From eefda2140e9dccd5a31f8cba2626476d48a8b3bd Mon Sep 17 00:00:00 2001
From: Uyanide <pywang0608@foxmail.com>
Date: Thu, 2 Apr 2026 10:48:26 +0200
Subject: [PATCH] feat: better cache stats

---
 lrx_cli/cache.py | 43 +++++++++++++++++++++++++++++++++++++++++++
 lrx_cli/cli.py   | 48 ++++++++++++++++++++++++++++++++++++++++--------
 pyproject.toml   |  2 +-
 uv.lock          |  2 +-
 4 files changed, 85 insertions(+), 10 deletions(-)

diff --git a/lrx_cli/cache.py b/lrx_cli/cache.py
index ef2e79d..c8f46a4 100644
--- a/lrx_cli/cache.py
+++ b/lrx_cli/cache.py
@@ -441,10 +441,53 @@ class CacheEngine:
                     "SELECT source, COUNT(*) FROM cache GROUP BY source"
                 ).fetchall()
             )
+            # Source × Status cross-tabulation
+            source_status = conn.execute(
+                "SELECT source, status, COUNT(*) FROM cache GROUP BY source, status"
+            ).fetchall()
+            # Confidence buckets (only for positive statuses)
+            confidence_rows = conn.execute(
+                "SELECT confidence FROM cache WHERE status IN (?, ?)",
+                (
+                    CacheStatus.SUCCESS_SYNCED.value,
+                    CacheStatus.SUCCESS_UNSYNCED.value,
+                ),
+            ).fetchall()
+
+        # Build source×status table: {source: {status: count}}
+        source_status_table: dict[str, dict[str, int]] = {}
+        for src, status, count in source_status:
+            source_status_table.setdefault(src, {})[status] = count
+
+        # Build confidence buckets
+        buckets = {
+            "legacy (NULL)": 0,
+            "0-24": 0,
+            "25-49": 0,
+            "50-79": 0,
+            "80-99": 0,
+            "100": 0,
+        }
+        for (conf,) in confidence_rows:
+            if conf is None:
+                buckets["legacy (NULL)"] += 1
+            elif conf >= 100:
+                buckets["100"] += 1
+            elif conf >= 80:
+                buckets["80-99"] += 1
+            elif conf >= 50:
+                buckets["50-79"] += 1
+            elif conf >= 25:
+                buckets["25-49"] += 1
+            else:
+                buckets["0-24"] += 1
+
         return {
             "total": total,
             "expired": expired,
             "active": total - expired,
             "by_status": by_status,
             "by_source": by_source,
+            "source_status": source_status_table,
+            "confidence_buckets": buckets,
         }
diff --git a/lrx_cli/cli.py b/lrx_cli/cli.py
index ff1ee9e..b1602de 100644
--- a/lrx_cli/cli.py
+++ b/lrx_cli/cli.py
@@ -354,14 +354,46 @@ def stats():
     print(f"Total entries : {s['total']}")
     print(f"Active        : {s['active']}")
     print(f"Expired       : {s['expired']}")
-    if s["by_status"]:
-        print("\nBy status:")
-        for status, count in s["by_status"].items():
-            print(f"  {status}: {count}")
-    if s["by_source"]:
-        print("\nBy source:")
-        for source, count in s["by_source"].items():
-            print(f"  {source}: {count}")
+
+    # Source × Status table
+    table = s.get("source_status", {})
+    if table:
+        all_statuses = sorted({st for row in table.values() for st in row})
+        # Short labels for column headers
+        short = {
+            "SUCCESS_SYNCED": "synced",
+            "SUCCESS_UNSYNCED": "unsynced",
+            "NOT_FOUND": "not_found",
+            "NETWORK_ERROR": "net_err",
+        }
+        headers = [short.get(st, st) for st in all_statuses]
+        sources = sorted(table.keys())
+        # Column widths
+        src_w = max(len(src) for src in sources)
+        src_w = max(src_w, 6)  # min width for "source" header
+        col_w = [max(len(h) if h else 0, 4) for h in headers]
+
+        print(
+            f"\n{'source':<{src_w}}  "
+            + "  ".join(f"{h:>{w}}" for h, w in zip(headers, col_w))
+        )
+        print("-" * src_w + "  " + "  ".join("-" * w for w in col_w))
+        for src in sources:
+            counts = [str(table[src].get(st, 0)) for st in all_statuses]
+            print(
+                f"{src:<{src_w}}  "
+                + "  ".join(f"{c:>{w}}" for c, w in zip(counts, col_w))
+            )
+
+    # Confidence distribution (positive entries only)
+    buckets = s.get("confidence_buckets", {})
+    non_empty = {k: v for k, v in buckets.items() if v > 0}
+    if non_empty:
+        label_w = max(len(k) for k in non_empty)
+        print("\nConfidence distribution (positive entries):")
+        for label, count in buckets.items():
+            if count > 0:
+                print(f"  {label:>{label_w}} : {count}")
 
 
 @cache_app.command
diff --git a/pyproject.toml b/pyproject.toml
index 4406265..b9772ec 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "lrx-cli"
-version = "0.3.0"
+version = "0.3.1"
 description = "Fetch line-synced lyrics for your music player."
 readme = "README.md"
 requires-python = ">=3.13"
diff --git a/uv.lock b/uv.lock
index 0d0e99d..4ccaaef 100644
--- a/uv.lock
+++ b/uv.lock
@@ -153,7 +153,7 @@ wheels = [
 
 [[package]]
 name = "lrx-cli"
-version = "0.3.0"
+version = "0.3.1"
 source = { editable = "." }
 dependencies = [
     { name = "cyclopts" },