Merge branch 'main' into main

Author: ForeverAngry

poetry.lock

@@ -22,8 +22,10 @@
     DelimitedList,
     Group,
     MatchFirst,
+    ParseException,
     ParserElement,
     ParseResults,
+    QuotedString,
     Suppress,
     Word,
     alphanums,
@@ -79,7 +81,16 @@
 LIKE = CaselessKeyword("like")
 
 unquoted_identifier = Word(alphas + "_", alphanums + "_$")
-quoted_identifier = Suppress('"') + unquoted_identifier + Suppress('"')
+quoted_identifier = QuotedString('"', escChar="\\", unquoteResults=True)
+
+
+@quoted_identifier.set_parse_action
+def validate_quoted_identifier(result: ParseResults) -> str:
+    if "." in result[0]:
+        raise ParseException("Expected '\"', found '.'")
+    return result[0]
+
+
 identifier = MatchFirst([unquoted_identifier, quoted_identifier]).set_results_name("identifier")
 column = DelimitedList(identifier, delim=".", combine=False).set_results_name("column")
 

pyiceberg/expressions/parser.py

@@ -439,6 +439,15 @@ def update_snapshot(self, snapshot_properties: Dict[str, str] = EMPTY_DICT) -> U
         """
         return UpdateSnapshot(self, io=self._table.io, snapshot_properties=snapshot_properties)
 
+    def update_statistics(self) -> UpdateStatistics:
+        """
+        Create a new UpdateStatistics to update the statistics of the table.
+
+        Returns:
+            A new UpdateStatistics
+        """
+        return UpdateStatistics(transaction=self)
+
     def append(self, df: pa.Table, snapshot_properties: Dict[str, str] = EMPTY_DICT) -> None:
         """
         Shorthand API for appending a PyArrow table to a table transaction.

pyiceberg/table/__init__.py

@@ -234,8 +234,8 @@ class BucketTransform(Transform[S, int]):
     _num_buckets: PositiveInt = PrivateAttr()
 
     def __init__(self, num_buckets: int, **data: Any) -> None:
-        self._num_buckets = num_buckets
         super().__init__(f"bucket[{num_buckets}]", **data)
+        self._num_buckets = num_buckets
 
     @property
     def num_buckets(self) -> int:

pyiceberg/transforms.py

@@ -230,9 +230,11 @@ def test_quoted_column_with_dots() -> None:
     with pytest.raises(ParseException) as exc_info:
         parser.parse("\"foo.bar\".baz = 'data'")
 
-    assert "Expected '\"', found '.'" in str(exc_info.value)
-
     with pytest.raises(ParseException) as exc_info:
         parser.parse("'foo.bar'.baz = 'data'")
 
     assert "Expected <= | <> | < | >= | > | == | = | !=, found '.'" in str(exc_info.value)
+
+
+def test_quoted_column_with_spaces() -> None:
+    assert EqualTo("Foo Bar", "data") == parser.parse("\"Foo Bar\" = 'data'")

tests/expressions/test_parser.py

@@ -82,3 +82,10 @@ def create_statistics_file(snapshot_id: int, type_name: str) -> StatisticsFile:
         update.remove_statistics(add_snapshot_id_1)
 
     assert len(tbl.metadata.statistics) == 1
+
+    with tbl.transaction() as txn:
+        with txn.update_statistics() as update:
+            update.set_statistics(statistics_file_snap_1)
+            update.set_statistics(statistics_file_snap_2)
+
+    assert len(tbl.metadata.statistics) == 2