Merge branch 'main' into shuowei-blob-image

Author: shuoweil

bigframes/core/compile/sqlglot/compiler.py

@@ -19,6 +19,7 @@
 
 import bigframes_vendored.sqlglot.expressions as sge
 
+from bigframes import dtypes
 from bigframes.core import (
     expression,
     guid,
@@ -277,6 +278,24 @@ def compile_explode(node: nodes.ExplodeNode, child: ir.SQLGlotIR) -> ir.SQLGlotI
     return child.explode(columns, offsets_col)
 
 
+@_compile_node.register
+def compile_fromrange(
+    node: nodes.FromRangeNode, start: ir.SQLGlotIR, end: ir.SQLGlotIR
+) -> ir.SQLGlotIR:
+    start_col_id = node.start.fields[0].id
+    end_col_id = node.end.fields[0].id
+
+    start_expr = expression_compiler.expression_compiler.compile_expression(
+        expression.DerefOp(start_col_id)
+    )
+    end_expr = expression_compiler.expression_compiler.compile_expression(
+        expression.DerefOp(end_col_id)
+    )
+    step_expr = ir._literal(node.step, dtypes.INT_DTYPE)
+
+    return start.resample(end, node.output_id.sql, start_expr, end_expr, step_expr)
+
+
 @_compile_node.register
 def compile_random_sample(
     node: nodes.RandomSampleNode, child: ir.SQLGlotIR

bigframes/core/compile/sqlglot/sqlglot_ir.py

@@ -134,18 +134,20 @@ def from_table(
         """
         version = (
             sge.Version(
-                this="TIMESTAMP",
-                expression=sge.Literal(this=system_time.isoformat(), is_string=True),
+                this=sge.Identifier(this="SYSTEM_TIME", quoted=False),
+                expression=sge.Literal.string(system_time.isoformat()),
                 kind="AS OF",
             )
             if system_time
             else None
         )
+        table_alias = next(uid_gen.get_uid_stream("bft_"))
         table_expr = sge.Table(
             this=sg.to_identifier(table_id, quoted=cls.quoted),
             db=sg.to_identifier(dataset_id, quoted=cls.quoted),
             catalog=sg.to_identifier(project_id, quoted=cls.quoted),
             version=version,
+            alias=sge.Identifier(this=table_alias, quoted=cls.quoted),
         )
         if sql_predicate:
             select_expr = sge.Select().select(sge.Star()).from_(table_expr)
@@ -410,6 +412,47 @@ def aggregate(
             new_expr = new_expr.where(condition, append=False)
         return SQLGlotIR(expr=new_expr, uid_gen=self.uid_gen)
 
+    def resample(
+        self,
+        right: SQLGlotIR,
+        array_col_name: str,
+        start_expr: sge.Expression,
+        stop_expr: sge.Expression,
+        step_expr: sge.Expression,
+    ) -> SQLGlotIR:
+        # Get identifier for left and right by pushing them to CTEs
+        left_select, left_id = self._select_to_cte()
+        right_select, right_id = right._select_to_cte()
+
+        # Extract all CTEs from the returned select expressions
+        _, left_ctes = _pop_query_ctes(left_select)
+        _, right_ctes = _pop_query_ctes(right_select)
+        merged_ctes = _merge_ctes(left_ctes, right_ctes)
+
+        generate_array = sge.func("GENERATE_ARRAY", start_expr, stop_expr, step_expr)
+
+        unnested_column_alias = sge.to_identifier(
+            next(self.uid_gen.get_uid_stream("bfcol_")), quoted=self.quoted
+        )
+        unnest_expr = sge.Unnest(
+            expressions=[generate_array],
+            alias=sge.TableAlias(columns=[unnested_column_alias]),
+        )
+
+        final_col_id = sge.to_identifier(array_col_name, quoted=self.quoted)
+
+        # Build final expression by joining everything directly in a single SELECT
+        new_expr = (
+            sge.Select()
+            .select(unnested_column_alias.as_(final_col_id))
+            .from_(sge.Table(this=left_id))
+            .join(sge.Table(this=right_id), join_type="cross")
+            .join(unnest_expr, join_type="cross")
+        )
+        new_expr = _set_query_ctes(new_expr, merged_ctes)
+
+        return SQLGlotIR(expr=new_expr, uid_gen=self.uid_gen)
+
     def insert(
         self,
         destination: bigquery.TableReference,

setup.py

@@ -122,12 +122,18 @@
     name=name,
     version=version_id,
     description=description,
+    download_url="https://github.com/googleapis/python-bigquery-dataframes/releases",
     long_description=readme,
     long_description_content_type="text/x-rst",
     author="Google LLC",
     author_email="bigframes-feedback@google.com",
     license="Apache 2.0",
-    url="https://github.com/googleapis/python-bigquery-dataframes",
+    url="https://dataframes.bigquery.dev",
+    project_urls={
+        "Source": "https://github.com/googleapis/python-bigquery-dataframes",
+        "Changelog": "https://dataframes.bigquery.dev/changelog.html",
+        "Issues": "https://github.com/googleapis/python-bigquery-dataframes/issues",
+    },
     classifiers=[
         release_status,
         "Intended Audience :: Developers",

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_binary_compiler/test_corr/out.sql

@@ -2,7 +2,7 @@ WITH `bfcte_0` AS (
   SELECT
     `int64_col`,
     `float64_col`
-  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types` AS `bft_0`
 ), `bfcte_1` AS (
   SELECT
     CORR(`int64_col`, `float64_col`) AS `bfcol_2`

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_binary_compiler/test_cov/out.sql

@@ -2,7 +2,7 @@ WITH `bfcte_0` AS (
   SELECT
     `int64_col`,
     `float64_col`
-  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types` AS `bft_0`
 ), `bfcte_1` AS (
   SELECT
     COVAR_SAMP(`int64_col`, `float64_col`) AS `bfcol_2`

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_nullary_compiler/test_row_number/out.sql

@@ -1,3 +1,3 @@
 SELECT
   ROW_NUMBER() OVER () - 1 AS `row_number`
-FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+FROM `bigframes-dev`.`sqlglot_test`.`scalar_types` AS `bft_0`

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_nullary_compiler/test_row_number_with_window/out.sql

@@ -1,3 +1,3 @@
 SELECT
   ROW_NUMBER() OVER (ORDER BY `int64_col` ASC NULLS LAST) - 1 AS `row_number`
-FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+FROM `bigframes-dev`.`sqlglot_test`.`scalar_types` AS `bft_0`

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_nullary_compiler/test_size/out.sql

@@ -1,7 +1,7 @@
 WITH `bfcte_0` AS (
   SELECT
     *
-  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types` AS `bft_0`
 ), `bfcte_1` AS (
   SELECT
     COUNT(1) AS `bfcol_32`

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_ordered_unary_compiler/test_array_agg/out.sql

@@ -1,7 +1,7 @@
 WITH `bfcte_0` AS (
   SELECT
     `int64_col`
-  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types` AS `bft_0`
 ), `bfcte_1` AS (
   SELECT
     ARRAY_AGG(`int64_col` IGNORE NULLS ORDER BY `int64_col` IS NULL ASC, `int64_col` ASC) AS `bfcol_1`

tests/unit/core/compile/sqlglot/aggregations/snapshots/test_ordered_unary_compiler/test_string_agg/out.sql

@@ -1,7 +1,7 @@
 WITH `bfcte_0` AS (
   SELECT
     `string_col`
-  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types`
+  FROM `bigframes-dev`.`sqlglot_test`.`scalar_types` AS `bft_0`
 ), `bfcte_1` AS (
   SELECT
     COALESCE(