bpo-39011: Preserve line endings within ElementTree attributes (GH-18468)

* bpo-39011: Preserve line endings within attributes

Line endings within attributes were previously normalized to "\n" in Py3.7/3.8.
This patch removes that normalization, as line endings which were
replaced by entity numbers should be preserved in original form.

diff --git a/Doc/whatsnew/3.9.rst b/Doc/whatsnew/3.9.rst
index 3beb721ed318e283b587020b2bbc4ebf27236b8e..6cd80ce8e4ff94cf575650e3530e2236ce890eeb 100644 (file)
--- a/Doc/whatsnew/3.9.rst
+++ b/Doc/whatsnew/3.9.rst
@@ -412,6 +412,15 @@ customization consistently by always using the value specified by
 case), and one used ``__VENV_NAME__`` instead.
 (Contributed by Brett Cannon in :issue:`37663`.)
 
+xml
+---
+
+White space characters within attributes are now preserved when serializing
+:mod:`xml.etree.ElementTree` to XML file. EOLNs are no longer normalized
+to "\n". This is the result of discussion about how to interpret
+section 2.11 of XML spec.
+(Contributed by Mefistotelis in :issue:`39011`.)
+
 
 Optimizations
 =============

diff --git a/Lib/test/test_xml_etree.py b/Lib/test/test_xml_etree.py
index 785edb737021a70e2e840114db15e8fb698e3b34..d01649d1c31b261c61c792ad64fcfa2f92c7e235 100644 (file)
--- a/Lib/test/test_xml_etree.py
+++ b/Lib/test/test_xml_etree.py
@@ -430,13 +430,14 @@ class ElementTreeTest(unittest.TestCase):
         self.assertEqual(ET.tostring(elem),
                 b'<test testa="testval" testb="test1" testc="test2">aa</test>')
 
+        # Test preserving white space chars in attributes
         elem = ET.Element('test')
         elem.set('a', '\r')
         elem.set('b', '\r\n')
         elem.set('c', '\t\n\r ')
-        elem.set('d', '\n\n')
+        elem.set('d', '\n\n\r\r\t\t  ')
         self.assertEqual(ET.tostring(elem),
-                b'<test a="&#10;" b="&#10;" c="&#09;&#10;&#10; " d="&#10;&#10;" />')
+                b'<test a="&#13;" b="&#13;&#10;" c="&#09;&#10;&#13; " d="&#10;&#10;&#13;&#13;&#09;&#09;  " />')
 
     def test_makeelement(self):
         # Test makeelement handling.

diff --git a/Lib/xml/etree/ElementTree.py b/Lib/xml/etree/ElementTree.py
index c8d898f32816dcae866aa3a077a17133ce21a765..da2bcad0b4d629880c9ced468927975df45def32 100644 (file)
--- a/Lib/xml/etree/ElementTree.py
+++ b/Lib/xml/etree/ElementTree.py
@@ -1057,15 +1057,15 @@ def _escape_attrib(text):
             text = text.replace(">", ">")
         if "\"" in text:
             text = text.replace("\"", """)
-        # The following business with carriage returns is to satisfy
-        # Section 2.11 of the XML specification, stating that
-        # CR or CR LN should be replaced with just LN
+        # Although section 2.11 of the XML specification states that CR or
+        # CR LN should be replaced with just LN, it applies only to EOLNs
+        # which take part of organizing file into lines. Within attributes,
+        # we are replacing these with entity numbers, so they do not count.
         # http://www.w3.org/TR/REC-xml/#sec-line-ends
-        if "\r\n" in text:
-            text = text.replace("\r\n", "\n")
+        # The current solution, contained in following six lines, was
+        # discussed in issue 17582 and 39011.
         if "\r" in text:
-            text = text.replace("\r", "\n")
-        #The following four lines are issue 17582
+            text = text.replace("\r", "
")
         if "\n" in text:
             text = text.replace("\n", "
")
         if "\t" in text:

diff --git a/Misc/NEWS.d/next/Library/2020-02-12-01-48-51.bpo-39011.hGve_t.rst b/Misc/NEWS.d/next/Library/2020-02-12-01-48-51.bpo-39011.hGve_t.rst
new file mode 100644 (file)
index 0000000..43962f0
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2020-02-12-01-48-51.bpo-39011.hGve_t.rst
@@ -0,0 +1,3 @@
+Normalization of line endings in ElementTree attributes was removed, as line
+endings which were replaced by entity numbers should be preserved in
+original form.