Planpipe

A production-ready C# solution for processing PDF construction plans, extracting features, deriving quantities, and comparing results with ground truth data.

Architecture

This solution follows clean architecture principles with clear separation of concerns:

• Planpipe.Core - Domain models, enums, and interfaces
• Planpipe.Infrastructure - EF Core DbContext, PostgreSQL integration, repositories, and PDF text extraction
• Planpipe.Application - Business logic services for feature extraction, quantity derivation, confidence scoring, and ground truth comparison
• Planpipe.Api - ASP.NET Core Web API with RESTful endpoints
• Planpipe.UnitTests - xUnit unit tests with Moq
• Planpipe.IntegrationTests - xUnit integration tests with Testcontainers

Technologies

• .NET 8 LTS
• Entity Framework Core 8.0 with PostgreSQL (Npgsql)
• PdfPig 0.1.8 for PDF text extraction
• xUnit, Moq for testing
• Testcontainers.PostgreSql for integration testing
• Docker & Docker Compose for containerization
• GitHub Actions for CI/CD

Getting Started

Prerequisites

• .NET 8 SDK
• Docker and Docker Compose
• PostgreSQL (or use Docker Compose)

Running with Docker Compose

docker-compose up -d

This will start:

• PostgreSQL database on port 5432
• API service on port 8080

Running Locally

1. Restore dependencies:

   dotnet restore

2. Update database connection string in src/Planpipe.Api/appsettings.json

3. Run migrations:

   cd src/Planpipe.Infrastructure
   dotnet ef database update --startup-project ../Planpipe.Api/Planpipe.Api.csproj

4. Run the API:

   dotnet run --project src/Planpipe.Api/Planpipe.Api.csproj

Running Tests

Unit tests:

dotnet test tests/Planpipe.UnitTests/Planpipe.UnitTests.csproj

Integration tests (requires Docker):

dotnet test tests/Planpipe.IntegrationTests/Planpipe.IntegrationTests.csproj

All tests:

dotnet test

API Endpoints

POST /api/runs

Upload and process a PDF file.

Request: multipart/form-data with PDF file Response: ProcessingRun object with status and ID

GET /api/runs

Retrieve all processing runs.

Response: List of ProcessingRun objects

GET /api/runs/{id}

Get details of a specific processing run including features, quantities, and comparison results.

Response: ProcessingRun object with related data

GET /api/runs/{id}/quantities

Get extracted quantities for a specific run.

Response: List of QuantityItem objects

GET /api/runs/{id}/comparison

Get comparison results against ground truth for a specific run.

Response: List of ComparisonResult objects

Pipeline Processing Flow

1. PDF Upload → Receives PDF file via API
2. Text Extraction → Extracts text from PDF using PdfPig
3. Feature Extraction → Uses regex patterns to extract:
   • Quantities (e.g., "150.5 m2", "25 m3")
   • Dimensions (e.g., "2400 x 1200 mm")
   • Item codes (e.g., "ABC1234")
4. Quantity Derivation → Groups and sums quantities by unit
5. Confidence Scoring → Calculates confidence based on corroborating features
6. Ground Truth Comparison → Compares extracted quantities with ground truth data
7. Results Storage → Saves all results to PostgreSQL database

CI/CD

The project includes a GitHub Actions workflow (.github/workflows/ci.yml) that:

• Builds the solution
• Runs unit tests
• Runs integration tests with PostgreSQL service container

Project Structure

Planpipe/
├── src/
│   ├── Planpipe.Core/              # Domain layer
│   │   ├── Models/
│   │   ├── Enums/
│   │   └── Interfaces/
│   ├── Planpipe.Infrastructure/    # Data access layer
│   │   ├── Data/
│   │   ├── Repositories/
│   │   ├── Services/
│   │   └── Migrations/
│   ├── Planpipe.Application/       # Business logic layer
│   │   └── Services/
│   └── Planpipe.Api/               # Presentation layer
│       └── Controllers/
├── tests/
│   ├── Planpipe.UnitTests/
│   └── Planpipe.IntegrationTests/
├── Dockerfile
├── docker-compose.yml
└── .github/workflows/ci.yml

License

MIT