From de2802b1ff2e91a9c52c6d43039b3bd96f63d713 Mon Sep 17 00:00:00 2001
From: "Jason R. Coombs" <jaraco@jaraco.com>
Date: Mon, 20 Feb 2023 11:27:01 -0500
Subject: [PATCH 1/4] Refactor zipfile._strip_extra to use higher level
 abstractions for extras instead of a heavy-state loop.

---
 Lib/zipfile/__init__.py | 59 ++++++++++++++++++++++++++---------------
 1 file changed, 37 insertions(+), 22 deletions(-)

diff --git a/Lib/zipfile/__init__.py b/Lib/zipfile/__init__.py
index 6e6211de6b1684..f342b1b4e9cf3b 100644
--- a/Lib/zipfile/__init__.py
+++ b/Lib/zipfile/__init__.py
@@ -188,28 +188,43 @@ class LargeZipFile(Exception):
 
 _DD_SIGNATURE = 0x08074b50
 
-_EXTRA_FIELD_STRUCT = struct.Struct('<HH')
-
-def _strip_extra(extra, xids):
-    # Remove Extra Fields with specified IDs.
-    unpack = _EXTRA_FIELD_STRUCT.unpack
-    modified = False
-    buffer = []
-    start = i = 0
-    while i + 4 <= len(extra):
-        xid, xlen = unpack(extra[i : i + 4])
-        j = i + 4 + xlen
-        if xid in xids:
-            if i != start:
-                buffer.append(extra[start : i])
-            start = j
-            modified = True
-        i = j
-    if not modified:
-        return extra
-    if start != len(extra):
-        buffer.append(extra[start:])
-    return b''.join(buffer)
+
+class _Extra(bytes):
+    FIELD_STRUCT = struct.Struct('<HH')
+
+    def __new__(cls, val, id=None):
+        return super().__new__(cls, val)
+
+    def __init__(self, val, id=None):
+        self.id = id
+
+    @classmethod
+    def read_one(cls, raw):
+        try:
+            xid, xlen = cls.FIELD_STRUCT.unpack(raw[:4])
+        except struct.error:
+            xid = None
+            xlen = 0
+        return cls(raw[:4+xlen], xid), raw[4+xlen:]
+
+    @classmethod
+    def split(cls, data):
+        while data:
+            extra, data = _Extra.read_one(data)
+            yield extra
+
+    @classmethod
+    def strip(cls, data, xids):
+        """Remove Extra fields with specified IDs."""
+        return b''.join(
+            ex
+            for ex in cls.split(data)
+            if ex.id not in xids
+        )
+
+
+_strip_extra = _Extra.strip
+
 
 def _check_zipfile(fp):
     try:

From 796f1dd02403e9358a2d7c6e58fbb88b03d50d1b Mon Sep 17 00:00:00 2001
From: "Jason R. Coombs" <jaraco@jaraco.com>
Date: Mon, 20 Feb 2023 12:00:16 -0500
Subject: [PATCH 2/4] Add blurb

---
 .../next/Library/2023-02-20-12-00-11.gh-issue-88233.o5Zb0t.rst  | 2 ++
 1 file changed, 2 insertions(+)
 create mode 100644 Misc/NEWS.d/next/Library/2023-02-20-12-00-11.gh-issue-88233.o5Zb0t.rst

diff --git a/Misc/NEWS.d/next/Library/2023-02-20-12-00-11.gh-issue-88233.o5Zb0t.rst b/Misc/NEWS.d/next/Library/2023-02-20-12-00-11.gh-issue-88233.o5Zb0t.rst
new file mode 100644
index 00000000000000..945f92d3dfa93b
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2023-02-20-12-00-11.gh-issue-88233.o5Zb0t.rst
@@ -0,0 +1,2 @@
+Refactored ``zipfile._strip_extra`` to use higher level abstactions for
+extras instead of a heavy-state loop.

From e6abc5543b9b35740a37ea3148cc900342aa429b Mon Sep 17 00:00:00 2001
From: "Jason R. Coombs" <jaraco@jaraco.com>
Date: Mon, 20 Feb 2023 12:02:37 -0500
Subject: [PATCH 3/4] Remove _strip_extra and use _Extra.strip directly.

---
 Lib/test/test_zipfile/test_core.py | 46 +++++++++++++++---------------
 Lib/zipfile/__init__.py            |  5 +---
 2 files changed, 24 insertions(+), 27 deletions(-)

diff --git a/Lib/test/test_zipfile/test_core.py b/Lib/test/test_zipfile/test_core.py
index e23f5c2a8556f2..fb95b453614aba 100644
--- a/Lib/test/test_zipfile/test_core.py
+++ b/Lib/test/test_zipfile/test_core.py
@@ -3023,14 +3023,14 @@ def test_no_data(self):
         b = s.pack(2, 0)
         c = s.pack(3, 0)
 
-        self.assertEqual(b'', zipfile._strip_extra(a, (self.ZIP64_EXTRA,)))
-        self.assertEqual(b, zipfile._strip_extra(b, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b'', zipfile._Extra.strip(a, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b, zipfile._Extra.strip(b, (self.ZIP64_EXTRA,)))
         self.assertEqual(
-            b+b"z", zipfile._strip_extra(b+b"z", (self.ZIP64_EXTRA,)))
+            b+b"z", zipfile._Extra.strip(b+b"z", (self.ZIP64_EXTRA,)))
 
-        self.assertEqual(b+c, zipfile._strip_extra(a+b+c, (self.ZIP64_EXTRA,)))
-        self.assertEqual(b+c, zipfile._strip_extra(b+a+c, (self.ZIP64_EXTRA,)))
-        self.assertEqual(b+c, zipfile._strip_extra(b+c+a, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b+c, zipfile._Extra.strip(a+b+c, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b+c, zipfile._Extra.strip(b+a+c, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b+c, zipfile._Extra.strip(b+c+a, (self.ZIP64_EXTRA,)))
 
     def test_with_data(self):
         s = struct.Struct("<HH")
@@ -3038,38 +3038,38 @@ def test_with_data(self):
         b = s.pack(2, 2) + b"bb"
         c = s.pack(3, 3) + b"ccc"
 
-        self.assertEqual(b"", zipfile._strip_extra(a, (self.ZIP64_EXTRA,)))
-        self.assertEqual(b, zipfile._strip_extra(b, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b"", zipfile._Extra.strip(a, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b, zipfile._Extra.strip(b, (self.ZIP64_EXTRA,)))
         self.assertEqual(
-            b+b"z", zipfile._strip_extra(b+b"z", (self.ZIP64_EXTRA,)))
+            b+b"z", zipfile._Extra.strip(b+b"z", (self.ZIP64_EXTRA,)))
 
-        self.assertEqual(b+c, zipfile._strip_extra(a+b+c, (self.ZIP64_EXTRA,)))
-        self.assertEqual(b+c, zipfile._strip_extra(b+a+c, (self.ZIP64_EXTRA,)))
-        self.assertEqual(b+c, zipfile._strip_extra(b+c+a, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b+c, zipfile._Extra.strip(a+b+c, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b+c, zipfile._Extra.strip(b+a+c, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b+c, zipfile._Extra.strip(b+c+a, (self.ZIP64_EXTRA,)))
 
     def test_multiples(self):
         s = struct.Struct("<HH")
         a = s.pack(self.ZIP64_EXTRA, 1) + b"a"
         b = s.pack(2, 2) + b"bb"
 
-        self.assertEqual(b"", zipfile._strip_extra(a+a, (self.ZIP64_EXTRA,)))
-        self.assertEqual(b"", zipfile._strip_extra(a+a+a, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b"", zipfile._Extra.strip(a+a, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b"", zipfile._Extra.strip(a+a+a, (self.ZIP64_EXTRA,)))
         self.assertEqual(
-            b"z", zipfile._strip_extra(a+a+b"z", (self.ZIP64_EXTRA,)))
+            b"z", zipfile._Extra.strip(a+a+b"z", (self.ZIP64_EXTRA,)))
         self.assertEqual(
-            b+b"z", zipfile._strip_extra(a+a+b+b"z", (self.ZIP64_EXTRA,)))
+            b+b"z", zipfile._Extra.strip(a+a+b+b"z", (self.ZIP64_EXTRA,)))
 
-        self.assertEqual(b, zipfile._strip_extra(a+a+b, (self.ZIP64_EXTRA,)))
-        self.assertEqual(b, zipfile._strip_extra(a+b+a, (self.ZIP64_EXTRA,)))
-        self.assertEqual(b, zipfile._strip_extra(b+a+a, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b, zipfile._Extra.strip(a+a+b, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b, zipfile._Extra.strip(a+b+a, (self.ZIP64_EXTRA,)))
+        self.assertEqual(b, zipfile._Extra.strip(b+a+a, (self.ZIP64_EXTRA,)))
 
     def test_too_short(self):
-        self.assertEqual(b"", zipfile._strip_extra(b"", (self.ZIP64_EXTRA,)))
-        self.assertEqual(b"z", zipfile._strip_extra(b"z", (self.ZIP64_EXTRA,)))
+        self.assertEqual(b"", zipfile._Extra.strip(b"", (self.ZIP64_EXTRA,)))
+        self.assertEqual(b"z", zipfile._Extra.strip(b"z", (self.ZIP64_EXTRA,)))
         self.assertEqual(
-            b"zz", zipfile._strip_extra(b"zz", (self.ZIP64_EXTRA,)))
+            b"zz", zipfile._Extra.strip(b"zz", (self.ZIP64_EXTRA,)))
         self.assertEqual(
-            b"zzz", zipfile._strip_extra(b"zzz", (self.ZIP64_EXTRA,)))
+            b"zzz", zipfile._Extra.strip(b"zzz", (self.ZIP64_EXTRA,)))
 
 
 if __name__ == "__main__":
diff --git a/Lib/zipfile/__init__.py b/Lib/zipfile/__init__.py
index f342b1b4e9cf3b..974aeaf5f9ffd5 100644
--- a/Lib/zipfile/__init__.py
+++ b/Lib/zipfile/__init__.py
@@ -223,9 +223,6 @@ def strip(cls, data, xids):
         )
 
 
-_strip_extra = _Extra.strip
-
-
 def _check_zipfile(fp):
     try:
         if _EndRecData(fp):
@@ -1950,7 +1947,7 @@ def _write_end_record(self):
             min_version = 0
             if extra:
                 # Append a ZIP64 field to the extra's
-                extra_data = _strip_extra(extra_data, (1,))
+                extra_data = _Extra.strip(extra_data, (1,))
                 extra_data = struct.pack(
                     '<HH' + 'Q'*len(extra),
                     1, 8*len(extra), *extra) + extra_data

From 980ae60b914acb0f4458fc2fcc5efb460df69090 Mon Sep 17 00:00:00 2001
From: "Jason R. Coombs" <jaraco@jaraco.com>
Date: Fri, 14 Jul 2023 12:52:25 -0400
Subject: [PATCH 4/4] Use memoryview to avoid unnecessary copies while
 splitting Extras.

---
 Lib/zipfile/__init__.py | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/Lib/zipfile/__init__.py b/Lib/zipfile/__init__.py
index cd1d7a0669867e..2c963de18e4f95 100644
--- a/Lib/zipfile/__init__.py
+++ b/Lib/zipfile/__init__.py
@@ -209,8 +209,10 @@ def read_one(cls, raw):
 
     @classmethod
     def split(cls, data):
-        while data:
-            extra, data = _Extra.read_one(data)
+        # use memoryview for zero-copy slices
+        rest = memoryview(data)
+        while rest:
+            extra, rest = _Extra.read_one(rest)
             yield extra
 
     @classmethod