use io_plugin example instead

Author: tswast

2025/polars_bigquery_gcloud_bigquery/Cargo.lock

@@ -8,9 +8,8 @@ name = "polars_bigquery_gcloud_bigquery"
 crate-type= ["cdylib"]
 
 [dependencies]
-pyo3 = { version = "0.23.0", features = ["extension-module", "abi3-py39"] }
-pyo3-polars = { version = "0.20.0", features = ["derive"] }
-serde = { version = "1", features = ["derive"] }
-polars = { version = "0.46.0", default-features = false }
-polars-arrow = { version = "0.46.0", default-features = false }
-
+polars = { version = "0.49.1", features = ["fmt", "dtype-date", "timezones", "lazy"], default-features = false }
+polars-arrow = { version = "0.49.1", default-features = false }
+pyo3 = { version = "0.25", features = ["abi3-py39"] }
+pyo3-polars = { version = "0.22.0", features = ["derive", "lazy"] }
+rand = { version = "0.8.5", features = [] }

2025/polars_bigquery_gcloud_bigquery/Cargo.toml

@@ -1,24 +1,45 @@
-from __future__ import annotations
-
-from pathlib import Path
-from typing import TYPE_CHECKING
-
+from .polars_bigquery_gcloud_bigquery import new_bernoulli, new_uniform, RandomSource
+from typing import Any, Iterator
+from polars.io.plugins import register_io_source
 import polars as pl
-from polars.plugins import register_plugin_function
-
-from polars_bigquery_gcloud_bigquery._internal import __version__ as __version__
-
-if TYPE_CHECKING:
-    from polars_bigquery_gcloud_bigquery.typing import IntoExprColumn
-
-LIB = Path(__file__).parent
-
 
-def pig_latinnify(expr: IntoExprColumn) -> pl.Expr:
-    return register_plugin_function(
-        args=[expr],
-        plugin_path=LIB,
-        function_name="pig_latinnify",
-        is_elementwise=True,
-    )
 
+def scan_random(samplers: list[Any], size: int = 1000) -> pl.LazyFrame:
+    def source_generator(
+        with_columns: list[str] | None,
+        predicate: pl.Expr | None,
+        n_rows: int | None,
+        batch_size: int | None,
+    ) -> Iterator[pl.DataFrame]:
+        """
+        Generator function that creates the source.
+        This function will be registered as IO source.
+        """
+
+        new_size = size
+        if n_rows is not None and n_rows < size:
+            new_size = n_rows
+
+        src = RandomSource(samplers, batch_size, new_size)
+        if with_columns is not None:
+            src.set_with_columns(with_columns)
+
+        # Set the predicate.
+        predicate_set = True
+        if predicate is not None:
+            try:
+                src.try_set_predicate(predicate)
+            except pl.exceptions.ComputeError:
+                predicate_set = False
+
+        while (out := src.next()) is not None:
+            # If the source could not apply the predicate
+            # (because it wasn't able to deserialize it), we do it here.
+            if not predicate_set and predicate is not None:
+                out = out.filter(predicate)
+
+            yield out
+
+    # create src again to compute the schema
+    src = RandomSource(samplers, 0, 0)
+    return register_io_source(io_source=source_generator, schema=src.schema())

2025/polars_bigquery_gcloud_bigquery/Makefile

@@ -4,17 +4,10 @@ build-backend = "maturin"
 
 [project]
 name = "polars-bigquery-gcloud-bigquery"
-requires-python = ">=3.8"
+version = "0.0.1"
+requires-python = ">=3.10"
 classifiers = [
   "Programming Language :: Rust",
   "Programming Language :: Python :: Implementation :: CPython",
   "Programming Language :: Python :: Implementation :: PyPy",
 ]
-dynamic = ["version"]
-
-[tool.maturin]
-module-name = "polars_bigquery_gcloud_bigquery._internal"
-
-[[tool.mypy.overrides]]
-module = "polars.utils.udfs"
-ignore_missing_imports = true

2025/polars_bigquery_gcloud_bigquery/polars_bigquery_gcloud_bigquery/__init__.py

@@ -1,3 +1,3 @@
 [toolchain]
-channel = "nightly"
+channel = "nightly-2025-05-21"