Port ArxivRetriever to LangChainJS

[bracesproul]

Discussed in #7077

^{Originally posted by boni-teppanyaki October 26, 2024}

Checked

• I searched existing ideas and did not find a similar one
• I added a very descriptive title
• I've clearly described the feature request and motivation for it

Feature request

Add an ArxivRetriever component to LangChainJS, enabling JavaScript-based applications to query and retrieve research papers directly from arXiv. This retriever would allow users to pull academic content by setting parameters such as titles, authors, topics and abstracts, as well as through natural language queries or article identifiers.

Motivation

Incorporation of arXiv will extend LangChainJS's utility for research, question-answering, summarization, and other use cases by retrieving and utilizing academic resources from the arXiv platform. Currently, arXiv support is available only in LangChain Python. We would like to port this feature so that we can use it in js version.

Proposal (If applicable)

We would like to develop this feature and contribute it as a community package.

cc @boni-teppanyaki