Merge pull request #2235 from mabel-dev/0.19.1-release

0.19.1 release

.gitignore

@@ -180,3 +180,4 @@ space_missions.parquet
 **.psv
 planets.parquet
 tmp/iceberg/**
+hits_split/*.parquet

opteryx/__init__.py

@@ -32,7 +32,7 @@
 getcontext().prec = 28
 
 # end-of-stream marker
-EOS = object()
+EOS: int = 0
 
 
 def is_mac() -> bool:  # pragma: no cover

opteryx/__version__.py

@@ -1,4 +1,4 @@
-__build__ = 976
+__build__ = 979
 
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -22,7 +22,7 @@ class VersionStatus(Enum):
 _major = 0
 _minor = 19
 _revision = 1
-_status = VersionStatus.ALPHA
+_status = VersionStatus.RELEASE
 
 __author__ = "@joocer"
 __version__ = f"{_major}.{_minor}.{_revision}" + (

opteryx/connectors/disk_connector.py

@@ -31,6 +31,8 @@
 # Define os.O_BINARY for non-Windows platforms if it's not already defined
 if not hasattr(os, "O_BINARY"):
     os.O_BINARY = 0  # Value has no effect on non-Windows platforms
+if not hasattr(os, "O_DIRECT"):
+    os.O_DIRECT = 0  # Value has no effect on non-Windows platforms
 
 
 def read_blob(
@@ -69,7 +71,7 @@ def read_blob(
     import mmap
 
     try:
-        file_descriptor = os.open(blob_name, os.O_RDONLY | os.O_BINARY)
+        file_descriptor = os.open(blob_name, os.O_RDONLY | os.O_BINARY | os.O_DIRECT)
         if hasattr(os, "posix_fadvise"):
             os.posix_fadvise(file_descriptor, 0, 0, os.POSIX_FADV_WILLNEED)
         size = os.fstat(file_descriptor).st_size

opteryx/operators/show_columns_node.py

@@ -77,18 +77,8 @@ def execute(self, morsel: pyarrow.Table, **kwargs) -> pyarrow.Table:
         if self._full or self._extended:
             # we're going to read the full table, so we can count stuff
 
-            if morsel == EOS:
-                dicts = self.collector.to_dicts()
-                dicts = [self.rename_column(d, self._column_map) for d in dicts]
-                self.seen = True
-                yield pyarrow.Table.from_pylist(dicts)
-                return
+            self.statistics.add_message("SHOW FULL/SHOW EXTENDED not implemented")
 
-            df = DataFrame.from_arrow(morsel)
-
-            if self.collector is None:
-                self.collector = df.profile
-            else:
-                self.collector += df.profile
-
-            yield None
+            self.seen = True
+            yield _simple_collector(self._schema)
+            return

tests/sql_battery/test_shapes_and_errors_battery.py

@@ -1033,10 +1033,10 @@
         ("SELECT name, id FROM $planets ORDER BY id LIMIT 100", 9, 2, None),
 
         ("SHOW COLUMNS FROM $satellites", 8, 4, None),
-        ("SHOW FULL COLUMNS FROM $satellites", 8, 12, None),
-        ("SHOW EXTENDED COLUMNS FROM $satellites", 8, 12, None),
-        ("SHOW EXTENDED COLUMNS FROM $planets", 20, 12, None),
-        ("SHOW EXTENDED COLUMNS FROM $astronauts", 19, 12, None),
+        ("SHOW FULL COLUMNS FROM $satellites", 8, 4, None),
+        ("SHOW EXTENDED COLUMNS FROM $satellites", 8, 4, None),
+        ("SHOW EXTENDED COLUMNS FROM $planets", 20, 4, None),
+        ("SHOW EXTENDED COLUMNS FROM $astronauts", 19, 4, None),
         ("SHOW COLUMNS FROM $satellites LIKE '%d'", 2, 4, UnsupportedSyntaxError),
         ("SHOW COLUMNS FROM testdata.partitioned.dated FOR '2024-02-03'", 8, 4, None),
 
@@ -2407,8 +2407,8 @@ def test_sql_battery(statement:str, rows:int, columns:int, exception: Optional[E
             print(">", err)
             failures.append((statement, err))
 
-#            print(opteryx.query(statement))
-#            raise err
+            print(opteryx.query(statement))
+            raise err
 
     print("--- ✅ \033[0;32mdone\033[0m")