From 0f3491b278c023191d1e6068baea36e09b033333 Mon Sep 17 00:00:00 2001
From: Peter Van Dyken <pvandyk2@uwo.ca>
Date: Fri, 26 Jul 2024 11:01:03 -0400
Subject: [PATCH] Deduplicate wildcards when parsing components

Specifying a wildcard multiple times leads to an unintuitive KeyError.
This includes re-specifying a wildcard defined in the config by the
developer.

Fix by using a set to deduplicate wildcards before parsing.

Resolves #422
---
 snakebids/core/input_generation.py      |  2 +-
 snakebids/tests/test_generate_inputs.py | 29 +++++++++++++++++++++++++
 2 files changed, 30 insertions(+), 1 deletion(-)

diff --git a/snakebids/core/input_generation.py b/snakebids/core/input_generation.py
index 2d8ee21f..c3f8f60f 100644
--- a/snakebids/core/input_generation.py
+++ b/snakebids/core/input_generation.py
@@ -599,7 +599,7 @@ def _get_component(
     for img in matching_files:
         wildcards: list[str] = [
             wildcard
-            for wildcard in component.get("wildcards", [])
+            for wildcard in set(component.get("wildcards", []))
             if wildcard in img.entities
         ]
         _logger.debug("Wildcards %s found entities for %s", wildcards, img.path)
diff --git a/snakebids/tests/test_generate_inputs.py b/snakebids/tests/test_generate_inputs.py
index 73963292..2f989a71 100644
--- a/snakebids/tests/test_generate_inputs.py
+++ b/snakebids/tests/test_generate_inputs.py
@@ -840,6 +840,35 @@ def test_filter_with_invalid_method_raises_error(self, tmpdir: Path, method: str
             generate_inputs(tmpdir, pybids_inputs)
 
 
+@settings(
+    deadline=800,
+    suppress_health_check=[
+        HealthCheck.function_scoped_fixture,
+        HealthCheck.too_slow,
+    ],
+    max_examples=1,
+)
+@given(dataset=sb_st.datasets_one_comp(unique=True))
+def test_duplicate_wildcards_does_not_create_error(
+    dataset: BidsDataset, bids_fs: Path, fakefs_tmpdir: Path
+):
+    root = tempfile.mkdtemp(dir=fakefs_tmpdir)
+    rooted = BidsDataset.from_iterable(
+        attrs.evolve(comp, path=os.path.join(root, comp.path))
+        for comp in dataset.values()
+    )
+    create_dataset(Path("/"), rooted)
+    config = create_snakebids_config(dataset)
+    wildcards = itx.first(config.values()).get("wildcards", [])
+    wildcards.append(wildcards[0])
+    reindexed = generate_inputs(
+        root,
+        config,
+    )
+    assert reindexed == rooted
+    assert reindexed.layout is not None
+
+
 class TestAbsentConfigEntries:
     def get_entities(self, root: Path):
         # Generate directory