Add expression interpreter

pyproject.toml

@@ -24,7 +24,7 @@ requires-python = ">=3.10"
 dependencies = [
     "acres >=0.5",
     "attrs >=24.1",
-    "bidsschematools >=1.1",
+    "bidsschematools[expressions]>=1.1",
     "nibabel>=5.3.0",
     "orjson>=3.11.3",
     "universal_pathlib >=0.2.1",
@@ -102,6 +102,10 @@ exclude = ".*"
 # Ruff configured in ruff.toml
 
 [dependency-groups]
+demo = [
+    "jupyterlab>=4.5.5",
+    "jupytext>=1.19.1",
+]
 dev = [
     "ipython>=8.37.0",
     "ruff>=0.15.5",

src/bids_validator/context.py

@@ -523,6 +523,56 @@ def sidecar(self) -> Namespace | None:
 
         return Namespace(sidecar)
 
+    def exists(self, arg: str | list[str] | None, rule: str | None = 'dataset') -> int:
+        """Count of files in an array that exist in the dataset. String is array with length 1."""
+        if arg is None:
+            return 0
+
+        prefix = UPath()
+        fileTree = self.file.parent if rule == 'file' else self.dataset.tree
+
+        if rule == 'stimuli':
+            prefix /= 'stimuli'
+        elif rule == 'subject':
+            if 'sub' not in self.entities:
+                return 0
+            prefix /= f'sub-{self.entities["sub"]}'
+
+        if isinstance(arg, str):
+            arg = [arg]
+
+        if rule == 'bids-uri':
+            valid_uris = (uri for uri in arg if uri.startswith('bids:') and uri.count(':') == 2)
+            uri_parts = [uri.split(':') for uri in valid_uris]
+
+            # Find number of files in this dataset that exist
+            # Once split these can be processed using dataset rule
+            dataset_files = self.exists([part[2] for part in uri_parts if part[1] == ''])
+
+            outside_files = [[part[1], part[2]] for part in uri_parts if part[1] != '']
+
+            # Find number of files in other dataset that exist
+            if outside_files:
+                # Resolve dataset links relative to the dataset path
+                dataset_links = getattr(self.dataset.dataset_description, 'DatasetLinks', {})
+                links = {
+                    key: (fileTree.path_obj / val).resolve() for key, val in dataset_links.items()
+                }
+
+                # For each uri with another dataset, generate the full path and sum the ones
+                # that exist
+                other_files = sum(
+                    UPath(links[ds_name], relpath).exists()
+                    for ds_name, relpath in outside_files
+                    if ds_name in links
+                )
+            else:
+                other_files = 0
+
+            return dataset_files + other_files
+        else:
+            return sum((prefix / file) in fileTree for file in arg)
+
 
 class Sessions:
     """Collections of sessions in subject."""