[Feat][CI] Building system of farsante

.github/workflows/release.yaml

@@ -0,0 +1,19 @@
+name: Build release
+
+on:
+  workflow_dispatch:
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 1
+
+      - name: Maturin build
+        uses: PyO3/maturin-action@v1
+        with:
+          command: build
+          args: --release
+

.gitignore

@@ -15,4 +15,7 @@ __pycache__/
 .pytest_cache/
 
 *.csv
-**/target
+**/target
+
+.env
+.venv

h2o-data-rust/Cargo.toml → Cargo.toml

@@ -1,10 +1,10 @@
 [package]
-name = "h2o_data_rust"
+name = "farsante"
 version = "0.1.0"
 edition = "2021"
 
 [lib]
-name = "h2o"
+name = "farsante"
 crate-type = ["cdylib"]
 
 [dependencies]

Makefile

@@ -0,0 +1,4 @@
+.PHONY: install_dev
+install_dev:
+	maturin develop --extras=test
+

pyproject.toml

@@ -1,25 +1,28 @@
-[tool.poetry]
+[build-system]
+requires = ["maturin>=1.0,<2.0"]
+build-backend = "maturin"
+
+[project]
 name = "farsante"
 version = "0.2.0"
 description = "Fake DataFrame generators for Pandas and PySpark"
-authors = ["MrPowers <[email protected]>"]
-
-[tool.poetry.dependencies]
-python = ">=3.8 <4.0"
-mimesis = "6.0.0"
-pyspark = "^3.3.1"
-pandas = ">=1.0.0"
-pyarrow = "11.0.0"
-maturin = "^1.3.2"
+authors = [{ name = "MrPowers", email = "[email protected]" }]
+requires-python = ">=3.8,<4.0"
+dependencies = [
+  "mimesis == 6.0.0",
+  "pyspark >= 3.3",
+  "pandas >=1.0.0",
+  "pyarrow >= 11.0.0",
+  "maturin >= 1.3.2",
+]
 
-[tool.poetry.dev-dependencies]
-pytest = "7.2.0"
-
-[build-system]
-requires = ["maturin>=1.0,<2.0"]
-build-backend = "maturin"
-# requires = ["poetry>=0.12"]
+[project.optional-dependencies]
+test = ["pytest >= 7.2"]
 
 [tool.maturin]
 # "extension-module" tells pyo3 we want to build an extension module (skips linking against libpython.so)
 features = ["pyo3/extension-module"]
+python-source = "python"
+python-packages = ["farsante"]
+module-name = "farsante_rs"
+bindings = "pyo3"

farsante/h2o_dataset_create.py → python/farsante/h2o_dataset_create.py

@@ -1,7 +1,8 @@
-from h2o import generate_csv_py
 from concurrent.futures import ProcessPoolExecutor
 from itertools import repeat
 
+from farsante_rs import generate_csv_py
+
 
 def pretty_sci(n: int) -> str:
     if n == 0:

farsante/h2o_dataset_create_all.py → python/farsante/h2o_dataset_create_all.py

@@ -1,6 +1,7 @@
-from farsante import generate_all_h2o_datasets
 import argparse
 
+from farsante import generate_all_h2o_datasets
+
 
 def main():
     parser = argparse.ArgumentParser(description="Generate all h2o datasets")

h2o-data-rust/src/lib.rs → src/lib.rs

@@ -24,6 +24,7 @@ fn hello_rust() -> () {
 }
 
 #[pymodule]
+#[pyo3(name = "farsante_rs")]
 fn h2o(_py: Python, m: &PyModule) -> PyResult<()> {
     m.add_function(wrap_pyfunction!(generate_csv_py, m)?)?;
     m.add_function(wrap_pyfunction!(hello_rust, m)?)?;