From 62262cca798dd63d44b11f3b332173b184642773 Mon Sep 17 00:00:00 2001
From: CentOS Sources <bugs@centos.org>
Date: May 02 2022 07:50:30 +0000
Subject: import rh-python38-python-lxml-4.4.1-8.el7


---

diff --git a/SOURCES/CVE-2021-43818.patch b/SOURCES/CVE-2021-43818.patch
new file mode 100644
index 0000000..18d9393
--- /dev/null
+++ b/SOURCES/CVE-2021-43818.patch
@@ -0,0 +1,127 @@
+diff --git a/src/lxml/html/clean.py b/src/lxml/html/clean.py
+index 15298b5..ee2f0f8 100644
+--- a/src/lxml/html/clean.py
++++ b/src/lxml/html/clean.py
+@@ -73,18 +73,25 @@ _looks_like_tag_content = re.compile(
+ 
+ # All kinds of schemes besides just javascript: that can cause
+ # execution:
+-_is_image_dataurl = re.compile(
+-    r'^data:image/.+;base64', re.I).search
++_find_image_dataurls = re.compile(
++    r'^data:image/(.+);base64,', re.I).findall
+ _is_possibly_malicious_scheme = re.compile(
+-    r'(?:javascript|jscript|livescript|vbscript|data|about|mocha):',
+-    re.I).search
++    r'(javascript|jscript|livescript|vbscript|data|about|mocha):',
++    re.I).findall
++# SVG images can contain script content
++_is_unsafe_image_type = re.compile(r"(xml|svg)", re.I).findall
++
+ def _is_javascript_scheme(s):
+-    if _is_image_dataurl(s):
+-        return None
+-    return _is_possibly_malicious_scheme(s)
++    is_image_url = False
++    for image_type in _find_image_dataurls(s):
++        is_image_url = True
++        if _is_unsafe_image_type(image_type):
++            return True
++    if is_image_url:
++        return False
++    return bool(_is_possibly_malicious_scheme(s))
+ 
+ _substitute_whitespace = re.compile(r'[\s\x00-\x08\x0B\x0C\x0E-\x19]+').sub
+-# FIXME: should data: be blocked?
+ 
+ # FIXME: check against: http://msdn2.microsoft.com/en-us/library/ms537512.aspx
+ _conditional_comment_re = re.compile(
+@@ -512,6 +519,8 @@ class Cleaner(object):
+             return True
+         if 'expression(' in style:
+             return True
++        if '@import' in style:
++            return True
+         if '</noscript' in style:
+             # e.g. '<noscript><style><a title="</noscript><img src=x onerror=alert(1)>">'
+             return True
+diff --git a/src/lxml/html/tests/test_clean.py b/src/lxml/html/tests/test_clean.py
+index d8df527..7021e48 100644
+--- a/src/lxml/html/tests/test_clean.py
++++ b/src/lxml/html/tests/test_clean.py
+@@ -1,3 +1,5 @@
++import base64
++import gzip
+ import unittest
+ from lxml.tests.common_imports import make_doctest
+ 
+@@ -89,6 +91,69 @@ class CleanerTest(unittest.TestCase):
+             b'<math><style>/* deleted */</style></math>',
+             lxml.html.tostring(clean_html(s)))
+ 
++    def test_sneaky_import_in_style(self):
++        # Prevent "@@importimport" -> "@import" replacement.
++        style_codes = [
++            "@@importimport(extstyle.css)",
++            "@ @  import import(extstyle.css)",
++            "@ @ importimport(extstyle.css)",
++            "@@  import import(extstyle.css)",
++            "@ @import import(extstyle.css)",
++            "@@importimport()",
++        ]
++        for style_code in style_codes:
++            html = '<style>%s</style>' % style_code
++            s = lxml.html.fragment_fromstring(html)
++
++            cleaned = lxml.html.tostring(clean_html(s))
++            self.assertEqual(
++                b'<style>/* deleted */</style>',
++                cleaned,
++                "%s  ->  %s" % (style_code, cleaned))
++
++    def test_svg_data_links(self):
++        # Remove SVG images with potentially insecure content.
++        svg = b'<svg onload="alert(123)" />'
++        svgz = gzip.compress(svg)
++        svg_b64 = base64.b64encode(svg).decode('ASCII')
++        svgz_b64 = base64.b64encode(svgz).decode('ASCII')
++        urls = [
++            "data:image/svg+xml;base64," + svg_b64,
++            "data:image/svg+xml-compressed;base64," + svgz_b64,
++        ]
++        for url in urls:
++            html = '<img src="%s">' % url
++            s = lxml.html.fragment_fromstring(html)
++
++            cleaned = lxml.html.tostring(clean_html(s))
++            self.assertEqual(
++                b'<img src="">',
++                cleaned,
++                "%s  ->  %s" % (url, cleaned))
++
++    def test_image_data_links(self):
++        data = b'123'
++        data_b64 = base64.b64encode(data).decode('ASCII')
++        urls = [
++            "data:image/jpeg;base64," + data_b64,
++            "data:image/apng;base64," + data_b64,
++            "data:image/png;base64," + data_b64,
++            "data:image/gif;base64," + data_b64,
++            "data:image/webp;base64," + data_b64,
++            "data:image/bmp;base64," + data_b64,
++            "data:image/tiff;base64," + data_b64,
++            "data:image/x-icon;base64," + data_b64,
++        ]
++        for url in urls:
++            html = '<img src="%s">' % url
++            s = lxml.html.fragment_fromstring(html)
++
++            cleaned = lxml.html.tostring(clean_html(s))
++            self.assertEqual(
++                html.encode("UTF-8"),
++                cleaned,
++                "%s  ->  %s" % (url, cleaned))
++
+     def test_formaction_attribute_in_button_input(self):
+         # The formaction attribute overrides the form's action and should be
+         # treated as a malicious link attribute
diff --git a/SPECS/python-lxml.spec b/SPECS/python-lxml.spec
index 5bb9083..9a11157 100644
--- a/SPECS/python-lxml.spec
+++ b/SPECS/python-lxml.spec
@@ -9,7 +9,7 @@
 
 Name:           %{?scl_prefix}python-%{modname}
 Version:        4.4.1
-Release:        7%{?dist}
+Release:        8%{?dist}
 Summary:        XML processing library combining libxml2/libxslt with the ElementTree API
 
 License:        BSD
@@ -28,6 +28,13 @@ Patch0:         CVE-2020-27783.patch
 # Fixed upstream: https://github.com/lxml/lxml/commit/2d01a1ba8984e0483ce6619b972832377f208a0d
 Patch1: CVE-2021-28957.patch
 
+# Fix for CVE-2021-43818: HTML Cleaner allows crafted
+# and SVG embedded scripts to pass through
+# Fixed upstream:
+# https://github.com/lxml/lxml/commit/12fa9669007180a7bb87d990c375cf91ca5b664a
+# https://github.com/lxml/lxml/commit/f2330237440df7e8f39c3ad1b1aa8852be3b27c0
+Patch2: CVE-2021-43818.patch
+
 %{?scl:Requires: %{scl}-runtime}
 %{?scl:BuildRequires: %{scl}-runtime}
 
@@ -88,6 +95,10 @@ set -ex
 
 
 %changelog
+* Wed Mar 16 2022 Charalampos Stratakis <cstratak@redhat.com> - 4.4.1-8
+- Security fix for CVE-2021-43818
+Resolves: rhbz#2032569
+
 * Mon Jul 19 2021 Charalampos Stratakis <cstratak@redhat.com> - 4.4.1-7
 - Security fix for CVE-2021-28957
 Resolves: rhbz#1941534