Struct PosPluginProcessor 

pub struct PosPluginProcessor { /* private fields */ }

NDPosPlugin processor: attaches position metadata to arrays from a position list.

Implementations

impl PosPluginProcessor

pub fn new(mode: PosMode) -> Self

pub fn load_positions_json(&mut self, json_str: &str) -> Result<usize, Error>

Load positions from a JSON string.

pub fn load_positions_xml(&mut self, xml_str: &str) -> Result<usize, String>

Load positions from an XML string (C++ NDPosPlugin pos_layout format).

Expected XML format (matching NDPosPluginFileReader):

<pos_layout>
  <dimensions>
    <dimension name="x"/>
    <dimension name="y"/>
  </dimensions>
  <positions>
    <position x="1" y="2"/>
    <position x="3" y="4"/>
  </positions>
</pos_layout>

Each <dimension name="N"/> declares an ordered dimension; each <position .../> carries one attribute per dimension, and the attribute value (parsed as f64) is stored under the dimension name. A position missing any declared dimension’s attribute is rejected (matching C addPosition returning asynError).

pub fn load_positions_auto(&mut self, content: &str) -> Result<usize, String>

Load positions from a string, auto-detecting format.

If the content starts with ‘<’ (after trimming whitespace), it is treated as XML. Otherwise, it is treated as JSON.

pub fn load_positions(&mut self, positions: Vec<HashMap<String, f64>>)

Load positions directly.

pub fn start(&mut self)

Start processing.

C writeInt32(NDPos_Running) resets ExpectedID to IDStart and does nothing else (NDPosPlugin.cpp:230-234) — in particular it does not clear MissingFrames/DuplicateFrames, which persist across runs until a client writes them (they are zeroed only in the constructor).

pub fn stop(&mut self)

Stop processing.

pub fn clear(&mut self)

Clear all positions.

pub fn missing_frames(&self) -> usize

pub fn duplicate_frames(&self) -> usize

pub fn remaining_positions(&self) -> usize

Trait Implementations

impl NDPluginProcess for PosPluginProcessor

fn process_array( &mut self, array: &NDArray, _pool: &NDArrayPool, ) -> ProcessResult

Process one array. Return output arrays and param updates.

fn plugin_type(&self) -> &str

Plugin type name for PLUGIN_TYPE param.

fn register_params(&mut self, base: &mut PortDriverBase) -> AsynResult<()>

Register plugin-specific params on the base. Called once during construction.

fn on_param_change( &mut self, reason: usize, params: &PluginParamSnapshot, ) -> ParamChangeResult

Called when a param changes. Reason is the param index. Return param updates to be written back to the port driver.

fn compression_aware(&self) -> bool

Whether this plugin can process compressed (codec != None) arrays (C++ compressionAware_, G3). Defaults to false: a plugin that operates on raw pixels must not be handed compressed bytes — the runtime drops compressed input and counts it into DroppedArrays. A codec/file plugin that understands compressed data overrides this.

fn does_array_callbacks(&self) -> bool

Whether this plugin delivers arrays to downstream plugins, i.e. the initial NDArrayCallbacks param value. Defaults to true: most plugins do array callbacks. Terminal plugins that never deliver downstream (NDPluginStdArrays, NDPluginAttribute, every NDPluginFile writer) override this to false so the param reflects the behaviour, matching C (e.g. NDPluginFile.cpp:948 setIntegerParam(NDArrayCallbacks, 0)).

fn array_data_handle( &self, ) -> Option<Arc<Mutex<RawMutex, Option<Arc<NDArray>>>>>

Return a handle to the latest NDArray data for array reads. Override this in plugins like NDPluginStdArrays that serve pixel data via readInt8Array/readInt16Array/etc.