Closes #2892: preserve iterparse events in case of SyntaxError.
diff --git a/Lib/test/test_xml_etree.py b/Lib/test/test_xml_etree.py
index 22fafa9..b7a996c 100644
--- a/Lib/test/test_xml_etree.py
+++ b/Lib/test/test_xml_etree.py
@@ -754,6 +754,7 @@
... print(action, elem.tag)
... except ET.ParseError as v:
... print(v)
+ end document
junk after document element: line 1, column 12
"""
diff --git a/Lib/xml/etree/ElementTree.py b/Lib/xml/etree/ElementTree.py
index ba33879..f94c48c 100644
--- a/Lib/xml/etree/ElementTree.py
+++ b/Lib/xml/etree/ElementTree.py
@@ -1250,6 +1250,7 @@
self._close_file = close_source
self._events = []
self._index = 0
+ self._error = None
self.root = self._root = None
self._parser = parser
# wire up the parser for event reporting
@@ -1291,24 +1292,31 @@
while 1:
try:
item = self._events[self._index]
- except IndexError:
- if self._parser is None:
- self.root = self._root
- if self._close_file:
- self._file.close()
- raise StopIteration
- # load event buffer
- del self._events[:]
- self._index = 0
- data = self._file.read(16384)
- if data:
- self._parser.feed(data)
- else:
- self._root = self._parser.close()
- self._parser = None
- else:
- self._index = self._index + 1
+ self._index += 1
return item
+ except IndexError:
+ pass
+ if self._error:
+ e = self._error
+ self._error = None
+ raise e
+ if self._parser is None:
+ self.root = self._root
+ if self._close_file:
+ self._file.close()
+ raise StopIteration
+ # load event buffer
+ del self._events[:]
+ self._index = 0
+ data = self._file.read(16384)
+ if data:
+ try:
+ self._parser.feed(data)
+ except SyntaxError as exc:
+ self._error = exc
+ else:
+ self._root = self._parser.close()
+ self._parser = None
def __iter__(self):
return self
diff --git a/Misc/NEWS b/Misc/NEWS
index 657c802..9e7a96c 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -66,6 +66,8 @@
Library
-------
+- Issue #2892: preserve iterparse events in case of SyntaxError.
+
- Issue #670664: Fix HTMLParser to correctly handle the content of
``<script>...</script>`` and ``<style>...</style>``.
diff --git a/Modules/_elementtree.c b/Modules/_elementtree.c
index 3061d8e..7887721 100644
--- a/Modules/_elementtree.c
+++ b/Modules/_elementtree.c
@@ -3000,6 +3000,7 @@
" self._file = file\n"
" self._events = []\n"
" self._index = 0\n"
+ " self._error = None\n"
" self.root = self._root = None\n"
" b = cElementTree.TreeBuilder()\n"
" self._parser = cElementTree.XMLParser(b)\n"
@@ -3008,24 +3009,31 @@
" while 1:\n"
" try:\n"
" item = self._events[self._index]\n"
- " except IndexError:\n"
- " if self._parser is None:\n"
- " self.root = self._root\n"
- " if self._close_file:\n"
- " self._file.close()\n"
- " raise StopIteration\n"
- " # load event buffer\n"
- " del self._events[:]\n"
- " self._index = 0\n"
- " data = self._file.read(16384)\n"
- " if data:\n"
- " self._parser.feed(data)\n"
- " else:\n"
- " self._root = self._parser.close()\n"
- " self._parser = None\n"
- " else:\n"
- " self._index = self._index + 1\n"
+ " self._index += 1\n"
" return item\n"
+ " except IndexError:\n"
+ " pass\n"
+ " if self._error:\n"
+ " e = self._error\n"
+ " self._error = None\n"
+ " raise e\n"
+ " if self._parser is None:\n"
+ " self.root = self._root\n"
+ " if self._close_file:\n"
+ " self._file.close()\n"
+ " raise StopIteration\n"
+ " # load event buffer\n"
+ " del self._events[:]\n"
+ " self._index = 0\n"
+ " data = self._file.read(16384)\n"
+ " if data:\n"
+ " try:\n"
+ " self._parser.feed(data)\n"
+ " except SyntaxError as exc:\n"
+ " self._error = exc\n"
+ " else:\n"
+ " self._root = self._parser.close()\n"
+ " self._parser = None\n"
" def __iter__(self):\n"
" return self\n"
"cElementTree.iterparse = iterparse\n"