From f7d561e42f97a437eb3250051ec8299309cd6a9d Mon Sep 17 00:00:00 2001
From: Abdessamad Derraz <3028866+Abdess@users.noreply.github.com>
Date: Wed, 21 Oct 2026 05:10:00 +0200
Subject: [PATCH] fix: store already-compressed pack members

Every member was deflated, archives and PNGs included. On the 2026-10-20
RetroArch pack the .zip, .7z, .chd, .png, .ogg and .pk4 members are 2.1 GB
that took 28 s to deflate again for 9.6 MB saved (0.45 %), and about 16 GB
for 72 MB across all packs. Those formats are now stored. The choice reads
the member name only, so a pack stays byte-reproducible.
---
 scripts/generate_pack.py        | 20 ++++++++++++++++--
 tests/test_deterministic_zip.py | 36 +++++++++++++++++++++++++++++++++
 2 files changed, 54 insertions(+), 2 deletions(-)

diff --git a/scripts/generate_pack.py b/scripts/generate_pack.py
index acf92ef9..8aa52869 100644
--- a/scripts/generate_pack.py
+++ b/scripts/generate_pack.py
@@ -105,6 +105,22 @@ DEFAULT_BIOS_DIR = "bios"
 
 
 
+# Formats that are already compressed: deflating them again took 28 s of the
+# RetroArch build for 2.1 GB of members and saved 9.6 MB (0.45 %), 16 GB and
+# about 72 MB across all packs. Measured per extension on the 2026-10-20
+# packs; .pk3 (3 %), .jpg (6 %), .avi (44 %) and .jar (60 %) still shrink
+# and stay deflated. Chosen by name alone, so a pack stays a function of its
+# inputs.
+_STORED_EXTENSIONS = frozenset({".7z", ".chd", ".ogg", ".pk4", ".png", ".zip"})
+
+
+def _member_compression(arcname: str) -> int:
+    """ZIP_STORED for a member whose format is already compressed."""
+    if os.path.splitext(arcname)[1].lower() in _STORED_EXTENSIONS:
+        return zipfile.ZIP_STORED
+    return zipfile.ZIP_DEFLATED
+
+
 def _add_pack_member(zf: zipfile.ZipFile, source_path: str, arcname: str) -> None:
     """Add a file to the pack with metadata that does not depend on the build.
 
@@ -117,7 +133,7 @@ def _add_pack_member(zf: zipfile.ZipFile, source_path: str, arcname: str) -> Non
     because some members are firmware images of several hundred megabytes.
     """
     info = zipfile.ZipInfo(filename=arcname, date_time=_FIXED_DATE_TIME)
-    info.compress_type = zipfile.ZIP_DEFLATED
+    info.compress_type = _member_compression(arcname)
     # The executable bit is the one permission worth carrying: the RetroDECK
     # pack ships the two Voxatron engine binaries, and a user who extracts
     # them at 644 cannot run them. Git records the bit, so reading it from
@@ -1218,7 +1234,7 @@ def _extract_zip_to_archive(
             member = zipfile.ZipInfo(
                 filename=target_path, date_time=_FIXED_DATE_TIME
             )
-            member.compress_type = zipfile.ZIP_DEFLATED
+            member.compress_type = _member_compression(target_path)
             member.external_attr = 0o100644 << 16
             target_zf.writestr(member, data)
 
diff --git a/tests/test_deterministic_zip.py b/tests/test_deterministic_zip.py
index 24ef3c0b..d5f09481 100644
--- a/tests/test_deterministic_zip.py
+++ b/tests/test_deterministic_zip.py
@@ -413,5 +413,41 @@ class ManifestInjectionIsAtomic(unittest.TestCase):
             )
 
 
+class PackMemberCompression(unittest.TestCase):
+    """A pack deflated every member, archives and PNGs included: 16 GB of
+    already-compressed data deflated again on each full build for 0.18 % of
+    the packs' size."""
+
+    def _pack(self, tmp: Path, name: str) -> Path:
+        from scripts.generate_pack import _add_pack_member
+
+        source = tmp / "member"
+        source.write_bytes(b"\x00" * 4096)
+        pack = tmp / name
+        with zipfile.ZipFile(pack, "w", zipfile.ZIP_DEFLATED) as zf:
+            for arcname in ("bios/a.zip", "bios/b.PNG", "bios/c.bin", "bios/d.pk3"):
+                _add_pack_member(zf, str(source), arcname)
+        return pack
+
+    def test_compressed_formats_are_stored_and_the_rest_deflated(self):
+        with tempfile.TemporaryDirectory() as tmp:
+            pack = self._pack(Path(tmp), "pack.zip")
+            with zipfile.ZipFile(pack) as zf:
+                kinds = {i.filename: i.compress_type for i in zf.infolist()}
+        self.assertEqual(kinds, {
+            "bios/a.zip": zipfile.ZIP_STORED,
+            "bios/b.PNG": zipfile.ZIP_STORED,
+            "bios/c.bin": zipfile.ZIP_DEFLATED,
+            "bios/d.pk3": zipfile.ZIP_DEFLATED,
+        })
+
+    def test_the_pack_stays_byte_reproducible(self):
+        with tempfile.TemporaryDirectory() as tmp:
+            first = self._pack(Path(tmp), "one.zip").read_bytes()
+            time.sleep(1.1)
+            second = self._pack(Path(tmp), "two.zip").read_bytes()
+        self.assertEqual(hashlib.sha256(first).hexdigest(), hashlib.sha256(second).hexdigest())
+
+
 if __name__ == "__main__":
     unittest.main()