Read and write to an abstract-level database using Web Streams. Compatible with browsers and Node.js.
📌 To instead consume data using Node.js streams, see
level-read-stream. On Node.js 16,level-read-streamis ~3 times faster thanlevel-web-streamand the performance of Web Streams isn't likely to improve anytime soon. On the other hand, Web Streams are a step towards a standard library for JavaScript (across Node.js, Deno and browsers).
const { EntryStream } = require('level-web-stream')
const { MemoryLevel } = require('memory-level')
const db = new MemoryLevel()
// Write sample data
await db.put('a', '1')
await db.put('b', '2')
await db.put('c', '3')
// Create a ReadableStream
const src = new EntryStream(db, {
gte: 'b'
})
// Pipe to a stream of choice
const dst = new WritableStream({
write ([key, value]) {
console.log('%s: %s', key, value)
}
})
await src.pipeTo(dst)Yields:
b: 2
c: 3
To only read keys or values rather than entries:
const { KeyStream, ValueStream } = require('level-web-stream')
await new KeyStream(db).pipeTo(new WritableStream({
write (key) {
console.log(key)
}
}))Note that WritableStream is a global in browsers. In Node.js it can be imported from stream/web.
const { EntryStream, BatchStream } = require('level-web-stream')
// Copy entries from db to db
const src = new EntryStream(db1)
const dst = new BatchStream(db2, { type: 'put' })
await src.pipeTo(dst)With npm do:
npm install level-web-stream
Create a ReadableStream that will yield entries. An entry is an array with two elements: a key and value. The db argument must be an abstract-level database. The optional options object may contain:
highWaterMark(number): the maximum number of entries to buffer internally before ceasing to read further entries. Default 1000.
Any other options are forwarded to db.iterator(options). The stream wraps that iterator. If you prefer to consume entries with for await...of then it's recommended to directly use db.iterator(). In either case, most databases will read from a snapshot (thus unaffected by simultaneous writes) as indicated by db.supports.snapshots.
Same as EntryStream but yields keys instead of entries, using db.keys() instead of db.iterator(). If only keys are needed, using KeyStream may increase performance because values won't have to be fetched.
Same as EntryStream but yields values instead of entries, using db.values() instead of db.iterator(). If only values are needed, using ValueStream may increase performance because keys won't have to be fetched.
Create a WritableStream that takes operations or entries, to be written to the database in batches of fixed size using db.batch(). If a batch fails the stream will be aborted but previous batches that did succeed will not be reverted.
An operation is an object containing:
- A
keyproperty (required) - A
typeproperty (optional, one of'put','del') - A
valueproperty (required if type is'put', ignored if type is'del') - Other operation properties accepted by
db.batch().
An entry is an array with two elements: a key and value. This allows piping a readable EntryStream into a writable BatchStream. Writing [key, value] is the same as writing { key, value }.
The db argument must be an abstract-level database. The optional options object may contain:
highWaterMark(number, default 500): the maximum number of operations to buffer internally before committing them to the database withdb.batch(). No data will be committed untilhighWaterMarkis reached or until the stream is closing. IncreasinghighWaterMarkcan improve throughput at the cost of memory usage and at risk of blocking the JavaScript event loop whiledb.batch()is copying data.type(string, default'put'): default operationtypeif not set on individual operations or entries (which can't set it). Must be'put'or'del'.
Any other options are forwarded to the options argument of db.batch(operations, options) thus following the same rules of precedence (of options versus operation properties).
Level/web-stream is an OPEN Open Source Project. This means that:
Individuals making significant and valuable contributions are given commit-access to the project to contribute as they see fit. This project is more like an open wiki than a standard guarded open source project.
See the Contribution Guide for more details.