Convert cmd arguments to strings, remove --vanilla, display executed command and release v0.0.6. (#7)

Author: tobiasraabe

.pre-commit-config.yaml

@@ -54,6 +54,7 @@ repos:
             flake8-print,
             flake8-pytest-style,
             flake8-todo,
+            flake8-typing-imports,
             flake8-unused-arguments,
             pep8-naming,
             pydocstyle,

CHANGES.rst

@@ -6,6 +6,13 @@ chronological order. Releases follow `semantic versioning <https://semver.org/>`
 all releases are available on `Anaconda.org <https://anaconda.org/pytask/pytask-r>`_.
 
 
+0.0.6 - 2021-01-16
+------------------
+
+- :gh:`7` converts command line arguments automatically to string and removes the
+  ``--vanilla`` option as the default.
+
+
 0.0.5 - 2020-10-30
 ------------------
 

README.rst

@@ -98,6 +98,7 @@ for a ``"source"`` key in the dictionary and, secondly, under the key ``0``.
 
 .. code-block:: python
 
+    @pytask.mark.r
     @pytask.mark.depends_on({"source": "script.r", "input": "input.rds"})
     def task_run_r_script():
         pass
@@ -106,6 +107,7 @@ for a ``"source"`` key in the dictionary and, secondly, under the key ``0``.
     # or
 
 
+    @pytask.mark.r
     @pytask.mark.depends_on({0: "script.r", "input": "input.rds"})
     def task_run_r_script():
         pass
@@ -114,6 +116,7 @@ for a ``"source"`` key in the dictionary and, secondly, under the key ``0``.
     # or two decorators for the function, if you do not assign a name to the input.
 
 
+    @pytask.mark.r
     @pytask.mark.depends_on({"source": "script.r"})
     @pytask.mark.depends_on("input.rds")
     def task_run_r_script():
@@ -123,13 +126,12 @@ for a ``"source"`` key in the dictionary and, secondly, under the key ``0``.
 Command Line Arguments
 ~~~~~~~~~~~~~~~~~~~~~~
 
-The decorator can be used to pass command line arguments to ``Rscript`` which is, by
-default, only the ``--vanilla`` flag. If you want to pass arguments to the script via
-the command line, use
+The decorator can be used to pass command line arguments to ``Rscript``. See the
+following example.
 
 .. code-block:: python
 
-    @pytask.mark.r(["--vanilla", "value"])
+    @pytask.mark.r("value")
     @pytask.mark.depends_on("script.r")
     @pytask.mark.produces("out.rds")
     def task_run_r_script():
@@ -140,7 +142,7 @@ And in your ``script.r``, you can intercept the value with
 .. code-block:: r
 
     args <- commandArgs(trailingOnly=TRUE)
-    arg <- args[1]  # ``arg`` holds ``"value"``
+    arg <- args[1]  # holds ``"value"``
 
 
 Parametrization
@@ -153,13 +155,23 @@ The following task executes two R scripts which produce different outputs.
 
 .. code-block:: python
 
+    from src.config import BLD, SRC
+
+
     @pytask.mark.r
     @pytask.mark.parametrize(
-        "depends_on, produces", [("script_1.r", "1.rds"), ("script_2.r", "2.rds")]
+        "depends_on, produces",
+        [(SRC / "script_1.r", BLD / "1.rds"), (SRC / "script_2.r", BLD / "2.rds")],
     )
     def task_execute_r_script():
         pass
 
+And the R script includes something like
+
+.. code-block:: r
+
+    args <- commandArgs(trailingOnly=TRUE)
+    produces <- args[1]  # holds the path
 
 If you want to pass different command line arguments to the same R script, you have to
 include the ``@pytask.mark.r`` decorator in the parametrization just like with
@@ -170,10 +182,7 @@ include the ``@pytask.mark.r`` decorator in the parametrization just like with
     @pytask.mark.depends_on("script.r")
     @pytask.mark.parametrize(
         "produces, r",
-        [
-            ("output_1.rds", (["--vanilla", "1"],)),
-            ("output_2.rds", (["--vanilla", "2"],)),
-        ],
+        [(BLD / "output_1.rds", "1"), (BLD / "output_2.rds", "2")],
     )
     def task_execute_r_script():
         pass
@@ -199,7 +208,7 @@ The plugin is a convenient wrapper around
 
     import subprocess
 
-    subprocess.run(["Rscript", "--vanilla", "script.r"], check=True)
+    subprocess.run(["Rscript", "script.r"], check=True)
 
 to which you can always resort to when the plugin does not deliver functionality you
 need.

setup.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.0.5
+current_version = 0.0.6
 parse = (?P<major>\d+)\.(?P<minor>\d+)(\.(?P<patch>\d+))(\-?((dev)?(?P<dev>\d+))?)
 serialize = 
 	{major}.{minor}.{patch}dev{dev}

setup.py

@@ -3,7 +3,7 @@
 
 setup(
     name="pytask-r",
-    version="0.0.5",
+    version="0.0.6",
     packages=find_packages(where="src"),
     package_dir={"": "src"},
     entry_points={"pytask": ["pytask_r = pytask_r.plugin"]},

src/pytask_r/__init__.py

@@ -1 +1 @@
-__version__ = "0.0.5"
+__version__ = "0.0.6"

src/pytask_r/collect.py

@@ -2,9 +2,9 @@
 import copy
 import functools
 import subprocess
-from pathlib import Path
 from typing import Iterable
 from typing import Optional
+from typing import Sequence
 from typing import Union
 
 from _pytask.config import hookimpl
@@ -24,15 +24,14 @@ def r(options: Optional[Union[str, Iterable[str]]] = None):
         One or multiple command line options passed to Rscript.
 
     """
-    if options is None:
-        options = ["--vanilla"]
-    elif isinstance(options, str):
-        options = [options]
+    options = [] if options is None else _to_list(options)
+    options = [str(i) for i in options]
     return options
 
 
 def run_r_script(r):
     """Run an R script."""
+    print("Executing " + " ".join(r) + ".")  # noqa: T001
     subprocess.run(r, check=True)
 
 
@@ -75,9 +74,8 @@ def pytask_collect_task_teardown(session, task):
 
 
 def _get_node_from_dictionary(obj, key, fallback=0):
-    if isinstance(obj, Path):
-        pass
-    elif isinstance(obj, dict):
+    """Get node from dictionary."""
+    if isinstance(obj, dict):
         obj = obj.get(key) or obj.get(fallback)
     return obj
 
@@ -99,4 +97,30 @@ def _prepare_cmd_options(session, task, args):
 
     """
     source = _get_node_from_dictionary(task.depends_on, session.config["r_source_key"])
-    return ["Rscript", source.value.as_posix(), *args]
+    return ["Rscript", source.path.as_posix(), *args]
+
+
+def _to_list(scalar_or_iter):
+    """Convert scalars and iterables to list.
+
+    Parameters
+    ----------
+    scalar_or_iter : str or list
+
+    Returns
+    -------
+    list
+
+    Examples
+    --------
+    >>> _to_list("a")
+    ['a']
+    >>> _to_list(["b"])
+    ['b']
+
+    """
+    return (
+        [scalar_or_iter]
+        if isinstance(scalar_or_iter, str) or not isinstance(scalar_or_iter, Sequence)
+        else list(scalar_or_iter)
+    )

src/pytask_r/parametrize.py

@@ -8,4 +8,4 @@ def pytask_parametrize_kwarg_to_marker(obj, kwargs):
     """Attach parametrized r arguments to the function with a marker."""
     if callable(obj):
         if "r" in kwargs:
-            mark.r(*kwargs.pop("r"))(obj)
+            mark.r(kwargs.pop("r"))(obj)

tests/test_collect.py

@@ -24,7 +24,7 @@ def task_dummy():
 @pytest.mark.parametrize(
     "r_args, expected",
     [
-        (None, ["--vanilla"]),
+        (None, []),
         ("--some-option", ["--some-option"]),
         (["--a", "--b"], ["--a", "--b"]),
     ],
@@ -132,7 +132,7 @@ def test_prepare_cmd_options(args, r_source_key):
     session.config = {"r_source_key": r_source_key}
 
     node = DummyClass()
-    node.value = Path("script.r")
+    node.path = Path("script.r")
     task = DummyClass()
     task.depends_on = {r_source_key: node}
     task.name = "task"

tests/test_parallel.py

@@ -82,15 +82,14 @@ def test_parallel_parametrization_over_source_file(runner, tmp_path):
 
     source = """
     import pytask
+    from pathlib import Path
+
+    SRC = Path(__file__).parent
 
     @pytask.mark.depends_on("script.r")
-    @pytask.mark.parametrize(
-        "produces, r",
-        [
-            ("output_1.rds", (["--vanilla", "1"],)),
-            ("output_2.rds", (["--vanilla", "2"],)),
-        ],
-    )
+    @pytask.mark.parametrize("produces, r", [
+        (SRC / "0.rds", (1, SRC / "0.rds")), (SRC / "1.rds", (1, SRC / "1.rds"))
+    ])
     def task_execute_r_script():
         pass
     """
@@ -99,8 +98,9 @@ def task_execute_r_script():
     r_script = """
     Sys.sleep(2)
     args <- commandArgs(trailingOnly=TRUE)
-    number <- args[2]
-    saveRDS(number, file=paste0("output_", number, ".rds"))
+    number <- args[1]
+    produces <- args[2]
+    saveRDS(number, file=produces)
     """
     tmp_path.joinpath("script.r").write_text(textwrap.dedent(r_script))
 
@@ -109,7 +109,7 @@ def task_execute_r_script():
     assert result.exit_code == 0
     duration_normal = time.time() - start
 
-    for name in ["output_1.rds", "output_2.rds"]:
+    for name in ["0.rds", "1.rds"]:
         tmp_path.joinpath(name).unlink()
 
     start = time.time()