Update type signatures in line with normality 3 logic

Author: pudo

LICENSE

@@ -1,6 +1,6 @@
 The MIT License (MIT)
 
-Copyright (c) 2024 OpenSanctions Datenbanken GmbH
+Copyright (c) 2025 OpenSanctions Datenbanken GmbH
 Copyright (c) 2018 Journalism Development Network, Inc.
 Copyright (c) 2016 Friedrich Lindenberg
 

fingerprints/cleanup.py

@@ -2,7 +2,7 @@
 import logging
 from typing import Optional
 from functools import lru_cache
-from normality import collapse_spaces, ascii_text, category_replace
+from normality import squash_spaces, ascii_text, category_replace
 
 from fingerprints.constants import WS, BRACKETED
 
@@ -68,23 +68,23 @@ def clean_name_ascii(text: Optional[str]) -> Optional[str]:
     It transliterates the text to ASCII, removes punctuation and symbols, converts the
     text to lowercase, replaces certain character categories, and collapses consecutive
     spaces.
-    
+
     Args:
         text (Optional[str]): The input text to be cleaned.
-    
+
     Returns:
         Optional[str]: The cleaned text, or None if the cleaned text is empty or too short.
     """
     # transliterate to ascii
-    text = ascii_text(text)
     if text is None:
         return None
+    text = ascii_text(text)
     # replace punctuation and symbols
     text = CHARACTERS_REMOVE_RE.sub("", text)
     text = text.lower()
     cleaned = category_replace(text)
-    cleaned = collapse_spaces(cleaned)
-    if cleaned is None or len(cleaned) < 2:
+    cleaned = squash_spaces(cleaned)
+    if len(cleaned) < 2:
         return None
     return cleaned
 
@@ -96,7 +96,7 @@ def clean_name_light(text: str) -> Optional[str]:
     text = CHARACTERS_REMOVE_RE.sub("", text)
     text = text.lower()
     cleaned = category_replace(text)
-    cleaned = collapse_spaces(cleaned)
-    if cleaned is None or len(cleaned) < 2:
+    cleaned = squash_spaces(cleaned)
+    if len(cleaned) < 2:
         return None
     return cleaned

fingerprints/fingerprint.py

@@ -1,6 +1,6 @@
 import logging
 from typing import Optional
-from normality import collapse_spaces, stringify
+from normality import squash_spaces, stringify
 
 from fingerprints.constants import WS
 from fingerprints.types import replace_types
@@ -26,16 +26,18 @@ def fingerprint(
 
     # Super hard-core string scrubbing
     text = clean_name_ascii(text)
+    if text is None:
+        return None
     text = replace_types(text)
 
     if keep_order:
-        text = collapse_spaces(text)
-    elif text is not None:
+        text = squash_spaces(text)
+    else:
         # final manicure, based on openrefine algo
         parts = [p for p in text.split(WS) if len(p)]
         text = WS.join(sorted(set(parts)))
 
-    if text is None or not len(text):
+    if not len(text):
         return None
 
     return text

fingerprints/types/__init__.py

@@ -1,22 +1,21 @@
-from typing import Optional
-from normality import collapse_spaces
+from normality import squash_spaces
 
 from fingerprints.cleanup import clean_name_ascii
 from fingerprints.types.replacer import get_replacer, NormFunc
 
 
-def replace_types(text: Optional[str]) -> Optional[str]:
+def replace_types(text: str) -> str:
     """Chomp down company types to a more convention form."""
     return get_replacer()(text)
 
 
-def remove_types(
-    text: Optional[str], clean: NormFunc = clean_name_ascii
-) -> Optional[str]:
+def remove_types(text: str, clean: NormFunc = clean_name_ascii) -> str:
     """Remove company type names from a piece of text.
 
     WARNING: This converts to ASCII by default, pass in a different
     `clean` function if you need a different behaviour."""
-    text = clean(text)
-    removed = get_replacer(clean, remove=True)(text)
-    return collapse_spaces(removed)
+    cleaned = clean(text)
+    if cleaned is None:
+        return ""
+    removed = get_replacer(clean, remove=True)(cleaned)
+    return squash_spaces(removed)

fingerprints/types/replacer.py

@@ -1,15 +1,15 @@
 import re
 import logging
 from functools import lru_cache
-from typing import Callable, Optional, Dict, Match
+from typing import Callable, Dict, Match, Optional
 
 from fingerprints.types.data import TYPES
 from fingerprints.constants import WS
 from fingerprints.cleanup import clean_name_ascii
 
 log = logging.getLogger(__name__)
-NormFunc = Callable[[Optional[str]], Optional[str]]
-ReplaceFunc = Callable[[Optional[str]], Optional[str]]
+NormFunc = Callable[[str], Optional[str]]
+ReplaceFunc = Callable[[str], str]
 
 
 class Replacer(object):
@@ -28,9 +28,7 @@ def get_canonical(self, match: Match[str]) -> str:
             return WS
         return self.replacements.get(match.group(1), match.group(1))
 
-    def __call__(self, text: Optional[str]) -> Optional[str]:
-        if text is None:
-            return None
+    def __call__(self, text: str) -> str:
         return self.matcher.sub(self.get_canonical, text)
 
 

pyproject.toml

@@ -17,7 +17,7 @@ classifiers = [
     "Programming Language :: Python :: 3.12",
 ]
 requires-python = ">= 3.10"
-dependencies = ["normality >= 2.5.0, < 3.0.0"]
+dependencies = ["normality >= 3.0.1, < 4.0.0"]
 
 [project.urls]
 Documentation = "https://github.com/opensanctions/fingerprints/"