fix: replace _converter_lock with timeout-based acquisition (C2)

A frozen conversion holding the lock indefinitely blocks all subsequent
jobs. Using lock.acquire(timeout=300) fails fast with a clear error
instead of waiting forever.

Ref #57 (C2)

document-parser/infra/local_converter.py

@@ -46,8 +46,10 @@ from infra.settings import settings
 
 logger = logging.getLogger(__name__)
 
-# Thread lock — DoclingConverter is not thread-safe
+# Thread lock — DoclingConverter is not thread-safe.
+# Uses a timeout to prevent a frozen conversion from blocking all others.
 _converter_lock = threading.Lock()
+_LOCK_TIMEOUT = 300  # seconds — fail fast rather than wait forever
 
 # US Letter page dimensions (points) — fallback when page size is unknown
 _DEFAULT_PAGE_WIDTH = 612.0
@@ -212,9 +214,17 @@ def _process_content_item(
 
 
 def _convert_sync(file_path: str, options: ConversionOptions) -> ConversionResult:
-    with _converter_lock:
+    acquired = _converter_lock.acquire(timeout=_LOCK_TIMEOUT)
+    if not acquired:
+        raise TimeoutError(
+            f"Could not acquire converter lock within {_LOCK_TIMEOUT}s — "
+            "a previous conversion may be frozen"
+        )
+    try:
         conv = _select_converter(options)
         result = conv.convert(file_path)
+    finally:
+        _converter_lock.release()
 
     doc = result.document
     page_count = len(doc.pages)