[split-upload 1/6] f2fd4fa backup via proxy

docs_repair_del.py

@@ -0,0 +1,56 @@
+# -*- coding: utf-8 -*-
+"""
+删除 A类(186) + B1类(36) 共 222 个已验证重复旧文件。
+铁律：仅当「新结构副本存在」时才删除旧文件，否则跳过并报错，绝不盲删。
+C类(39) 与 b2(9) 不在此脚本处理范围。
+"""
+import json, os
+
+ROOT = "docs"
+
+d = json.load(open('docs_cmp_result2.json', encoding='utf-8'))
+content_dup = d['content_dup']      # [old, new]
+name_diff = d['name_diff']          # [old, om, os, new, nm, ns]
+
+# b2 旧路径集合（已抢救，不可删）
+b3 = json.load(open('docs_cmp_result3.json', encoding='utf-8'))
+b2_old = set()
+for row in b3.get('b2', []):
+    b2_old.add(row[0])
+
+def norm(p):
+    return os.path.normpath(os.path.join(ROOT, p))
+
+delete_list = []   # (old_abs, reason, new_abs)
+# A类
+for old, new in content_dup:
+    delete_list.append((norm(old), 'A-content_dup', norm(new)))
+# B1类 = name_diff 中不在 b2 的
+b1_count = 0
+for row in name_diff:
+    old, new = row[0], row[3]
+    if old in b2_old:
+        continue
+    delete_list.append((norm(old), 'B1-name_diff', norm(new)))
+    b1_count += 1
+
+print("待删总数:", len(delete_list), "(A=%d, B1=%d)" % (len(content_dup), b1_count))
+
+deleted, skipped = [], []
+for old_abs, reason, new_abs in delete_list:
+    if not os.path.isfile(new_abs):
+        skipped.append((old_abs, "NEW_MISSING:" + new_abs))
+        continue
+    if not os.path.isfile(old_abs):
+        skipped.append((old_abs, "OLD_MISSING"))
+        continue
+    try:
+        os.remove(old_abs)
+        deleted.append((old_abs, reason))
+    except Exception as e:
+        skipped.append((old_abs, "ERR:" + str(e)))
+
+print("DELETED:", len(deleted))
+print("SKIPPED:", len(skipped))
+for p, r in skipped:
+    print("  SKIP", p, "::", r)