fix: Ignore ncx and use lxml-xml option to suppress warning

p0n1 · Jun 28, 2024 · 74d1644 · 74d1644
1 parent c11d9a3
commit 74d1644
Showing 1 changed file with 2 additions and 2 deletions.
diff --git a/audiobook_generator/book_parsers/epub_book_parser.py b/audiobook_generator/book_parsers/epub_book_parser.py
@@ -16,7 +16,7 @@ class EpubBookParser(BaseBookParser):
     def __init__(self, config: GeneralConfig):
         super().__init__(config)
         logger.setLevel(config.log)
-        self.book = epub.read_epub(self.config.input_file)
+        self.book = epub.read_epub(self.config.input_file, {"ignore_ncx": True})
 
     def __str__(self) -> str:
         return super().__str__()
@@ -44,7 +44,7 @@ def get_chapters(self, break_string) -> List[Tuple[str, str]]:
         chapters = []
         for item in self.book.get_items_of_type(ebooklib.ITEM_DOCUMENT):
             content = item.get_content()
-            soup = BeautifulSoup(content, "lxml")
+            soup = BeautifulSoup(content, "lxml-xml")
             raw = soup.get_text(strip=False)
             logger.debug(f"Raw text: <{raw[:]}>")