SCHEMA: Add index_columns metadata, render in tables

src/schema/README.md

@@ -742,6 +742,7 @@ rule is applied.
     | `selectors`          | List of expressions; any evaluating false indicate rule does not apply                                         |
     | `columns`            | Object with keys that may be found in `objects.columns`, values either a requirement level or an object        |
     | `initial_columns`    | An optional list of columns that must be the first N columns of a file                                         |
+    | `index_columns`      | An optional list of columns that uniquely identify a row.                                                      |
     | `additional_columns` | Indicates whether additional columns may be defined. One of `allowed`, `allowed_if_defined` and `not_allowed`. |
 
 The following tables demonstrate how mutual exclusive, required fields, may be

src/schema/rules/tabular_data/derivatives/common_derivatives.yaml

@@ -9,3 +9,4 @@ SegmentationLookup:
     abbreviation: optional
     color: optional
     mapping: optional
+  index_columns: [index]

src/schema/rules/tabular_data/eeg.yaml

@@ -20,6 +20,7 @@ EEGChannels:
     notch: optional
     status: optional
     status_description: optional
+  index_columns: [name__channels]
   additional_columns: allowed_if_defined
 
 EEGElectrodes:
@@ -40,4 +41,5 @@ EEGElectrodes:
     type__electrodes: recommended
     material: recommended
     impedance: recommended
+  index_columns: [name__electrodes]
   additional_columns: allowed_if_defined

src/schema/rules/tabular_data/ieeg.yaml

@@ -29,6 +29,7 @@ iEEGChannels:
     notch: optional
     status: optional
     status_description: optional
+  index_columns: [name__channels]
   additional_columns: allowed_if_defined
 
 iEEGElectrodes:
@@ -60,4 +61,5 @@ iEEGElectrodes:
     type__electrodes: optional
     impedance: optional
     dimension: optional
+  index_columns: [name__electrodes]
   additional_columns: allowed_if_defined

src/schema/rules/tabular_data/meg.yaml

@@ -20,4 +20,5 @@ MEGChannels:
     software_filters: optional
     status: optional
     status_description: optional
+  index_columns: [name__channels]
   additional_columns: allowed_if_defined

src/schema/rules/tabular_data/modality_agnostic.yaml

@@ -15,23 +15,19 @@ Participants:
     handedness: recommended
     strain: recommended
     strain_rrid: recommended
+  index_columns: [participant_id]
   additional_columns: allowed
 
 Samples:
   selectors:
     - path == "samples.tsv"
   columns:
-    sample_id:
-      level: required
-      description_addendum: |
-        The combination of `sample_id` and `participant_id` MUST be unique.
-    participant_id:
-      level: required
-      description_addendum: |
-        The combination of `sample_id` and `participant_id` MUST be unique.
+    sample_id: required
+    participant_id: required
     sample_type: required
     pathology: recommended
     derived_from: recommended
+  index_columns: [sample_id, participant_id]
   additional_columns: allowed
 
 Scans:
@@ -46,6 +42,7 @@ Scans:
       description_addendum: |
         There MUST be exactly one row for each file.
     acq_time__scans: optional
+  index_columns: [filename]
   additional_columns: allowed
 
 Sessions:
@@ -61,4 +58,5 @@ Sessions:
         There MUST be exactly one row for each session.
     acq_time__sessions: optional
     pathology: recommended
+  index_columns: [session_id]
   additional_columns: allowed

tools/schemacode/bidsschematools/render/tables.py

@@ -179,6 +179,7 @@ def _make_table_from_rule(
             # Since only one table may be provided for columns, we can just use this directly.
             additional_columns = table_schema.get("additional_columns", "not_allowed")
             initial_columns = table_schema.get("initial_columns", [])
+            index_columns = table_schema.get("index_columns", [])
         else:
             raise ValueError(f"Unsupported 'table_type': '{table_type}'")
 
@@ -214,6 +215,14 @@ def _make_table_from_rule(
                 # Typically begins with "if"
                 level = f"{level} {level_addendum}"
 
+        if table_type == "columns" and element in index_columns:
+            if len(index_columns) == 1:
+                msg = f"Values in `{schema.objects.columns[element].name}`"
+            else:
+                cols = [f"`{schema.objects.columns[col].name}`" for col in index_columns]
+                msg = f"The combination of {', '.join(cols[:-1])} and {cols[-1]}"
+            description_addendum += f"\n\n\n\n{msg} MUST be unique."
+
         if table_type == "columns" and initial_columns:
             if element in initial_columns:
                 order = initial_columns.index(element) + 1