()- Check if entry is a fileisDirectory()- Check if entry is a directory
Example of unreliable usage:
// DON'T rely on timestamps
const stats = await fs.stat("file.txt");
console.log(stats.mtime); // May be incorrect!
// DO rely on size and type checks
console.log(stats.size); // Reliable
console.log(stats.isFile()); // ReliableAll files and directories effectively have read/write permissions in the
sandboxed environment (644) and are owned by the current (non root) user
(uid=1000, gid=1000).
File system watching is not supported.
The following operations are unavailable: - `watch()` - Watch for file changes - `watchFile()` - Poll for file changes - `unwatchFile()` - Stop watching - `FSWatcher` classWorkaround: Use polling with stat() if you need to detect changes, but
be mindful of performance implications and CPU time limits.
CPU time: File I/O counts toward the 30-second CPU time limit per request.
Best practices:
- Stream large files instead of reading entirely into memory
- Clean up temporary files to avoid storage bloat
- See Limits for more details on resource constraints
Here are practical examples showing common file system patterns in EdgeScripting.
import * as fs from "node:fs/promises";
import * as BunnySDK from "@bunny.net/edgescript-sdk";
BunnySDK.net.http.serve(async (request: Request) => {
try {
// Read file as UTF-8 string
const content = await fs.readFile("/tmp/data.txt", "utf-8");
return new Response(content, {
headers: { "content-type": "text/plain" }
});
} catch (error) {
return new Response(`Error reading file: ${error.message}`, {
status: 500
});
}
});import * as fs from "node:fs/promises";
import * as BunnySDK from "@bunny.net/edgescript-sdk";
BunnySDK.net.http.serve(async (request: Request) => {
const data = await request.json();
try {
// Write JSON data to file
await fs.writeFile(
"/tmp/output.json",
JSON.stringify(data, null, 2),
"utf-8"
);
return new Response("File written successfully", { status: 201 });
} catch (error) {
return new Response(`Error writing file: ${error.message}`, {
status: 500
});
}
});import * as fs from "node:fs/promises";
import * as BunnySDK from "@bunny.net/edgescript-sdk";
BunnySDK.net.http.serve(async (request: Request) => {
try {
// Create directory (recursive option creates parent directories)
await fs.mkdir("/tmp/myapp/data", { recursive: true });
// Write a file in the new directory
await fs.writeFile("/tmp/myapp/data/config.json", "{}", "utf-8");
// List directory contents
const files = await fs.readdir("/tmp/myapp/data");
return new Response(JSON.stringify({ files }), {
headers: { "content-type": "application/json" }
});
} catch (error) {
return new Response(`Error: ${error.message}`, { status: 500 });
}
});import * as fs from "node:fs/promises";
import * as BunnySDK from "@bunny.net/edgescript-sdk";
BunnySDK.net.http.serve(async (request: Request) => {
const filepath = "/tmp/cache/data.txt";
try {
// Check if file exists by trying to access it
await fs.access(filepath);
// File exists, read it
const content = await fs.readFile(filepath, "utf-8");
return new Response(content, {
headers: { "x-cache": "hit" }
});
} catch (error) {
if (error.code === "ENOENT") {
// File doesn't exist, create it
const defaultContent = "Default cached data";
await fs.writeFile(filepath, defaultContent, "utf-8");
return new Response(defaultContent, {
status: 201,
headers: { "x-cache": "miss" }
});
}
// Re-throw other errors
throw error;
}
});import * as fs from "node:fs/promises";
import * as BunnySDK from "@bunny.net/edgescript-sdk";
// Global promise to track the download operation
let downloadPromise: Promise<void> | null = null;
const filePath = "/tmp/large-file.txt";
async function downloadLoremIpsum() {
try {
const response = await fetch('https://lorem-api.com/api/lorem', {
method: 'GET',
headers: {
'Content-Type': 'application/json'
}
});
if (!response.ok) {
throw new Error(`HTTP error! status: ${response.status}`);
}
const content = await response.text();
// Calculate how many times we need to repeat to reach ~10MB
const targetSize = 10 * 1024 * 1024; // 10MB in bytes
const contentSize = Buffer.byteLength(content, 'utf-8');
const repeatCount = Math.ceil(targetSize / contentSize);
// Create the repeated content
let finalContent = '';
for (let i = 0; i < repeatCount; i++) {
finalContent += content;
}
// Write to file
await fs.writeFile(filePath, finalContent, 'utf-8');
} catch (error) {
// Clear the promise on error so it can be retried
downloadPromise = null;
throw error;
}
}
async function ensureFileExists() {
try {
// Check if file exists
await fs.access(filePath);
} catch (error) {
// File doesn't exist, initiate download
if (!downloadPromise) {
// Start download and store the promise
downloadPromise = downloadLoremIpsum();
}
// Wait for the download to complete (whether we started it or another request did)
await downloadPromise;
}
}
BunnySDK.net.http.serve(async (request: Request) => {
try {
// Ensure file exists (download if necessary, blocking until complete)
await ensureFileExists();
// Create SSE stream
const stream = new ReadableStream({
async start(controller) {
const encoder = new TextEncoder();
let fileHandle;
try {
// Open file for reading
fileHandle = await fs.open(filePath, "r");
// Get file stats to know the total size
const stats = await fileHandle.stat();
const totalSize = stats.size;
// Send initial event with file info
controller.enqueue(encoder.encode(`data: ${JSON.stringify({
type: "start",
totalSize,
chunkSize: 1024,
timestamp: new Date().toISOString()
})}\n\n`));
const buffer = new Uint8Array(1024);
let position = 0;
let chunkNumber = 0;
// Read file 1KB at a time
while (position < totalSize) {
const { bytesRead } = await fileHandle.read(buffer, 0, 1024, position);
if (bytesRead === 0) break; // End of file
// Convert to string
const content = new TextDecoder().decode(buffer.slice(0, bytesRead));
// Send chunk as SSE event
controller.enqueue(encoder.encode(`data: ${JSON.stringify({
type: "chunk",
chunkNumber,
bytesRead,
position,
content,
progress: ((position + bytesRead) / totalSize * 100).toFixed(2) + "%",
timestamp: new Date().toISOString()
})}\n\n`));
position += bytesRead;
chunkNumber++;
// Small delay to prevent overwhelming the client
await new Promise(resolve => setTimeout(resolve, 10));
}
// Send completion event
controller.enqueue(encoder.encode(`data: ${JSON.stringify({
type: "complete",
totalChunks: chunkNumber,
totalBytesRead: position,
timestamp: new Date().toISOString()
})}\n\n`));
} catch (error) {
// Send error event
controller.enqueue(encoder.encode(`data: ${JSON.stringify({
type: "error",
message: error.message,
timestamp: new Date().toISOString()
})}\n\n`));
} finally {
// Close file handle
await fileHandle?.close();
controller.close();
}
}
});
return new Response(stream, {
headers: {
"Content-Type": "text/event-stream",
"Cache-Control": "no-cache",
"Connection": "keep-alive",
"X-Accel-Buffering": "no"
}
});
} catch (error) {
return new Response(`Error: ${error.message}`, { status: 500 });
}
});- Node.js File System Documentation - Complete Node.js fs module reference
- Node.js fs/promises API - Promise-based file system API
- Node.js FileHandle Class - Advanced file operations
- Node.js File System Flags - Available flags for file operations
- MDN File API - Web standard File interface
- MDN Blob - Binary data objects
- MDN TextEncoder - Encoding strings to bytes
- MDN TextDecoder - Decoding bytes to strings
- EdgeScripting Limits - Resource limits and quotas