fix: return Option<&[u8]> instead of &Option<Vec<u8>> for key_metadata

* test: use `None` instead of `Some` for key_metadata fields

Author: feniljain

crates/iceberg/src/expr/visitors/expression_evaluator.rs

@@ -338,7 +338,7 @@ mod tests {
             nan_value_counts: HashMap::new(),
             lower_bounds: HashMap::new(),
             upper_bounds: HashMap::new(),
-            key_metadata: Some(vec![]),
+            key_metadata: None,
             split_offsets: vec![],
             equality_ids: vec![],
             sort_order_id: None,
@@ -361,7 +361,7 @@ mod tests {
             nan_value_counts: HashMap::new(),
             lower_bounds: HashMap::new(),
             upper_bounds: HashMap::new(),
-            key_metadata: Some(vec![]),
+            key_metadata: None,
             split_offsets: vec![],
             equality_ids: vec![],
             sort_order_id: None,

crates/iceberg/src/expr/visitors/inclusive_metrics_evaluator.rs

@@ -1991,7 +1991,7 @@ mod test {
             nan_value_counts: Default::default(),
             lower_bounds: Default::default(),
             upper_bounds: Default::default(),
-            key_metadata: Some(vec![]),
+            key_metadata: None,
             split_offsets: vec![],
             equality_ids: vec![],
             sort_order_id: None,
@@ -2012,7 +2012,7 @@ mod test {
             nan_value_counts: Default::default(),
             lower_bounds: Default::default(),
             upper_bounds: Default::default(),
-            key_metadata: Some(vec![]),
+            key_metadata: None,
             split_offsets: vec![],
             equality_ids: vec![],
             sort_order_id: None,
@@ -2069,7 +2069,7 @@ mod test {
             ]),
 
             column_sizes: Default::default(),
-            key_metadata: Some(vec![]),
+            key_metadata: None,
             split_offsets: vec![],
             equality_ids: vec![],
             sort_order_id: None,
@@ -2095,7 +2095,7 @@ mod test {
             upper_bounds: HashMap::from([(3, Datum::string("dC"))]),
 
             column_sizes: Default::default(),
-            key_metadata: Some(vec![]),
+            key_metadata: None,
             split_offsets: vec![],
             equality_ids: vec![],
             sort_order_id: None,
@@ -2122,7 +2122,7 @@ mod test {
             upper_bounds: HashMap::from([(3, Datum::string("3str3"))]),
 
             column_sizes: Default::default(),
-            key_metadata: Some(vec![]),
+            key_metadata: None,
             split_offsets: vec![],
             equality_ids: vec![],
             sort_order_id: None,
@@ -2149,7 +2149,7 @@ mod test {
             upper_bounds: HashMap::from([(3, Datum::string("イロハニホヘト"))]),
 
             column_sizes: Default::default(),
-            key_metadata: Some(vec![]),
+            key_metadata: None,
             split_offsets: vec![],
             equality_ids: vec![],
             sort_order_id: None,

crates/iceberg/src/spec/manifest.rs

@@ -1134,8 +1134,8 @@ impl DataFile {
         &self.upper_bounds
     }
     /// Get the Implementation-specific key metadata for the data file.
-    pub fn key_metadata(&self) -> &Option<Vec<u8>> {
-        &self.key_metadata
+    pub fn key_metadata(&self) -> Option<&[u8]> {
+        self.key_metadata.as_ref().map(|x| &x as &[u8])
     }
     /// Get the split offsets of the data file.
     /// For example, all row group offsets in a Parquet file.
@@ -1628,7 +1628,7 @@ mod tests {
                         nan_value_counts: HashMap::new(),
                         lower_bounds: HashMap::new(),
                         upper_bounds: HashMap::new(),
-                        key_metadata: Some(Vec::new()),
+                        key_metadata: None,
                         split_offsets: vec![4],
                         equality_ids: Vec::new(),
                         sort_order_id: None,
@@ -1784,7 +1784,7 @@ mod tests {
                     nan_value_counts: HashMap::new(),
                     lower_bounds: HashMap::new(),
                     upper_bounds: HashMap::new(),
-                    key_metadata: Some(vec![]),
+                    key_metadata: None,
                     split_offsets: vec![4],
                     equality_ids: vec![],
                     sort_order_id: None,
@@ -1851,7 +1851,7 @@ mod tests {
                     nan_value_counts: HashMap::new(),
                     lower_bounds: HashMap::from([(1,Datum::int(1)),(2,Datum::string("a")),(3,Datum::string("AC/DC"))]),
                     upper_bounds: HashMap::from([(1,Datum::int(1)),(2,Datum::string("a")),(3,Datum::string("AC/DC"))]),
-                    key_metadata: Some(vec![]),
+                    key_metadata: None,
                     split_offsets: vec![4],
                     equality_ids: vec![],
                     sort_order_id: Some(0),
@@ -1931,7 +1931,7 @@ mod tests {
                         (2, Datum::string("a")),
                         (3, Datum::string("x"))
                         ]),
-                        key_metadata: Some(vec![]),
+                        key_metadata: None,
                         split_offsets: vec![4],
                         equality_ids: vec![],
                         sort_order_id: Some(0),
@@ -2006,7 +2006,7 @@ mod tests {
                         (2, Datum::int(2)),
                         (3, Datum::string("x"))
                     ]),
-                    key_metadata: Some(vec![]),
+                    key_metadata: None,
                     split_offsets: vec![4],
                     equality_ids: vec![],
                     sort_order_id: None,
@@ -2076,7 +2076,7 @@ mod tests {
                         (1, Datum::long(1)),
                         (2, Datum::int(2)),
                     ]),
-                    key_metadata: Some(vec![]),
+                    key_metadata: None,
                     split_offsets: vec![4],
                     equality_ids: vec![],
                     sort_order_id: None,