FlatBuffers is the format chosen for the ESS messaging system.
We would like to be able to read any message in the system at any time, therefore:
All schemas that we use for the transmission of data are collected in this repository.
The names of the schema files in this repository are prefixed by their unique
4-character file_identifier
. This file_identifier
must be set in the
schema definition file as:
file_identifier = "abcd";
The file identifiers (also called "schema id") must be unique on the network.
The root_type
should include the schema version number. For example, the f144 schema has
the root_type
:
root_type f144_LogData;
The naming convention for new identifiers and a table of existing identifiers follows later in this README. Please add your own (new schema) with file identifier to that table.
Please, avoid changes which break binary compatibility. FlatBuffers documentation contains good information about how to maintain binary compatibility. If you need to make breaking changes to schemas that are not under development, acquire a new schema id.
Schemas that are under development should be clearly marked as such in the schema file and in the Schema ids below to warn users of possible loss of backwards compatibility.
If you feel that you may need a lot of schema ids, you can use a single schema and work with the flat buffers union data type in your root element.
- Completely new schemas should have an ID comprising of two characters plus 00, e.g. hs00
- When updating an existing schema with a breaking change then the new schema should have the same ID but with the number incremented, e.g. hs00 -> hs01
- For older schema which don't end with two numbers, propose a new name which matches the convention.
- Prefix your schema files in this repository with the chosen schema id to more easily prevent id collision.
- Tables should use UpperCamelCase.
- Fields should use snake_case.
- Try to keep names consistent with equivalent fields in existing schema, e.g.:
timestamp
for timestampsource_name
for a string indicating origin/source of data in flatbufferservice_id
for a string indicating the name of the service that created the flatbuffer
- Do not use unsigned integers unless required for your application.
ID | File name | Description |
---|---|---|
f140 | f140_general.fbs |
[OBSOLETE] Can encode an arbitrary EPICS PV |
f141 | f141_ntarraydouble.fbs |
[OBSOLETE] A simple array of double, testing file writing |
f142 | f142_logdata.fbs |
(DEPRECATED) For log data, for example forwarded EPICS PV update [superseded by f144] |
f143 | f143_structure.fbs |
[OBSOLETE] Arbitrary nested data |
f144 | f144_logdata.fbs |
Controls related log data, typically from EPICS or NICOS. Note: not to be used for array data |
ev42 | ev42_events.fbs |
Multi-institution neutron event data for a single pulse |
ev43 | ev43_events.fbs |
Multi-institution neutron event data from multiple pulses |
ev44 | ev44_events.fbs |
Multi-institution neutron event data for both single and multiple pulses |
an44 | an44_events.fbs |
ANSTO-specific variant of ev44 |
ar51 | ar51_readout_data.fbs |
Streaming raw ESS detector readout data |
is84 | is84_isis_events.fbs |
ISIS specific addition to event messages |
ba57 | ba57_run_info.fbs |
[OBSOLETE] Run start/stop information for Mantid [superseded by pl72] |
df12 | df12_det_spec_map.fbs |
Detector-spectrum map for Mantid |
senv | senv_data.fbs |
(DEPRECATED) Used for storing for waveforms from DG ADC readout system. |
se00 | se00_data.fbs |
Used for storing arrays with optional timestamps, for example waveform data. Replaces senv. |
NDAr | NDAr_NDArray_schema.fbs |
(DEPRECATED) Holds binary blob of data with n dimensions |
ADAr | ADAr_area_detector_array.fbs |
(DEPRECATED) EPICS area detector array data [superseded by ad00] |
ad00 | ad00_area_detector_array.fbs |
EPICS area detector array data |
mo01 | mo01_nmx.fbs |
Daquiri monitor data: pre-binned histograms, raw hits and NMX tracks |
ns10 | ns10_cache_entry.fbs |
(DEPRECATED) NICOS cache entry |
ns11 | ns11_typed_cache_entry.fbs |
(DEPRECATED) NICOS cache entry with typed data (not used) |
hs00 | hs00_event_histogram.fbs |
(DEPRECATED) Event histogram stored in n dim array |
hs01 | hs01_event_histogram.fbs |
Event histogram stored in n dim array |
dtdb | dtdb_adc_pulse_debug.fbs |
Debug fields that can be added to the ev42 schema |
ep00 | ep00_epics_connection_info.fbs |
(DEPRECATED) Status of the EPICS connection |
ep01 | ep01_epics_connection.fbs |
Status or event of EPICS connection. Replaces ep00 |
json | json_json.fbs |
Carries a JSON payload |
tdct | tdct_timestamps.fbs |
Timestamps from a device (e.g. a chopper) |
pl72 | pl72_run_start.fbs |
File writing, run start message for file writer and Mantid |
6s4t | 6s4t_run_stop.fbs |
File writing, run stop message for file writer and Mantid |
answ | answ_action_response.fbs |
Holds the result of a command to the filewriter |
wrdn | wrdn_finished_writing.fbs |
Message from the filewriter when it is done writing a file |
x5f2 | x5f2_status.fbs |
Status update and heartbeat message for any software |
rf5k | rf5k_forwarder_config.fbs |
(DEPRECATED) Configuration update for Forwarder [superseded by fc00] |
fc00 | fc00_forwarder_config.fbs |
Configuration update for Forwarder |
al00 | al00_alarm.fbs |
Generic alarm schema for EPICS, NICOS, etc. |
da00 | da00_dataarray.fbs |
Pseudo-scipp DataArray with time-dependent and constant Variables |