Advance simulation parity and 3-axis tutorial coverage

scripts/generate_sim_case_catalog.py

@@ -6,6 +6,7 @@ from __future__ import annotations
 import argparse
 import hashlib
 import json
+import subprocess
 from pathlib import Path
 from typing import Any
 
@@ -17,6 +18,12 @@ HAL_REPORT = Path("coverage/hal-component-manifest.report.json")
 OUTPUT = Path("web/workspace/sim-case-catalog.mjs")
 REPORT = Path("coverage/sim-case-catalog.report.json")
 SCHEMA = "linuxcnc-sim-case-catalog.v1"
+RUNTIME_FIXTURE_FILES = (
+    "linuxcnc/configs/sim/axis/minimal_xyz_cmds.hal",
+    "linuxcnc/configs/sim/axis/ja_tests/xz/xz_cmds.hal",
+    "linuxcnc/configs/sim/axis/ja_tests/xz/xz_both_cmds.hal",
+    "linuxcnc/configs/sim/axis/ja_tests/xz/xz_both_auto_cmds.hal",
+)
 
 
 def canonical(value: Any) -> str:
@@ -32,6 +39,44 @@ def artifact(root: Path, path: str) -> dict[str, Any]:
     return {"path": path, "status": "missing"} if not value.is_file() else {"path": path, "bytes": value.stat().st_size, "sha256": sha256(value.read_bytes()), "status": "present"}
 
 
+def source_files(root: Path, directory: Path) -> list[Path]:
+    """Index source files without native runtime outputs.
+
+    Native LinuxCNC creates ignored parameter files, backups, bytecode, and
+    generated HAL fixtures in the source tree.  A filesystem glob would make
+    the browser manifest depend on the last native run, so use the submodule's
+    tracked set whenever Git metadata is available.  The small set of
+    LinuxCNC runtime HAL fixtures is added explicitly below because those
+    files are consumed by the loader's existing runtime-fixture contract.
+    """
+    base = root / directory
+    git_root = root / "linuxcnc"
+    if (git_root / ".git").exists():
+        prefix = directory.as_posix().removeprefix("linuxcnc/")
+        tracked_result = subprocess.run(
+            ["git", "-C", str(git_root), "ls-files", "--cached", "--", prefix],
+            check=True,
+            capture_output=True,
+            text=True,
+        )
+        tracked = set(tracked_result.stdout.splitlines())
+        result = subprocess.run(
+            ["git", "-C", str(git_root), "ls-files", "--cached", "--others", "--exclude-standard", "--", prefix],
+            check=True,
+            capture_output=True,
+            text=True,
+        )
+        return [
+            git_root / line
+            for line in result.stdout.splitlines()
+            # Native startup leaves empty tool.tbl files behind; only tracked
+            # tables are source inputs, while missing tables are materialized
+            # by the Web loader at their configured path.
+            if (git_root / line).is_file() and (not line.endswith(".tbl") or line in tracked)
+        ]
+    return [path for path in sorted(base.rglob("*")) if path.is_file() and "__pycache__" not in path.parts and path.suffix != ".pyc"]
+
+
 def ini_summary(path: Path) -> dict[str, str | None]:
     section = ""
     values: dict[str, str] = {}
@@ -61,7 +106,7 @@ def generate(root: Path, output: Path, report_output: Path = REPORT) -> dict[str
         if path.startswith("linuxcnc/configs/sim/"):
             matrix_by_path.setdefault(path, row)
     cases = []
-    for path in sorted((root / SIM_ROOT).rglob("*.ini")):
+    for path in sorted(path for path in source_files(root, SIM_ROOT) if path.suffix == ".ini"):
         relative = path.relative_to(root).as_posix()
         row = matrix_by_path.get(relative, {})
         summary = ini_summary(path)
@@ -81,12 +126,13 @@ def generate(root: Path, output: Path, report_output: Path = REPORT) -> dict[str
             "unresolved_count": len(row.get("unresolved_resources", [])),
             "source_sha256": sha256(path.read_bytes()),
         })
-    file_index = [path.relative_to(root).as_posix() for path in sorted((root / SIM_ROOT).rglob("*")) if path.is_file()]
+    file_index = [path.relative_to(root).as_posix() for path in sorted(source_files(root, SIM_ROOT))]
     for support_root in SUPPORT_ROOTS:
-        file_index.extend(path.relative_to(root).as_posix() for path in sorted((root / support_root).rglob("*")) if path.is_file())
+        file_index.extend(path.relative_to(root).as_posix() for path in sorted(source_files(root, support_root)))
     shared_asset = root / "linuxcnc/linuxcnc.gif"
     if shared_asset.is_file():
         file_index.append(shared_asset.relative_to(root).as_posix())
+    file_index.extend(path for path in RUNTIME_FIXTURE_FILES if (root / path).is_file())
     file_index = sorted(set(file_index))
     hal_report = json.loads((root / HAL_REPORT).read_text(encoding="utf-8"))
     hal_components = sorted({str(row["component"]) for row in hal_report.get("component_manifest", []) if row.get("component")})