Subtheme of barrio

8.9 KiB

Raw Permalink Blame History

EventStream

[] (http://ci.testling.com/dominictarr/event-stream)

Streams are node's best and most misunderstood idea, and EventStream is a toolkit to make creating and working with streams easy.

Normally, streams are only used for IO,
but in event stream we send all kinds of objects down the pipe.
If your application's input and output are streams,
shouldn't the throughput be a stream too?

The EventStream functions resemble the array functions,
because Streams are like Arrays, but laid out in time, rather than in memory.

All the event-stream functions return instances of Stream.

event-stream creates 0.8 streams , which are compatible with 0.10 streams.

NOTE: I shall use the term "through stream" to refer to a stream that is writable and readable.

simple example:


//pretty.js

if(!module.parent) {
  var es = require('event-stream')
  var inspect = require('util').inspect

  process.stdin                        //connect streams together with `pipe`
    .pipe(es.split())                  //split stream to break on newlines
    .pipe(es.map(function (data, cb) { //turn this async function into a stream
      cb(null
        , inspect(JSON.parse(data)))   //render it nicely
    }))
    .pipe(process.stdout)              // pipe it to stdout !
}

run it ...

curl -sS registry.npmjs.org/event-stream | node pretty.js

node Stream documentation

through (write?, end?)

Re-emits data synchronously. Easy way to create synchronous through streams. Pass in optional write and end methods. They will be called in the context of the stream. Use this.pause() and this.resume() to manage flow. Check this.paused to see current flow state. (write always returns !this.paused)

this function is the basis for most of the synchronous streams in event-stream.


es.through(function write(data) {
    this.emit('data', data)
    //this.pause() 
  },
  function end () { //optional
    this.emit('end')
  })

map (asyncFunction)

Create a through stream from an asynchronous function.

var es = require('event-stream')

es.map(function (data, callback) {
  //transform data
  // ...
  callback(null, data)
})

Each map MUST call the callback. It may callback with data, with an error or with no arguments,

callback() drop this data.
this makes the map work like filter,
note:callback(null,null) is not the same, and will emit null
callback(null, newData) turn data into newData
callback(error) emit an error for this item.

Note: if a callback is not called, map will think that it is still being processed,
every call must be answered or the stream will not know when to end.

Also, if the callback is called more than once, every call but the first will be ignored.

mapSync (syncFunction)

Same as map, but the callback is called synchronously. Based on es.through

split (matcher)

Break up a stream and reassemble it so that each line is a chunk. matcher may be a String, or a RegExp

Example, read every line in a file ...

fs.createReadStream(file, {flags: 'r'})
  .pipe(es.split())
  .pipe(es.map(function (line, cb) {
    //do something with the line 
    cb(null, line)
  }))

split takes the same arguments as string.split except it defaults to '\n' instead of ',', and the optional limit parameter is ignored. String#split

join (separator)

Create a through stream that emits separator between each chunk, just like Array#join.

(for legacy reasons, if you pass a callback instead of a string, join is a synonym for es.wait)

merge (stream1,...,streamN) or merge (streamArray)

concat → merge

Merges streams into one and returns it. Incoming data will be emitted as soon it comes into - no ordering will be applied (for example: data1 data1 data2 data1 data2 - where data1 and data2 is data from two streams). Counts how many streams were passed to it and emits end only when all streams emitted end.

es.merge(
  process.stdout,
  process.stderr
).pipe(fs.createWriteStream('output.log'));

It can also take an Array of streams as input like this:

es.merge([
  fs.createReadStream('input1.txt'),
  fs.createReadStream('input2.txt')
]).pipe(fs.createWriteStream('output.log'));

replace (from, to)

Replace all occurrences of from with to. from may be a String or a RegExp.
Works just like string.split(from).join(to), but streaming.

parse

Convenience function for parsing JSON chunks. For newline separated JSON, use with es.split. By default it logs parsing errors by console.error; for another behaviour, transforms created by es.parse({error: true}) will emit error events for exceptions thrown from JSON.parse, unmodified.

fs.createReadStream(filename)
  .pipe(es.split()) //defaults to lines.
  .pipe(es.parse())

stringify

convert javascript objects into lines of text. The text will have whitespace escaped and have a \n appended, so it will be compatible with es.parse

objectStream
  .pipe(es.stringify())
  .pipe(fs.createWriteStream(filename))

readable (asyncFunction)

create a readable stream (that respects pause) from an async function.
while the stream is not paused,
the function will be polled with (count, callback),
and this will be the readable stream.


es.readable(function (count, callback) {
  if(streamHasEnded)
    return this.emit('end')
  
  //...
  
  this.emit('data', data) //use this way to emit multiple chunks per call.
      
  callback() // you MUST always call the callback eventually.
             // the function will not be called again until you do this.
})

you can also pass the data and the error to the callback.
you may only call the callback once.
calling the same callback more than once will have no effect.

readArray (array)

Create a readable stream from an Array.

Just emit each item as a data event, respecting pause and resume.

  var es = require('event-stream')
    , reader = es.readArray([1,2,3])

  reader.pipe(...)

If you want the stream behave like a 0.10 stream you will need to wrap it using Readable.wrap() function. Example:

	var s = new stream.Readable({objectMode: true}).wrap(es.readArray([1,2,3]));

writeArray (callback)

create a writeable stream from a callback,
all data events are stored in an array, which is passed to the callback when the stream ends.

  var es = require('event-stream')
    , reader = es.readArray([1, 2, 3])
    , writer = es.writeArray(function (err, array){
      //array deepEqual [1, 2, 3]
    })

  reader.pipe(writer)

pause ()

A stream that buffers all chunks when paused.

  var ps = es.pause()
  ps.pause() //buffer the stream, also do not allow 'end' 
  ps.resume() //allow chunks through

duplex (writeStream, readStream)

Takes a writable stream and a readable stream and makes them appear as a readable writable stream.

It is assumed that the two streams are connected to each other in some way.

(This is used by pipeline and child.)

  var grep = cp.exec('grep Stream')

  es.duplex(grep.stdin, grep.stdout)

child (child_process)

Create a through stream from a child process ...

  var cp = require('child_process')

  es.child(cp.exec('grep Stream')) // a through stream

wait (callback)

waits for stream to emit 'end'. joins chunks of a stream into a single string or buffer. takes an optional callback, which will be passed the complete string/buffer when it receives the 'end' event.

also, emits a single 'data' event.


readStream.pipe(es.wait(function (err, body) {
  // have complete text here.
}))

Other Stream Modules

These modules are not included as a part of EventStream but may be useful when working with streams.

reduce (syncFunction, initial)

Like Array.prototype.reduce but for streams. Given a sync reduce function and an initial value it will return a through stream that emits a single data event with the reduced value once the input stream ends.

var reduce = require("stream-reduce");
process.stdin.pipe(reduce(function(acc, data) {
  return acc + data.length;
}, 0)).on("data", function(length) {
  console.log("stdin size:", length);
});

8.9 KiB Raw Permalink Blame History