ocr-stringdist 1.0.0__tar.gz → 1.0.1__tar.gz

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/CHANGELOG.md

@@ -5,6 +5,17 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [1.0.1] - 2025-09-21
+
+### Fixed
+
+- Fix critical bug in `WeightedLevenshtein.from_dict` when using insertion costs.
+
+### Added
+
+- `from ocr_stringdist import EditOperation`
+- `EditOperation.as_dict()`
+
 ## [1.0.0] - 2025-09-20
 
 ### Changed

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/Cargo.lock

@@ -74,7 +74,7 @@ dependencies = [
 
 [[package]]
 name = "ocr_stringdist"
-version = "1.0.0"
+version = "1.0.1"
 dependencies = [
  "pyo3",
  "rayon",

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "ocr_stringdist"
-version = "1.0.0"
+version = "1.0.1"
 edition = "2021"
 description = "String distances considering OCR errors."
 authors = ["Niklas von Moers <niklasvmoers@protonmail.com>"]

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ocr-stringdist
-Version: 1.0.0
+Version: 1.0.1
 Classifier: Programming Language :: Rust
 Classifier: Programming Language :: Python :: Implementation :: CPython
 Classifier: License :: OSI Approved :: MIT License

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/docs/source/examples.rst

@@ -77,4 +77,5 @@ The custom costs can be learned from a dataset of pairs of (OCR output, ground t
     distance = learned_wl.distance("Hay", "Hey")
     print(f"Distance with learned costs: {distance}")  # < 1.0
 
-Note that this only supports learning from character-level edits, not multi-character tokens.
+Note that this by default only supports learning from character-level edits.
+If multi-character tokens are to be considered, an `initial_model` that's already configured to know specific multi-character edits needs to be provided.

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/docs/source/index.rst

@@ -4,7 +4,7 @@
 
 A Python library to learn, model, explain and correct OCR errors using a fast string distance engine.
 
-:Repository: https://niklasvonm.github.io/ocr-stringdist/
+:Repository: https://github.com/NiklasvonM/ocr-stringdist
 :Current version: |release|
 
 .. image:: https://img.shields.io/badge/PyPI-Package-blue

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/python/ocr_stringdist/__init__.py

@@ -1,10 +1,12 @@
 from .default_ocr_distances import ocr_distance_map
+from .edit_operation import EditOperation
 from .learner import CostLearner
 from .levenshtein import WeightedLevenshtein
 from .matching import find_best_candidate
 
 __all__ = [
     "ocr_distance_map",
+    "EditOperation",
     "CostLearner",
     "WeightedLevenshtein",
     "find_best_candidate",

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/python/ocr_stringdist/edit_operation.py

@@ -1,5 +1,5 @@
-from dataclasses import dataclass
-from typing import Literal, Optional
+from dataclasses import asdict, dataclass
+from typing import Any, Literal, Optional
 
 OperationType = Literal["substitute", "insert", "delete", "match"]
 
@@ -14,3 +14,6 @@ class EditOperation:
     source_token: Optional[str]
     target_token: Optional[str]
     cost: float
+
+    def to_dict(self) -> dict[str, Any]:
+        return asdict(self)

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/python/ocr_stringdist/levenshtein.py

@@ -191,6 +191,7 @@ class WeightedLevenshtein:
         For the counterpart, see :meth:`WeightedLevenshtein.to_dict`.
 
         :param data: A dictionary with (not necessarily all of) the following keys:
+
                      - "substitution_costs": {"from": str, "to": str, "cost": float}
                      - "substitution_costs": dict[str, float]
                      - "deletion_costs": dict[str, float]
@@ -206,7 +207,7 @@ class WeightedLevenshtein:
 
         return cls(
             substitution_costs=sub_costs,
-            insertion_costs=data.get("substitution_costs"),
+            insertion_costs=data.get("insertion_costs"),
             deletion_costs=data.get("deletion_costs"),
             symmetric_substitution=data.get("symmetric_substitution", True),
             default_substitution_cost=data.get("default_substitution_cost", 1.0),

{ocr_stringdist-1.0.0 → ocr_stringdist-1.0.1}/python/tests/test_weighted_levenshtein.py

@@ -549,3 +549,13 @@ def test_costs_above_default_cost() -> None:
     )
     actual_cost = wl.distance("a", "b")
     assert actual_cost == configured_cost
+
+
+def test_serialization() -> None:
+    wl_orig = WeightedLevenshtein(
+        substitution_costs={("a", "b"): 0.5},
+        insertion_costs={"a": 0.5},
+        deletion_costs={"b": 0.5},
+    )
+    wl_loaded = WeightedLevenshtein.from_dict(wl_orig.to_dict())
+    assert wl_loaded == wl_orig