Pure Python FIX protocol engine — mirrors the QuickFIX architecture (QuickFIX/n, QuickFIX/J).
Features
- Full FIX session layer (Logon, Logout, Heartbeat, TestRequest, ResendRequest, SequenceReset, Reject) with gap-fill, out-of-order buffering, hard-reset recovery, and logon/logout timeouts
- Hardened session recovery: byte-faithful resend (repeating groups survive replay), automatic ResendRequest retry for stalled gaps, duplicate-connection refusal, and spec-compliant
Rejects for malformed admin fields instead of dropped connections - Message encode/decode (checksum, BodyLength, field-order validation) and repeating groups, with public enumeration accessors for walking an arbitrary decoded message
- FIX versions 4.0–4.4 plus FIX 5.0 over FIXT.1.1 (FIX.5.0 / SP1 / SP2), with
DefaultApplVerIDnegotiation and a combined transport + application dictionary — all QuickFIX dictionaries bundled - Opt-in per-session validation against a QuickFIX
DataDictionary— required fields and out-of-range enum values, with spec-compliantRejects - Async transport:
SocketAcceptor+SocketInitiatorwith auto-reconnect - Dynamic sessions — add/remove sessions at runtime without a restart; acceptor sessions share the listening port, and wildcard acceptor templates (
AcceptorTemplate=Y) admit unknown counterparties on the fly, with an approval hook to vet them (see docs/dynamic-sessions.md) - Session schedules (
StartTime/EndTime/StartDay/EndDay/TimeZone) — daily or weekly active windows with sequence reset at each new period - Durable
FileStore(fsync + atomic sequence-number writes) andMemoryStore MessageCrackerdispatch mixin- Lightweight browser GUI via aiohttp (
fixcore-gui) — session management, message builder, live message log
Quick start
pip install fixcore-engine
pip install "fixcore-engine[gui]" # includes aiohttp for the GUI
Validation (opt-in)
Validation is off by default (raw FIX passthrough). Enable it per session via QuickFIX-compatible settings:
[SESSION]
UseDataDictionary=Y ; load the spec: group-aware decode + (by default) validation
DataDictionary=FIX44.xml ; bundled spec name (FIX40/41/42/43/44.xml), or a path to your own
ValidateIncomingMessage=N ; keep group-aware decode but DON'T reject non-conforming inbound (default Y)
ValidateFieldsOutOfRange=Y ; also reject out-of-range enum values (e.g. Side=Z)
UseDataDictionary=Y loads the spec so repeating groups decode into proper
instances. By default it also validates inbound app messages (required fields,
known MsgType) and Rejects violations. Set ValidateIncomingMessage=N to keep
the group-aware decode while delivering non-conforming messages as-is — useful
for permissive/testing peers that need to inspect off-spec traffic rather than
reject it.
FIX 5.0 (FIXT.1.1)
FIX 5.0 splits the session transport from the application layer. Use
BeginString=FIXT.1.1 and select the application version — fixcore combines the
transport and application dictionaries automatically:
[SESSION]
BeginString=FIXT.1.1
UseDataDictionary=Y
AppDataDictionary=FIX50SP2.xml ; FIX50.xml / FIX50SP1.xml / FIX50SP2.xml
DefaultApplVerID=FIX.5.0SP2 ; negotiated on Logon (tag 1137); also accepts 7/8/9
TransportDataDictionary defaults to FIXT11.xml. If AppDataDictionary is
omitted it is inferred from DefaultApplVerID. See
docs/fixt-5.0.md.
Reading messages
Field access is by tag, with a mapping protocol alongside the QuickFIX-style methods:
msg[55] = "AAPL" # set_field
symbol = msg[55] # get_field (KeyError if absent)
side = msg.get(54, "?") # get_field_or
if 55 in msg: ... # has_field
Repeating groups decode into Group instances when the session has a
DataDictionary (UseDataDictionary=Y). Fetch them by count tag:
for party in msg.get_groups(453): # NoPartyIDs
print(party.get_field(448)) # PartyID
To walk a message generically — a message log, a flattener, a correlator that sees arbitrary MsgTypes — enumerate instead of asking by tag:
def flatten(msg):
# Every field in the message as {tag: value}; last value wins.
flat = dict(msg.header.items())
flat.update(msg.body)
flat.update(msg.trailer.items())
for instances in msg.groups().values(): # count_tag -> [Group, ...]
for inst in instances:
flat.update(flatten_group(inst))
return flat
def flatten_group(group):
flat = dict(group.items()) # incl. nested count tags
for instances in group.nested_groups().values():
for inst in instances:
flat.update(flatten_group(inst))
return flat
Message.groups() and Group.nested_groups() return a read-only view of
live state (they run per message on the event loop, so nothing is copied):
the mapping rejects mutation, but the Group instances it exposes are the
real ones — add instances with add_group, which keeps the count field in
sync. Group.items() mirrors Header.items() / Trailer.items(), returning
(tag, value) pairs in insertion order.
Configuration
Every recognised setting (identity, connection, timers, sequence reset, session schedules, validation) is documented in docs/configs.md. Session-schedule semantics and edge cases are in docs/session-schedule.md.
Development
pip install -e ".[dev,gui]"
pytest
fixcore-gui
Metadata
Release files for fixcore-engine 0.15.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| fixcore_engine-0.15.0.tar.gz | 497.8 kB | With provenance | Details for fixcore_engine-0.15.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| fixcore_engine-0.15.0-py3-none-any.whl | Python 3 | none | any | Details for fixcore_engine-0.15.0-py3-none-any.whl |
Total release size: 962.7 kB
Release files / fixcore_engine-0.15.0.tar.gz
| Download URL | fixcore_engine-0.15.0.tar.gz |
|---|---|
| Size | 497.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
112d63c7c61926f163bcd1e476ac3e62218c3be3448e7c19afdc45916756d6cc
|
|
BLAKE2b-256 checksum How to use checksums |
dfb31226bf101a0d1fc23d53dc19be690be5c23e6b0d189bb06ec7c28e573ad6
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Aug 26, 2026.
Transparency logRelease files / fixcore_engine-0.15.0-py3-none-any.whl
| Download URL | fixcore_engine-0.15.0-py3-none-any.whl |
|---|---|
| Size | 464.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
bdf0f93ccc4e3cf5ec58432896ea74f415f6049867dc6690db70810fd59df6d0
|
|
BLAKE2b-256 checksum How to use checksums |
0b661cf1ebeebe5cd2d57493127e25f16aa58c8f1672985804b3d85d2b4d4ed3
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Aug 26, 2026.
Transparency logSecurity for fixcore-engine 0.15.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
Full provenance
Both files in this release have been uploaded with provenance metadata.
Attestation for fixcore_engine-0.15.0.tar.gz
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Aug 26, 2026.
Transparency logA note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.