From 9cba2be54e0494bca40fd03970a35b0c4d46bbd8 Mon Sep 17 00:00:00 2001
From: Jeff Smith <jeff@unbiasedgeek.com>
Date: Mon, 30 Mar 2026 09:57:31 -0600
Subject: [PATCH] feat: add disk usage summary

Uses du to show per-directory disk usage and highlights the top 5
largest directories.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
---
 luminos_lib/disk.py | 63 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 63 insertions(+)
 create mode 100644 luminos_lib/disk.py

diff --git a/luminos_lib/disk.py b/luminos_lib/disk.py
new file mode 100644
index 0000000..da069cc
--- /dev/null
+++ b/luminos_lib/disk.py
@@ -0,0 +1,63 @@
+"""Disk usage summary using du."""
+
+import subprocess
+
+
+def get_disk_usage(target, show_hidden=False):
+    """Get per-directory disk usage via du.
+
+    Returns a list of dicts: {path, size_bytes, size_human}.
+    """
+    cmd = ["du", "-b", "--max-depth=2", target]
+
+    try:
+        result = subprocess.run(
+            cmd, capture_output=True, text=True, timeout=30,
+        )
+    except (subprocess.TimeoutExpired, FileNotFoundError):
+        return []
+
+    entries = []
+    for line in result.stdout.strip().split("\n"):
+        if not line.strip():
+            continue
+        parts = line.split("\t", 1)
+        if len(parts) != 2:
+            continue
+        try:
+            size = int(parts[0])
+        except ValueError:
+            continue
+        path = parts[1]
+
+        # Skip hidden directories if not requested
+        if not show_hidden:
+            segments = path.replace(target, "").split("/")
+            if any(s.startswith(".") and s != "." for s in segments):
+                continue
+
+        entries.append({
+            "path": path,
+            "size_bytes": size,
+            "size_human": _human_size(size),
+        })
+
+    return entries
+
+
+def _human_size(nbytes):
+    """Convert bytes to human-readable string."""
+    for unit in ("B", "KB", "MB", "GB", "TB"):
+        if nbytes < 1024:
+            if unit == "B":
+                return f"{nbytes} {unit}"
+            return f"{nbytes:.1f} {unit}"
+        nbytes /= 1024
+    return f"{nbytes:.1f} PB"
+
+
+def top_directories(usage, n=5):
+    """Return the top n largest directories from usage data."""
+    # Exclude the root entry (last line of du is usually the total)
+    sorted_entries = sorted(usage, key=lambda x: x["size_bytes"], reverse=True)
+    return sorted_entries[:n]