abstract-extract

abstract-extract

Typed Python utilities for retrieving and normalizing scholarly article metadata and abstracts from Scopus and Crossref.

The package provides a small, explicit API for querying Scopus, following cursor-based pagination, converting raw Scopus records into typed Article objects, and retrieving abstracts from Crossref by DOI.

Features

• Scopus search with explicit request timeouts.
• Cursor-based retrieval of complete Scopus result sets.
• Optional author and date-range filtering.
• Typed immutable Article model.
• Normalization of raw Scopus records.
• Crossref abstract lookup by DOI.
• Injectable HTTP sessions for testing and connection reuse.
• Strict static typing with mypy.
• Ruff linting and formatting.
• Deterministic unit tests with no live API calls.
• CI-enforced test coverage.

Requirements

• Python 3.12
• Poetry
• A Scopus API key for Scopus requests

Crossref requests do not require a Scopus API key.

Installation

Clone the repository and install the project with Poetry:

git clone https://github.com/DiogoRibeiro7/abstract_extract.git
cd abstract_extract
poetry install

The package can then be used inside the Poetry environment:

poetry run python

Scopus credentials

Do not store API keys in source code.

Copy the example environment file:

cp .env.example .env

Set your Scopus key locally:

SCOPUS_API_KEY=your-scopus-api-key

The package itself accepts the API key explicitly. Applications can load it from their preferred environment/configuration mechanism.

For example:

import os

from abstract_extract import fetch_from_scopus

api_key = os.environ["SCOPUS_API_KEY"]

response = fetch_from_scopus(
    query="changepoint detection",
    api_key=api_key,
    max_results=10,
)

Usage

Fetch one page from Scopus

from abstract_extract import fetch_from_scopus

response = fetch_from_scopus(
    query="bayesian changepoint detection",
    api_key="your-api-key",
    max_results=25,
)

fetch_from_scopus returns the raw Scopus JSON response as a dictionary.

Fetch all Scopus results

from abstract_extract import fetch_all_from_scopus

entries = fetch_all_from_scopus(
    query="time series forecasting",
    api_key="your-api-key",
    author="Jane Doe",
    start_date="2020-01-01",
    end_date="2026-01-01",
)

Pagination is handled automatically using the Scopus cursor returned by each page.

If a date range is used, both start_date and end_date must be supplied.

Normalize Scopus entries

from abstract_extract import process_scopus_entries

articles = process_scopus_entries(entries)

for article in articles:
    print(article.title)
    print(article.doi)
    print(article.authors)

Each item is returned as an immutable Article:

from abstract_extract import Article

article = Article(
    title="Example paper",
    abstract="Example abstract",
    publication_date="2026-01-01",
    authors=("Alice Example", "Bob Example"),
    doi="10.1000/example",
)

Normalize a single Scopus response

For a raw response returned by fetch_from_scopus:

from abstract_extract import fetch_from_scopus, process_scopus_response

response = fetch_from_scopus(
    query="uncertainty quantification",
    api_key="your-api-key",
)

articles = process_scopus_response(response)

Retrieve an abstract from Crossref

from abstract_extract import get_abstract_from_doi

abstract = get_abstract_from_doi("10.1000/example")

if abstract is not None:
    print(abstract)

Crossref may return abstracts containing JATS/XML markup. The package currently returns that value unchanged.

Public API

The supported top-level interface is:

from abstract_extract import (
    Article,
    fetch_all_from_scopus,
    fetch_from_scopus,
    get_abstract_from_doi,
    process_scopus_entries,
    process_scopus_response,
)

HTTP and response-shape errors are intentionally propagated to callers rather than converted into sentinel values. This keeps failure handling explicit.

Development

Install all dependencies:

poetry install

Run the same checks enforced by CI:

poetry run ruff check .
poetry run ruff format --check .
poetry run mypy src
poetry run pytest --cov=abstract_extract --cov-report=term-missing

The coverage configuration enforces a minimum total coverage of 90%.

Project structure

.
├── .github/
│   └── workflows/
│       └── ci.yml
├── src/
│   └── abstract_extract/
│       ├── __init__.py
│       ├── crossref.py
│       ├── models.py
│       └── scopus.py
├── tests/
├── .env.example
├── pyproject.toml
└── LICENSE

Security

Credentials must be supplied at runtime and must not be committed to the repository.

If a credential is accidentally exposed in Git history, removing it from the current source tree is not sufficient. Revoke or rotate the credential at its provider.

License

This project is licensed under the MIT License. See LICENSE.