perf(cleaning): combine string cleaning rules into single-pass execution

Author: Medalcode

src/cleaning.py

@@ -143,25 +143,28 @@ def clean_data(
         df = df.dropna(how="all")
         logger.info(f"Dropped all-null rows → {len(df):,} rows remaining")
 
-    # 3. Strip whitespace from strings
-    # NOTE: We use a value-level map instead of col.str.strip() to avoid
-    # silently converting non-string values (e.g. ints in object columns)
-    # to NaN. Only actual str instances are stripped; all other types are
-    # left untouched.
-    if config.get("strip_strings", True):
-        obj_cols = _cat_cols(df)
-        df[obj_cols] = df[obj_cols].apply(
-            lambda col: col.map(lambda v: v.strip() if isinstance(v, str) else v)
-        )
-        logger.debug(f"Stripped whitespace from {len(obj_cols)} string columns")
-
-    # 4. Lowercase string columns
-    # Same rationale: apply lower() only to actual str values.
-    if config.get("lowercase_strings", False):
+    # 3 & 4. String cleaning (strip whitespace / lowercase)
+    # NOTE: We use value-level transformation to preserve non-string types
+    # (e.g. ints/floats in object columns) without coercing to NaN.
+    do_strip = config.get("strip_strings", True)
+    do_lower = config.get("lowercase_strings", False)
+
+    if do_strip or do_lower:
         obj_cols = _cat_cols(df)
-        df[obj_cols] = df[obj_cols].apply(
-            lambda col: col.map(lambda v: v.lower() if isinstance(v, str) else v)
-        )
+        if obj_cols:
+            def _clean_val(v):
+                if isinstance(v, str):
+                    if do_strip:
+                        v = v.strip()
+                    if do_lower:
+                        v = v.lower()
+                return v
+
+            df[obj_cols] = df[obj_cols].apply(lambda col: col.map(_clean_val))
+            if do_strip:
+                logger.debug(f"Stripped whitespace from {len(obj_cols)} string columns")
+            if do_lower:
+                logger.debug(f"Lowercased {len(obj_cols)} string columns")
 
     # 5. Drop duplicates
     if config.get("drop_duplicates", True):