RTMA/Dragonfly client written in python with no external dependencies. Based on and compatible with Dragonfly Messaging

pyrtma is available on PyPI

pip install pyrtma --upgrade

This is only necessary for individuals who would like to contribute to pyrtma.

From pyrtma git repo directory:

pip install --upgrade pip setuptools
pip install -e .

python -m pyrtma.manager

or

message_manager

Message definitions are created in a .yaml file.

The ruamel.yaml parser library is used internally (https://yaml.readthedocs.io/en/latest/)

Notes about yaml format:

• Whitespace sensitive. Use either 2 or 4 spaces for tab not '\t'
• Key-values must be separated by a colon followed by a space, (Key: Value, not Key:Value)
• Must follow the top-level headers shown below.
• Unused sections should be marked null
• Names must start with letter. (no _ or numeric prefixes allowed)
• Use (.yaml) extension not (.yml)

List of supported native data types:

• char
• byte
• float
• double
• int8
• int16
• int32
• int64
• uint8
• uint16
• uint32
• uint64

Below is an example:

# message.yaml

imports: null

# Constant values and expressions
constants: 
    STR_SIZE: 32
    LONG_STRING: STR_SIZE * 2

# Constant string values
string_constants:
    default_msg: "hello_world"

host_ids: null

module_ids:
    PERSON_PUBLISHER: 212
    PERSON_SUBSCRIBER: 214

# Alias a type by another name
aliases:
    AGE_TYPE: int32

# Non-message structured data (no id field)
struct_defs:
    TEST_STRUCT:
        fields:
            value_str: char[STR_SIZE]
            value_int: int32

# Message defintions with user assigned id field
message_defs:
    PERSON_MESSAGE:
        id: 1234
        fields:
            name: char[STR_SIZE]
            age: AGE_TYPE

    ANOTHER_EXAMPLE:
        id: 5678
        fields:
            value_struct: TEST_STRUCT
            value_float: float
            value_double: double

    # Example signal definition
    USER_SIGNAL:
        id: 2468
        fields: null

    # Example using a nested message defintion 
    PERSON_LIST:
        id: 1357
        fields:
            person: PERSON_MESSAGE[32]

    # Example reusing a message definition by another name
    EMPLOYEES:
        id: 1368
        fields: PERSON_LIST

    # A block of message ids can be reserved by a file for future use
    # Ranges are inclusive on both ends
    # Note that ':' cannot be used to indicate a range, as this will cause
    # the yaml parser to throw an error
    _RESERVED_:
        id: [1000, 1002 - 1008, 1009 to 1012]

Run the following command to compile the yaml file into Python, C, Matlab, or Javascript files. This will output a message.(py|h|m|js) file.

python -m pyrtma.compile -i examples/msg_defs/message.yaml --py --c --mat --js

or

rtma_compiler -i examples/msg_defs/message.yaml --py --c --mat --js

The msg_defs directory should now have message def files created for each language.

The rtma objects are compiled into objects suitable for each language.

See /examples/example.py for pub/sub demo app

Compile the example message defintions:

python -m pyrtma.compile -i ./examples/msg_defs/message.yaml --py

Start the demo MessageManager server

python -m pyrtma.manager

Start the publisher in one console:

python ./examples/example.py --pub

Start the subscriber in another:

python ./examples/example.py --sub

Clients in javascript require the web_manager server to convert websocket messages to rtma messages.

Run:

python -m pyrtma.web_manager -m <MM_IP> -p <WEBSOCKET_PORT> -d <DEFS_FILE>

or

web_manager -m <MM_IP> -p <WEBSOCKET_PORT> -d <DEFS_FILE>

to launch the web_manager which will listen to websocket connects on port <WEBSOCKET_PORT> and forward to message manager at <MM_IP>, using pyrtma message definitions defined in <DEFS_FILE>.

See rtma-js for developing rtma clients in javascript.

See examples/data_logger_example.py for an example on how to configure data collection for a DataLogger.

To start up the logging module:

python -m pyrtma.data_logger --help

or

data_logger --help

There is also an interactive command line tool to control and configure the DataLogger manually:

python -m pyrtma.data_logger.cli