feat: add anyvlm database

korikuzma · korikuzma · commit 4ce22dea12bb · 2025-12-11T14:23:13.000-05:00
close #23
diff --git a/.env.example b/.env.example
@@ -0,0 +1,13 @@
+###############
+# BASIC SETUP #
+###############
+
+# Storage configuration -- see "Configuration" -> "Object Storage" in the docs
+ANYVLM_STORAGE_URI=postgresql://anyvlm:anyvlm-pw@localhost:5435/anyvlm
+
+############
+# OPTIONAL #
+############
+
+## Testing - see "Contributing" -> "Testing" in the docs
+ANYVLM_TEST_STORAGE_URI=postgresql://anyvlm_test:anyvlm-test-pw@localhost:5436/anyvlm_test
diff --git a/.github/workflows/python-package.yaml b/.github/workflows/python-package.yaml
@@ -47,6 +47,7 @@ jobs:
         run: uv run pytest
         env:
           ANYVLM_ANYVAR_TEST_STORAGE_URI: postgresql://postgres:postgres@localhost:5432/postgres
+          ANYVLM_TEST_STORAGE_URI: postgresql://postgres:postgres@localhost:5432/postgres
   lint:
     name: lint
     runs-on: ubuntu-latest
diff --git a/compose.yaml b/compose.yaml
@@ -0,0 +1,29 @@
+services:
+
+  anyvlm_db:
+    image: postgres:17
+    volumes:
+      - anyvlm_vol:/var/lib/postgresql/data
+    ports:
+      - 127.0.0.1:5435:5432
+    environment:
+      - POSTGRES_DB=anyvlm
+      - POSTGRES_USER=anyvlm
+      - POSTGRES_PASSWORD=anyvlm-pw
+
+  anyvlm_test_db:
+    image: postgres:17
+    volumes:
+      - anyvlm_test_vol:/var/lib/postgresql/data
+    ports:
+      - 127.0.0.1:5436:5432
+    environment:
+      - POSTGRES_DB=anyvlm_test
+      - POSTGRES_USER=anyvlm_test
+      - POSTGRES_PASSWORD=anyvlm-test-pw
+
+volumes:
+  anyvlm_vol:
+    external: true
+  anyvlm_test_vol:
+    external: true
diff --git a/docs/Makefile b/docs/Makefile
@@ -0,0 +1,20 @@
+# Minimal makefile for Sphinx documentation
+#
+
+# You can set these variables from the command line, and also
+# from the environment for the first two.
+SPHINXOPTS    ?=
+SPHINXBUILD   ?= sphinx-build
+SOURCEDIR     = source
+BUILDDIR      = build
+
+# Put it first so that "make" without argument is like "make help".
+help:
+	@$(SPHINXBUILD) -M help "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)
+
+.PHONY: help Makefile
+
+# Catch-all target: route all unknown targets to Sphinx using the new
+# "make mode" option.  $(O) is meant as a shortcut for $(SPHINXOPTS).
+%: Makefile
+	@$(SPHINXBUILD) -M $@ "$(SOURCEDIR)" "$(BUILDDIR)" $(SPHINXOPTS) $(O)
diff --git a/docs/source/conf.py b/docs/source/conf.py
@@ -0,0 +1,62 @@
+# Configuration file for the Sphinx documentation builder.
+#
+# For the full list of built-in configuration values, see the documentation:
+# https://www.sphinx-doc.org/en/master/usage/configuration.html
+
+# -- Project information -----------------------------------------------------
+# https://www.sphinx-doc.org/en/master/usage/configuration.html#project-information
+
+project = "anyvlm"
+author = "GenomicMedLab"
+html_title = "AnyVLM"
+
+# -- General configuration ---------------------------------------------------
+# https://www.sphinx-doc.org/en/master/usage/configuration.html#general-configuration
+
+extensions = [
+    "sphinx_rtd_theme",
+    "sphinx.ext.autodoc",
+    "sphinx_autodoc_typehints",
+    "sphinx.ext.linkcode",
+    "sphinx_copybutton",
+    "sphinx.ext.autosummary",
+    "sphinx_github_changelog",
+]
+
+templates_path = ["_templates"]
+exclude_patterns = []
+
+# -- Options for HTML output -------------------------------------------------
+# https://www.sphinx-doc.org/en/master/usage/configuration.html#options-for-html-output
+
+html_theme = "sphinx_rtd_theme"
+html_theme_options = {
+    "collapse_navigation": False,
+}
+
+# -- autodoc things ----------------------------------------------------------
+import os
+import sys
+
+sys.path.insert(0, os.path.abspath("../../"))
+autodoc_preserve_defaults = True
+
+# -- get version -------------------------------------------------------------
+from anyvlm import __version__  # noqa: E402
+
+version = release = __version__
+
+
+# -- linkcode ----------------------------------------------------------------
+def linkcode_resolve(domain, info):
+    if domain != "py":
+        return None
+    if not info["module"]:
+        return None
+    filename = info["module"].replace(".", "/")
+    return f"https://github.com/genomicmedlab/anyvlm/blob/main/src/{filename}.py"
+
+
+# -- code block style --------------------------------------------------------
+pygments_style = "default"
+pygements_dark_style = "monokai"
diff --git a/docs/source/configuration/docker_compose.rst b/docs/source/configuration/docker_compose.rst
@@ -0,0 +1,34 @@
+Configuring Docker Compose
+!!!!!!!!!!!!!!!!!!!!!!!!!!
+
+This page describes how to use the provided ``compose.yaml`` file to start AnyVLM alongside its dependencies, and highlights some configuration options you can customize for your environment.
+
+Overview
+--------
+
+The compose file defines four main services:
+
+* ``anyvlm_db`` - PostgreSQL database for AnyVLM
+* ``anyvlm_test_db`` - PostgreSQL database for AnyVLM Tests
+
+It also defines Docker volumes:
+
+* ``anyvlm_vol`` - storage for the AnyVLM PostgreSQL data directory
+* ``anyvlm_test_vol`` - storage for the AnyVLM PostgreSQL Test data directory
+
+These volumes are declared as ``external: true``, so it must exist before
+you run ``docker compose up``. For example:
+
+.. code-block:: bash
+
+   docker volume create anyvlm_vol
+   docker volume create anyvlm_test_vol
+
+Running the stack
+-----------------
+
+After creating the external volumes and configuring any optional environment variables, you can start the stack with:
+
+.. code-block:: bash
+
+   docker compose up
diff --git a/docs/source/configuration/dotenv_example.rst b/docs/source/configuration/dotenv_example.rst
@@ -0,0 +1,4 @@
+Example .env File
+!!!!!!!!!!!!!!!!!
+
+.. literalinclude:: ../../../.env.example
diff --git a/docs/source/configuration/index.rst b/docs/source/configuration/index.rst
@@ -0,0 +1,16 @@
+Configuration
+!!!!!!!!!!!!!
+
+This section details AnyVLM configuration. It is broken down into the following subsections:
+
+* :doc:`Object Storage <storage>`: define database connection, alter table names, and set parameters for bulk processing
+* :doc:`Example .env file <dotenv_example>`: use a ``.env`` file to declare environment variables when running REST API service
+* :doc:`Docker Compose <docker_compose>`: edit the provided Docker Compose file to tailor it to your needs
+
+.. toctree::
+   :maxdepth: 2
+   :hidden:
+
+   Storage<storage>
+   Example .env file<dotenv_example>
+   Docker Compose<docker_compose>
diff --git a/docs/source/configuration/storage.rst b/docs/source/configuration/storage.rst
@@ -0,0 +1,16 @@
+Object Storage Configuration
+!!!!!!!!!!!!!!!!!!!!!!!!!!!!
+
+Storage Connection
+==================
+
+Use the ``ANYVLM_STORAGE_URI`` environment variable to pass a `libpq connection string <https://www.postgresql.org/docs/current/libpq.html>`_ to the PostgreSQL connection constructor.
+
+.. list-table::
+   :widths: 30 70
+   :header-rows: 1
+
+   * - Environment Variable
+     - Default Value
+   * - ``ANYVLM_STORAGE_URI``
+     - ``"postgresql://postgres@localhost:5432/anyvlm"``
diff --git a/docs/source/index.rst b/docs/source/index.rst
@@ -0,0 +1,9 @@
+AnyVLM
+!!!!!!
+
+.. toctree::
+   :maxdepth: 2
+   :caption: Contents
+   :hidden:
+
+   Configuration<configuration/index>
diff --git a/pyproject.toml b/pyproject.toml
@@ -46,6 +46,15 @@ dev = [
     "fastapi[standard]",
     "seqrepo-rest-service",  # for generating SeqRepo-based tests fixtures
 ]
+docs = [
+    "sphinx==8.2.3",
+    "sphinx_rtd_theme==3.0.2",
+    "sphinx-autodoc-typehints==3.2.0",
+    "sphinx-autobuild==2024.10.3",
+    "sphinx-copybutton==0.5.2",
+    "sphinxext-opengraph==0.10.0",
+    "sphinx-github-changelog==1.7.1",
+]
 
 [project.urls]
 Homepage = "https://github.com/genomicmedlab/anyvlm"
@@ -75,6 +84,7 @@ pythonpath = ["src"]
 [tool.ruff]
 src = ["src"]
 include = ["src/**/*.py", "tests/**/*.py"]
+exclude = ["docs/source/"]
 
 [tool.ruff.lint]
 select = ["ALL"]
diff --git a/src/anyvlm/storage/__init__.py b/src/anyvlm/storage/__init__.py
@@ -0,0 +1,7 @@
+"""Provide tools and implementations of AnyVLM storage backends."""
+
+from .base_storage import Storage
+
+DEFAULT_STORAGE_URI = "postgresql://postgres@localhost:5432/anyvlm"
+
+__all__ = ["DEFAULT_STORAGE_URI", "Storage"]
diff --git a/src/anyvlm/storage/base_storage.py b/src/anyvlm/storage/base_storage.py
@@ -0,0 +1,32 @@
+"""Provide base storage implementation."""
+
+from abc import ABC, abstractmethod
+
+from ga4gh.va_spec.base import CohortAlleleFrequencyStudyResult
+
+
+class StorageError(Exception):
+    """Base AnyLM storage error."""
+
+
+class Storage(ABC):
+    """Abstract base class for interacting with storage backends."""
+
+    @abstractmethod
+    def __init__(self, *args, **kwargs) -> None:
+        """Initialize the storage backend."""
+
+    @abstractmethod
+    def close(self) -> None:
+        """Close the storage backend."""
+
+    @abstractmethod
+    def wipe_db(self) -> None:
+        """Wipe all data from the storage backend."""
+
+    @abstractmethod
+    def add_allele_frequency(self, caf: CohortAlleleFrequencyStudyResult) -> None:
+        """Add allele frequency data to the database. Will skip conflicts.
+
+        :param caf: Cohort allele frequency study result object to insert into the DB
+        """
diff --git a/src/anyvlm/storage/mapper_registry.py b/src/anyvlm/storage/mapper_registry.py
@@ -0,0 +1,51 @@
+"""Central registry for all object mappers."""
+
+from typing import TypeVar
+
+from ga4gh.va_spec.base import CohortAlleleFrequencyStudyResult
+
+from anyvlm.storage import orm
+from anyvlm.storage.mappers import AlleleFrequencyMapper, BaseMapper
+
+T = TypeVar("T")
+
+
+class MapperRegistry:
+    """Central registry for all object mappers."""
+
+    def __init__(self) -> None:
+        """Initialize the MapperRegistry with known mappers."""
+        self.anyvlm_to_db_mapping = {
+            CohortAlleleFrequencyStudyResult: orm.AlleleFrequencyData
+        }
+
+        self._mappers: dict[type, BaseMapper] = {
+            orm.AlleleFrequencyData: AlleleFrequencyMapper()
+        }
+
+    def get_mapper(self, entity_type: type[T]) -> BaseMapper:
+        """Get mapper for the given entity type."""
+        mapper = self._mappers.get(entity_type)
+        if mapper is None:
+            raise ValueError(f"No mapper registered for type: {entity_type}")
+        return mapper
+
+    def from_db_entity(self, db_entity):  # noqa: ANN201, ANN001
+        """Convert any DB entity to its corresponding VA-Spec model."""
+        mapper = self.get_mapper(type(db_entity))
+        return mapper.from_db_entity(db_entity)
+
+    def to_db_entity(self, anyvlm_entity) -> orm.Base:  # noqa: ANN001
+        """Convert any VA-Spec model to its corresponding DB entity."""
+        db_type = self.anyvlm_to_db_mapping.get(type(anyvlm_entity))
+        if db_type is None:
+            raise ValueError(
+                f"No DB entity type mapped for VRS model: {type(anyvlm_entity)}"
+            )
+
+        mapper = self.get_mapper(db_type)
+        return mapper.to_db_entity(anyvlm_entity)
+
+
+# Global registry instance
+mapper_registry = MapperRegistry()
diff --git a/src/anyvlm/storage/mappers.py b/src/anyvlm/storage/mappers.py
diff --git a/src/anyvlm/storage/orm.py b/src/anyvlm/storage/orm.py
diff --git a/src/anyvlm/storage/postgres.py b/src/anyvlm/storage/postgres.py
diff --git a/tests/__init__.py b/tests/__init__.py
diff --git a/tests/conftest.py b/tests/conftest.py
diff --git a/tests/integration/storage/test_postgres_integration.py b/tests/integration/storage/test_postgres_integration.py
diff --git a/tests/unit/storage/test_postgres_unit.py b/tests/unit/storage/test_postgres_unit.py
