SQL/node_modules/fs-capacitor

FS Capacitor

FS Capacitor is a filesystem buffer for finite node streams. It supports simultaneous read/write, and can be used to create multiple independent readable streams, each starting at the beginning of the buffer.

This is useful for file uploads and other situations where you want to avoid delays to the source stream, but have slow downstream transformations to apply:

import fs from "fs";
import http from "http";
import WriteStream from "fs-capacitor";

http.createServer((req, res) => {
  const capacitor = new WriteStream();
  const destination = fs.createReadStream("destination.txt");

  // pipe data to the capacitor
  req.pipe(capacitor);

  // read data from the capacitor
  capacitor
    .createReadStream()
    .pipe(/* some slow Transform streams here */)
    .pipe(destination);

  // read data from the very beginning
  setTimeout(() => {
    capacitor.createReadStream().pipe(/* elsewhere */);

    // you can destroy a capacitor as soon as no more read streams are needed
    // without worrying if existing streams are fully consumed
    capacitor.destroy();
  }, 100);
});

It is especially important to use cases like graphql-upload where server code may need to stash earler parts of a stream until later parts have been processed, and needs to attach multiple consumers at different times.

FS Capacitor creates its temporary files in the directory ideneified by os.tmpdir() and attempts to remove them:

• after readStream.destroy() has been called and all read streams are fully consumed or destroyed
• before the process exits

Please do note that FS Capacitor does NOT release disk space as data is consumed, and therefore is not suitable for use with infinite streams or those larger than the filesystem.

API

WriteStream

WriteStream inherets all the methods of fs.WriteStream

new WriteStream()

Create a new WriteStream instance.

.createReadStream(): ReadStream

Create a new ReadStream instance attached to the WriteStream instance.

Once a WriteStream is fully destroyed, calling .createReadStream() will throw a ReadAfterDestroyedError error.

As soon as a ReadStream ends or is closed (such as by calling readStream.destroy()), it is detached from its WriteStream.

.destroy(error?: ?Error): void

• If error is present, WriteStreams still attached are destroyed with the same error.
• If error is null or undefined, destruction of underlying resources is delayed until no ReadStreams are attached the WriteStream instance.

ReadStream

ReadStream inherets all the methods of fs.ReadStream.