---

part1/mutator_corpus_splicing.py

@@ -0,0 +1,69 @@
+import random
+import re
+
+_CORPUS: list[str] = []
+
+
+def _load_corpus() -> None:
+  candidate_paths = [
+    '/fuzzer/additional.sql',
+  ]
+  for p in candidate_paths:
+    if not p:
+      continue
+    try:
+      with open(p, 'r', encoding='latin-1') as f:
+        content = f.read()
+    except OSError:
+      continue
+    chunks = [c.strip() for c in content.split('---') if c.strip()]
+    chunks = [
+      c for c in chunks if any(not ln.strip().startswith('--') and ln.strip()
+                               for ln in c.split('\n'))
+    ]
+    if chunks:
+      _CORPUS.extend(chunks)
+      return
+
+
+_load_corpus()
+
+
+def mut_splice_append(s: str) -> str:
+  """Append a random chunk from the corpus"""
+
+  if not _CORPUS:
+    return s
+
+  return s + '\n' + random.choice(_CORPUS)
+
+
+def mut_splice_prepend(s: str) -> str:
+  """Prepend a random chunk from the corpus"""
+
+  if not _CORPUS:
+    return s
+
+  return random.choice(_CORPUS) + '\n' + s
+
+
+def mut_splice_replace(s: str) -> str:
+  """Replace s etirely with a random corpus chunk"""
+
+  if not _CORPUS:
+    return s
+
+  return random.choice(_CORPUS)
+
+
+def mut_splice_intersperse(s: str) -> str:
+  """Insert a random corpus chunk between two random statements of s"""
+
+  if not _CORPUS:
+    return s
+  parts = re.split(r'(?<=;)\s*\n', s)
+  if len(parts) < 2:
+    return s + '\n' + random.choice(_CORPUS)
+  i = random.randint(1, len(parts) - 1)
+
+  return '\n'.join(parts[:i] + [random.choice(_CORPUS)] + parts[i:])