fix: handle pathlib.Path in get_pdf_name to avoid UnboundLocalError

pageindex/index/utils.py

@@ -741,13 +741,15 @@ def sanitize_filename(filename, replacement='-'):
 
 def get_pdf_name(pdf_path):
     # Extract PDF name
-    if isinstance(pdf_path, str):
+    if isinstance(pdf_path, (str, Path)):
         pdf_name = os.path.basename(pdf_path)
     elif isinstance(pdf_path, BytesIO):
         pdf_reader = PyPDF2.PdfReader(pdf_path)
         meta = pdf_reader.metadata
         pdf_name = meta.title if meta and meta.title else 'Untitled'
         pdf_name = sanitize_filename(pdf_name)
+    else:
+        pdf_name = 'Untitled'
     return pdf_name
 
 

tests/test_get_pdf_name.py

@@ -0,0 +1,21 @@
+from pathlib import Path
+
+from pageindex.index.utils import get_pdf_name
+
+
+def test_get_pdf_name_str():
+    assert get_pdf_name("/some/path/document.pdf") == "document.pdf"
+
+
+def test_get_pdf_name_pathlib_path():
+    # Previously raised UnboundLocalError — Path is neither str nor BytesIO
+    assert get_pdf_name(Path("/some/path/document.pdf")) == "document.pdf"
+
+
+def test_get_pdf_name_pathlib_relative():
+    assert get_pdf_name(Path("reports/annual.pdf")) == "annual.pdf"
+
+
+def test_get_pdf_name_unknown_type_returns_untitled():
+    # Any unrecognised type should not crash — returns a safe fallback
+    assert get_pdf_name(12345) == "Untitled"