From 8915fad472adee40fdf1ff6736e3a75565b0f5ec Mon Sep 17 00:00:00 2001
From: Parfait Gasana <parfait.gasana@gmail.com>
Date: Fri, 24 Jun 2022 22:49:30 -0500
Subject: [PATCH] BUG: read_xml iterparse doesn't handle multiple toplevel
 elements with lxml parser

---
 pandas/io/xml.py                |   2 +-
 pandas/tests/io/xml/test_xml.py | 109 +++++++++++++++++++++++++++++++-
 2 files changed, 109 insertions(+), 2 deletions(-)

diff --git a/pandas/io/xml.py b/pandas/io/xml.py
index 78fbeaad09300..9398e995c81ce 100644
--- a/pandas/io/xml.py
+++ b/pandas/io/xml.py
@@ -693,7 +693,7 @@ def _iterparse_nodes(self) -> list[dict[str, str | None]]:
                     row = None
 
                 elem.clear()
-                while elem.getprevious() is not None:
+                while elem.getprevious() is not None and elem.getparent() is not None:
                     del elem.getparent()[0]
 
         if dicts == []:
diff --git a/pandas/tests/io/xml/test_xml.py b/pandas/tests/io/xml/test_xml.py
index eb2230bbf7fd5..b89adf85d8e26 100644
--- a/pandas/tests/io/xml/test_xml.py
+++ b/pandas/tests/io/xml/test_xml.py
@@ -1271,7 +1271,7 @@ def test_wrong_dict_value(datapath, parser):
         read_xml(filename, parser=parser, iterparse={"book": "category"})
 
 
-def test_bad_xml(datapath, parser):
+def test_bad_xml(parser):
     bad_xml = """\
 <?xml version='1.0' encoding='utf-8'?>
   <row>
@@ -1312,6 +1312,113 @@ def test_bad_xml(datapath, parser):
             )
 
 
+def test_comment(parser):
+    xml = """\
+<!-- comment before root -->
+<shapes>
+  <!-- comment within root -->
+  <shape>
+    <name>circle</name>
+    <type>2D</type>
+  </shape>
+  <shape>
+    <name>sphere</name>
+    <type>3D</type>
+    <!-- comment within child -->
+  </shape>
+  <!-- comment within root -->
+</shapes>
+<!-- comment after root -->"""
+
+    df_xpath = read_xml(xml, xpath=".//shape", parser=parser)
+
+    df_iter = read_xml_iterparse(
+        xml, parser=parser, iterparse={"shape": ["name", "type"]}
+    )
+
+    df_expected = DataFrame(
+        {
+            "name": ["circle", "sphere"],
+            "type": ["2D", "3D"],
+        }
+    )
+
+    tm.assert_frame_equal(df_xpath, df_expected)
+    tm.assert_frame_equal(df_iter, df_expected)
+
+
+def test_dtd(parser):
+    xml = """\
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE non-profits [
+    <!ELEMENT shapes (shape*) >
+    <!ELEMENT shape ( name, type )>
+    <!ELEMENT name (#PCDATA)>
+]>
+<shapes>
+  <shape>
+    <name>circle</name>
+    <type>2D</type>
+  </shape>
+  <shape>
+    <name>sphere</name>
+    <type>3D</type>
+  </shape>
+</shapes>"""
+
+    df_xpath = read_xml(xml, xpath=".//shape", parser=parser)
+
+    df_iter = read_xml_iterparse(
+        xml, parser=parser, iterparse={"shape": ["name", "type"]}
+    )
+
+    df_expected = DataFrame(
+        {
+            "name": ["circle", "sphere"],
+            "type": ["2D", "3D"],
+        }
+    )
+
+    tm.assert_frame_equal(df_xpath, df_expected)
+    tm.assert_frame_equal(df_iter, df_expected)
+
+
+def test_processing_instruction(parser):
+    xml = """\
+<?xml version="1.0" encoding="UTF-8"?>
+<?xml-stylesheet type="text/xsl" href="style.xsl"?>
+<?display table-view?>
+<?sort alpha-ascending?>
+<?textinfo whitespace is allowed ?>
+<?elementnames <shape>, <name>, <type> ?>
+<shapes>
+  <shape>
+    <name>circle</name>
+    <type>2D</type>
+  </shape>
+  <shape>
+    <name>sphere</name>
+    <type>3D</type>
+  </shape>
+</shapes>"""
+
+    df_xpath = read_xml(xml, xpath=".//shape", parser=parser)
+
+    df_iter = read_xml_iterparse(
+        xml, parser=parser, iterparse={"shape": ["name", "type"]}
+    )
+
+    df_expected = DataFrame(
+        {
+            "name": ["circle", "sphere"],
+            "type": ["2D", "3D"],
+        }
+    )
+
+    tm.assert_frame_equal(df_xpath, df_expected)
+    tm.assert_frame_equal(df_iter, df_expected)
+
+
 def test_no_result(datapath, parser):
     filename = datapath("io", "data", "xml", "books.xml")
     with pytest.raises(