jspurefix

1. fast 100% native clean fix engine for Node JS
2. supports tls encrypted sessions over tcp
3. represent data dictionary as quickfix or repo notation
4. compile interface types against definitions
5. ascii / fixml supported
6. parses repeat groups, components and raw data fields
7. views make convinient ways of accessing data
8. complete trade capture sample to get started
9. a skeleton example shows connection, login and session only
10. parse fix logs into human readable format - JSON, tokens
11. socket session management for login, heartbeat etc
12. implement httpInitiator or acceptor

Typescript FIX Engine for Node JS

This fix engine provides a fast easy API to parse or send finacial protocol FIX messages. It is implemented entirely in typescript and runs in Node JS. Messages of any complexity can be handled providing they are backed by a suitable data dictionary. All structures within a message will be resolved for easy access - groups of components containing groups etc.

Installing

compiled typescript is now included. Install the package from npm:

  npm install jspurefix
  cd node_modules/jspurefix && npm run unzip-repo

see our demo application

  https://github.com/TimelordUK/jspf-demo

if you wish to generate custom messages this is easily doing using quick fix XML format - see quickfix demo

edit the dictionary file data/generate and run the generator.

  npm run generate

check the messages generated in src\types

  https://github.com/TimelordUK/jspf-md-demo

import types for usage with a client

import {
  AsciiSession,
  MsgView,
  IJsFixConfig,
  IJsFixLogger,
  Dictionary,
  MsgType,
  IJsFixConfig,
  initiator,
  acceptor,
  makeConfig
} from 'jspurefix'
// types for the specific FIX dialect
import {
  ITradeCaptureReport,
  ITradeCaptureReportRequest,
  ITradeCaptureReportRequestAck
} from 'jspurefix/dist/types/FIX4.4/repo'

see example trade-capture-client - use

getting started

clone from git

unix

  git clone https://github.com/TimelordUK/jspurefix.git
  cd jspurefix
  script/build.sh

  npm install
  npm run unzip-repo
  ./node_modules/.bin/tsc --version
  ./node_modules/.bin/tsc
  npm run tcp-tc

windows

  git clone https://github.com/TimelordUK/jspurefix.git
  cd jspurefix
  script\build.cmd

or

  git clone https://github.com/TimelordUK/jspurefix.git
  cd jspurefix
  npm install
  npm run unzip-repo
  node_modules/.bin/tsc
  npm run tcp-tc

Run Sample

The code provided in src/sample/tcp/trade_capture is a good place to start in building an application with this library. In this case both client and server are run together communicating over a socket. In reality a client is more likely connecting to an external acceptor such as CME, ICE.

There is also a skeleton application which shows all application code stripped away to just enough to create a connection, login and see the session.

Both examples will automatically close down after ~1 minute or can be terminated CTRL-C

ensure the repo dictionary has been unpacked above. This provides the data dictionary used to parse messages.

run examples such as the trade capture, simple skeleton or http.

npm run tcp-tc

npm run tcp-sk

included is an example fixml over http application where an order is submitted and execution report returned.

npm run http-oms

tls SSL encryption

see example data\session\test-initiator-tls.json

to run the provided example tls trade capture a script such as script\getKey.ps1 can be run to generate self certified certificates. This is a powershell script that requires openssl in the tunnel e.g. /mingw64/bin/openssl. the ca field below is only required when using self certified and will not be needed for a third party vendor where certificates are provided. Set enableTrace flag whilst diagnosing the session.

{
  "application": {
    "reconnectSeconds": 10,
    "type": "initiator",
    "name": "test_client",
    "tcp": {
      "host" : "localhost",
      "port": 2344,
      "tls": {
        "timeout": 10000,
        "sessionTimeout": 10000,
        "enableTrace": true,
        "key": "data/session/certs/client/client.key",
        "cert": "data/session/certs/client/client.crt",
        "ca": [
          "data/session/certs/ca/ca.crt"
        ]
      }
    },
    "protocol": "ascii",
    "dictionary": "repo44"
  },
  "Username": "js-tls-client",
  "Password": "pwd-tls-client",
  "EncryptMethod": 0,
  "ResetSeqNumFlag": true,
  "LastSentSeqNum": 10,
  "LastReceivedSeqNum": 11,
  "HeartBtInt": 30,
  "SenderCompId": "init-tls-comp",
  "TargetCompID": "accept-tls-comp",
  "TargetSubID": "fix",
  "BeginString": "FIX4.4",
  "BodyLengthChars": 6
}

configure the body length field padding width

see example initiator file data\session\test-initiator-tls.json.

i.e. include field BodyLengthChars which defaults to 7 characters if omitted.

{
  "BodsyLengthChars": 6
}

launching without resetting message sequences

If the message sequences are persisted over multiple sessions and are not reset on logon (ie. "ResetSeqNumFlag": false,), then the previously used sequence numbers can be set as follows:

{
  "ResetSeqNumFlag": false,
  "LastSentSeqNum": 10,
  "LastReceivedSeqNum": 11
}

resending messages

By default, the library will not resend past messages as this requires persisting messages which depending on the volume, may also require a database. If you want to support resending you must override AsciiSession.onResendRequest() with a resending logic. Additionally, make sure to include the original message sequence and the duplicate message flag in the FIX object:

{
...messageBodyData,
    StandardHeader: { PossDupFlag: true, MsgSeqNum: sequenceNumber},
}

Example

{
  "ClOrdID": "acceptor-order-id",
  "HandlInst": "2",
  "OrdType": "2",
  "Side": "2",
  "TransactTime": "2021-08-03T08:23:57.041Z",
  "Symbol": "some ticker",
  "StandardHeader": {
    "PossDupFlag": true,
    "MsgSeqNum": 2
  }
}

Unit Tests

there is a comprehensive suite of tests which can be run

npm t

PASS  src/test/elastic-buffer.test.ts
RUNS  src/test/session.test.ts
PASS  src/test/encode-proxy.test.tsst.ts
PASS  src/test/execution-report.test.ts
PASS  src/test/view-decode.test.ts
PASS  src/test/ascii-encoder.test.ts
PASS  src/test/ascii-parser.test.ts
PASS  src/test/includes.test.ts
PASS  src/test/fixml-alloc-parse.test.ts (9.433s)
PASS  src/test/repo-full-fixml-msg.test.ts (6.025s)
PASS  src/test/fixml-mkt-data-settle-parse.test.ts (6.021s)
PASS  src/test/qf-full-msg.test.ts
PASS  src/test/logon.test.ts
PASS  src/test/fixml-mkt-data-fut-parse.test.ts (7.761s)
PASS  src/test/time-formatter.test.ts
PASS  src/test/ascii-segment.test.ts
PASS  src/test/session-state.test.ts
PASS  src/test/fixml-tc-bi-lateral-parse.test.ts (7.534s)
PASS  src/test/ascii-tag-pos.test.ts
PASS  src/test/fix-log-replay.test.ts
PASS  src/test/repo-full-ascii-msg.test.ts
PASS  src/test/session.test.ts (52.637s)

Test Suites: 21 passed, 21 total
Tests:       204 passed, 204 total
Snapshots:   0 total
Time:        54.606s, estimated 65s
Ran all test suites.

Dictionary Definitions

base definitions on existing template e.g. quickfix format FIX44.xml create an alias in data/dictionary.json compile interfaces

npm run cmd -- --dict=repo42 --compile

use the alias in a session file e.g. data/session/test-httpInitiator.json

sample trade-capture-client.ts

the method onApplicationMsg is called when a message is received. In this case the client has inherited from AsciiSession which carries out the session management.

  constructor (public readonly config: IJsFixConfig) {
  super(config)
  this.logReceivedMsgs = true
  this.reports = new Dictionary<ITradeCaptureReport>()
  this.fixLog = config.logFactory.plain(`jsfix.${config.description.application.name}.txt`)
  this.logger = config.logFactory.logger(`${this.me}:TradeCaptureClient`)
}

protected onApplicationMsg (msgType: string, view: MsgView): void {
  this.logger.info(`${view.toJson()}`)
  switch (msgType) {
  case MsgType.TradeCaptureReport: {
      // create an object and cast to the interface
      const tc: ITradeCaptureReport = view.toObject()
      this.reports.addUpdate(tc.TradeReportID, tc)
      this.logger.info(`[reports: ${this.reports.count()}] received tc ExecID = ${tc.ExecID} TradeReportID  = ${tc.TradeReportID} Symbol = ${tc.Instrument.Symbol} ${tc.LastQty} @ ${tc.LastPx}`)
      break
    }

  case MsgType.TradeCaptureReportRequestAck: {
      const tc: ITradeCaptureReportRequestAck = view.toObject()
      this.logger.info(`received tcr ack ${tc.TradeRequestID} ${tc.TradeRequestStatus}`)
      break
    }
  }
}

the client onReady method is called when a connection is made and logon established and confirmed. In this case, client sends a trade capture request to the server.

  protected onReady (view: MsgView): void {
  this.logger.info('ready')
  const tcr: ITradeCaptureReportRequest = TradeFactory.tradeCaptureReportRequest('all-trades', new Date())
// send request to server
this.send(MsgType.TradeCaptureReportRequest, tcr)
const logoutSeconds = 32
this.logger.info(`will logout after ${logoutSeconds}`)
setTimeout(() => {
  this.done()
}, logoutSeconds * 1000)
}

working with Views

see src/test/view-decode.test.ts

note that a view can only be used within a callback context unless it is cloned. Once returned, the memory is re-used for next message. It is intended to convert to an object or parsed into an application specific message.

fetch a group view

  const noMDEntriesView: MsgView = view.getView('NoMDEntries')
const mmEntryView: MsgView = noMDEntriesView.getGroupInstance(1)

const mmEntryExpireTimeAsString: string = mmEntryView.getString('ExpireTime')
expect(mmEntryExpireTimeAsString).toEqual('20180608-20:53:14.000')
expect(mmEntryView.getString(126)).toEqual('20180608-20:53:14.000')

fetch single tags

  const erView: MsgView = views[0]
expect(erView.getString(35)).toEqual('8')
expect(erView.getString('MsgType')).toEqual('8')
expect(erView.getString(8)).toEqual('FIX4.4')
expect(erView.getTyped(9)).toEqual(6542)
expect(erView.getTyped('TotNumReports')).toEqual(19404)
expect(erView.getTyped('StrikePrice')).toEqual(52639)

fetch repeated tag

  const erView: MsgView = views[0]
expect(erView.getStrings('PartyID')).toEqual(['magna.', 'iaculis', 'vitae,'])

fetch a set of tags

  const [a, b, c, d] = view.getTypedTags([8, 9, 35, 49])
expect(a).toEqual('FIX4.4')
expect(b).toEqual(2955)
expect(c).toEqual('W')
expect(d).toEqual('sender-10')

convert view into an object which can be used alongside an interface for intellisense.

  import { ITradeCaptureReport } from '../../../types/FIX4.4/repo/trade_capture_report'
import { ITradeCaptureReportRequest } from '../../../types/FIX4.4/repo/trade_capture_report_request'

const tc: ITradeCaptureReport = view.toObject()

from data/examples/FIX.4.4/quickfix/execution-report

get first in group fetched from object where group is array

  import { IUndInstrmtGrp } from '../types/FIX4.4/quickfix/set/und_instrmt_grp'
import { IUnderlyingInstrument } from '../types/FIX4.4/quickfix/set/underlying_instrument'
const erView: MsgView = views[0]
const undInstrmtGrpView: MsgView = erView.getView('UndInstrmtGrp')
const undInstrmtGrpViewAsObject: IUndInstrmtGrp = undInstrmtGrpView.toObject()
expect(undInstrmtGrpViewAsObject.NoUnderlyings.length).toEqual(2)
const underlying0: IUnderlyingInstrument = undInstrmtGrpViewAsObject.NoUnderlyings[0].UnderlyingInstrument
expect(underlying0.UnderlyingSymbol).toEqual('massa.')

fetch nested structure in one call

  const legGrpView = view.getView('InstrmtLegGrp.NoLegs')
expect(legGrpView).toBeTruthy()
const legGrp: IInstrumentLeg[] = legGrpView.toObject()
expect(legGrp).toBeTruthy()
expect(Array.isArray(legGrp))
expect(legGrp.length).toEqual(2)

get a tokenised view of tags in view

console.log(view.toString())

get a component from parent - this is very low cost

 const instrumentView: MsgView = view.getView('Instrument')
const instrumentObject: IInstrument = view.getView('Instrument').toObject()

FIXML

Please see sample code src/sample/http for an example of how fixml can be used. Note regardless of using AsciiChars or Fixml application code looks very similar for example the client in this case.

  protected onReady (view: MsgView): void {
  this.logger.info('onReady')
  const logoutSeconds = this.logoutSeconds
  const req = this.factory.createOrder('IBM', Side.Buy, 10000, 100.12)
  this.send('NewOrderSingle', req)
  this.logger.info(`will logout after ${logoutSeconds}`)
  setTimeout(() => {
  this.done()
}, 11 * 1000)
}

public createOrder (symbol: string, side: Side, qty: number, price: number): INewOrderSingle {
  const id: number = this.id++
  return {
    ClOrdID: `Cli${id}`,
    Account: this.account,
    Side: side,
    Price: price,
    OrdType: OrdType.Limit,
    OrderQtyData: {
      OrderQty: qty
    } as IOrderQtyData,
    Instrument: {
      Symbol: symbol,
      SecurityID: '459200101',
      SecurityIDSource: SecurityIDSource.IsinNumber
    } as IInstrument,
    TimeInForce: TimeInForce.GoodTillCancelGtc
  } as INewOrderSingle
}

this renders to this message sent over http

<FIXML>
    <Order ID="Cli1" Acct="TradersRUs" Side="1" Typ="2" Px="100.12" TmInForce="1">
        <Hdr SID="accept-comp" TID="init-comp" SSub="user123" TSub="INC"/>
        <Instrmt Sym="IBM" ID="459200101" Src="4"/>
        <OrdQty Qty="10000"/>
    </Order>
</FIXML>

the server receives this message and sends back an execution report :-

  protected onApplicationMsg (msgType: string, view: MsgView): void {
  // dispatch messages
  this.logger.info(view.toJson())
  switch (msgType) {
  case 'Order': {
      const order: INewOrderSingle = view.toObject()
      this.logger.info(`received order id ${order.ClOrdID}`)
      const fill: IExecutionReport = this.factory.fillOrder(order)
      this.send('ExecutionReport', fill)
    }
  }
}

public fillOrder (order: INewOrderSingle): IExecutionReport {
  const id: number = this.execId++
  return {
    ClOrdID: order.ClOrdID,
    OrdType: order.OrdType,
    TransactTime: new Date(),
    AvgPx: order.Price,
    LeavesQty: 0,
    LastPx: order.Price,
    ExecType: ExecType.OrderStatus,
    OrdStatus: OrdStatus.Filled,
    ExecID: `exec${id}`,
    Side: order.Side,
    Price: order.Price,
    OrderQtyData: {
      OrderQty: order.OrderQtyData.OrderQty
    } as IOrderQtyData,
    Instrument: {
      Symbol: order.Instrument.Symbol,
      SecurityID: order.Instrument.SecurityID,
      SecurityIDSource: SecurityIDSource.IsinNumber
    } as IInstrument
  } as IExecutionReport
}

the fixml is sent back to the client :-

<FIXML>
    <ExecRpt ID="Cli1" ExecID="exec1" ExecTyp="I" Stat="2" Side="1" Typ="2" Px="100.12" LastPx="100.12" LeavesQty="0" AvgPx="100.12" TxnTm="2018-10-07T12:16:12.584">
        <Hdr SID="accept-comp" TID="init-comp" TSub="fix"/>
        <Instrmt Sym="IBM" ID="459200101" Src="4"/>
        <OrdQty Qty="10000"/>
    </ExecRpt>
</FIXML>

Performance

These messages have been randomly generated with command line tool. They are syntactically valid.

data/examples/FIX.4.4/repo/execution-report/fix.txt

npm run repo44-bench-er

performance on Windows Intel Core I7-4770 @ 3.5 GHz

[8]: repeats = 250000, fields = 58, length = 604 chars, elapsed ms 3658, 14.632 micros per msg
[8]: iterations = 80000, fields = 646, length = 6572 chars, elapsed ms 16499, 206.23749999999998 micros per msg

performance on Windows 12th Gen Intel(R) Core(TM) i7-12700H 2.30 GHz

[8]: iterations = 80000, fields = 646, length = 6572 chars, elapsed ms 7476, 93.45 micros per msg

data/examples/FIX.4.4/repo/security-definition/fix.txt

npm run repo44-bench-sd

performance on Windows Intel Core I7-4770 @ 3.5 GHz

[d]: repeats = 150000, fields = 223, length = 2233 chars, elapsed ms 7962, 53.080000000000005 micros per msg
d]: iterations = 150000, fields = 229, length = 2466 chars, elapsed ms 8672, 57.81333333333333 micros per msg

performance on Windows 12th Gen Intel(R) Core(TM) i7-12700H 2.30 GHz

[d]: iterations = 150000, fields = 229, length = 2466 chars, elapsed ms 4628, 30.85333333333333 micros per msg

data/examples/FIX.4.4/repo/trade-capture/fix.txt

npm run repo44-bench-tc

performance on Windows Intel Core I7-4770 @ 3.5 GHz

[AE]: repeats = 30000, fields = 613, length = 5818 chars, elapsed ms 5206, 173.53333333333333 micros per msg
[AE]: iterations = 30000, fields = 578, length = 5741 chars, elapsed ms 5245, 174.83333333333334 micros per msg```

performance on Windows 12th Gen Intel(R) Core(TM) i7-12700H 2.30 GHz

[AE]: iterations = 30000, fields = 578, length = 5741 chars, elapsed ms 2725, 90.83333333333333 micros per msg```

data/examples/FIX.4.4/quickfix/heartbeat/fix.txt

npm run qf-bench-hb

performance on Windows Intel Core I7-4770 @ 3.5 GHz

[0]: iterations = 250000, fields = 10, length = 131 chars, elapsed ms 950, 3.8 micros per msg

performance on Windows 12th Gen Intel(R) Core(TM) i7-12700H 2.30 GHz

[0]: iterations = 250000, fields = 10, length = 131 chars, elapsed ms 468, 1.8719999999999999 micros per msg

data/examples/FIX.4.4/quickfix/logon/fix.txt

npm run qf-bench-lo

performance on Windows Intel Core I7-4770 @ 3.5 GHz

[A]: iterations = 250000, fields = 22, length = 214 chars, elapsed ms 1466, 5.864 micros per msg

performance on Windows 12th Gen Intel(R) Core(TM) i7-12700H 2.30 GHz

[A]: iterations = 250000, fields = 22, length = 214 chars, elapsed ms 693, 2.7720000000000002 micros per msg

Log parsing

the command line tool jsfix can be used to parse any fix log providing an appropriate dictionary is provided.

parsing fields

npm run cmd -- --dict=repo44 --fix=data/examples/FIX.4.4/jsfix.test_client.txt --delimiter="|" --type=AD --tokens

[0] 8 (BeginString) = FIX4.4, [1] 9 (BodyLength) = 0000135
[2] 35 (MsgType) = AD[TradeCaptureReportRequest], [3] 49 (SenderCompID) = init-comp
[4] 56 (TargetCompID) = accept-comp, [5] 34 (MsgSeqNum) = 2
[6] 57 (TargetSubID) = fix, [7] 52 (SendingTime) = 20180923-16:07:04.763
[8] 568 (TradeRequestID) = all-trades, [9] 569 (TradeRequestType) = 0[AllTrades]
[10] 263 (SubscriptionRequestType) = 1[SnapshotAndUpdates], [11] 580 (NoDates) = 1
[12] 75 (TradeDate) = 20180923, [13] 10 (CheckSum) = 250

stats on entire file

npm run cmd -- --dict=repo44 --fix=data/examples/FIX.4.4/jsfix.test_client.txt --delimiter="|" --stats

messages 13 elapsed ms 8
{
  "0": 1,
  "5": 2,
  "A": 2,
  "AD": 1,
  "AQ": 2,
  "AE": 5
}

benchmark parsing repeated reads of file

npm run cmd -- --dict=repo44 --fix=data/examples/FIX.4.4/jsfix.test_client.txt --delimiter="|" --stats --repeats=20

messages 13 elapsed ms 0
{
  "0": 1,
  "5": 2,
  "A": 2,
  "AD": 1,
  "AQ": 2,
  "AE": 5
}

parse message type in a file

npm run cmd -- --dict=repo44 --fix=data/examples/FIX.4.4/jsfix.test_client.txt --delimiter="|" --type=AD --objects

{
  "StandardHeader": {
    "BeginString": "FIX4.4",
    "BodyLength": 135,
    "MsgType": "AD",
    "SenderCompID": "init-comp",
    "TargetCompID": "accept-comp",
    "MsgSeqNum": 2,
    "TargetSubID": "fix",
    "SendingTime": "2018-09-23T16:07:04.763Z"
  },
  "TradeRequestID": "all-trades",
  "TradeRequestType": 0,
  "SubscriptionRequestType": "1",
  "TrdCapDtGrp": [
    {
      "TradeDate": "2018-09-22T23:00:00.000Z"
    }
  ],
  "StandardTrailer": {
    "CheckSum": "250"
  }
}

as above where --type=AE --objects

{
  "StandardHeader": {
    "BeginString": "FIX4.4",
    "BodyLength": 213,
    "MsgType": "AE",
    "SenderCompID": "accept-comp",
    "TargetCompID": "init-comp",
    "MsgSeqNum": 4,
    "TargetSubID": "fix",
    "SendingTime": "2018-09-23T16:07:04.986Z"
  },
  "TradeReportID": "100001",
  "TradeReportTransType": 0,
  "TradeReportType": 0,
  "TrdType": 0,
  "ExecID": "600001",
  "OrdStatus": "2",
  "PreviouslyReported": false,
  "Instrument": {
    "Symbol": "Gold",
    "SecurityID": "Gold.INC"
  },
  "LastQty": 107,
  "LastPx": 45.38,
  "TradeDate": "2018-09-22T23:00:00.000Z",
  "TransactTime": "2018-09-23T16:07:04.776Z",
  "StandardTrailer": {
    "CheckSum": "54"
  }
}

show all structures within a message

npm run cmd -- --dict=repo44 --fix=data/examples/FIX.4.4/jsfix.test_client.txt --delimiter="|" --type=AD --structures

[
  {
    "name": "StandardHeader",
    "depth": 2,
    "startTag": 8,
    "startPosition": 0,
    "endTag": 52,
    "endPosition": 7,
    "delimiterTag": 0,
    "delimiterPositions": []
  },
  {
    "name": "TrdCapDtGrp",
    "depth": 1,
    "startTag": 580,
    "startPosition": 11,
    "endTag": 75,
    "endPosition": 12,
    "delimiterTag": 75,
    "delimiterPositions": [
      12
    ]
  },
  {
    "name": "StandardTrailer",
    "depth": 1,
    "startTag": 10,
    "startPosition": 13,
    "endTag": 10,
    "endPosition": 13,
    "delimiterTag": 0,
    "delimiterPositions": []
  },
  {
    "name": "TradeCaptureReportRequest",
    "depth": 1,
    "startTag": 8,
    "startPosition": 0,
    "endTag": 10,
    "endPosition": 13,
    "delimiterTag": 0,
    "delimiterPositions": []
  },
  {
    "name": "StandardTrailer",
    "depth": 0,
    "startTag": 10,
    "startPosition": 14,
    "endTag": 10,
    "endPosition": 13,
    "delimiterTag": 0,
    "delimiterPositions": []
  }
]