Backend Dev Task

Solar-layout helper that turns a list of panel coordinates into two precise maps: where to bolt mounts onto rafters and where to join adjacent modules. Plus a visualization so you can verify the result quickly.

How It Works

• We read panel top-left coordinates (defaults mirror the PDF sample data).
• LayoutCalculator builds the context: generates a rafter grid, normalizes panel dimensions, and then runs two specialized calculators.
• MountCalculator finds usable rafters inside each panel, enforces limits, and emits mounting points on the top and bottom edges.
• JointCalculator looks for horizontally/vertically adjacent panels and adds unique connectors that can serve up to four panels at once.
• The result (LayoutResult) contains two coordinate arrays (mounts, joints). The CLI writes them to JSON, and RenderLayout can draw a PNG with panels/rafters/joints.

Requirements

• Python 3.10+
• (Optional) virtual environment

Installation

python -m venv .venv
.venv\Scripts\Activate.ps1
pip install -e .[test]

Running The Service

The CLI runs the algorithm against the inputs defined in the config (defaults come from config.yaml). Treat it as a reference snapshot or point it to your own YAML file to recompute different geometries.

python -m backend_dev_task.cli

Pass a custom configuration or output directory:

python -m backend_dev_task.cli --config config.yaml --output-dir output

The command stores its output in output/layout_<timestamp>.json, e.g. output/layout_20251203_101500.json. Inside you get both mount and joint arrays for whatever pipeline or report you want to plug them into.

Optional plotting (requires pip install -e .[viz] to pull in matplotlib and pillow):

python -m backend_dev_task.cli --config config.yaml --show-plot

Every plot run also writes layout_<timestamp>.png alongside the JSON file. Provide a custom directory via --plot-path path/to/dir or a filename (the timestamp is appended automatically to avoid overwriting previous renders).

Configuration

Settings live in config.yaml. All keys are optional—omit a section to fall back to defaults. Editing this file is the quickest way to experiment with rafter spacing, validation thresholds, and layout tolerances.

panel:
  width: 44.7
  height: 71.1
rafters:
  spacing: 16.0
  edge_clearance: 2.0
mounts:
  span_limit: 48.0
  cantilever_limit: 16.0
  edge_clearance: 2.0
joints:
  horizontal_gap_threshold: 1.0
  vertical_tolerance: 0.5
validation:
  allow_negative_coordinates: false
  allow_duplicates: false
  allow_overlaps: false
  coordinate_tolerance: 0.0001

Use the validation block to relax or tighten input checks (negative coordinates, duplicate panels, overlap detection, tolerance for float comparisons). The PDF profile keeps every guard enabled.

The same YAML structure can be used anywhere the code accepts a LayoutConfig (e.g., LayoutCalculator(config=LoadConfig(Path("my_config.yaml")))).

Running Tests (pytest)

Execute all tests with:

python -m pytest

You can also run a single file:

python -m pytest tests/test_layout.py

Library Usage

from backend_dev_task.layout_service import LayoutCalculator

calculator = LayoutCalculator()
layout = calculator.CalculateLayout([
  {"x": 0, "y": 0},
  {"x": 45.05, "y": 0},
])
print(len(layout.mounts), len(layout.joints))

layout exposes two collections (mounts, joints) represented as lists of (x, y) coordinates. Feed them to downstream calculations (costing, CAD export, reporting, etc.).

Visualization Module

The helper backend_dev_task.visualization.RenderLayout renders panels, rafters, mounts, and joints using matplotlib. Import it manually or rely on the CLI flags to generate PNG files / show the chart once matplotlib is installed via the viz extra.