ROB: Fixing infinite loop in ArrayObject read_from_stream (py-pdf#2928)

Fixing infinite loop in ArrayObject read_from_stream if stream ends unexpectedly.

Author: jakep-allenai

pypdf/generic/_data_structures.py

@@ -251,6 +251,8 @@ def read_from_stream(
             tok = stream.read(1)
             while tok.isspace():
                 tok = stream.read(1)
+            if tok == b"":
+                break
             if tok == b"%":
                 stream.seek(-1, 1)
                 skip_over_comment(stream)

tests/test_text_extraction.py

@@ -260,3 +260,15 @@ def test_layout_mode_space_vertically_font_height_weight():
                 "PDF extracted text differs from expected value.\n\n"
                 "Expected:\n\n%r\n\nExtracted:\n\n%r\n\n" % (pdftext, text)
         )
+
+
+@pytest.mark.enable_socket
+def test_infinite_loop_arrays():
+    """Tests for #2928"""
+    url = "https://github.com/user-attachments/files/17576546/arrayabruptending.pdf"
+    name = "arrayabruptending.pdf"
+    reader = PdfReader(BytesIO(get_data_from_url(url, name=name)))
+
+    page = reader.pages[0]
+    extracted = page.extract_text()
+    assert "RNA structure comparison" in extracted