Switch to latest libxml2 2.9.12+ (unreleased) that has fixes for traversing lxml's fake root trees.

4 files changed, 21 insertions, 4 deletions

diff --git a/.github/workflows/wheels.yml b/.github/workflows/wheels.yml
index bfd8e9ef..6117f9e6 100644
--- a/.github/workflows/wheels.yml
+++ b/.github/workflows/wheels.yml
@@ -118,7 +118,7 @@ jobs:
         python_version: ["2.7", "3.6", "3.7", "3.8", "3.9", "3.10", "pypy-3.7-v7.3.3", "pypy-3.8-v7.3.7"]
 
     runs-on: ${{ matrix.os }}
-    env: { LIBXML2_VERSION: 2.9.10, LIBXSLT_VERSION: 1.1.34, MACOSX_DEPLOYMENT_TARGET: 10.14 }
+    env: { LIBXML2_VERSION: 2.9.12, LIBXSLT_VERSION: 1.1.34, MACOSX_DEPLOYMENT_TARGET: 10.14 }
 
     steps:
     - uses: actions/checkout@v2
diff --git a/CHANGES.txt b/CHANGES.txt
index 72a123b6..f0fa06ba 100644
--- a/CHANGES.txt
+++ b/CHANGES.txt
@@ -19,6 +19,11 @@ Bugs fixed
   as integers or float values in Python 3.6 and later. It now adheres to the number
   format of the XML spec again.
 
+Other changes
+-------------
+
+* Wheels include libxml2 2.9.12+ and libxslt 1.1.34.
+
 
 4.6.4 (2021-11-01)
 ==================
diff --git a/Makefile b/Makefile
index 555d851e..dec41378 100644
--- a/Makefile
+++ b/Makefile
@@ -13,7 +13,7 @@ CYTHON_WITH_COVERAGE?=$(shell $(PYTHON) -c 'import Cython.Coverage; import sys;
 CYTHON3_WITH_COVERAGE?=$(shell $(PYTHON3) -c 'import Cython.Coverage; import sys; assert not hasattr(sys, "pypy_version_info")' >/dev/null 2>/dev/null && echo " --coverage" || true)
 
 PYTHON_BUILD_VERSION ?= *
-MANYLINUX_LIBXML2_VERSION=2.9.10
+MANYLINUX_LIBXML2_VERSION=2.9.12
 MANYLINUX_LIBXSLT_VERSION=1.1.34
 MANYLINUX_CFLAGS=-O3 -g1 -pipe -fPIC -flto
 MANYLINUX_LDFLAGS=-flto
diff --git a/buildlibxml.py b/buildlibxml.py
index 086d9115..08b465de 100644
--- a/buildlibxml.py
+++ b/buildlibxml.py
@@ -121,6 +121,7 @@ def get_prebuilt_libxml2xslt(download_dir, static_include_dirs, static_library_d
 ## Routines to download and build libxml2/xslt from sources:
 
 LIBXML2_LOCATION = 'http://xmlsoft.org/sources/'
+LIBXSLT_LOCATION = 'http://xmlsoft.org/sources/'
 LIBICONV_LOCATION = 'https://ftp.gnu.org/pub/gnu/libiconv/'
 ZLIB_LOCATION = 'https://zlib.net/'
 match_libfile_version = re.compile('^[^-]*-([.0-9-]+)[.].*').match
@@ -214,7 +215,15 @@ def download_libxml2(dest_dir, version=None):
     #version_re = re.compile(r'LATEST_LIBXML2_IS_([0-9.]+[0-9](?:-[abrc0-9]+)?)')
     version_re = re.compile(r'libxml2-([0-9.]+[0-9]).tar.gz')
     filename = 'libxml2-%s.tar.gz'
-    return download_library(dest_dir, LIBXML2_LOCATION, 'libxml2',
+
+    if version == "2.9.12":
+        # Temporarily using the latest master (2.9.12+) until there is a release that supports lxml again.
+        from_location = "https://gitlab.gnome.org/GNOME/libxml2/-/archive/dea91c97debeac7c1aaf9c19f79029809e23a353/"
+        version = "dea91c97debeac7c1aaf9c19f79029809e23a353"
+    else:
+        from_location = LIBXML2_LOCATION
+
+    return download_library(dest_dir, from_location, 'libxml2',
                             version_re, filename, version=version)
 
 
@@ -223,7 +232,7 @@ def download_libxslt(dest_dir, version=None):
     #version_re = re.compile(r'LATEST_LIBXSLT_IS_([0-9.]+[0-9](?:-[abrc0-9]+)?)')
     version_re = re.compile(r'libxslt-([0-9.]+[0-9]).tar.gz')
     filename = 'libxslt-%s.tar.gz'
-    return download_library(dest_dir, LIBXML2_LOCATION, 'libxslt',
+    return download_library(dest_dir, LIBXSLT_LOCATION, 'libxslt',
                             version_re, filename, version=version)
 
 
@@ -441,6 +450,9 @@ def build_libxml2xslt(download_dir, build_dir,
     except Exception:
         pass # this isn't required, so ignore any errors
     if not has_current_lib("libxml2", libxml2_dir):
+        if not os.path.exists(os.path.join(libxml2_dir, "configure")):
+            # Allow building from git sources by running autoconf etc.
+            libxml2_configure_cmd[0] = "./autogen.sh"
         cmmi(libxml2_configure_cmd, libxml2_dir, multicore, **call_setup)
 
     # Fix up libxslt configure script (needed up to and including 1.1.34)