diff --git a/pypdf/_reader.py b/pypdf/_reader.py
index 54fb33f1b..034a0d091 100644
--- a/pypdf/_reader.py
+++ b/pypdf/_reader.py
@@ -677,6 +677,14 @@ def _read_standard_xref_table(self, stream: StreamType) -> None:
             read_non_whitespace(stream)
             stream.seek(-1, 1)
             size = cast(int, read_object(stream, self))
+            if not isinstance(size, int):
+                logger_warning(
+                    "Invalid/Truncated xref table. Rebuilding it.",
+                    __name__,
+                )
+                self._rebuild_xref_table(stream)
+                stream.read()
+                return
             read_non_whitespace(stream)
             stream.seek(-1, 1)
             cnt = 0
@@ -815,6 +823,9 @@ def _read_xref_tables_and_trailers(
 
     def _read_xref(self, stream: StreamType) -> Optional[int]:
         self._read_standard_xref_table(stream)
+        if stream.read(1) == b"":
+            return None
+        stream.seek(-1, 1)
         read_non_whitespace(stream)
         stream.seek(-1, 1)
         new_trailer = cast(Dict[str, Any], read_object(stream, self))
diff --git a/tests/test_reader.py b/tests/test_reader.py
index f8be72dc1..a346ca76d 100644
--- a/tests/test_reader.py
+++ b/tests/test_reader.py
@@ -1508,3 +1508,11 @@ def test_corrupted_xref():
     name = "iss2516.pdf"
     reader = PdfReader(BytesIO(get_data_from_url(url, name=name)))
     assert reader.root_object["/Type"] == "/Catalog"
+
+
+@pytest.mark.enable_socket()
+def test_truncated_xref(caplog):
+    url = "https://github.com/py-pdf/pypdf/files/14843553/002-trivial-libre-office-writer-broken.pdf"
+    name = "iss2575.pdf"
+    PdfReader(BytesIO(get_data_from_url(url, name=name)))
+    assert "Invalid/Truncated xref table. Rebuilding it." in caplog.text