minipass-json-stream

Like JSONStream, but using Minipass streams

install

npm install minipass-json-stream

example

1
2const request = require('request')
3const JSONStream = require('minipass-json-stream')
4const es = require('event-stream')
5
6request({url: 'http://isaacs.couchone.com/registry/_all_docs'})
7  .pipe(JSONStream.parse('rows.*'))
8  .pipe(es.mapSync(function (data) {
9    console.error(data)
10    return data
11  }))

new JSONStream(options)

Create a new stream. This is a minipass stream that is always set in objectMode. It emits objects parsed out of string/buffer JSON input that match the supplied path option.

JSONStream.parse(path)

Return a new JSONStream object to stream values that match a path.

(Equivalent to new JSONStream({path}).)

1JSONStream.parse('rows.*.doc')

The .. operator is the recursive descent operator from JSONPath, which will match a child at any depth (see examples below).

If your keys have keys that include . or * etc, use an array instead. ['row', true, /^doc/].

If you use an array, RegExps, booleans, and/or functions. The .. operator is also available in array representation, using {recurse: true}. any object that matches the path will be emitted as 'data' (and piped down stream)

If path is empty or null, no 'data' events are emitted.

If you want to have keys emitted, you can prefix your * operator with $: obj.$* - in this case the data passed to the stream is an object with a key holding the key and a value property holding the data.

Examples

query a couchdb view:

1curl -sS localhost:5984/tests/_all_docs&include_docs=true

you will get something like this:

1{"total_rows":129,"offset":0,"rows":[
2  { "id":"change1_0.6995461115147918"
3  , "key":"change1_0.6995461115147918"
4  , "value":{"rev":"1-e240bae28c7bb3667f02760f6398d508"}
5  , "doc":{
6      "_id":  "change1_0.6995461115147918"
7    , "_rev": "1-e240bae28c7bb3667f02760f6398d508","hello":1}
8  },
9  { "id":"change2_0.6995461115147918"
10  , "key":"change2_0.6995461115147918"
11  , "value":{"rev":"1-13677d36b98c0c075145bb8975105153"}
12  , "doc":{
13      "_id":"change2_0.6995461115147918"
14    , "_rev":"1-13677d36b98c0c075145bb8975105153"
15    , "hello":2
16    }
17  },
18]}

we are probably most interested in the rows.*.doc

create a JSONStream that parses the documents from the feed like this:

1var stream = JSONStream.parse(['rows', true, 'doc']) //rows, ANYTHING, doc
2
3stream.on('data', function(data) {
4  console.log('received:', data);
5});
6
7//emits anything from _before_ the first match
8stream.on('header', function (data) {
9  console.log('header:', data) // => {"total_rows":129,"offset":0}
10})

awesome!

In case you wanted the contents the doc emitted:

1// equivalent to: 'rows.*.doc.$*'
2var stream = JSONStream.parse([
3  'rows',
4  true,
5  'doc',
6  {emitKey: true}
7]) //rows, ANYTHING, doc, items in docs with keys
8
9stream.on('data', function(data) {
10  console.log('key:', data.key);
11  console.log('value:', data.value);
12});

You can also emit the path:

1var stream = JSONStream.parse([
2  'rows',
3  true,
4  'doc',
5  {emitPath: true}
6]) //rows, ANYTHING, doc, items in docs with keys
7
8stream.on('data', function(data) {
9  console.log('path:', data.path);
10  console.log('value:', data.value);
11});

recursive patterns (..)

JSONStream.parse('docs..value') (or JSONStream.parse(['docs', {recurse: true}, 'value']) using an array) will emit every value object that is a child, grand-child, etc. of the docs object. In this example, it will match exactly 5 times at various depth levels, emitting 0, 1, 2, 3 and 4 as results.

1{
2  "total": 5,
3  "docs": [
4    {
5      "key": {
6        "value": 0,
7        "some": "property"
8      }
9    },
10    {"value": 1},
11    {"value": 2},
12    {"blbl": [{}, {"a":0, "b":1, "value":3}, 10]},
13    {"value": 4}
14  ]
15}

JSONStream.parse(pattern, map)

(Equivalent to new JSONStream({ pattern, map }))

provide a function that can be used to map or filter the json output. map is passed the value at that node of the pattern, if map return non-nullish (anything but null or undefined) that value will be emitted in the stream. If it returns a nullish value, nothing will be emitted.

JSONStream also emits 'header' and 'footer' events, the 'header' event contains anything in the output that was before the first match, and the 'footer', is anything after the last match.

Acknowlegements

This module is a fork of JSONStream by Dominic Tarr, modified and redistributed under the terms of the MIT license.

this module depends on https://github.com/creationix/jsonparse by Tim Caswell