Getting to the way it's supposed to be!

modules/ufbx/misc/split_content_test_gen.py

@@ -0,0 +1,37 @@
+import transmute_fbx as tfbx
+from dataclasses import replace
+
+Node = tfbx.Node
+Value = tfbx.Value
+
+divisor = 1
+
+def replace_content(node: Node) -> Node:
+    global divisor
+    if node.name == b"Content":
+        value = node.values[0]
+        values = []
+        data = value.value
+        size = len(data)
+        step = max(1, size // divisor)
+        for base in range(0, size, step):
+            values.append(Value(b"R", data[base:base+step]))
+        divisor *= 4
+        return replace(node, values=values)
+    return replace(node, children=[replace_content(n) for n in node.children])
+
+if __name__ == "__main__":
+    import argparse
+    parser = argparse.ArgumentParser(usage="transmute_fbx.py src -o dst -v 7400 -f binary-be")
+    parser.add_argument("src", help="Source file to read")
+    parser.add_argument("dst", help="Output filename")
+    argv = parser.parse_args()
+
+    with open(argv.src, "rb") as f:
+        fbx = tfbx.parse_fbx(f)
+
+    fbx = replace(fbx, root=replace_content(fbx.root))
+
+    with open(argv.dst, "wb") as f:
+        bf = tfbx.BinaryFormat(fbx.version, False, 1)
+        tfbx.binary_dump_root(f, fbx.root, bf, fbx.footer)