Version 0.15.0

• Polish docs (automate page, with pinned colrev version; functional overview)
• Extend tests (covering code in the paper; dist smoke test)
• JSON: make SearchQuery(query=...) the (safer) default

Version 0.14.1

• Fix broken sdist: caused wheels to ship without package code.

Version 0.14.0

• Documentation and demo notebook: Revised the documentation and extended the demo notebook.
• Generic query syntax: Make fields case insensitive.
• Linter message refinement: Formatting (cli) and internal handling revised. WOS_0005deprecated.
• Extend QueryErrorCode.JOURNAL_FILTER_IN_SUBQUERY: Added ISSN and ISBN fields.
• Dependency: Removed packaging dependency.

Version 0.13.0

• Versioned platform architecture & upgrades: Introduced version-aware parser registry that selects list or string parsers per platform version and defaults to the latest registered release. Auto-discovery for parser, serializer, and translator implementations. Added an upgrade pipeline that routes through the generic query as an intermediate representation, exposed through a new upgrade CLI subcommand.
• CLI improvements: Rebuilt the CLI around explicit translate, lint, and upgrade sub-commands with improved error handling and user feedback, including success messages emitted by the linter workflow.
• Search file handling: Refactored SearchFile to replace filepath with search_results_path, derive the default history path, ensure directories exist when saving, and exclude private attributes from serialization output.
• Linter updates: Added the deprecated-syntax warning (LINT_2001) to guide users toward upgrading queries that rely on legacy syntax.
• Documentation: Documented the syntax upgrade workflow and versioning policy for database-specific queries, including CLI examples.

Version 0.12.0

• Platform Support:

  • PubMed: #26
  • Web of Science: #20
  • EBSCOHost: #19
    → Implemented full platform support including:
    • Parsers (supporting query-string and list formats)
    • Linters
    • Serializers
    • Translators
      → Enhanced parsing capabilities with artificial parentheses to correctly reflect operator precedence.
      → Linters now provide categorized messages across six dimensions: parsing, structure, terms, fields, databases, and quality.

• Refactoring and Internals:

  • Refactored core query classes and methods (query_and, query_or, query_near, etc.).
  • Improved tokenization logic and consolidated linter messages into structured groups.
  • Removed deprecated linter-mode and unreachable code paths.
  • Migrated to uv for dependency management and streamlined environment setup.

• New Features and Improvements:

  • Introduced a query database for programmatic retrieval and sharing of query examples.
  • Achieved comprehensive unit test coverage, exceeding 95% of the codebase.
  • Expanded and improved documentation, including developer guidelines and platform-specific usage notes.

Version 0.11.0

• Refactoring (AND/OR/NOT-Queries, method signatures, public to private methods)
• Rename methods: query.translate_xy() to query.write(syntax=xy), query.print_query_xy() to query.to_string(syntax=xy)
• Integrate Query tree into node attribute
• Make dev dependencies optional
• Add tests on GitHub actions
• Extract constants
• Code skeleton for Parsers and Linters
• Documentation: Sphinx
• SearchFields instead of str
• Binder demo
• Implement Query.selects()
• Drop ABC inheritance
• Add is_term() and is_operator()

Release 0.10.0

First basic implementation of a search query translator.

Included databases for translation:

• PubMed
• Web of Science - Core Collection
• IEEE Xplore

Version 0.1.0

Package for managing literature search queries.