Refactor PostgreSQLExecutor to support Windows compatibility for process management

- Convert _get_base_command method to BASE_PROC_START_COMMAND class attribute

- Use unified command format without single quotes around PostgreSQL config values

- Add _windows_terminate_process method for graceful Windows process termination

- Update stop() method to use list args instead of shell=True for security

- Add platform-specific termination logic with killpg fallback

- Add comprehensive Windows compatibility test suite

- Merge latest changes from main branch

Author: tboy1337

.github/workflows/automerge.yml

@@ -13,7 +13,7 @@ on:
 
 jobs:
   automerge:
-    uses: fizyk/actions-reuse/.github/workflows/shared-automerge.yml@v4.1.1
+    uses: fizyk/actions-reuse/.github/workflows/shared-automerge.yml@v4.2.1
     secrets:
       app_id: ${{ secrets.MERGE_APP_ID }}
       private_key: ${{ secrets.MERGE_APP_PRIVATE_KEY }}

.github/workflows/build.yml

@@ -8,4 +8,4 @@ on:
 
 jobs:
   build:
-    uses: fizyk/actions-reuse/.github/workflows/shared-pypi.yml@v4.1.1
+    uses: fizyk/actions-reuse/.github/workflows/shared-pypi.yml@v4.2.1

.github/workflows/pre-commit.yml

@@ -9,4 +9,4 @@ on:
 
 jobs:
   pre-commit:
-    uses: fizyk/actions-reuse/.github/workflows/shared-pre-commit.yml@v4.1.1
+    uses: fizyk/actions-reuse/.github/workflows/shared-pre-commit.yml@v4.2.1

.github/workflows/pypi.yml

@@ -5,7 +5,7 @@ on:
       - v*
 jobs:
   build-n-publish:
-    uses: fizyk/actions-reuse/.github/workflows/shared-pypi.yml@v4.1.1
+    uses: fizyk/actions-reuse/.github/workflows/shared-pypi.yml@v4.2.1
     with:
       publish: true
     secrets:

.pre-commit-config.yaml

@@ -25,7 +25,7 @@ repos:
       - id: debug-statements
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.14.11
+    rev: v0.14.14
     hooks:
       - id: ruff-check
         args: [--fix, --exit-non-zero-on-fix, --respect-gitignore, --show-fixes]

CHANGES.rst

@@ -3,6 +3,66 @@ CHANGELOG
 
 .. towncrier release notes start
 
+pytest-postgresql 8.0.0 (2026-01-23)
+====================================
+
+Breaking changes
+----------------
+
+- Refactor ``DatabaseJanitor`` to use explicit template management. This includes a new ``as_template`` flag and making ``dbname`` a required parameter. (`#890 <https://github.com/dbfixtures/pytest-postgresql/issues/890>`__)
+- Bump the minimum supported pytest version to 8.2.
+
+  The previous minimum was about two years old, and older pytest versions
+  can be flaky with fixture chaining that relies on `getfixturevalue` on
+  Python 3.12-3.13 when used alongside xdist. (`#890 <https://github.com/dbfixtures/pytest-postgresql/issues/890>`__)
+- Support only PostgreSQL version 14 and up. (`#1250 <https://github.com/dbfixtures/pytest-postgresql/issues/1250>`__)
+- Dropped support for Python 3.9 as it has reached End of Life
+
+
+Bugfixes
+--------
+
+- Fix issue where `--postgresql-drop-test-database` was applied in process fixture.
+
+  It was already being read in client and noprocess fixtures. (`#1148 <https://github.com/dbfixtures/pytest-postgresql/issues/1148>`__)
+
+
+Documentation
+-------------
+
+- Add a Mermaid sequence diagram to the documentation to illustrate fixture chaining and hierarchical cloning. (`#890 <https://github.com/dbfixtures/pytest-postgresql/issues/890>`__)
+- Improve code examples by adding types (`#1233 <https://github.com/dbfixtures/pytest-postgresql/issues/1233>`__)
+- Add plugins' architecture diagram and display it in project's README. (`#1244 <https://github.com/dbfixtures/pytest-postgresql/issues/1244>`__)
+- Improved README.rst:
+  - Added a "Quick Start" section with a simple example.
+  - Clarified `psycopg` 3 requirement for the plugin vs. `psycopg` 2 support for the application.
+  - Updated PostgreSQL versioning notes (>= 10).
+  - Explained the binary discovery mechanism (`pg_config` vs. fallbacks).
+  - Promoted `DatabaseJanitor` as an advanced API.
+  - Fixed broken table layouts and corrected various typos.
+
+
+Features
+--------
+
+- Add ``depends_on`` parameter to ``postgresql_noproc`` factory to allow hierarchical cloning and chaining of process fixtures. (`#890 <https://github.com/dbfixtures/pytest-postgresql/issues/890>`__)
+- Convert Config kept in TypedDict into a dataclass (`#1225 <https://github.com/dbfixtures/pytest-postgresql/issues/1225>`__)
+- Marked support for Python 3.14
+
+
+Miscellaneus
+------------
+
+- `#1127 <https://github.com/dbfixtures/pytest-postgresql/issues/1127>`__, `#1218 <https://github.com/dbfixtures/pytest-postgresql/issues/1218>`__
+- Add new release workflow, which will allow releasing new versions directly from GitHub.
+- Drop PostgreSQL 13 and add PostgreSQL 18 to the CI
+- Update github-actions for actions-reuse 4.x
+- Updates to pre-commit
+
+  Replaced black with ruff-format, Added https://github.com/fizyk/pyproject-validator,
+  Moved mypy configuration into pyproject.toml, extended line length to 120 characters.
+
+
 7.0.2 (2025-05-17)
 ==================
 

Pipfile

@@ -7,7 +7,7 @@ name = "pypi"
 pytest = "==9.0.2"
 port-for = "==1.0.0"
 mirakuru = "==3.0.1"
-packaging = "==25.0"
+packaging = "==26.0"
 psycopg = "==3.3.2"
 
 [dev-packages]
@@ -17,5 +17,5 @@ pytest-cov = "==7.0.0"
 pytest-xdist = "==3.8.0"
 mock = "==5.2.0"
 mypy = {version = "==1.19.1", markers="implementation_name == 'cpython'"}
-types-setuptools = "==80.9.0.20251223"
+types-setuptools = "==80.10.0.20260124"
 tbump = "==6.11.0"

README.rst

@@ -146,6 +146,33 @@ To connect to an external server (e.g., running in Docker), use the ``postgresql
 
 By default, it connects to ``127.0.0.1:5432``.
 
+Chaining fixtures
+-----------------
+
+You can chain multiple ``postgresql_noproc`` fixtures to layer your data pre-population. Each fixture in the chain will create its own template database based on the previous one.
+
+.. code-block:: python
+
+    from pytest_postgresql import factories
+
+    # 1. Start with a process or a no-process base
+    base_proc = factories.postgresql_proc(load=[load_schema])
+
+    # 2. Add a layer with some data
+    seeded_noproc = factories.postgresql_noproc(depends_on="base_proc", load=[load_data])
+
+    # 3. Add another layer with more data
+    more_seeded_noproc = factories.postgresql_noproc(depends_on="seeded_noproc", load=[load_more_data])
+
+    # 4. Use the final layer in your test
+    client = factories.postgresql("more_seeded_noproc")
+
+
+
+.. image:: https://raw.githubusercontent.com/dbfixtures/pytest-postgresql/main/docs/images/architecture_chaining.svg
+    :alt: Fixture Chaining Diagram
+    :align: center
+
 Configuration
 =============
 

docs/architecture_chaining.mmd

@@ -0,0 +1,22 @@
+sequenceDiagram
+        participant Test as Test
+        participant ProcF as base_proc Fixture
+        participant NoProc1 as seeded_noproc Fixture
+        participant NoProc2 as more_seeded_noproc Fixture
+        participant DB as PostgreSQL DB
+
+        Test->>ProcF: request base_proc
+        ProcF->>DB: init database & run load_schema
+        ProcF-->>Test: return PostgreSQLExecutor
+
+        Test->>NoProc1: request seeded_noproc (depends_on=base_proc)
+        NoProc1->>ProcF: read connection/template info
+        NoProc1->>DB: create layered DB / run load_data
+        NoProc1-->>Test: return NoopExecutor
+
+        Test->>NoProc2: request more_seeded_noproc (depends_on=seeded_noproc)
+        NoProc2->>NoProc1: read connection/template info
+        NoProc2->>DB: run load_more_data on layered DB
+        NoProc2-->>Test: return NoopExecutor
+
+        Test->>Test: validate tables and data across layers