types: drop redundant casts in html_cleanup (bs4 stubs present)

Author: dmsfiris

pyproject.toml

@@ -11,7 +11,8 @@ requires-python = ">=3.10"
 authors = [{ name = "Your Name", email = "you@example.com" }]
 license = { file = "LICENSE" }
 dependencies = [
-  "pdfminer.six>=20221105"
+  "pdfminer.six>=20221105",
+  "beautifulsoup4>=4.12",
 ]
 classifiers = [
   "License :: OSI Approved :: Apache Software License",
@@ -42,6 +43,7 @@ dev = [
   "mkdocs>=1.5",
   "mkdocs-material>=9.5",
   "build>=1.2.1",
+  "types-beautifulsoup4>=4.12",
 ]
 
 [project.scripts]

src/fek_extractor/utils/html_cleanup.py

@@ -271,7 +271,7 @@ def tidy_article_html(html: str) -> str:
     if not html or not isinstance(html, str):
         return html or ""
 
-    soup = BeautifulSoup(html, "html.parser")
+    soup: BeautifulSoup = BeautifulSoup(html, "html.parser")
 
     # Use the body if BS4 wrapped our fragment; else use soup itself
     root_pe: PageElement = soup.body if soup.body else soup