diff --git a/image_prediction/image_extractor/extractors/parsable.py b/image_prediction/image_extractor/extractors/parsable.py
index 8e80390..5fe0a96 100644
--- a/image_prediction/image_extractor/extractors/parsable.py
+++ b/image_prediction/image_extractor/extractors/parsable.py
@@ -65,7 +65,7 @@ def take_right(pair: Either):
 
 
 def format_context(context):
-    return f"Reason: {context['reason']}. Metadata: {EnumFormatter()(context['metadata'])}"
+    return f"Reason: {context['reason'].rstrip('.')}. Metadata: {EnumFormatter()(context['metadata'])}"
 
 
 def extract_pages(doc, page_range):
@@ -131,16 +131,29 @@ def xref_to_maybe_image(doc, xref) -> Either:
 
 
 def make_maybe_image_metadata_pair(image: Either, metadata: Either):
-    """
-    Reference:
-        haskell.org/tutorial/monads.html
-        (>>) :: m a -> m b -> m b
-    """
+    """Reference: haskell.org/tutorial/monads.html"""
 
     def context(value):
         return {"reason": value, "metadata": metadata.either(bottom, identity)}
 
-    return Right(make_image_metadata_pair).amap(image).amap(metadata).either(left(context), right(identity))
+    # What is most readable?
+
+    # 1)
+    # return Either.apply(make_image_metadata_pair).to_arguments(image, metadata).either(left(context), right(identity))
+
+    # 2)
+    # m (a -> b -> c) -> m a -> m b -> m c
+    # return Right(make_image_metadata_pair).amap(image).amap(metadata).either(left(context), right(identity))
+
+    # 3)
+    return (
+        image.bind(right(make_image_metadata_pair))  # m a >>= (m (a -> b -> c)) -> m (b -> c)
+        .amap(metadata)  # m (b -> c) <*> m b -> m c
+        .either(
+            left(context),
+            right(identity),
+        )
+    )
 
 
 @curry(2)