Accepting request 1129127 from devel:languages:python

OBS-URL: https://build.opensuse.org/request/show/1129127
OBS-URL: https://build.opensuse.org/package/show/openSUSE:Factory/python-lxml?expand=0&rev=105

libxml2212-tests.patch

@@ -0,0 +1,41 @@
+From 2a6770566ab57d601abc7c2f49a8051b9d97b64c Mon Sep 17 00:00:00 2001
+From: Nick Wellnhofer <wellnhofer@aevum.de>
+Date: Tue, 31 Oct 2023 12:36:02 +0100
+Subject: [PATCH] Make Unicode recovery test work with libxml2 2.12 (GH-383)
+
+When encountering encoding errors, libxml2 no longer switches to ISO-8859-1 since version 2.12.
+---
+ src/lxml/parser.pxi            | 2 +-
+ src/lxml/tests/test_unicode.py | 6 +++++-
+ 2 files changed, 6 insertions(+), 2 deletions(-)
+
+diff --git a/src/lxml/parser.pxi b/src/lxml/parser.pxi
+index 4b7b52065..8ceec7d25 100644
+--- a/src/lxml/parser.pxi
++++ b/src/lxml/parser.pxi
+@@ -693,7 +693,7 @@ cdef xmlDoc* _handleParseResult(_ParserContext context,
+             # An encoding error occurred and libxml2 switched from UTF-8
+             # input to (undecoded) Latin-1, at some arbitrary point in the
+             # document.  Better raise an error than allowing for a broken
+-            # tree with mixed encodings.
++            # tree with mixed encodings. This is fixed in libxml2 2.12.
+             well_formed = 0
+         elif recover or (c_ctxt.wellFormed and
+                          c_ctxt.lastError.level < xmlerror.XML_ERR_ERROR):
+diff --git a/src/lxml/tests/test_unicode.py b/src/lxml/tests/test_unicode.py
+index 6d4ee9c0f..3636539b2 100644
+--- a/src/lxml/tests/test_unicode.py
++++ b/src/lxml/tests/test_unicode.py
+@@ -167,7 +167,11 @@ def test_illegal_utf8(self):
+     def test_illegal_utf8_recover(self):
+         data = _bytes('<test>\x80\x80\x80</test>', encoding='iso8859-1')
+         parser = etree.XMLParser(recover=True)
+-        self.assertRaises(etree.XMLSyntaxError, etree.fromstring, data, parser)
++        if etree.LIBXML_VERSION >= (2, 12, 0):
++            tree = etree.fromstring(data, parser)
++            self.assertEqual('\ufffd\ufffd\ufffd', tree.text)
++        else:
++            self.assertRaises(etree.XMLSyntaxError, etree.fromstring, data, parser)
+ 
+     def _test_encoding(self, encoding, xml_encoding_name=None):
+         foo = """<?xml version='1.0' encoding='%s'?>\n<tag attrib='123'></tag>""" % (

python-lxml.changes

@@ -1,3 +1,8 @@
+-------------------------------------------------------------------
+Mon Nov 27 12:36:51 UTC 2023 - Markéta Machová <mmachova@suse.com>
+
+- Add libxml2212-tests.patch to fix tests with new libxml2
+
 -------------------------------------------------------------------
 Wed Sep  6 20:04:29 UTC 2023 - Dirk Müller <dmueller@suse.com>
 

python-lxml.spec

@@ -42,6 +42,9 @@ Patch3:         ISO-Schematron-schema-optional.patch
 # PATCH-FIX-UPSTREAM remove-ISO-Schematron-schema.patch gl#fedora/legal/fedora-license-data/-#154 mcepl@suse.com
 # Actually remove the schema
 Patch4:         remove-ISO-Schematron-schema.patch
+# PATCH-FIX-UPSTREAM libxml2212-tests.patch https://github.com/lxml/lxml/pull/383
+# Make Unicode recovery test work with upcoming libxml2 2.12
+Patch5:         libxml2212-tests.patch
 BuildRequires:  %{python_module Cython >= 0.29.7 with %python-Cython < 3}
 BuildRequires:  %{python_module base}
 BuildRequires:  %{python_module cssselect >= 0.9.1}