Fixed warnings from pylint

db.py

@@ -1,3 +1,7 @@
+#pylint: disable=missing-class-docstring,invalid-name,broad-exception-caught
+"""
+Database module for semantic document search tool.
+"""
 import pickle
 from pathlib import Path
 from dataclasses import dataclass
@@ -6,13 +10,13 @@ from concurrent.futures import ThreadPoolExecutor, as_completed
 
 import numpy as np
 import pymupdf
-import ollama  # TODO split to another file
+import ollama
 
 #
 # Types
 #
 
-type Vector = np.NDArray  # np.NDArray[np.float32] ?
+type Vector = np.NDArray
 type VectorBytes = bytes
 
 
@@ -90,12 +94,29 @@ def _vectorize_record(record: Record) -> tuple[Record, Vector]:
     return record, _embed(record.text)
 
 
+def test_embedding() -> bool:
+    """
+    Test if embedding functionality is available and working.
+    
+    Returns:
+        bool: True if embedding is working, False otherwise
+    """
+    try:
+        _ = _embed("Test.")
+        return True
+    except Exception:
+        return False
+
+
 #
 # High-level (exported) functions
 #
 
 
 def create_dummy() -> Database:
+    """
+    Create a dummy database for testing purposes.
+    """
     db_length: Final[int] = 10
     vectors = [np.array([i, 2 * i, 3 * i, 4 * i]) for i in range(db_length)]
     records = {
@@ -250,13 +271,11 @@ def add_document(db: Database | Path, file: Path, max_workers: int = 4) -> None:
         records: list[Record] = []
         chunk_size = 1024
 
-        for page_num in range(len(doc)):
-            page = doc[page_num]
+        for page_num, page in enumerate(doc):
             text = page.get_text().strip()
             if not text:
                 print(f"  Page {page_num + 1}: Skipped (empty)")
                 continue
-
             # Simple chunking - split text into chunks of specified size
             for chunk_idx, i in enumerate(range(0, len(text), chunk_size)):
                 chunk = text[i : i + chunk_size]
@@ -267,14 +286,14 @@ def add_document(db: Database | Path, file: Path, max_workers: int = 4) -> None:
                     )
         doc.close()
     except Exception as e:
-        raise RuntimeError(f"Error processing PDF {file}: {e}")
+        raise RuntimeError(f"Error processing PDF {file}: {e}") from e
 
     # Process chunks in parallel
     print(f"Processing {len(records)} chunks with {max_workers} workers...")
 
     db.documents.append(file)
 
-    # TODO measure with GIL disabled to check if multithreading actually helps
+    # NOTE this will only help with GIL disabled
     with ThreadPoolExecutor(max_workers=max_workers) as pool:
         futures = [pool.submit(_vectorize_record, r) for r in records]
         for f in as_completed(futures):