From e0e96d10d7ecdd767d36c7abe97f4be1e9ba04e5 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Andr=C3=A9=20Padez?= <andre.padez@gmail.com>
Date: Sat, 29 Aug 2026 01:58:30 +0100
Subject: [PATCH] slim-embedded-art: hash the audio region from mutagen own tag
 parse

A tagger that writes a non-standard size field puts the header claim and
the real tag end in different places. mutagen preserves audio from where
it parsed the tag to end; the guard hashed from where the header claimed,
and refused good rewrites across a whole album. Both hashes now use the
same authority. Failed files stay untouched and a re-run picks them up.
---
 scripts/slim-embedded-art.py | 31 ++++++++++++++++++++++---------
 1 file changed, 22 insertions(+), 9 deletions(-)

diff --git a/scripts/slim-embedded-art.py b/scripts/slim-embedded-art.py
index 5bec3cf..1cee63e 100755
--- a/scripts/slim-embedded-art.py
+++ b/scripts/slim-embedded-art.py
@@ -41,16 +41,29 @@ EMBED_TARGET = 80_000
 PROCESS_OVER = 100_000
 
 
-def audio_region_hash(path):
-    """SHA-256 of everything after the ID3v2 tag - the audio, byte-for-byte."""
+def audio_region_hash(path, offset=None):
+    """SHA-256 of everything after the ID3v2 tag - the audio, byte-for-byte.
+
+    The offset comes from mutagen's OWN parse of the tag when available. Reading the header's
+    syncsafe size field ourselves disagreed with mutagen on files whose tagger wrote a
+    non-standard size (Shamrocks and Shenanigans, 2026-08-29): mutagen preserved the audio from
+    where IT parsed the tag to end, our hash read from where the header claimed, and the guard
+    refused a perfectly good rewrite.
+    """
+    if offset is None:
+        try:
+            offset = ID3(path).size
+        except Exception:
+            offset = 0
+        if not offset:
+            with open(path, "rb") as f:
+                head = f.read(10)
+                if head[:3] == b"ID3":
+                    offset = 10 + ((head[6] << 21) | (head[7] << 14) | (head[8] << 7) | head[9])
+                    if head[5] & 0x10:  # v2.4 footer
+                        offset += 10
     with open(path, "rb") as f:
-        head = f.read(10)
-        off = 0
-        if head[:3] == b"ID3":
-            off = 10 + ((head[6] << 21) | (head[7] << 14) | (head[8] << 7) | head[9])
-            if head[5] & 0x10:  # v2.4 footer
-                off += 10
-        f.seek(off)
+        f.seek(offset)
         h = hashlib.sha256()
         while chunk := f.read(1 << 20):
             h.update(chunk)