CHANGELOG

Changes in v2.0.0

Breaking Changes

  • Package renamed from scholarly to scholarly2. Update all imports: from scholarly import … → from scholarly2 import …

Features

  • Add scholarly.load_socks5_proxy_file(path: str) -> bool — explicit runtime loading of a SOCKS5 proxy file from any path, independent of the automatic .env.socks5 discovery at import time.

Deprecations

  • Mark ProxyGenerator.Luminati(), ScraperAPI(), FreeProxies(), and SingleProxy() as deprecated compatibility methods. Recommend SOCKS5 workflows only: Socks5Proxies(), Socks5ProxyFile(), or scholarly.load_socks5_proxy_file(path).

Changes

  • scholarly.use_proxy(pg) now reuses the primary proxy generator for all requests when no secondary proxy generator is supplied. It no longer auto-enables FreeProxies() as an implicit secondary path.

—

Changes in v1.7.0

Features

  • Add a new citation entry to pub fetched from an author profile with formatted citation entry #423.

Bugfixes

  • Fix pprint failures on Windows #413.

  • Thoroughly handle 1000 or more publications that are available (or not) according to public access mandates #414.

  • Fix errors in download_mandates_csv that may occassionally occur for agencies without a policy link #413.

Changes in v1.6.3

Bugfix

  • search_pubs method did not respect include_last_year, which is now fixed #420, #421.

Enhancement

  • Unit tests involving funding agency mandates are a bit more robust.

Changes in v1.6.2

Bugfix

  • Fix an error in the workflow publishing to PyPI.

Changes in v1.6.1

Bugfix

  • Handle 1000 or more publications that are available (or not) according to public access mandates #414.

Enhancement

  • Fetch 20+ coauthors without requiring geckodriver/chrome-driver to be installed #411.

Changes in v1.6.0

Features

  • Download table of funding agencies as a CSV file with URL to the funding mandates included

  • Downlad top-ranking journals in general, under sub-categories and in different languages as a CSV file

Bugfixes

  • #392

  • #394

Changes in v1.5.1

Feature

  • Support chromium (chrome-driver) as an alternative to geckodriver #387

Improvements

  • Firefox/Geckodriver operates in headless mode

  • Increase test coverage to include all public APIs

  • Clean up legacy code and improve coding styles

  • Remove the use of deprecated functions in dependency packages

Bugfix

  • Stop attempting to reuse a closed webdriver

Changes in v1.5.0

Features

  • Fetch the public access mandates information from a Scholar profile and mark the publications whether or not they satisfy the open-access mandate.

  • Fetch an author’s organization identifer from their Scholar profile

  • Search for all authors affiliated with an organization

  • Fetch homepage URL from a Scholar profile

Enhancements

  • Make FreeProxies more robust

  • Stop the misleading traceback error message #313

Bugfix

  • Fix bug in exception handling #366

—

Changes in v1.4.4

Bugfix

  • Fix a bug that would have prevented setting up ScraperAPI with exactly 1000 successful requests during the first week of the trial #356

Enhancement

  • Use FreeProxy instead of premium proxy servers when possible

—

Changes in v1.4.3

Bugfixes

  • Fill the complete title of publications even if it appears truncated

  • ## Robustly handle exceptions when more than 20 coauthors of a scholar cannot be fetched

Changes in v1.4.2

Bugfix

  • ## ScraperAPI proxy works reliably

Changes in v1.4.0

Features

  • Fetch the complete list of coauthors #322

  • Fetch all citeids for a given publication #324

  • Make scholarly objects inherently serializable #325

  • Expose scholarly specific exceptions #327

Bugfixes

  • Test Tor on macOS and skip the test if tor is not installed #323

  • Get cites_id and citedby_url without having to fill the publication #328

—

Changes in v1.3.0

Features

  • Make the Author and Publication objects serializable

  • Make cites_id a list to allow for multiple values

  • Fetch all (more than 20) coauthors from a Scholar profile