fix: Perform choices validation after parse/mapping time.

CHANGELOG.md

@@ -2,6 +2,10 @@
 
 ## 0.25
 
+### 0.25.1
+
+- fix: Perform choices validation after parse/mapping time.
+
 ### 0.25.0
 
 - feat: Add `Arg.propagate`.

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "cappa"
-version = "0.25.0"
+version = "0.25.1"
 description = "Declarative CLI argument parser."
 
 urls = {repository = "https://github.com/dancardin/cappa"}

src/cappa/arg.py

@@ -15,7 +15,7 @@
 from cappa.completion.completers import complete_choices
 from cappa.completion.types import Completion
 from cappa.env import Env
-from cappa.parse import evaluate_parse, parse_value
+from cappa.parse import Parser, evaluate_parse, parse_literal, parse_value
 from cappa.type_view import Empty, EmptyType, TypeView
 from cappa.typing import (
     Doc,
@@ -582,7 +582,18 @@ def infer_parse(arg: Arg, type_view: TypeView) -> Callable:
     else:
         parse = parse_value(type_view)
 
-    return evaluate_parse(parse, type_view)
+    # This is the original arg choices, e.g. explicitly provided. Inferred choices
+    # need to be handled internally to the parsers.
+    if arg.choices:
+        if not isinstance(parse, Sequence):
+            parse = [parse]
+
+        literal_type = typing.Literal[tuple(arg.choices)]  # type: ignore
+        parse = [*parse, parse_literal(literal_type)]  # type: ignore
+
+    return evaluate_parse(
+        typing.cast(Union[Sequence[Parser], Parser], parse), type_view
+    )
 
 
 def infer_help(arg: Arg, fallback_help: str | None) -> str | None:

src/cappa/argparse.py

@@ -250,8 +250,8 @@ def add_argument(
     elif is_positional and not arg.required:
         kwargs["nargs"] = "?"
 
-    if arg.choices:
-        kwargs["choices"] = arg.choices
+    # if arg.choices:
+    #     kwargs["choices"] = arg.choices
 
     deprecated_kwarg = add_deprecated_kwarg(arg)
     kwargs.update(deprecated_kwarg)

src/cappa/parse.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import enum
 import functools
 import types
 import typing
@@ -92,6 +93,9 @@ def parse_value(typ: MaybeTypeView) -> Parser[T]:
     if type_view.is_none_type:
         return parse_none
 
+    if type_view.is_subclass_of(enum.Enum):
+        return parse_enum(type_view)
+
     if type_view.is_subclass_of(datetime):
         return datetime.fromisoformat  # type: ignore
 
@@ -130,14 +134,24 @@ def literal_mapper(value):
             if raw_value == value:
                 return type_arg
 
-        options = ", ".join(f"'{t}'" for t in type_view.args)
-        raise ValueError(
-            f"Invalid choice: '{value}' (choose from literal values {options})"
-        )
+        raise choices_error(type_view.args, value)
 
     return literal_mapper
 
 
+def parse_enum(typ):
+    type_view = _as_type_view(typ)
+    choices = tuple(v.value for v in type_view.annotation)
+
+    def enum_mapper(value):
+        try:
+            return type_view.annotation(value)
+        except ValueError:
+            raise choices_error(choices, value)
+
+    return enum_mapper
+
+
 def parse_list(typ: MaybeTypeView[list[T]]) -> Parser[list[T]]:
     """Create a value parser for a list of given type `of_type`."""
     type_view = _as_type_view(typ)
@@ -278,3 +292,8 @@ def sequence_parsers(value):
         return result
 
     return sequence_parsers
+
+
+def choices_error(choices, value):
+    options = ", ".join(f"{t!r}" for t in choices)
+    return ValueError(f"Invalid choice: '{value}' (choose from {options})")

src/cappa/parser.py

@@ -578,15 +578,6 @@ def consume_arg(
     if orig_num_args == 1:
         if result:
             result = result[0]
-            if arg.choices and result not in arg.choices:
-                choices = ", ".join(f"'{c}'" for c in arg.choices)
-                raise BadArgumentError(
-                    f"Invalid choice: '{result}' (choose from {choices})",
-                    value=result,
-                    command=parse_state.current_command,
-                    arg=arg,
-                )
-
             if parse_state.provide_completions and not parse_state.has_values():
                 if arg.completion:
                     completions: list[Completion] | list[FileCompletion] = (
@@ -645,6 +636,7 @@ def consume_arg(
         ParseState: parse_state,
         Arg: arg,
         Value: Value(result),
+        typing.Any: result,
     }
     if option:
         fulfilled_deps[RawOption] = option

tests/arg/test_choices.py

@@ -72,3 +72,18 @@ class ArgTest:
 
     result = capsys.readouterr().out
     assert "Valid options: one, two." not in result
+
+
+@backends
+def test_literal_parse_sequence(backend):
+    @dataclass
+    class LiteralParse:
+        log_level: Annotated[
+            str,
+            cappa.Arg(
+                short="-L", long=True, choices=["DEBUG"], parse=[str.upper, str.strip]
+            ),
+        ] = "DEBUG"
+
+    result = parse(LiteralParse, "--log-level", "  debug  ", backend=backend)
+    assert result == LiteralParse(log_level="DEBUG")

tests/arg/test_literal.py

@@ -11,31 +11,52 @@
 from tests.utils import backends, parse
 
 
-@dataclass
-class ArgTest:
-    name: Union[Literal["one"], Literal["two"], Literal["three"], Literal[4]]
-
-
 @backends
 def test_valid(backend):
+    @dataclass
+    class ArgTest:
+        name: Literal["one", "two"]
+
     test = parse(ArgTest, "two", backend=backend)
     assert test.name == "two"
 
 
 @backends
 def test_valid_int(backend):
+    @dataclass
+    class ArgTest:
+        name: Literal["one", 4]
+
     test = parse(ArgTest, "4", backend=backend)
     assert test.name == 4
 
 
 @backends
 def test_invalid(backend):
+    @dataclass
+    class ArgTest:
+        name: Literal["one", "two", "three", 4]
+
+    with pytest.raises(cappa.Exit) as e:
+        parse(ArgTest, "thename", backend=backend)
+
+    message = str(e.value.message).lower()
+    assert "invalid choice: 'thename' (choose from 'one', 'two', 'three', 4)" in message
+
+
+@backends
+def test_unioned_literals(backend):
+    @dataclass
+    class ArgTest:
+        name: Union[Literal["one"], Literal["two"], Literal["three"], Literal[4]]
+
     with pytest.raises(cappa.Exit) as e:
         parse(ArgTest, "thename", backend=backend)
 
     message = str(e.value.message).lower()
     assert (
-        "invalid choice: 'thename' (choose from 'one', 'two', 'three', '4')" in message
+        "invalid value for 'name': possible variants\n - literal['one']: invalid choice: 'thename' (choose from 'one')\n - literal['two']: invalid choice: 'thename' (choose from 'two')\n - literal['three']: invalid choice: 'thename' (choose from 'three')\n - literal[4]: invalid choice: 'thename' (choose from 4)"
+        == message
     )
 
 
@@ -55,8 +76,21 @@ class Args:
     err = textwrap.dedent(
         """\
         Invalid value for '-f': Possible variants
-         - set[Literal['one', 'two']]: Invalid choice: 'three' (choose from literal values 'one', 'two')
+         - set[Literal['one', 'two']]: Invalid choice: 'three' (choose from 'one', 'two')
          - list[int]: invalid literal for int() with base 10: 'three'
          - <no value>"""
     )
     assert str(e.value.message).lower() == err.lower()
+
+
+@backends
+def test_literal_parse(backend):
+    @dataclass
+    class LiteralParse:
+        log_level: Annotated[
+            Literal["TRACE", "DEBUG", "INFO"],
+            cappa.Arg(short="-L", long=True, parse=str.upper),
+        ] = "INFO"
+
+    result = parse(LiteralParse, "--log-level=debug", backend=backend)
+    assert result == LiteralParse(log_level="DEBUG")

tests/arg/test_literal_intermixed_union.py

@@ -37,7 +37,7 @@ def test_invalid_string(backend):
     err = textwrap.dedent(
         """\
         Invalid value for 'name': Possible variants
-         - Literal['one']: Invalid choice: 'thename' (choose from literal values 'one')
+         - Literal['one']: Invalid choice: 'thename' (choose from 'one')
          - int: invalid literal for int() with base 10: 'thename'"""
     )
     assert err in str(e.value.message)

tests/arg/test_literal_mulitple_variant.py

@@ -26,6 +26,4 @@ def test_literal(backend):
         parse(ArgTest, "thename", backend=backend)
 
     message = str(e.value.message).lower()
-    assert (
-        "invalid choice: 'thename' (choose from 'one', 'two', 'three', '4')" in message
-    )
+    assert "invalid choice: 'thename' (choose from 'one', 'two', 'three', 4)" in message

tests/arg/test_literal_sequence.py

@@ -1,7 +1,6 @@
 from __future__ import annotations
 
 from dataclasses import dataclass
-from typing import Union
 
 import pytest
 from typing_extensions import Annotated, Literal
@@ -13,15 +12,15 @@
 @dataclass
 class ArgTest:
     list: Annotated[
-        list[Union[Literal["one"], Literal["two"], Literal["three"], Literal[4]]],
+        list[Literal["one", "two", "three", 4]],
         cappa.Arg(short=True, default=[]),
     ]
     tuple: Annotated[
-        tuple[Union[Literal["one"], Literal["two"], Literal["three"], Literal[4]], ...],
+        tuple[Literal["one", "two", "three", 4], ...],
         cappa.Arg(short=True, default=()),
     ]
     set: Annotated[
-        set[Union[Literal["one"], Literal["two"], Literal["three"], Literal[4]]],
+        set[Literal["one", "two", "three", 4]],
         cappa.Arg(short=True, default=set()),
     ]
 
@@ -50,4 +49,4 @@ def test_invalid(backend):
         parse(ArgTest, "-l", "one", "-l", "wat", backend=backend)
 
     message = str(e.value.message).lower()
-    assert "invalid choice: 'wat' (choose from 'one', 'two', 'three', '4')" in message
+    assert "invalid choice: 'wat' (choose from 'one', 'two', 'three', 4)" in message