Upload fso_content_shatterer.py with huggingface_hub

fso_content_shatterer.py

@@ -0,0 +1,79 @@
+import hashlib
+import re
+
+class ContentShatterer:
+    """
+    Law XII Component: The Atomic Shatterer
+    Breaks down flat files into multi-dimensional topological atoms.
+    """
+    def __init__(self, m=256, k=4):
+        self.m = m
+        self.k = k
+
+    def _get_coord(self, atom_data, target_fiber):
+        """Map an atom's data to a Z_m^k coordinate on a specific fiber."""
+        h = hashlib.sha256(atom_data.encode('utf-8') if isinstance(atom_data, str) else atom_data).digest()
+        coords = []
+        for i in range(self.k - 1):
+            coords.append(h[i % len(h)] % self.m)
+        
+        # Closure Lemma (Law III)
+        w = (target_fiber - sum(coords)) % self.m
+        coords.append(w)
+        return tuple(coords)
+
+    def shatter(self, filename, content):
+        """Shatters content based on file type."""
+        atoms = []
+        
+        if filename.endswith(('.py', '.js', '.c', '.rs', '.go')):
+            # Fiber 1: Code / Logic
+            fiber = 1
+            # Simple logical block splitting (e.g., by double newlines or common keywords)
+            blocks = re.split(r'\n\s*\n', content.decode('utf-8', errors='ignore'))
+            for b in blocks:
+                if b.strip():
+                    atoms.append({
+                        "data": b.strip(),
+                        "fiber": fiber,
+                        "coord": self._get_coord(b.strip(), fiber),
+                        "type": "logic_block"
+                    })
+                    
+        elif filename.endswith(('.json', '.csv', '.txt', '.pdf', '.md')):
+            # Fiber 2: Knowledge / Data
+            fiber = 2
+            text = content.decode('utf-8', errors='ignore')
+            # Split by sentences (simple version)
+            sentences = re.split(r'(?<=[.!?])\s+', text)
+            for s in sentences:
+                if s.strip():
+                    atoms.append({
+                        "data": s.strip(),
+                        "fiber": fiber,
+                        "coord": self._get_coord(s.strip(), fiber),
+                        "type": "knowledge_atom"
+                    })
+        else:
+            # Fiber 3: Aesthetics / Raw Binary
+            fiber = 3
+            # Chunking binary into 64-byte atoms
+            chunk_size = 64
+            for i in range(0, len(content), chunk_size):
+                chunk = content[i:i+chunk_size]
+                atoms.append({
+                    "data": chunk,
+                    "fiber": fiber,
+                    "coord": self._get_coord(chunk, fiber),
+                    "type": "aesthetic_chunk"
+                })
+        
+        return atoms
+
+if __name__ == "__main__":
+    shatterer = ContentShatterer()
+    test_text = "TGI is the future of intelligence. It maps raw data into geometric tori. Determinism is the key."
+    atoms = shatterer.shatter("test.txt", test_text.encode())
+    print(f"Shattered test.txt into {len(atoms)} atoms.")
+    for a in atoms:
+        print(f"  Atom: '{a['data'][:30]}...' -> Coord: {a['coord']}")