feat: Scopus API SearchSource

[geritwagner]

Description

Extend the colrev.scopus package by implementing searches of the scopus API. Scopus provides access to over 90 Million records, supports Boolean searches, and thereby offers a valuable resource for literature searches.

Preferred Solution

The existing colrev.scopus package should be extended to support automated API searches. The API requires authentication based on an API key. Based on a given query, results should be retrieved and stored. Search updates should use unique identifiers to update existing records (if necessary) and add new records. Records that were already retrieved should not be added redundantly. Ideally, a unit test is implemented to ensure reliable operation.

Implementation Notes

• The packages docs page explains the steps of developing CoLRev packages.
• Packages like colrev.ieee implement methods for handling API keys.
• The search-feed offers functionality for storing records.
• CEP003 describes principles for SearchSources.
• CEP002 describes the standard data schema for records.
• Implement unit tests simulating the retrieval process with a simple example. The colrev.crossref package implements similar unit tests.

User Story

1. The user initializes a CoLRev project using colrev init.
2. He/she conducts a search operation with colrev search -a colrev.scopus, specifying search parameters relevant to their topic.
3. The package retrieves records based on the parameters and saves them in the data/search directory of the project.
4. When users run colrev load, the records from the search directory are added to the data/records.bib file, which integrates all search results in the project.
5. When users run colrev search again, records are retrieved from the API, updating the existing records.

Useful Links

• Elsevier Developer Portal
• Scopus Search API Documentation

Expected Effort

2 months, 3-4 people.