Inclusion config

src/b2aiprep/commands.py

@@ -638,7 +638,7 @@ def validate_bundled_dataset(dataset_path, config_dir):
         with open(config_dir / "participants_to_remove.json") as f:
             participants_to_remove = set(json.load(f))
 
-        with open(config_dir / "sensitive_audio_tasks.json") as f:
+        with open(config_dir / "audio_tasks_to_include.json") as f:
             sensitive_tasks = set(json.load(f))
             sensitive_tasks_normalized = {
                 normalize_task_label(task)
@@ -809,7 +809,7 @@ def deidentify_bids_dataset(
     - participants_to_remove.json
     - audio_to_remove.json
     - id_remapping.json
-    - sensitive_audio_tasks.json
+    - audio_tasks_to_include.json
     """
     bids_path = Path(bids_path)
     deidentify_config_dir = Path(deidentify_config_dir)

src/b2aiprep/prepare/dataset.py

@@ -1701,18 +1701,18 @@ def load_audio_filestems_to_remove(publish_config_dir: Path) -> t.List[str]:
         return data
 
     @staticmethod
-    def load_sensitive_audio_tasks(deidentify_config_dir: Path) -> t.List[str]:
+    def load_audio_tasks_to_include(deidentify_config_dir: Path) -> t.List[str]:
         """Load list of audio tasks that are sensitive from JSON file."""
-        sensitive_audio_tasks_path = deidentify_config_dir / "sensitive_audio_tasks.json"
-        if not sensitive_audio_tasks_path.exists():
+        audio_tasks_to_include_path = deidentify_config_dir / "audio_tasks_to_include.json"
+        if not audio_tasks_to_include_path.exists():
             # If file doesn't exist, raise an error
-            raise FileNotFoundError(f"Sensitive audio tasks file {sensitive_audio_tasks_path} does not exist.")
+            raise FileNotFoundError(f"Sensitive audio tasks file {audio_tasks_to_include_path} does not exist.")
 
-        with open(sensitive_audio_tasks_path, 'r') as f:
+        with open(audio_tasks_to_include_path, 'r') as f:
             data = json.load(f)
 
         if not isinstance(data, list):
-            raise ValueError(f"Sensitive audio tasks file {sensitive_audio_tasks_path} should contain a list of audio task names.")
+            raise ValueError(f"Sensitive audio tasks file {audio_tasks_to_include_path} should contain a list of audio task names.")
 
         return data
 
@@ -1741,7 +1741,8 @@ def deidentify(self, outdir: t.Union[str, Path], deidentify_config_dir: Path, sk
         participant_ids_to_remap = BIDSDataset.load_remap_id_list(deidentify_config_dir)
         participant_ids_to_remove = BIDSDataset.load_participant_ids_to_remove(deidentify_config_dir)
         audio_filestems_to_remove = BIDSDataset.load_audio_filestems_to_remove(deidentify_config_dir)
-        sensitive_audio_tasks = BIDSDataset.load_sensitive_audio_tasks(deidentify_config_dir)
+        #sensitive_audio_tasks = BIDSDataset.load_sensitive_audio_tasks(deidentify_config_dir)
+        audio_tasks_to_include = BIDSDataset.load_audio_tasks_to_include(deidentify_config_dir)
         participant_session_id_to_remap = BIDSDataset.map_sequential_session_ids(self.data_path)
 
         # Allow users to specify filestems either in the original ID space (pre-deidentify)
@@ -1784,7 +1785,7 @@ def deidentify(self, outdir: t.Union[str, Path], deidentify_config_dir: Path, sk
                 outdir, 
                 participant_ids_to_remove, 
                 audio_filestems_to_remove,
-                sensitive_audio_tasks, 
+                audio_tasks_to_include, 
                 participant_ids_to_remap, 
                 participant_session_id_to_remap,
                 max_workers=max_workers,
@@ -1799,7 +1800,7 @@ def deidentify(self, outdir: t.Union[str, Path], deidentify_config_dir: Path, sk
                 outdir, 
                 participant_ids_to_remove, 
                 audio_filestems_to_remove,
-                sensitive_audio_tasks, 
+                audio_tasks_to_include, 
                 participant_ids_to_remap, 
                 participant_session_id_to_remap,
                 max_workers=max_workers,
@@ -1843,7 +1844,6 @@ def _apply_exclusion_list_to_filepaths(
         exclusion = set(exclusion_list)
         if len(exclusion) == 0:
             return paths
-
         def _canonical_recording_stem(stem: str) -> str:
             """Canonicalize a BIDS-like recording stem for matching.
 
@@ -1894,6 +1894,29 @@ def _canonical_recording_stem(stem: str) -> str:
             )
         return paths
 
+    @staticmethod
+    def _grab_filepaths_list_to_include(
+        paths: t.List[Path],
+        inclusion_list: t.List[str],
+    ) -> t.List[Path]:
+        """Grabs filepaths based on overlap with a specified inclusion list."""
+        n = len(paths)
+        inclusion = set(inclusion_list)
+        if len(inclusion) == 0:
+            return paths
+
+        new_paths = []
+        for file in paths:
+            for incl in inclusion_list:
+                if normalize_task_label(incl) in normalize_task_label(file.stem):
+                    new_paths.append(file)
+
+        if len(new_paths) < n:
+            _LOGGER.info(
+                f"Removed {n - len(new_paths)} records due to not being part of inclusion list: {inclusion_list}."
+            )
+        return new_paths
+
     @staticmethod
     def _expand_filestems_for_deidentification(
         filestems: t.Iterable[str],
@@ -1996,7 +2019,7 @@ def _deidentify_audio_files(
         outdir: Path,
         exclude_participant_ids: t.List[str] = [],
         exclude_audio_filestems: t.List[str] = [],
-        sensitive_audio_task_list: t.List[str] = [],
+        audio_tasks_to_include_list: t.List[str] = [],
         participant_ids_to_remap: t.Dict[str, str] = {},
         participant_session_id_to_remap: t.Dict[str, str] = {},
         max_workers: int = 16,
@@ -2041,9 +2064,9 @@ def _deidentify_audio_files(
             audio_paths, exclusion_list=['audio-check'], exclusion_type='filestem_contains'
         )
 
-        sensitive_audio_task_list = [f"task-{normalize_task_label(task)}" for task in sensitive_audio_task_list]
-        audio_paths = BIDSDataset._apply_exclusion_list_to_filepaths(
-            audio_paths, exclusion_list=sensitive_audio_task_list, exclusion_type='filestem_contains'
+        sensitive_audio_task_list = [f"task-{normalize_task_label(task)}" for task in audio_tasks_to_include_list]
+        audio_paths = BIDSDataset._grab_filepaths_list_to_include(
+            audio_paths, inclusion_list=sensitive_audio_task_list
         )
 
         _LOGGER.info(f"Copying {len(audio_paths)} recordings.")
@@ -2085,7 +2108,7 @@ def _deidentify_feature_files(
         outdir: Path,
         exclude_participant_ids: t.List[str] = [],
         exclude_audio_filestems: t.List[str] = [],
-        sensitive_audio_task_list: t.List[str] = [],
+        audio_tasks_to_include_list: t.List[str] = [],
         participant_ids_to_remap: t.Dict[str, str] = {},
         participant_session_id_to_remap: t.Dict[str, str] = {},
         max_workers: int = 16,
@@ -2130,7 +2153,7 @@ def _deidentify_feature_files(
             paths, exclusion_list=['audio-check'], exclusion_type='filestem_contains'
         )
 
-        sensitive_task_labels = {normalize_task_label(t) for t in sensitive_audio_task_list}
+        audio_task_labels = {normalize_task_label(t) for t in audio_tasks_to_include_list}
 
         _LOGGER.info(f"Copying {len(paths)} feature files.")
 
@@ -2157,7 +2180,7 @@ def process_feature_file(features_path):
 
             # if it is not sensitive and we want to keep features, move all features over
             task_name = BIDSDataset._extract_task_name_from_path(features_path)
-            if normalize_task_label(task_name) not in sensitive_task_labels:
+            if normalize_task_label(task_name) in audio_task_labels:
                 shutil.copy(features_path, output_path)
             else:
                 features = torch.load(features_path, weights_only=False, map_location=torch.device('cpu'))

tests/conftest.py

@@ -22,7 +22,7 @@ def setup_publish_config(tmp_path):
         "audio_filestems_to_remove.json": [],
         "id_remapping.json": {},
         "participants_to_remove.json": [],
-        "sensitive_audio_tasks.json": [],
+        "audio_tasks_to_include.json": [],
     }
 
     for filename, content in defaults.items():