morganstanley / pyfixmsg Goto Github PK

View Code? Open in Web Editor NEW

54.0 17.0 28.0 64 KB

pyfixmsg is a library for parsing, manipulating and serialising FIX messages, primarily geared towards testing.

License: Apache License 2.0

Python 100.00%

fix quickfix codec

pyfixmsg's People

Contributors

Stargazers

Watchers

pyfixmsg's Issues

If component tag is located inside group, parsing of raw message works incorrect

The spec has the following message:

<message name="QuoteRequest" msgtype="R" msgcat="app">
      <field name="QuoteReqID" required="Y" />
      <group name="NoRelatedSym" required="Y">
          <component name="Instrument" required="Y" />
          <field name="QuoteRequestType" required="N" />
          <field name="OrderQty" required="Y" />
      </group>
    </message>

and the component looks like:

<component name="Instrument">
      <field name="Symbol" required="Y" />
      <field name="Product" required="Y" />
    </component>

when trying to parse the raw message, I see this

>>> messages[2].strip()
b'8=FIX.4.3\x019=151\x0135=R\x0134=3\x0149=9222237_TEST\x0150=FX\x0152=20200616-18:25:26\x0156=SAXOQ
UOTE_TEST\x01131=1\x01146=3\x0155=AUD/CA\x01460=4\x01303=2\x0155=AUD/CHF\x01460=4\x01303=2\x0155=EUR
/ZAR\x01460=4\x01303=2\x0110=193\x01'
>>>
>>> codec.parse(messages[2].strip(), separator=chr(1))
{8: 'FIX.4.3', 9: '151', 35: 'R', 34: '3', 49: '9222237_TEST', 50: 'FX', 52: '20200616-18:25:26', 56
: 'SAXOQUOTE_TEST', 131: '1', 146: [{55: 'AUD/CA'}], 460: '4', 303: '2', 55: 'EUR/ZAR', 10: '193'}
>>>

Only field "Symbol" has been added into the group

But if I replace the component by the suitable fields like this

<message name="QuoteRequest" msgtype="R" msgcat="app">
      <field name="QuoteReqID" required="Y" />
      <group name="NoRelatedSym" required="Y">
          <field name="Symbol" required="Y" />
          <field name="Product" required="Y" />
          <field name="QuoteRequestType" required="N" />
          <field name="OrderQty" required="Y" />
      </group>
    </message>

all works as expected

>>> codec2.parse(messages[2].strip(), separator=chr(1))
{8: 'FIX.4.3', 9: '151', 35: 'R', 34: '3', 49: '9222237_TEST', 50: 'FX', 52: '20200616-18:25:26', 56
: 'SAXOQUOTE_TEST', 131: '1', 146: [{55: 'AUD/CA', 460: '4', 303: '2'}, {55: 'AUD/CHF', 460: '4', 30
3: '2'}, {55: 'EUR/ZAR', 460: '4', 303: '2'}], 10: '193'}
>>>

and all fields are placed into the group

ValueError when encountering XML comments (using lxml)

Unlike xml.etree.ElementTree.parse(), lxml.etree.parse() reads XML comments and stores them as lxml.etree._Comment objects, with tag == lxml.etree.Comment (the function which creates lxml.etree._Comment objects). This results in the error below:

Traceback (most recent call last):
File "./error_example.py", line 32, in
main()
File "./error_example.py", line 16, in main
spec = FixSpec('FIXFX.xml')
File "../pyfixmsg/reference.py", line 164, in init
(MessageType(e, self) for e in self.tree.findall('messages/message'))}
File "../pyfixmsg/reference.py", line 163, in
self.msg_types = {m.msgtype.encode('ascii'): m for m in
File "../pyfixmsg/reference.py", line 164, in
(MessageType(e, self) for e in self.tree.findall('messages/message'))}
File "../pyfixmsg/reference.py", line 278, in init
self.composition = _extract_composition(element, spec)
File "../pyfixmsg/reference.py", line 198, in _extract_composition
raise ValueError("Could not process element '{}'".format(elem.tag))
ValueError: Could not process element '<built-in function Comment>'

Any plans for supporting Fix 5.0 ?

The package works really well for Fix 4.2 and 4.4 which use one single spec file. However Fix above 4.4 uses TransportDataDictionary and AppDataDictionary for two seperate fix spec file as mentioned here http://www.quickfixengine.org/quickfix/doc/html/configuration.html. Is there any plans to support this ?.

Could not pip install?

I could not pip install pyfixmsg on MacOS M1. Not sure if it is about the OS, architecture, or the library itself

Parsing works incorrect, if repeated group started with other rgroup

The test below shows the problem:

    def test_nested_rgroup_ext(self, spec):
        codec = Codec(spec=spec)
        msg = b'35=CUS\x0134=3\x01131=some text\x01658=99\x01' \
              b'295=2\x01683=2\x01688=1\x01689=1\x01688=2\x01689=2\x01683=1' \
              b'\x01688=3\x01689=3\x0110=193\x01'
        msg = codec.parse(msg, separator=chr(1))
        assert {35: 'CUS',
                34: '3',
                131: 'some text',
                658: '99',
                295: [
                    {683: [{688: '1', 689: '1'}, {688: '2', 689: '2'}]},
                    {683: [{688: '3', 689: '3'}]}
                ],
                10: '193'} == msg

The actual parsed message look like,
{35: 'CUS', 34: '3', 131: 'some text', 658: '99', 295: [{683: '2'}], 688: '3', 689: '3', 683: '1', 10: '193'}

morganstanley / pyfixmsg Goto Github PK

pyfixmsg's People

Contributors

Stargazers

Watchers

Forkers

pyfixmsg's Issues

If component tag is located inside group, parsing of raw message works incorrect

ValueError when encountering XML comments (using lxml)

Any plans for supporting Fix 5.0 ?

Could not pip install?

Parsing works incorrect, if repeated group started with other rgroup

Parsing work incorrect, if first tag of group is group tag

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent