feat(cadfs): expand selector provenance coverage

cadfs_to_cdsl/operation_registry.py

@@ -0,0 +1,88 @@
+"""Source-derived FeatureScript operation coverage inventory.
+
+The registry records what the current source corpus actually contains.  It is
+deliberately independent from conversion diagnostics: an old output directory
+cannot make an operation appear supported, and a planned operation cannot be
+reported as observed merely because its name is listed in the roadmap.
+"""
+
+from __future__ import annotations
+
+from collections import Counter, defaultdict
+from pathlib import Path
+from typing import Any, Iterable
+
+from .dataset import Sample
+from .featurescript_parser import parse_featurescript
+
+
+REGISTRY_SCHEMA = "cadfs_to_cdsl.operation_registry.v1"
+
+# These are roadmap work items, not parser aliases or support declarations.
+P3_OPERATION_WORK_ITEMS = (
+    "draft",
+    "thicken",
+    "split",
+    "moveFace",
+    "deleteFace",
+    "replaceFace",
+    "transform",
+    "derive",
+    "import",
+    "bend_add",
+)
+
+
+def build_operation_registry(samples: Iterable[Sample]) -> dict[str, Any]:
+    """Parse every available FeatureScript source and inventory feature calls.
+
+    Parse failures and missing FeatureScript modalities are recorded separately
+    instead of being treated as a zero-count result.  Sample examples remain
+    bounded only for report size; the count is computed from every feature.
+    """
+    counts: Counter[str] = Counter()
+    sample_ids: dict[str, set[str]] = defaultdict(set)
+    parse_failures: list[dict[str, str]] = []
+    source_sample_count = 0
+
+    for sample in samples:
+        source_path = sample.files.get("featurescript")
+        if not source_path:
+            continue
+        source_sample_count += 1
+        try:
+            model = parse_featurescript(Path(source_path).read_text(encoding="utf-8"), sample.sample_id)
+        except Exception as error:
+            parse_failures.append({"sample_id": sample.sample_id, "error": f"{type(error).__name__}: {error}"})
+            continue
+        for feature in model.features:
+            counts[feature.operation] += 1
+            sample_ids[feature.operation].add(sample.sample_id)
+
+    operations: dict[str, dict[str, Any]] = {}
+    for operation in sorted(counts):
+        operations[operation] = {
+            "state": "observed",
+            "feature_count": counts[operation],
+            "sample_count": len(sample_ids[operation]),
+            "sample_ids": sorted(sample_ids[operation]),
+        }
+    for operation in P3_OPERATION_WORK_ITEMS:
+        if operation in operations:
+            operations[operation]["roadmap_work_item"] = True
+        else:
+            operations[operation] = {
+                "state": "not_observed_in_current_source",
+                "feature_count": 0,
+                "sample_count": 0,
+                "sample_ids": [],
+                "roadmap_work_item": True,
+            }
+
+    return {
+        "schema": REGISTRY_SCHEMA,
+        "source_sample_count": source_sample_count,
+        "parse_failure_count": len(parse_failures),
+        "parse_failures": parse_failures,
+        "operations": operations,
+    }