v1.2.1

v1.2.1 (2026-02-05)

Bug Fixes

• _score was named score (056bad8)

• Add_blocks_to_dataset does not use index (8fed724)

• Crash on sort (3e4f0ab)

• Flatten does not raise anything (40affb0)

• Must_not_be_different was ignored by MixedBlocker (5807571)

• Normalize crash on checking if array is NaN (3d3f06e)

• Normalize would skip normalization on lists (fb8092f)

• Parse_list and normalize should be better at handling null values (4e7e435)

Refactoring

• Replace must_not_be_different logic with dropna and duplicated combination; fix: preprocessed does not always run on right columns; perf: add second dropna for overlap since empty lists were not considered as NaN's (this might be a huge improvement) (f18d8cd)

• Typo (5e44994)

• Use .empty instead of len()==0 (3fa86f8)

Testing

• 100% (920a7a5)

• Add a dozen series of tests for near-complete coverage (0183c51)

• Style (a70cda2)

---

Detailed Changes: v1.2.0...v1.2.1

v1.2.0

v1.2.0 (2026-02-04)

Bug Fixes

• Parse_list crashes on strings that do not represent lists (83b1932)

• Parse_list use startswith instead of endswith (b37e017)

• Switche must_not_be_different and normalize_strings (b4441c4)

Code Style

• Reformat (460808d)

• Reformat (c502c19)

• Remove obsolete comments (9e72260)

Documentation

• Add typehints (86a443c)

• Fix discarded reference (41e8def)

• Fix motives (a128219)

• Fix obsolete type in docstring (3643a00)

• Fix rendering issue in notebook (add6ae5)

• Rebuild docs (49ce632)

• Update example notebook with new motive system (60a3cc9)

Features

• BREAKING CHANGES various improvements and bugfixes (d08173d)

Performance Improvements

• Move dropna higher in the block logic (8cdd3bd)

Refactoring

• Make preprocessing in .block more compact and pandas-esque (9521338)

• Motive as list instead of string (c0e1891)

• Motives are now instances of own classes (9da6796)

• Remove remove_value_if_appears_only_once since it was redundant with df.duplicated (9fce21d)

Testing

• Fix checks depending on (random) ordering of motives (4822c87)

• Update tests to new Motives (41e154e)

---

Detailed Changes: v1.1.0...v1.2.0

v1.1.0

v1.1.0 (2026-02-02)

Bug Fixes

• Crash on multiple OR (c05683e)

• Overlap motive message (7306e13)

Build System

• Add ruff to cicd (afb3fed)

• Add ruff to cicd (212b248)

• Add versioning to workflow (ea64a3d)

• Remove obsolete comments (022b55a)

Code Style

• Formatting (1ec4d4d)

• Remove obsolete comment (d2d80e7)

• Ruff (1523cc6)

Documentation

• Add must_not_be_different to example notebook (987978e)

• Add scoring to example notebook (5a9211f)

• Run notebook (3793a13)

• Update readme (c51220e)

Features

• Add eq for AndNode and OrNode (c526ca9)

• Add eq for blockers (ad91e93)

• Add scoring function (eab75ad)

• Revamp motives structure (a381fc3)

Refactoring

• Add underscore in front of new column names; perf: do not add temp column to df in must_not_be_different_apply (043ac2d)

• Move adding motive logic to own function (80b3561)

• Move helpers to utils; docs: add docstrings and typehints to utils (1877168)

• Move must_not_be_different logic to own function (ec7b2f2)

• Move overlap coords generation logic to own function (eb7d1ea)

• Rename block and motive to new names (c6f6fd6)

• Rename Node to BlockerNode; fix: motives handling in AndNode (4e14e6d)

• Ruff (c2799e4)

Testing

• Add tests for MixedBlocker (5dec1a9)

---

Detailed Changes: v1.0.0...v1.1.0