ProcessingContext

Overview

The ProcessingContext is a crucial component that maintains state during pipeline execution. It:

• Holds input/output directories, well filter, and configuration

• Stores processing results

• Serves as a communication mechanism between steps

Creating a Context

The context is typically created by the pipeline, but you can create it manually for advanced usage:

from ezstitcher.core.pipeline import ProcessingContext

# Create a processing context
context = ProcessingContext(
    input_dir="path/to/input",
    output_dir="path/to/output",
    well_filter=["A01", "B02"],
    orchestrator=orchestrator,  # Reference to the PipelineOrchestrator
    # Additional attributes can be added as kwargs
    positions_file="path/to/positions.csv",
    custom_parameter=42
)

Accessing Context Attributes

Context attributes can be accessed directly:

# Access standard attributes
print(context.input_dir)
print(context.well_filter)

# Access custom attributes
print(context.positions_file)
print(context.custom_parameter)

Accessing the Orchestrator

The context provides access to the orchestrator, allowing steps to use plate-specific services. Here’s how it’s done in the actual Step class:

# From ezstitcher/core/steps.py - Step.process method
def process(self, context: 'ProcessingContext') -> 'ProcessingContext':
    """
    Process the step with the given context.
    """
    logger.info("Processing step: %s", self.name)

    # Get directories and microscope handler
    input_dir = self.input_dir
    output_dir = self.output_dir
    well_filter = self.well_filter or context.well_filter
    orchestrator = context.orchestrator  # Required, will raise AttributeError if missing
    microscope_handler = orchestrator.microscope_handler

    if not input_dir:
        raise ValueError("Input directory must be specified")

    # ... rest of the method ...

Steps use the orchestrator’s high-level methods for specialized operations:

# From ezstitcher/core/steps.py - PositionGenerationStep.process method
def process(self, context):
    # Get required objects from context
    well = context.well_filter[0] if context.well_filter else None
    orchestrator = context.orchestrator  # Required, will raise AttributeError if missing
    input_dir = self.input_dir or context.input_dir
    positions_dir = self.output_dir or context.output_dir

    # Call the generate_positions method
    positions_file, reference_pattern = orchestrator.generate_positions(well, input_dir, positions_dir)

    # Store in context
    context.positions_dir = positions_dir
    context.reference_pattern = reference_pattern
    return context

# From ezstitcher/core/steps.py - ImageStitchingStep.process method
def process(self, context):
    # Get orchestrator from context
    orchestrator = getattr(context, 'orchestrator', None)
    if not orchestrator:
        raise ValueError("ImageStitchingStep requires an orchestrator in the context")

    # Call the stitch_images method
    orchestrator.stitch_images(
        well=context.well,
        input_dir=context.input_dir,
        output_dir=context.output_dir,
        positions_file=positions_file
    )

    return context