Add Chromia vectorstore integration

Description

Adds integration with Chromia as a new vectorstore option in LangChain.js. Chromia is a relational blockchain database that can be used for vector similarity search.

Key Features

• Full vectorstore implementation supporting:
  • Document/vector addition
  • Similarity search
  • Document deletion
  • Filtering support (interface added but not yet implemented)
• Built on top of postchain-client for Chromia interaction
• Complete TypeScript types and documentation

Usage Example

import { Chromia } from '@langchain/community/vectorstores/chromia';
import { OpenAIEmbeddings } from '@langchain/openai';

const embeddings = new OpenAIEmbeddings({
  model: "text-embedding-3-small", 
});

const vectorStore = new Chromia(embeddings, {
  client: postchainClient,
  numDimensions: embeddings.dimensions,
});

// Add documents
await vectorStore.addDocuments([
  { pageContent: "sample text", metadata: { source: "doc1" } }
]);

// Search
const results = await vectorStore.similaritySearch("query", 1);

Dependencies

• Requires postchain-client as a peer dependency

Testing

• Added unit tests for core vectorstore operations
• Tested with sample documents and queries
• Verified error handling for edge cases

Documentation

• Added comprehensive JSDoc comments
• Included usage examples in class documentation
• Added type definitions for all public interfaces

Todo

• Implement filter support in similarity search
• Add batch operations optimization
• Add more advanced search options