An isomorphic, streamable and datastore agnostic library for efficient and flexible file storage management.







An isomorphic library for efficient and flexible file storage management. It uses chunk-based storage, stream support, and batch operations for optimized performance and data integrity. The library is agnostic to the storage backend, accepting any datastore through dependency injection.
- Write, read and delete file operations.
- Chunk-based storage for better memory utilization.
- Pluggable storage backends.
- Stream support for write and read operations.
- Support for batch operations to commit multiple changes atomically.
Install the package with:
``bash`
npm install @plaindb/drive
First, import the Drive library.
`js`
import Drive from '@plaindb/drive'`
orjs`
const Drive = require('@plaindb/drive')
You can create an instance of Drive using your own storage backend.
`js`
const Drive = require('@plaindb/drive')
const storage = / your storage instance here /
const drive = new Drive(storage)
The Drive constructor accepts an optional second argument to specify various settings:
`js`
new Drive(storage, {
chunkSize: 8192, // Size of each file chunk in bytes (16kb default)
hashLength: 8 // Length of the hash generated from each chunk (32 default)
})
For standard file operations, you can use the following methods:
- drive.writeFile(fid, buffer): Writes a buffer to the file.drive.readFile(fid)
- : Reads a file and returns a buffer.drive.deleteFile(fid)
- : Deletes the file.
Example:
`js`
const buffer = Buffer.from('Hello, world!')
await drive.writeFile('fileID', buffer)
const readBuffer = await drive.readFile('fileID')
console.log(readBuffer.toString())
await drive.deleteFile('fileID')
For handling large files, you can use streams:
- drive.createWriteStream(fid, opts): Creates a writable stream.drive.createReadStream(fid, opts)
- : Creates a readable stream.
Example:
`js
const ws = drive.createWriteStream('fileID')
ws.write('Hello, ')
ws.write('world!')
ws.end()
const rs = drive.createReadStream('fileID')
rs.on('data', chunk => {
console.log(chunk.toString())
})
`
- drive.commit(ops, chunks): Commits a batch of operations atomically.drive.sub(path, opts)
- : Creates a sub-storage for scoped file operations.
Builds the operations required to write a file to the storage but does not commit them.
Writes a file to the storage.
Builds the operations required to delete a file but does not commit them.
Deletes a file from the storage.
Commits a batch of operations atomically.
Reads a file from the storage.
Returns a writable stream for the specified file.
Returns a readable stream for the specified file.
Creates a sub-storage and returns a new Drive instance for it.
In order to run the test suite, simply clone the repository and install its dependencies:
`bash`
git clone https://gitlab.com/frenware/framework/plaindb/drive.git
cd drive
npm install
To run the tests:
`bash`
npm test
Thank you! Please see our contributing guidelines for details.
If you find this project useful and want to help support further development, please send us some coin. We greatly appreciate any and all contributions. Thank you!
Bitcoin (BTC):
``
1JUb1yNFH6wjGekRUW6Dfgyg4J4h6wKKdF
Monero (XMR):
```
46uV2fMZT3EWkBrGUgszJCcbqFqEvqrB4bZBJwsbx7yA8e2WBakXzJSUK8aqT4GoqERzbg4oKT2SiPeCgjzVH6VpSQ5y7KQ
@plaindb/drive is MIT licensed.