Fix critical race conditions and unsafe disk report command

services/disk_report.py

@@ -1,11 +1,48 @@
 import os
+import re
 from typing import Any
 
 from services.runner import run_cmd
 
 
 def _top_dirs_cmd(path: str, limit: int) -> list[str]:
-    return ["bash", "-lc", f"du -xhd1 {path} 2>/dev/null | sort -h | tail -n {limit}"]
+    _ = limit
+    return ["du", "-x", "-h", "-d", "1", path]
+
+
+_SIZE_RE = re.compile(r"^\s*([0-9]+(?:\.[0-9]+)?)([KMGTP]?)(i?B?)?$", re.IGNORECASE)
+
+
+def _size_to_bytes(value: str) -> float:
+    m = _SIZE_RE.match(value.strip())
+    if not m:
+        return -1.0
+    num = float(m.group(1))
+    unit = (m.group(2) or "").upper()
+    mul = {
+        "": 1,
+        "K": 1024,
+        "M": 1024**2,
+        "G": 1024**3,
+        "T": 1024**4,
+        "P": 1024**5,
+    }.get(unit, 1)
+    return num * mul
+
+
+def _format_top_dirs(raw: str, limit: int) -> str:
+    rows: list[tuple[float, str]] = []
+    for line in raw.splitlines():
+        line = line.strip()
+        if not line:
+            continue
+        parts = line.split(maxsplit=1)
+        if len(parts) != 2:
+            continue
+        size, name = parts
+        rows.append((_size_to_bytes(size), f"{size}\t{name}"))
+    rows.sort(key=lambda x: x[0])
+    return "\n".join(line for _sz, line in rows[-max(1, limit):])
 
 
 async def build_disk_report(cfg: dict[str, Any], mount: str, usage: int) -> str:
@@ -15,24 +52,27 @@ async def build_disk_report(cfg: dict[str, Any], mount: str, usage: int) -> str:
 
     rc, out = await run_cmd(_top_dirs_cmd(mount, limit), timeout=30)
     if rc == 0 and out.strip():
+        top_out = _format_top_dirs(out, limit)
         lines.append("")
         lines.append("Top directories:")
-        lines.append(out.strip())
+        lines.append(top_out)
 
     docker_dir = cfg.get("disk_report", {}).get("docker_dir", "/var/lib/docker")
     if docker_dir and os.path.exists(docker_dir):
         rc2, out2 = await run_cmd(_top_dirs_cmd(docker_dir, limit), timeout=30)
         if rc2 == 0 and out2.strip():
+            top_out2 = _format_top_dirs(out2, limit)
             lines.append("")
             lines.append(f"Docker dir: {docker_dir}")
-            lines.append(out2.strip())
+            lines.append(top_out2)
 
     logs_dir = cfg.get("disk_report", {}).get("logs_dir", "/var/log")
     if logs_dir and os.path.exists(logs_dir):
         rc3, out3 = await run_cmd(_top_dirs_cmd(logs_dir, limit), timeout=30)
         if rc3 == 0 and out3.strip():
+            top_out3 = _format_top_dirs(out3, limit)
             lines.append("")
             lines.append(f"Logs dir: {logs_dir}")
-            lines.append(out3.strip())
+            lines.append(top_out3)
 
     return "\n".join(lines)