sayou-assembler 0.3.0

Assembler components for the Sayou Data Platform

sayou-assembler

The Knowledge Builder for Sayou Fabric.

sayou-assembler is the construction site of the data pipeline. It takes the standardized SayouNode objects produced by the Wrapper and assembles them into target-specific formats ready for loading into databases.

It bridges the gap between the abstract data model and concrete storage technologies (Graph DBs, Vector DBs).

💡 Core Philosophy

"Build Once, Deploy Anywhere."

The Assembler decouples the structure of knowledge from the syntax of the database.

• Graph Builder: Constructs a topology of Nodes and Edges (including automatic reverse linking).
• Vector Builder: Transforms nodes into embeddings and metadata payloads.
• Query Builder: Generates specific query languages like Cypher or SQL.

📦 Installation

pip install sayou-assembler

⚡ Quick Start

The AssemblerPipeline converts standardized data into database payloads.

from sayou.assembler.pipeline import AssemblerPipeline

def run_demo():
    # 1. Initialize (Inject embedding function for Vector Builder)
    pipeline = AssemblerPipeline()
    pipeline.initialize(embedding_fn=my_embedding_func)

    # 2. Input Data (from sayou-wrapper)
    wrapper_output = {
        "nodes": [
            {"node_id": "doc_1", "node_class": "sayou:Document", "attributes": {"text": "..."}},
            # ...
        ]
    }

    # 3. Build for Graph DB (Nodes + Edges)
    graph_data = pipeline.run(wrapper_output, strategy="graph")
    print(f"Edges created: {len(graph_data['edges'])}")
    
    # 4. Build for Vector DB (Embeddings)
    vector_data = pipeline.run(wrapper_output, strategy="vector")
    print(f"Vectors created: {len(vector_data)}")

if __name__ == "__main__":
    def my_embedding_func(text): return [0.1, 0.2, 0.3] # Dummy
    run_demo()

🔑 Key Components

Builders

• GraphBuilder: Converts nodes into a generic graph structure (Dictionary). Automatically creates reverse edges (e.g., hasChild from hasParent) to ensure bi-directional traversability.
• VectorBuilder: Extracts text from nodes, computes embeddings, and formats payloads for Vector DBs.

Plugins

• CypherBuilder: Generates Neo4j MERGE queries to insert nodes and relationships idempotently.

🤝 Contributing

We welcome builders for other targets (e.g., SqlBuilder, GremlinBuilder).

📜 License

Apache 2.0 License © 2025 Sayouzone