Merge branch 'v1.4-andium'

Author: evertlammerts

.github/workflows/code_quality.yml

@@ -0,0 +1,45 @@
+name: Code Quality Checks
+on:
+  workflow_dispatch:
+    inputs:
+      git_ref:
+        type: string
+        description: Git ref of the DuckDB python package
+        required: false
+  workflow_call:
+    inputs:
+      git_ref:
+        type: string
+        description: Git ref of the DuckDB python package
+        required: false
+
+defaults:
+  run:
+    shell: bash
+
+jobs:
+  run_checks:
+    name: Run linting and formatting
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          ref: ${{ inputs.git_ref }}
+          fetch-depth: 0
+          persist-credentials: false
+
+      - name: Install Astral UV
+        uses: astral-sh/setup-uv@v6
+        with:
+          version: "0.7.14"
+          python-version: 3.9
+
+      - name: pre-commit (cache)
+        uses: actions/cache@v4
+        with:
+          path: ~/.cache/pre-commit
+          key: pre-commit-${{ hashFiles('.pre-commit-config.yaml') }}
+
+      - name: pre-commit (--all-files)
+        run: |
+          uvx pre-commit run --show-diff-on-failure --color=always --all-files

.github/workflows/coverage.yml

@@ -151,4 +151,4 @@ jobs:
           echo "### C++ Coverage Summary" >> $GITHUB_STEP_SUMMARY
           echo '```' >> $GITHUB_STEP_SUMMARY
           echo "$SUMMARY_CPP" >> $GITHUB_STEP_SUMMARY
-          echo '```' >> $GITHUB_STEP_SUMMARY
+          echo '```' >> $GITHUB_STEP_SUMMARY

.github/workflows/on_pr.yml

@@ -23,9 +23,14 @@ jobs:
     name: Make sure submodule is in a sane state
     uses: ./.github/workflows/submodule_sanity.yml
 
+  code_quality:
+    name: Code-quality checks
+    needs: submodule_sanity_guard
+    uses: ./.github/workflows/code_quality.yml
+
   packaging_test:
     name: Build a minimal set of packages and run all tests on them
-    needs: submodule_sanity_guard
+    needs: code_quality
     # Skip packaging tests for draft PRs
     if: ${{ github.event_name != 'pull_request' || github.event.pull_request.draft == false }}
     uses: ./.github/workflows/packaging.yml
@@ -36,7 +41,7 @@ jobs:
 
   coverage_test:
     name: Run coverage tests
-    needs: submodule_sanity_guard
+    needs: code_quality
     # Only run coverage test for draft PRs
     if: ${{ github.event_name == 'pull_request' && github.event.pull_request.draft == true }}
     uses: ./.github/workflows/coverage.yml

.github/workflows/on_push.yml

@@ -18,8 +18,13 @@ concurrency:
   cancel-in-progress: true
 
 jobs:
+  code_quality:
+    name: Code-quality checks
+    uses: ./.github/workflows/code_quality.yml
+
   test:
     name: Run coverage tests
+    needs: code_quality
     uses: ./.github/workflows/coverage.yml
     with:
       git_ref: ${{ github.ref }}

.pre-commit-config.yaml

@@ -0,0 +1,28 @@
+repos:
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    # Ruff version.
+    rev: v0.13.0
+    hooks:
+      # Run the linter.
+      - id: ruff-check
+      # Run the formatter.
+      - id: ruff-format
+
+  - repo: https://github.com/pre-commit/mirrors-clang-format
+    rev: v18.1.8    # pick the version of clang-format you want
+    hooks:
+      - id: clang-format
+        files: \.(c|cpp|cc|h|hpp|cxx|hxx)$
+
+  - repo: https://github.com/cheshirekow/cmake-format-precommit
+    rev: v0.6.13
+    hooks:
+      - id: cmake-format
+
+  - repo: local
+    hooks:
+      - id: post-checkout-submodules
+        name: Update submodule post-checkout
+        entry: .githooks/post-checkout
+        language: script
+        stages: [ post-checkout ]

CMakeLists.txt

@@ -48,50 +48,84 @@ duckdb_add_library(duckdb_target)
 
 # Bundle in INTERFACE library
 add_library(_duckdb_dependencies INTERFACE)
-target_link_libraries(_duckdb_dependencies INTERFACE
-        pybind11::pybind11
-        duckdb_target
-)
+target_link_libraries(_duckdb_dependencies INTERFACE pybind11::pybind11
+                                                     duckdb_target)
 # Also add include directory
-target_include_directories(_duckdb_dependencies INTERFACE
-        $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/src/duckdb_py/include>
+target_include_directories(
+  _duckdb_dependencies
+  INTERFACE $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/src/duckdb_py/include>
 )
 
 # ────────────────────────────────────────────
 # Descend into the real DuckDB‑Python sources
 # ────────────────────────────────────────────
 add_subdirectory(src/duckdb_py)
 
-pybind11_add_module(_duckdb
-        $<TARGET_OBJECTS:python_src>
-        $<TARGET_OBJECTS:python_arrow>
-        $<TARGET_OBJECTS:python_common>
-        $<TARGET_OBJECTS:python_functional>
-        $<TARGET_OBJECTS:python_jupyter>
-        $<TARGET_OBJECTS:python_native>
-        $<TARGET_OBJECTS:python_numpy>
-        $<TARGET_OBJECTS:python_pandas>
-        $<TARGET_OBJECTS:python_pybind11>
-        $<TARGET_OBJECTS:python_connection>
-        $<TARGET_OBJECTS:python_expression>
-        $<TARGET_OBJECTS:python_relation>
-        $<TARGET_OBJECTS:python_type>
-)
+pybind11_add_module(
+  _duckdb
+  $<TARGET_OBJECTS:python_src>
+  $<TARGET_OBJECTS:python_arrow>
+  $<TARGET_OBJECTS:python_common>
+  $<TARGET_OBJECTS:python_functional>
+  $<TARGET_OBJECTS:python_jupyter>
+  $<TARGET_OBJECTS:python_native>
+  $<TARGET_OBJECTS:python_numpy>
+  $<TARGET_OBJECTS:python_pandas>
+  $<TARGET_OBJECTS:python_pybind11>
+  $<TARGET_OBJECTS:python_connection>
+  $<TARGET_OBJECTS:python_expression>
+  $<TARGET_OBJECTS:python_relation>
+  $<TARGET_OBJECTS:python_type>)
 # add _duckdb_dependencies
 target_link_libraries(_duckdb PRIVATE _duckdb_dependencies)
 
+# ────────────────────────────────────────────
+# Controlling symbol export
+#
+# We want to export exactly two symbols: - PyInit__duckdb: this allows CPython
+# to load the module - duckdb_adbc_init: the DuckDB ADBC driver
+#
+# The export of symbols on OSX and Linux is controlled by: - Visibility
+# annotations in the code (for this lib we use the PYBIND11_EXPORT macro) -
+# Telling the linker which symbols we want exported, which we do below
+#
+# For Windows, we rely on just the visbility annotations.
+# ────────────────────────────────────────────
+set_target_properties(
+  _duckdb
+  PROPERTIES CXX_VISIBILITY_PRESET hidden
+             C_VISIBILITY_PRESET hidden
+             VISIBILITY_INLINES_HIDDEN ON)
+
+if(APPLE)
+  target_link_options(
+    _duckdb PRIVATE "LINKER:-exported_symbol,_duckdb_adbc_init"
+    "LINKER:-exported_symbol,_PyInit__duckdb")
+elseif(UNIX AND NOT APPLE)
+  target_link_options(
+    _duckdb PRIVATE "LINKER:--export-dynamic-symbol=duckdb_adbc_init"
+    "LINKER:--export-dynamic-symbol=PyInit__duckdb")
+elseif(WIN32)
+  target_link_options(_duckdb PRIVATE "/EXPORT:duckdb_adbc_init"
+                      "/EXPORT:PyInit__duckdb")
+endif()
+
 # ────────────────────────────────────────────
 # Put the object file in the correct place
 # ────────────────────────────────────────────
 
-# If we're not building through scikit-build-core then we have to set a different dest dir
+# If we're not building through scikit-build-core then we have to set a
+# different dest dir
 include(GNUInstallDirs)
 if(DEFINED SKBUILD_PLATLIB_DIR)
   set(_DUCKDB_PY_INSTALL_DIR "${SKBUILD_PLATLIB_DIR}")
 elseif(DEFINED Python_SITEARCH)
   set(_DUCKDB_PY_INSTALL_DIR "${Python_SITEARCH}")
 else()
-  message(WARNING "Could not determine Python install dir. Falling back to CMAKE_INSTALL_LIBDIR.")
+  message(
+    WARNING
+      "Could not determine Python install dir. Falling back to CMAKE_INSTALL_LIBDIR."
+  )
   set(_DUCKDB_PY_INSTALL_DIR "${CMAKE_INSTALL_LIBDIR}")
 endif()
 

adbc_driver_duckdb/__init__.py

@@ -19,12 +19,11 @@
 
 import enum
 import functools
+import importlib
 import typing
 
 import adbc_driver_manager
 
-__all__ = ["StatementOptions", "connect"]
-
 
 class StatementOptions(enum.Enum):
     """Statement options specific to the DuckDB driver."""
@@ -36,12 +35,16 @@ class StatementOptions(enum.Enum):
 def connect(path: typing.Optional[str] = None) -> adbc_driver_manager.AdbcDatabase:
     """Create a low level ADBC connection to DuckDB."""
     if path is None:
-        return adbc_driver_manager.AdbcDatabase(driver=_driver_path(), entrypoint="duckdb_adbc_init")
-    return adbc_driver_manager.AdbcDatabase(driver=_driver_path(), entrypoint="duckdb_adbc_init", path=path)
+        return adbc_driver_manager.AdbcDatabase(driver=driver_path(), entrypoint="duckdb_adbc_init")
+    return adbc_driver_manager.AdbcDatabase(driver=driver_path(), entrypoint="duckdb_adbc_init", path=path)
 
 
 @functools.cache
-def _driver_path() -> str:
-    import duckdb
-
-    return duckdb.duckdb.__file__
+def driver_path() -> str:
+    """Get the path to the DuckDB ADBC driver."""
+    duckdb_module_spec = importlib.util.find_spec("_duckdb")
+    if duckdb_module_spec is None:
+        msg = "Could not find duckdb shared library. Did you pip install duckdb?"
+        raise ImportError(msg)
+    print(f"Found duckdb shared library at {duckdb_module_spec.origin}")
+    return duckdb_module_spec.origin

adbc_driver_duckdb/dbapi.py

@@ -15,14 +15,13 @@
 # specific language governing permissions and limitations
 # under the License.
 
-"""
-DBAPI 2.0-compatible facade for the ADBC DuckDB driver.
-"""
+"""DBAPI 2.0-compatible facade for the ADBC DuckDB driver."""
 
 import typing
 
 import adbc_driver_manager
 import adbc_driver_manager.dbapi
+
 import adbc_driver_duckdb
 
 __all__ = [

cmake/compiler_launcher.cmake

@@ -8,19 +8,25 @@ include(CMakeParseArguments)
 # Function to look for ccache and sccache to speed up builds, if available
 # ────────────────────────────────────────────
 function(setup_compiler_launcher_if_available)
-    if(NOT DEFINED CMAKE_C_COMPILER_LAUNCHER AND NOT DEFINED ENV{CMAKE_C_COMPILER_LAUNCHER})
-        find_program(COMPILER_LAUNCHER NAMES ccache sccache)
-        if(COMPILER_LAUNCHER)
-            message(STATUS "Using ${COMPILER_LAUNCHER} as C compiler launcher")
-            set(CMAKE_C_COMPILER_LAUNCHER "${COMPILER_LAUNCHER}" CACHE STRING "" FORCE)
-        endif()
+  if(NOT DEFINED CMAKE_C_COMPILER_LAUNCHER AND NOT DEFINED
+                                               ENV{CMAKE_C_COMPILER_LAUNCHER})
+    find_program(COMPILER_LAUNCHER NAMES ccache sccache)
+    if(COMPILER_LAUNCHER)
+      message(STATUS "Using ${COMPILER_LAUNCHER} as C compiler launcher")
+      set(CMAKE_C_COMPILER_LAUNCHER
+          "${COMPILER_LAUNCHER}"
+          CACHE STRING "" FORCE)
     endif()
+  endif()
 
-    if(NOT DEFINED CMAKE_CXX_COMPILER_LAUNCHER AND NOT DEFINED ENV{CMAKE_CXX_COMPILER_LAUNCHER})
-        find_program(COMPILER_LAUNCHER NAMES ccache sccache)
-        if(COMPILER_LAUNCHER)
-            message(STATUS "Using ${COMPILER_LAUNCHER} as C++ compiler launcher")
-            set(CMAKE_CXX_COMPILER_LAUNCHER "${COMPILER_LAUNCHER}" CACHE STRING "" FORCE)
-        endif()
+  if(NOT DEFINED CMAKE_CXX_COMPILER_LAUNCHER
+     AND NOT DEFINED ENV{CMAKE_CXX_COMPILER_LAUNCHER})
+    find_program(COMPILER_LAUNCHER NAMES ccache sccache)
+    if(COMPILER_LAUNCHER)
+      message(STATUS "Using ${COMPILER_LAUNCHER} as C++ compiler launcher")
+      set(CMAKE_CXX_COMPILER_LAUNCHER
+          "${COMPILER_LAUNCHER}"
+          CACHE STRING "" FORCE)
     endif()
+  endif()
 endfunction()