Truncate - length > width

Author: rambleraptor

pyiceberg/transforms.py

@@ -812,6 +812,10 @@ def project(self, name: str, pred: BoundPredicate) -> UnboundPredicate | None:
             if isinstance(pred, BoundLiteralPredicate):
                 return _truncate_number(name, pred, self.transform(field_type))
         elif isinstance(field_type, (BinaryType, StringType)):
+            if isinstance(pred, BoundNotStartsWith) and len(pred.literal.value) > self.width:
+                # A prefix longer than the width can't be projected: the truncated partition
+                # holds both matching and non-matching rows, so it cannot be pruned.
+                return None
             if isinstance(pred, BoundLiteralPredicate):
                 return _truncate_array(name, pred, self.transform(field_type))
 

tests/test_transforms.py

@@ -1026,10 +1026,21 @@ def test_projection_truncate_string_starts_with(bound_reference_str: BoundRefere
 
 
 def test_projection_truncate_string_not_starts_with(bound_reference_str: BoundReference) -> None:
+    # shorter than width: projects to not-starts-with on the untruncated prefix
     assert TruncateTransform(2).project(
-        "name", BoundNotStartsWith(term=bound_reference_str, literal=literal("hello"))
+        "name", BoundNotStartsWith(term=bound_reference_str, literal=literal("h"))
+    ) == NotStartsWith(term="name", literal=literal("h"))
+
+    # equal to width: projects to not-starts-with on the full prefix
+    assert TruncateTransform(2).project(
+        "name", BoundNotStartsWith(term=bound_reference_str, literal=literal("he"))
     ) == NotStartsWith(term="name", literal=literal("he"))
 
+    # longer than width: can't be projected, so the partition is always read
+    assert (
+        TruncateTransform(2).project("name", BoundNotStartsWith(term=bound_reference_str, literal=literal("hello"))) is None
+    )
+
 
 def _test_projection(lhs: UnboundPredicate | None, rhs: UnboundPredicate | None) -> None:
     assert type(lhs) is type(lhs), f"Different classes: {type(lhs)} != {type(rhs)}"