Move on stop_criterion to CLI

Author: riesentoaster

src/fandango/cli/parser.py

@@ -460,9 +460,9 @@ def _populate_fuzz_parser(parser: argparse.ArgumentParser) -> None:
     parser.add_argument(
         "--stop-criterion",
         type=str,
+        default=None,
         dest="stop_criterion",
-        default="lambda t: False",
-        help='stop criterion to be used. This is a lambda function which is run on every new solution. Example: `lambda t: t.to_string().startswith("abc")`',
+        help='stop criterion to be used. This is a lambda function which is run on every new solution. Example: `lambda t: t.to_string().startswith("abc")`. Is `eval()`ed, so be careful!',
     )
     parser.add_argument(
         "--stop-after-seconds",

src/fandango/cli/utils.py

@@ -93,7 +93,9 @@ def make_fandango_settings(
     _copy_setting(args, settings, "max_repetitions")
     _copy_setting(args, settings, "max_nodes")
     _copy_setting(args, settings, "max_node_rate")
-    _copy_setting(args, settings, "stop_criterion")
+    if hasattr(args, "stop_criterion") and args.stop_criterion is not None:
+        # previously is a str, we eval it into a function
+        settings["stop_criterion"] = eval(args.stop_criterion)
     _copy_setting(args, settings, "stop_after_seconds")
 
     if hasattr(args, "start_symbol") and args.start_symbol is not None:

src/fandango/evolution/evaluation.py

@@ -30,7 +30,7 @@ def __init__(
         diversity_k: int,
         diversity_weight: float,
         warnings_are_errors: bool = False,
-        stop_criterion: Optional[Callable[..., bool]] = None,
+        stop_criterion: Optional[Callable[[DerivationTree], bool]] = None,
     ):
         self._grammar = grammar
         self._soft_constraints: list[SoftValue] = []
@@ -258,9 +258,7 @@ def evaluate_individual(
         self._fitness_cache[key] = (fitness, failing_trees, suggestion)
         return fitness, failing_trees, suggestion
 
-    def evaluate_population(
-        self, population: list[DerivationTree]
-    ) -> Generator[
+    def evaluate_population(self, population: list[DerivationTree]) -> Generator[
         DerivationTree,
         None,
         list[tuple[DerivationTree, float, list[FailingTree], Suggestion]],
@@ -442,9 +440,7 @@ def evaluate_individual(
         self._fitness_cache[key] = (fitness, failing_trees, suggestion)
         return fitness, failing_trees, suggestion
 
-    def evaluate_population(
-        self, population: list[DerivationTree]
-    ) -> Generator[
+    def evaluate_population(self, population: list[DerivationTree]) -> Generator[
         DerivationTree,
         None,
         list[tuple[DerivationTree, float, list[FailingTree], Suggestion]],