mayutils 2.0.0__tar.gz → 3.0.0__tar.gz

{mayutils-2.0.0 → mayutils-3.0.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: mayutils
-Version: 2.0.0
+Version: 3.0.0
 Summary: Utilities for Python from Mayuran Visakan
 Author: Mayuran Visakan
 Author-email: Mayuran Visakan <mayuran.k.v@gmail.com>
@@ -25,6 +25,7 @@ License: MIT License
          LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
          OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
          SOFTWARE.
+Requires-Dist: jinja2>=3.1
 Requires-Dist: pydantic>=2.11.7
 Requires-Dist: pydantic-settings>=2.10.1
 Requires-Dist: mayutils[async,cli,console,dataframes,datetime,filesystem,financials,google,mathematics,microsoft,notebook,numerics,pandas,pdf,plotting,secrets,snowflake,sql,statistics,streamlit,tui,web] ; extra == 'all'
@@ -150,7 +151,7 @@ Description-Content-Type: text/markdown
 
 # mayutils
 
-[![CI](https://github.com/mayuran-visakan/mayutils/actions/workflows/ci.yaml/badge.svg)](https://github.com/mayuran-visakan/mayutils/actions/workflows/ci.yaml) [![Documentation](https://img.shields.io/badge/Documentation-mkdocs%20material-indigo.svg)](https://mayuran-visakan.github.io/mayutils/) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](./LICENSE)
+[![CI](https://img.shields.io/github/actions/workflow/status/mayurankv/mayutils/ci.yaml?branch=main&style=for-the-badge&logo=githubactions&logoColor=white&label=CI)](https://github.com/mayurankv/mayutils/actions/workflows/ci.yaml) [![PyPI](https://img.shields.io/pypi/v/mayutils?style=for-the-badge&logo=pypi&logoColor=white&label=PyPI&color=3775A9)](https://pypi.org/project/mayutils/) [![Python](https://img.shields.io/badge/python-3.13%2B-3776AB?style=for-the-badge&logo=python&logoColor=white)](https://www.python.org/downloads/) [![License](https://img.shields.io/badge/license-MIT-green?style=for-the-badge&logo=opensourceinitiative&logoColor=white)](./LICENSE) [![Docs](https://img.shields.io/badge/docs-mkdocs--material-3F51B5?style=for-the-badge&logo=materialformkdocs&logoColor=white)](https://mayurankv.github.io/mayutils/)
 
 Utilities for Python — plotting helpers, dataframe adapters, Snowflake/SQL glue, PowerPoint/PDF export, notebook display tweaks, OAuth helpers, and a fistful of miscellaneous object helpers. Heavy dependencies are grouped behind extras so the core install stays small.
 
@@ -206,7 +207,7 @@ Full details: [docs/guides/dependency-groups.md](docs/guides/dependency-groups.m
 - [Documentation](docs/guides/documentation.md)
 - [Roadmap](docs/roadmap.md) — translated from the legacy `.todo` file
 - [Changelog](docs/changelog.md)
-- [API Reference](https://mayuran-visakan.github.io/mayutils/reference/) — auto-generated from docstrings
+- [API Reference](https://mayurankv.github.io/mayutils/reference/) — auto-generated from docstrings
 
 ## Contributing
 

{mayutils-2.0.0 → mayutils-3.0.0}/README.md

@@ -1,6 +1,6 @@
 # mayutils
 
-[![CI](https://github.com/mayuran-visakan/mayutils/actions/workflows/ci.yaml/badge.svg)](https://github.com/mayuran-visakan/mayutils/actions/workflows/ci.yaml) [![Documentation](https://img.shields.io/badge/Documentation-mkdocs%20material-indigo.svg)](https://mayuran-visakan.github.io/mayutils/) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](./LICENSE)
+[![CI](https://img.shields.io/github/actions/workflow/status/mayurankv/mayutils/ci.yaml?branch=main&style=for-the-badge&logo=githubactions&logoColor=white&label=CI)](https://github.com/mayurankv/mayutils/actions/workflows/ci.yaml) [![PyPI](https://img.shields.io/pypi/v/mayutils?style=for-the-badge&logo=pypi&logoColor=white&label=PyPI&color=3775A9)](https://pypi.org/project/mayutils/) [![Python](https://img.shields.io/badge/python-3.13%2B-3776AB?style=for-the-badge&logo=python&logoColor=white)](https://www.python.org/downloads/) [![License](https://img.shields.io/badge/license-MIT-green?style=for-the-badge&logo=opensourceinitiative&logoColor=white)](./LICENSE) [![Docs](https://img.shields.io/badge/docs-mkdocs--material-3F51B5?style=for-the-badge&logo=materialformkdocs&logoColor=white)](https://mayurankv.github.io/mayutils/)
 
 Utilities for Python — plotting helpers, dataframe adapters, Snowflake/SQL glue, PowerPoint/PDF export, notebook display tweaks, OAuth helpers, and a fistful of miscellaneous object helpers. Heavy dependencies are grouped behind extras so the core install stays small.
 
@@ -56,7 +56,7 @@ Full details: [docs/guides/dependency-groups.md](docs/guides/dependency-groups.m
 - [Documentation](docs/guides/documentation.md)
 - [Roadmap](docs/roadmap.md) — translated from the legacy `.todo` file
 - [Changelog](docs/changelog.md)
-- [API Reference](https://mayuran-visakan.github.io/mayutils/reference/) — auto-generated from docstrings
+- [API Reference](https://mayurankv.github.io/mayutils/reference/) — auto-generated from docstrings
 
 ## Contributing
 

{mayutils-2.0.0 → mayutils-3.0.0}/pyproject.toml

@@ -36,6 +36,7 @@ testing = ["pytest>=8.4.2"]
 [project]
 authors = [{email = "mayuran.k.v@gmail.com", name = "Mayuran Visakan"}]
 dependencies = [
+  "jinja2>=3.1",
   "pydantic>=2.11.7",
   "pydantic-settings>=2.10.1"
 ]
@@ -45,7 +46,7 @@ maintainers = [{email = "mayuran.k.v@gmail.com", name = "Mayuran Visakan"}]
 name = "mayutils"
 readme = "README.md"
 requires-python = ">=3.13,<4.0"
-version = "2.0.0"
+version = "3.0.0"
 
 [project.optional-dependencies]
 all = [
@@ -208,7 +209,7 @@ pre_bump_hooks = [
 ]
 tag_format = "v$version"
 update_changelog_on_bump = true
-version = "2.0.0"
+version = "3.0.0"
 version_files = ["pyproject.toml:^version"]
 
 [tool.mdformat]

{mayutils-2.0.0 → mayutils-3.0.0}/src/mayutils/__init__.py

@@ -44,15 +44,15 @@ def setup(
     >>> from mayutils import setup
     >>> setup(logging=False, plotly=False, notebook=False, pandas=False)
     """
-    from mayutils.core.extras import format_missing_extra_hint  # noqa: PLC0415
-    from mayutils.environment.logging import Logger  # noqa: PLC0415
+    from mayutils.core.extras import format_missing_extra_hint
+    from mayutils.environment.logging import Logger
 
     if logging:
         Logger.configure()
 
     if plotly:
         try:
-            from mayutils.visualisation.graphs.plotly.templates import register_templates  # noqa: PLC0415
+            from mayutils.visualisation.graphs.plotly.templates import register_templates
 
             register_templates()
         except ImportError as err:
@@ -63,7 +63,7 @@ def setup(
 
     if notebook:
         try:
-            from mayutils.visualisation.notebook import Notebook  # noqa: PLC0415
+            from mayutils.visualisation.notebook import Notebook
 
             Notebook.setup()
         except ImportError as err:
@@ -76,7 +76,7 @@ def setup(
 
     if pandas:
         try:
-            from mayutils.objects.dataframes import setup_pandas  # noqa: PLC0415
+            from mayutils.objects.dataframes import setup_pandas
 
             setup_pandas()
         except ImportError as err:

{mayutils-2.0.0 → mayutils-3.0.0}/src/mayutils/data/live.py

@@ -4,22 +4,20 @@ from __future__ import annotations
 
 from typing import TYPE_CHECKING, Any, Self, cast
 
-from mayutils.core.extras import may_require_extras
 from mayutils.data.queries import QUERIES_FOLDERS
 from mayutils.data.read import render_query
 from mayutils.environment.logging import Logger
 from mayutils.objects.dataframes.backends import Backend, BackendOperations, DataFrames, default_backend
-from mayutils.objects.datetime import DateTime, Interval
-
-with may_require_extras():
-    import pandas as pd
 
 if TYPE_CHECKING:
+    from collections.abc import Mapping
     from pathlib import Path
 
+    import pandas as pd
+
     from mayutils.data.read import QueryReader
-    from mayutils.objects.datetime import Duration
-    from mayutils.objects.types import SQL, SupportsStr
+    from mayutils.objects.datetime import DateTime, Duration, Interval
+    from mayutils.objects.types import SQL
 
 
 logger = Logger.spawn()
@@ -29,14 +27,19 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
     """
     Incrementally pull rows by tracking ``max(cursor_column)``.
 
-    The query template must contain a ``{cursor}`` placeholder. On each
-    update the cursor is formatted into the template and only rows past
-    the previous cursor are fetched.
+    The query template must contain a ``{{ cursor }}`` placeholder. On
+    each update the cursor is rendered into the template and only rows
+    past the previous cursor are fetched. Results are returned exactly
+    as the *reader* produces them: unlike
+    :func:`mayutils.data.read.read_query`, no automatic temporal column
+    parsing is applied, keeping the schema stable across incremental
+    fetches that are concatenated together.
 
     Parameters
     ----------
     query
-        SQL string or path to a ``.sql`` template containing ``{cursor}``.
+        SQL string or path to a ``.sql`` template containing
+        ``{{ cursor }}``.
     cursor_column
         Column whose maximum is tracked as the cursor.
     initial_cursor
@@ -55,9 +58,10 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         :meth:`~datetime.datetime.strftime` format for datetime cursors.
     queries_folders
         Directories searched when *query* is a filename.
-    **fixed_format_kwargs
-        Extra keyword arguments substituted into the query template on
-        every call.
+    template_kwargs
+        Jinja2 template variables rendered into the query template on
+        every call.  The key ``cursor`` is injected per-call by
+        :meth:`fetch` and overrides any value stored here.
 
     See Also
     --------
@@ -82,7 +86,7 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         update_frequency: Duration | None = None,
         time_format: str = "%Y-%m-%d %H:%M:%S",
         queries_folders: tuple[Path, ...] = QUERIES_FOLDERS,
-        **fixed_format_kwargs: SupportsStr,
+        template_kwargs: Mapping[str, object] | None = None,
     ) -> None:
         """
         Initialise the streaming query and perform the first fetch.
@@ -94,7 +98,7 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         ----------
         query
             SQL string or path to a ``.sql`` template containing
-            ``{cursor}``.
+            ``{{ cursor }}``.
         cursor_column
             Column whose maximum is tracked as the cursor.
         initial_cursor
@@ -115,9 +119,10 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
             cursors.
         queries_folders
             Directories searched when *query* is a filename.
-        **fixed_format_kwargs
-            Extra keyword arguments substituted into the query template
-            on every call.
+        template_kwargs
+            Jinja2 template variables rendered into the query template
+            on every call.  The key ``cursor`` is injected per-call by
+            :meth:`fetch` and overrides any value stored here.
 
         See Also
         --------
@@ -127,7 +132,7 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         --------
         >>> from mayutils.data.live import StreamingQuery  # doctest: +SKIP
         >>> sq = StreamingQuery(  # doctest: +SKIP
-        ...     "SELECT * FROM t WHERE id > {cursor}",
+        ...     "SELECT * FROM t WHERE id > {{ cursor }}",
         ...     cursor_column="id",
         ...     initial_cursor=0,
         ...     reader=my_reader,
@@ -142,11 +147,13 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         self.update_frequency = update_frequency
         self.time_format = time_format
         self.queries_folders = queries_folders
-        self.fixed_format_kwargs = fixed_format_kwargs
+        self.template_kwargs: dict[str, object] = dict(template_kwargs or {})
         self.initial_cursor = initial_cursor
 
         self.validate_retention()
 
+        from mayutils.objects.datetime import DateTime
+
         self.cursor_value: Any = initial_cursor
         self.cursor_is_datetime: bool = hasattr(initial_cursor, "strftime")
         self._data: DataFrameType = self.fetch()
@@ -208,6 +215,8 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         if not self.should_update(force=force):
             return self
 
+        from mayutils.objects.datetime import DateTime
+
         snapshot = self.data
         try:
             delta = self.fetch()
@@ -248,6 +257,8 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         --------
         >>> sq.reset()  # doctest: +SKIP
         """
+        from mayutils.objects.datetime import DateTime
+
         self._data = self.fetch()
         self.last_updated = DateTime.now()
 
@@ -274,7 +285,7 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         --------
         >>> delta = sq.fetch()  # doctest: +SKIP
         """
-        data = self.read_query(cursor=self.cursor)
+        data = self.read_query(template_kwargs={"cursor": self.cursor})
 
         if len(data) != 0:
             self.cursor_value = BackendOperations.max(data, self.cursor_column, backend=self.backend)
@@ -313,21 +324,23 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         self,
         *,
         default_suffix: str = "sql",
-        **extra_kwargs: SupportsStr,
+        template_kwargs: Mapping[str, object] | None = None,
     ) -> DataFrameType:
         """
         Render and execute the SQL query template.
 
-        Combines *fixed_format_kwargs* with any additional keyword
-        arguments, renders the template, and passes it to *reader*.
+        Merges the per-call *template_kwargs* over the mapping stored at
+        construction (per-call keys win), renders the template, and
+        passes it to *reader*.
 
         Parameters
         ----------
         default_suffix
             File extension appended when *query* is a filename without
             one.
-        **extra_kwargs
-            Additional format arguments merged into the template.
+        template_kwargs
+            Per-call Jinja2 template variables merged over the stored
+            mapping.
 
         Returns
         -------
@@ -339,14 +352,13 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
 
         Examples
         --------
-        >>> df = sq.read_query(cursor="0")  # doctest: +SKIP
+        >>> df = sq.read_query(template_kwargs={"cursor": "0"})  # doctest: +SKIP
         """
         rendered = render_query(
             self.query,
             queries_folders=self.queries_folders,
             default_suffix=default_suffix,
-            **self.fixed_format_kwargs,
-            **extra_kwargs,
+            template_kwargs={**self.template_kwargs, **(template_kwargs or {})},
         )
 
         return self.reader(
@@ -387,6 +399,8 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         if force or self.update_frequency is None:
             return True
 
+        from mayutils.objects.datetime import DateTime
+
         return (DateTime.now() - self.last_updated) > self.update_frequency
 
     def apply_retention(
@@ -409,6 +423,8 @@ class StreamingQuery[DataFrameType: DataFrames = pd.DataFrame]:
         if self.max_age is not None and len(self.data) > 0:
             newest = BackendOperations.max(self.data, self.cursor_column, backend=self.backend)
             if hasattr(newest, "strftime"):
+                from mayutils.objects.datetime import DateTime
+
                 try:
                     cutoff = DateTime.parse(cast("DateTime", newest).strftime(self.time_format)) - self.max_age
 
@@ -464,9 +480,13 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
     """
     Manage a sliding or expanding time window over a SQL query.
 
-    The query template must contain ``{start_timestamp}`` and
-    ``{end_timestamp}`` placeholders. On each update, only the delta
-    since the previous window end is fetched.
+    The query template must contain ``{{ start_timestamp }}`` and
+    ``{{ end_timestamp }}`` placeholders. On each update, only the
+    delta since the previous window end is fetched. Results are
+    returned exactly as the *reader* produces them: unlike
+    :func:`mayutils.data.read.read_query`, no automatic temporal column
+    parsing is applied, keeping the schema stable across windowed
+    fetches that are concatenated together.
 
     When *deduplicate* is ``True``, rows are deduped on *index_column*
     after each concat (keeps last), which handles re-fetched open
@@ -476,7 +496,7 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
     ----------
     query
         SQL string or path to a ``.sql`` template containing
-        ``{start_timestamp}`` and ``{end_timestamp}``.
+        ``{{ start_timestamp }}`` and ``{{ end_timestamp }}``.
     index_column
         Column used for deduplication and rolling-window filtering.
     start_timestamp
@@ -499,9 +519,11 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         :meth:`~datetime.datetime.strftime` format for window boundaries.
     queries_folders
         Directories searched when *query* is a filename.
-    **fixed_format_kwargs
-        Extra keyword arguments substituted into the query template on
-        every call.
+    template_kwargs
+        Jinja2 template variables rendered into the query template on
+        every call.  The keys ``start_timestamp`` and ``end_timestamp``
+        are injected per-call by :meth:`fetch` and override any values
+        stored here.
 
     See Also
     --------
@@ -528,7 +550,7 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         update_frequency: Duration | None = None,
         time_format: str = "%Y-%m-%d",
         queries_folders: tuple[Path, ...] = QUERIES_FOLDERS,
-        **fixed_format_kwargs: SupportsStr,
+        template_kwargs: Mapping[str, object] | None = None,
     ) -> None:
         """
         Initialise the windowed query and perform the first fetch.
@@ -540,7 +562,7 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         ----------
         query
             SQL string or path to a ``.sql`` template containing
-            ``{start_timestamp}`` and ``{end_timestamp}``.
+            ``{{ start_timestamp }}`` and ``{{ end_timestamp }}``.
         index_column
             Column used for deduplication and rolling-window filtering.
         start_timestamp
@@ -565,9 +587,11 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
             boundaries.
         queries_folders
             Directories searched when *query* is a filename.
-        **fixed_format_kwargs
-            Extra keyword arguments substituted into the query template
-            on every call.
+        template_kwargs
+            Jinja2 template variables rendered into the query template
+            on every call.  The keys ``start_timestamp`` and
+            ``end_timestamp`` are injected per-call by :meth:`fetch`
+            and override any values stored here.
 
         See Also
         --------
@@ -577,7 +601,7 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         --------
         >>> from mayutils.data.live import WindowedQuery  # doctest: +SKIP
         >>> wq = WindowedQuery(  # doctest: +SKIP
-        ...     "SELECT * FROM t WHERE ts BETWEEN '{start_timestamp}' AND '{end_timestamp}'",
+        ...     "SELECT * FROM t WHERE ts BETWEEN '{{ start_timestamp }}' AND '{{ end_timestamp }}'",
         ...     index_column="ts",
         ...     start_timestamp=DateTime.now(),
         ...     reader=my_reader,
@@ -594,11 +618,13 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         self.update_frequency = update_frequency
         self.time_format = time_format
         self.queries_folders = queries_folders
-        self.fixed_format_kwargs = fixed_format_kwargs
+        self.template_kwargs: dict[str, object] = dict(template_kwargs or {})
         self.start_timestamp = start_timestamp
 
         self.validate_retention()
 
+        from mayutils.objects.datetime import DateTime, Interval
+
         self._interval: Interval[DateTime] = Interval(
             start=start_timestamp,
             end=DateTime.now(),
@@ -738,6 +764,8 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         if not self.should_update(force=force):
             return self
 
+        from mayutils.objects.datetime import DateTime
+
         snapshot = self.data
         snapshot_interval = self.interval
         try:
@@ -794,6 +822,8 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         --------
         >>> wq.reset()  # doctest: +SKIP
         """
+        from mayutils.objects.datetime import DateTime, Interval
+
         if start_timestamp is not None:
             self.start_timestamp = start_timestamp
 
@@ -840,10 +870,14 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         """
         if initial:
             return self.read_query(
-                start_timestamp=self._interval.start.strftime(format=self.time_format),
-                end_timestamp=self._interval.end.strftime(format=self.time_format),
+                template_kwargs={
+                    "start_timestamp": self._interval.start.strftime(format=self.time_format),
+                    "end_timestamp": self._interval.end.strftime(format=self.time_format),
+                },
             )
 
+        from mayutils.objects.datetime import DateTime, Interval
+
         now = DateTime.now()
         previous_end = self._interval.end
 
@@ -858,8 +892,10 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
             )
 
         delta = self.read_query(
-            start_timestamp=previous_end.strftime(format=self.time_format),
-            end_timestamp=now.strftime(format=self.time_format),
+            template_kwargs={
+                "start_timestamp": previous_end.strftime(format=self.time_format),
+                "end_timestamp": now.strftime(format=self.time_format),
+            },
         )
 
         self._interval = Interval(start=new_start, end=now, absolute=True)
@@ -871,21 +907,23 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         /,
         *,
         default_suffix: str = "sql",
-        **extra_kwargs: SupportsStr,
+        template_kwargs: Mapping[str, object] | None = None,
     ) -> DataFrameType:
         """
         Render and execute the SQL query template.
 
-        Combines *fixed_format_kwargs* with any additional keyword
-        arguments, renders the template, and passes it to *reader*.
+        Merges the per-call *template_kwargs* over the mapping stored at
+        construction (per-call keys win), renders the template, and
+        passes it to *reader*.
 
         Parameters
         ----------
         default_suffix
             File extension appended when *query* is a filename without
             one.
-        **extra_kwargs
-            Additional format arguments merged into the template.
+        template_kwargs
+            Per-call Jinja2 template variables merged over the stored
+            mapping.
 
         Returns
         -------
@@ -898,16 +936,17 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         Examples
         --------
         >>> df = wq.read_query(  # doctest: +SKIP
-        ...     start_timestamp="2024-01-01",
-        ...     end_timestamp="2024-01-02",
+        ...     template_kwargs={
+        ...         "start_timestamp": "2024-01-01",
+        ...         "end_timestamp": "2024-01-02",
+        ...     },
         ... )
         """
         rendered = render_query(
             self.query,
             queries_folders=self.queries_folders,
             default_suffix=default_suffix,
-            **self.fixed_format_kwargs,
-            **extra_kwargs,
+            template_kwargs={**self.template_kwargs, **(template_kwargs or {})},
         )
 
         return self.reader(
@@ -948,6 +987,8 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         if force or self.update_frequency is None:
             return True
 
+        from mayutils.objects.datetime import DateTime
+
         return (DateTime.now() - self.last_updated) > self.update_frequency
 
     def apply_retention(
@@ -970,6 +1011,8 @@ class WindowedQuery[DataFrameType: DataFrames = pd.DataFrame]:
         if self.max_age is not None and len(self.data) > 0:
             newest = BackendOperations.max(self.data, self.index_column, backend=self.backend)
             if hasattr(newest, "strftime"):
+                from mayutils.objects.datetime import DateTime
+
                 try:
                     cutoff = DateTime.parse(cast("DateTime", newest).strftime(self.time_format)) - self.max_age