Merge branch 'dc/truncate' of https://github.com/dancardin/fakesnow i…

…nto dc/truncate

.cruft.json

@@ -1,6 +1,6 @@
 {
   "template": "[email protected]:tekumara/python-typed-template.git",
-  "commit": "0f4759d4049e0ec77f9b351e39197cc01e0cb770",
+  "commit": "49cb14fd299526b393f6257238c2baf1f3b59725",
   "checkout": null,
   "context": {
     "cookiecutter": {

.github/workflows/auto-merge.yml

@@ -0,0 +1,20 @@
+name: Dependabot Auto-merge
+on:
+  pull_request:
+    types: [opened]
+
+permissions:
+  contents: write
+  pull-requests: write
+
+jobs:
+  dependabot:
+    runs-on: ubuntu-latest
+    if: github.event.pull_request.user.login == 'dependabot[bot]' && github.repository == 'tekumara/fakesnow'
+    steps:
+      - name: Enable auto-merge for Dependabot PRs
+        # don't include verbose dependabot PR description as body when merging
+        run: gh pr merge --auto --squash "$PR_URL" --body ""
+        env:
+          PR_URL: ${{github.event.pull_request.html_url}}
+          GH_TOKEN: ${{secrets.GITHUB_TOKEN}}

.github/workflows/ci.yml

@@ -12,6 +12,8 @@ jobs:
     - name: Set up Python
       id: setup-python
       uses: actions/setup-python@v5
+    - name: Install uv
+      uses: astral-sh/setup-uv@v3
     - name: Cache virtualenv
       id: cache-venv
       uses: actions/cache@v4

.github/workflows/release-please.yml

@@ -21,9 +21,9 @@ jobs:
           # https://github.com/apps/potatobot-prime
           app-id: ${{ secrets.APP_ID }}
           private-key: ${{ secrets.APP_PRIVATE_KEY }}
-      - uses: google-github-actions/release-please-action@v4
+      - uses: googleapis/release-please-action@v4
         with:
           # use app token so that PRs and releases created by release-please trigger
           # additional workflows. They'll also be authored by the app.
-          # see https://github.com/google-github-actions/release-please-action#github-credentials
+          # see https://github.com/googleapis/release-please-action#github-credentials
           token: ${{ steps.app-token.outputs.token }}

.github/workflows/release.yml

@@ -18,8 +18,8 @@ jobs:
       - name: Set up Python 3.9
         id: setup-python
         uses: actions/setup-python@v5
-        with:
-          python-version: 3.9
+      - name: Install uv
+        uses: astral-sh/setup-uv@v3
       - name: Cache virtualenv
         id: cache-venv
         uses: actions/cache@v4

.pre-commit-config.yaml

@@ -1,8 +1,8 @@
 # unless otherwise specified, hooks run on push only
-default_stages: [push]
+default_stages: [pre-push]
 repos:
   - repo: https://github.com/crate-ci/typos
-    rev: v1.23.2
+    rev: v1.27.0
     hooks:
       - id: typos
   # formatters and linters are available in the virtualenv so they can be run from the makefile & vscode

Makefile-common.mk

@@ -1,4 +1,4 @@
-MAKEFLAGS += --warn-undefined-variables
+MAKEFLAGS += --warn-undefined-variables --check-symlink-times
 SHELL = /bin/bash -o pipefail
 .DEFAULT_GOAL := help
 .PHONY: help clean install format check pyright test dist hooks install-hooks
@@ -11,7 +11,8 @@ venv ?= .venv
 # this is a symlink so we set the --check-symlink-times makeflag above
 python := $(venv)/bin/python
 # use uv if present, else fall back to pip
-pip = $(shell command -v uv >/dev/null && echo "uv pip" || echo "$(venv)/bin/pip")
+# set VIRTUAL_ENV to avoid uv installing into a different activated venv
+pip = $(shell command -v uv >/dev/null && echo "VIRTUAL_ENV=$(venv) uv pip" || echo "$(venv)/bin/pip")
 
 $(python): $(if $(value CI),|,) .python-version
 # create venv using system python even when another venv is active
@@ -20,6 +21,7 @@ $(python): $(if $(value CI),|,) .python-version
 	$(pip) install --upgrade pip~=24.0
 
 $(venv): $(if $(value CI),|,) pyproject.toml $(python)
+# TODO use uv sync
 	$(pip) install -e '.[dev,server$(if $(value CI),,,notebook)]'
 	touch $(venv)
 
@@ -52,9 +54,9 @@ test: $(venv)
 
 ## build python distribution
 dist: $(venv)
-	# start with a clean slate (see setuptools/#2347)
-	rm -rf dist *.egg-info
-	$(venv)/bin/python -m build --sdist --wheel
+# start with a clean slate (see setuptools/#2347)
+	rm -rf build dist *.egg-info
+	$(venv)/bin/python -m build --wheel
 
 ## publish to pypi
 publish: $(venv)

fakesnow/cursor.py

@@ -162,6 +162,7 @@ def _transform(self, expression: exp.Expression) -> exp.Expression:
             .transform(transforms.extract_comment_on_columns)
             .transform(transforms.information_schema_fs_columns_snowflake)
             .transform(transforms.information_schema_fs_tables_ext)
+            .transform(transforms.information_schema_fs_views)
             .transform(transforms.drop_schema_cascade)
             .transform(transforms.tag)
             .transform(transforms.semi_structured_types)

fakesnow/info_schema.py

@@ -102,7 +102,7 @@
 # replicates https://docs.snowflake.com/sql-reference/info-schema/views
 SQL_CREATE_INFORMATION_SCHEMA_VIEWS_VIEW = Template(
     """
-create view if not exists ${catalog}.information_schema.views AS
+create view if not exists ${catalog}.information_schema._fs_views AS
 select
     database_name as table_catalog,
     schema_name as table_schema,

fakesnow/transforms.py

@@ -471,7 +471,6 @@ def flatten(expression: exp.Expression) -> exp.Expression:
 
     See https://docs.snowflake.com/en/sql-reference/functions/flatten
 
-    TODO: return index.
     TODO: support objects.
     """
     if (
@@ -483,20 +482,34 @@ def flatten(expression: exp.Expression) -> exp.Expression:
     ):
         explode_expression = expression.this.this.expression
 
-        return exp.Lateral(
-            this=exp.Unnest(
+        value = exp.Cast(
+            this=explode_expression,
+            to=exp.DataType(
+                this=exp.DataType.Type.ARRAY,
+                expressions=[exp.DataType(this=exp.DataType.Type.JSON, nested=False, prefix=False)],
+                nested=True,
+            ),
+        )
+
+        return exp.Subquery(
+            this=exp.Select(
                 expressions=[
-                    exp.Cast(
-                        this=explode_expression,
-                        to=exp.DataType(
-                            this=exp.DataType.Type.ARRAY,
-                            expressions=[exp.DataType(this=exp.DataType.Type.JSON, nested=False, prefix=False)],
-                            nested=True,
+                    exp.Unnest(
+                        expressions=[value],
+                        alias=exp.Identifier(this="VALUE", quoted=False),
+                    ),
+                    exp.Alias(
+                        this=exp.Sub(
+                            this=exp.Anonymous(
+                                this="generate_subscripts", expressions=[value, exp.Literal(this="1", is_string=False)]
+                            ),
+                            expression=exp.Literal(this="1", is_string=False),
                         ),
-                    )
+                        alias=exp.Identifier(this="INDEX", quoted=False),
+                    ),
                 ],
             ),
-            alias=exp.TableAlias(this=alias.this, columns=[exp.Identifier(this="VALUE", quoted=False)]),
+            alias=exp.TableAlias(this=alias.this),
         )
 
     return expression
@@ -622,6 +635,20 @@ def information_schema_fs_tables_ext(expression: exp.Expression) -> exp.Expressi
     return expression
 
 
+def information_schema_fs_views(expression: exp.Expression) -> exp.Expression:
+    """Use information_schema._fs_views to return Snowflake's version instead of duckdb's."""
+
+    if (
+        isinstance(expression, exp.Select)
+        and (tbl_exp := expression.find(exp.Table))
+        and tbl_exp.name.upper() == "VIEWS"
+        and tbl_exp.db.upper() == "INFORMATION_SCHEMA"
+    ):
+        tbl_exp.set("this", exp.Identifier(this="_FS_VIEWS", quoted=False))
+
+    return expression
+
+
 def integer_precision(expression: exp.Expression) -> exp.Expression:
     """Convert integers to bigint.
 

package.json

@@ -1,5 +1,5 @@
 {
     "devDependencies": {
-        "pyright": "1.1.382"
+        "pyright": "1.1.387"
     }
 }

pyproject.toml

@@ -8,7 +8,7 @@ classifiers = ["License :: OSI Approved :: MIT License"]
 keywords = ["snowflake", "snowflakedb", "fake", "local", "mock", "testing"]
 requires-python = ">=3.9"
 dependencies = [
-    "duckdb~=1.0.0",
+    "duckdb~=1.1.3",
     "pyarrow",
     "snowflake-connector-python",
     "sqlglot~=25.24.1",
@@ -28,13 +28,13 @@ dev = [
     "pandas-stubs",
     # include compatible version of pandas, and secure-local-storage for token caching
     "snowflake-connector-python[pandas, secure-local-storage]",
-    "pre-commit~=3.4",
+    "pre-commit~=4.0",
     "pyarrow-stubs==10.0.1.9",
     "pytest~=8.0",
     "pytest-asyncio",
-    "ruff~=0.6.3",
+    "ruff~=0.7.2",
     "twine~=5.0",
-    "snowflake-sqlalchemy~=1.5.0",
+    "snowflake-sqlalchemy~=1.6.1",
 ]
 # for debugging, see https://duckdb.org/docs/guides/python/jupyter.html
 notebook = ["duckdb-engine", "ipykernel", "jupysql"]
@@ -92,7 +92,9 @@ select = [
     "SIM",  # flake8-simplify
     "S113", # request-without-timeout
     "A",    # flake8-builtins
+    "ASYNC", # flake8-async
 ]
+
 ignore = [
     # allow untyped self and cls args
     "ANN101",

tests/conftest.py

@@ -44,7 +44,7 @@ def dcur(conn: snowflake.connector.SnowflakeConnection) -> Iterator[snowflake.co
 
 @pytest.fixture
 def snowflake_engine(_fakesnow: None) -> Engine:
-    return create_engine("snowflake://user:password@account/db1/schema1")  # type: ignore
+    return create_engine("snowflake://user:password@account/db1/schema1")
 
 
 @pytest.fixture(scope="session")

tests/test_fakes.py

@@ -7,6 +7,7 @@
 import re
 import tempfile
 from decimal import Decimal
+from typing import cast
 
 import pandas as pd
 import pytest
@@ -713,6 +714,17 @@ def test_flatten(cur: snowflake.connector.cursor.SnowflakeCursor):
     assert cur.fetchall() == [(1, '"banana"'), (2, '"coconut"'), (2, '"durian"')]
 
 
+def test_flatten_index(cur: snowflake.connector.cursor.SnowflakeCursor):
+    cur.execute(
+        """
+        select id, f.value::varchar as v, f.index as i
+        from (select column1 as id, column2 as col from (values (1, 's1,s3,s2'), (2, 's2,s1'))) as t
+        , lateral flatten(input => split(t.col, ',')) as f order by id;
+        """
+    )
+    assert cur.fetchall() == [(1, "s1", 0), (1, "s3", 1), (1, "s2", 2), (2, "s2", 0), (2, "s1", 1)]
+
+
 def test_flatten_value_cast_as_varchar(cur: snowflake.connector.cursor.SnowflakeCursor):
     cur.execute(
         """
@@ -1084,44 +1096,33 @@ def test_show_keys(dcur: snowflake.connector.cursor.SnowflakeCursor):
 
 def test_show_objects(dcur: snowflake.connector.cursor.SnowflakeCursor):
     dcur.execute("create table example(x int)")
-    dcur.execute("create view view1 as select * from example")
-    dcur.execute("show terse objects in db1.schema1")
+    dcur.execute("create schema schema2")
+    dcur.execute("create view schema2.view1 as select * from schema1.example")
+
     objects = [
         {
             "created_on": datetime.datetime(1970, 1, 1, 0, 0, tzinfo=pytz.utc),
-            "name": "EXAMPLE",
-            "kind": "TABLE",
             "database_name": "DB1",
+            "kind": "TABLE",
+            "name": "EXAMPLE",
             "schema_name": "SCHEMA1",
         },
         {
             "created_on": datetime.datetime(1970, 1, 1, 0, 0, tzinfo=pytz.utc),
-            "name": "VIEW1",
-            "kind": "VIEW",
             "database_name": "DB1",
-            "schema_name": "SCHEMA1",
+            "kind": "VIEW",
+            "name": "VIEW1",
+            "schema_name": "SCHEMA2",
         },
     ]
-    assert dcur.fetchall() == objects
+
+    dcur.execute("show terse objects in db1.schema1")
+    assert dcur.fetchall() == [objects[0]]
 
     dcur.execute("show terse objects in database")
-    assert dcur.fetchall() == [
-        *objects,
-        {
-            "created_on": datetime.datetime(1970, 1, 1, 0, 0, tzinfo=pytz.utc),
-            "name": "databases",
-            "kind": "VIEW",
-            "database_name": "DB1",
-            "schema_name": "information_schema",
-        },
-        {
-            "created_on": datetime.datetime(1970, 1, 1, 0, 0, tzinfo=pytz.utc),
-            "name": "views",
-            "kind": "VIEW",
-            "database_name": "DB1",
-            "schema_name": "information_schema",
-        },
-    ]
+    rows: list[dict] = cast(list[dict], dcur.fetchall())
+    assert [r for r in rows if r["schema_name"] != "information_schema"] == objects
+
     assert [r.name for r in dcur.description] == ["created_on", "name", "kind", "database_name", "schema_name"]
 
     dcur.execute("show objects").fetchall()
@@ -1148,16 +1149,16 @@ def test_show_schemas(dcur: snowflake.connector.cursor.SnowflakeCursor):
     assert dcur.fetchall() == [
         {
             "created_on": datetime.datetime(1970, 1, 1, 0, 0, tzinfo=pytz.utc),
-            "name": "SCHEMA1",
-            "kind": None,
             "database_name": "DB1",
+            "kind": None,
+            "name": "SCHEMA1",
             "schema_name": None,
         },
         {
             "created_on": datetime.datetime(1970, 1, 1, 0, 0, tzinfo=pytz.utc),
-            "name": "information_schema",
-            "kind": None,
             "database_name": "DB1",
+            "kind": None,
+            "name": "information_schema",
             "schema_name": None,
         },
     ]
@@ -1171,9 +1172,9 @@ def test_show_tables(dcur: snowflake.connector.cursor.SnowflakeCursor):
     objects = [
         {
             "created_on": datetime.datetime(1970, 1, 1, 0, 0, tzinfo=pytz.utc),
-            "name": "EXAMPLE",
-            "kind": "TABLE",
             "database_name": "DB1",
+            "kind": "TABLE",
+            "name": "EXAMPLE",
             "schema_name": "SCHEMA1",
         },
     ]

tests/test_sqlalchemy.py

@@ -10,7 +10,7 @@ def test_engine(snowflake_engine: Engine):
     with snowflake_engine.connect() as conn:
         conn.execute(TextClause("CREATE VIEW foo AS SELECT * FROM information_schema.databases"))
 
-        result = conn.execute("SELECT database_name FROM foo")
+        result = conn.execute(TextClause("SELECT database_name FROM foo"))
         assert result
         assert result.fetchall() == [("DB1",)]