Make LZMADecompressor.needs_input public; simplify implementation

Calling decompress() with one argument is a maintenance burden,
so mark it as deprecated in 3.16.

Author: encukou

Lib/_py_warnings.py

@@ -873,7 +873,8 @@ def wrapper(*args, **kwargs):
 _DEPRECATED_MSG = "{name!r} is deprecated and slated for removal in Python {remove}"
 
 
-def _deprecated(name, message=_DEPRECATED_MSG, *, remove, _version=sys.version_info):
+def _deprecated(name, message=_DEPRECATED_MSG, *, remove, _version=sys.version_info,
+                stacklevel=3):
     """Warn that *name* is deprecated or should be removed.
 
     RuntimeError is raised if *remove* specifies a major/minor tuple older than
@@ -889,7 +890,7 @@ def _deprecated(name, message=_DEPRECATED_MSG, *, remove, _version=sys.version_i
         raise RuntimeError(msg)
     else:
         msg = message.format(name=name, remove=remove_formatted)
-        _wm.warn(msg, DeprecationWarning, stacklevel=3)
+        _wm.warn(msg, DeprecationWarning, stacklevel=stacklevel)
 
 
 # Private utility function called by _PyErr_WarnUnawaitedCoroutine

Lib/test/test_zipfile/test_core.py

@@ -33,7 +33,9 @@
     with_source_date_epoch, without_source_date_epoch,
 )
 from test.support.import_helper import ensure_lazy_imports
-from test.support.warnings_helper import check_no_resource_warning
+from test.support.warnings_helper import (
+    check_no_resource_warning, ignore_warnings,
+)
 
 
 TESTFN2 = TESTFN + "2"
@@ -4920,8 +4922,7 @@ class MonkeypatchedDecompressorTests(unittest.TestCase):
     # Some third-party projects monkey-patch _get_decompressor() to add
     # additional compression schemes. This can break at any time as the
     # internal compressor objects change.
-    # To protect users, we try to keep this case working (until it becomes
-    # too big of a burden, or we make the API public).
+    # To protect users, we try to keep this case working (see ).
     # See also: GH-156002 and GH-113756.
     COMPRESSION = 99
 
@@ -4972,7 +4973,9 @@ def test_roundtrip_monkeypatched_decompressor(self):
         with zipfile.ZipFile(buf, "w", compression=self.COMPRESSION) as zf:
             zf.writestr("member", data)
         self.assertIn(data.swapcase(), buf.getvalue())
-        with zipfile.ZipFile(io.BytesIO(buf.getvalue())) as zf:
+        with (ignore_warnings(category=DeprecationWarning,
+                              message='.*two argumentzs.*'),
+              zipfile.ZipFile(io.BytesIO(buf.getvalue())) as zf):
             self.assertEqual(zf.read("member"), data)
             with zf.open("member") as f:
                 self.assertEqual(f.read(100), data[:100])

Lib/zipfile/__init__.py

@@ -802,7 +802,7 @@ def unused_data(self):
             return b''
 
     @property
-    def _needs_input(self):
+    def needs_input(self):
         # While the LZMA properties header is still being buffered, more input
         # is required; afterwards defer to the wrapped decompressor so a bounded
         # decompress() call can be drained across reads.
@@ -893,25 +893,6 @@ def _get_compressor(compress_type, compresslevel=None):
         return None
 
 
-def _decompressor_needs_input(decompressor, default):
-    # bz2/zstd expose the stdlib decompressor's public needs_input; the LZMA
-    # wrapper keeps it private (_needs_input) to avoid adding public API.
-    # A decompressor with neither attribute reports *default*.
-    needs_input = getattr(decompressor, "needs_input", None)
-    if needs_input is None:
-        needs_input = getattr(decompressor, "_needs_input", default)
-    return needs_input
-
-
-def _decompressor_bounds_output(decompressor):
-    # The stdlib bzip2/LZMA/Zstandard decompressors report needs_input and
-    # accept decompress(data, max_length).  A third-party decompressor
-    # installed by replacing _get_decompressor() may support neither; it is
-    # then read unbounded, as before the bounded-decompression fix.
-    return (hasattr(decompressor, "needs_input")
-            or hasattr(decompressor, "_needs_input"))
-
-
 def _get_decompressor(compress_type):
     _check_compression(compress_type)
     if compress_type == ZIP_STORED:
@@ -1019,7 +1000,7 @@ def __init__(self, fileobj, mode, zipinfo, pwd=None,
         self._compress_left = zipinfo.compress_size
         self._left = zipinfo.file_size
 
-        self._set_decompressor()
+        self._decompressor = _get_decompressor(self._compress_type)
 
         self._eof = False
         self._readbuffer = b''
@@ -1202,11 +1183,6 @@ def read1(self, n):
                     break
         return buf
 
-    def _set_decompressor(self):
-        self._decompressor = _get_decompressor(self._compress_type)
-        self._decompress_bounded = _decompressor_bounds_output(
-            self._decompressor)
-
     def _read1(self, n):
         # Read up to n compressed bytes with at most one read() system call,
         # decrypt and decompress them.
@@ -1224,7 +1200,7 @@ def _read1(self, n):
         else:
             # bzip2/lzma/zstd: a bounded decompress() call may leave input
             # buffered inside the decompressor; drain that before reading more.
-            if _decompressor_needs_input(self._decompressor, default=True):
+            if getattr(self._decompressor, "needs_input", True):
                 data = self._read2(n)
             else:
                 data = b''
@@ -1239,18 +1215,27 @@ def _read1(self, n):
                          not self._decompressor.unconsumed_tail)
             if self._eof:
                 data += self._decompressor.flush()
-        elif self._decompress_bounded:
+        else:
             # Bound the output of a single decompress() call (mirroring the
             # DEFLATE path above) so that a small compressed member cannot
             # expand into one unbounded read.
-            data = self._decompressor.decompress(data, max(n, self.MIN_READ_SIZE))
+            try:
+                data = self._decompressor.decompress(data, max(n, self.MIN_READ_SIZE))
+            except TypeError:
+                # See MonkeypatchedDecompressorTests in test_core.py
+                warnings._deprecated(
+                    'one-argument decompress()',
+                    'The decompress() method of '
+                    + type(self._decompressor).__name__
+                    + ' should take two arguments, data and max_length.'
+                    + ' One-argument calls will stop working before'
+                    + ' Python 3.21.',
+                    remove=(3, 21),
+                    stacklevel=4)
+                data = self._decompressor.decompress(data)
             self._eof = (self._decompressor.eof or
                          self._compress_left <= 0 and
-                         _decompressor_needs_input(self._decompressor,
-                                                   default=True))
-        else:
-            data = self._decompressor.decompress(data)
-            self._eof = self._decompressor.eof or self._compress_left <= 0
+                         getattr(self._decompressor, "needs_input", True))
 
         data = data[:self._left]
         self._left -= len(data)
@@ -1339,7 +1324,7 @@ def seek(self, offset, whence=os.SEEK_SET):
             self._left = self._orig_file_size
             self._readbuffer = b''
             self._offset = 0
-            self._set_decompressor()
+            self._decompressor = _get_decompressor(self._compress_type)
             self._eof = False
             read_offset = new_pos
             if self._decrypter is not None:

Misc/NEWS.d/next/Library/2026-09-08-13-06-29.gh-issue-156002.vmOC8T.rst

@@ -1,5 +1,6 @@
 :mod:`zipfile` again reads members through a third-party decompressor
-installed by replacing the private ``_get_decompressor()``, unbounded as
-before the bounded-decompression fix, instead of raising
-:exc:`AttributeError`. The bound still applies to the standard library's
-bzip2, LZMA and Zstandard decompressors.
+installed by monkey-patching the private ``_get_decompressor()`` to return an
+object that only implements old BZ2Decompressor API from Python 3.3.
+Calling to decompress() with one argument is deprecated.
+Note that decompressors without ``needs_input`` and two-argument
+``decompress()`` are vulnerable to :cve:`2026-15310`.