dev-python/lxml: Backport py3.12 unicode fix

Signed-off-by: Violet Purcell <vimproved@inventati.org>
Closes: https://github.com/gentoo/gentoo/pull/31327
Signed-off-by: Sam James <sam@gentoo.org>

2 files changed, 149 insertions, 0 deletions

diff --git a/dev-python/lxml/files/lxml-4.9.2-py3.12-backport-unicode-fix.patch b/dev-python/lxml/files/lxml-4.9.2-py3.12-backport-unicode-fix.patch
new file mode 100644
index 000000000000..d1ca31d1b909
--- /dev/null
+++ b/dev-python/lxml/files/lxml-4.9.2-py3.12-backport-unicode-fix.patch
@@ -0,0 +1,33 @@
+Upstream: https://github.com/lxml/lxml/commit/a1f2231dfdd31a37f2c55bc9ae4ac33e92221227
+
+From a1f2231dfdd31a37f2c55bc9ae4ac33e92221227 Mon Sep 17 00:00:00 2001
+From: Stefan Behnel <stefan_ml@behnel.de>
+Date: Fri, 12 May 2023 09:49:34 +0200
+Subject: [PATCH] Adapt unicode parsing to Py3.12.
+
+---
+ src/lxml/includes/etree_defs.h | 10 ++++++++++
+ 1 file changed, 10 insertions(+)
+
+diff --git a/src/lxml/includes/etree_defs.h b/src/lxml/includes/etree_defs.h
+index e671fa85d..d42f2158c 100644
+--- a/src/lxml/includes/etree_defs.h
++++ b/src/lxml/includes/etree_defs.h
+@@ -120,6 +120,16 @@ static PyObject* PyBytes_FromFormat(const char* format, ...) {
+ #  define _lx_PySlice_GetIndicesEx(o, l, b, e, s, sl) PySlice_GetIndicesEx(((PySliceObject*)o), l, b, e, s, sl)
+ #endif
+ 
++#if PY_VERSION_HEX >= 0x030B00A1
++/* Python 3.12 doesn't have wstr Unicode strings any more. */
++#undef PyUnicode_GET_DATA_SIZE
++#define PyUnicode_GET_DATA_SIZE(ustr)  (0)
++#undef PyUnicode_AS_DATA
++#define PyUnicode_AS_DATA(ustr)  (NULL)
++#undef PyUnicode_IS_READY
++#define PyUnicode_IS_READY(ustr)  (1)
++#endif
++
+ #ifdef WITHOUT_THREADING
+ #  undef PyEval_SaveThread
+ #  define PyEval_SaveThread() (NULL)
+
diff --git a/dev-python/lxml/lxml-4.9.2-r1.ebuild b/dev-python/lxml/lxml-4.9.2-r1.ebuild
new file mode 100644
index 000000000000..9b26e534fcf8
--- /dev/null
+++ b/dev-python/lxml/lxml-4.9.2-r1.ebuild
@@ -0,0 +1,116 @@
+# Copyright 1999-2023 Gentoo Authors
+# Distributed under the terms of the GNU General Public License v2
+
+EAPI=8
+
+DISTUTILS_EXT=1
+DISTUTILS_USE_PEP517=setuptools
+PYTHON_COMPAT=( python3_{10..12} pypy3 )
+
+inherit distutils-r1 optfeature toolchain-funcs
+
+DESCRIPTION="A Pythonic binding for the libxml2 and libxslt libraries"
+HOMEPAGE="
+	https://lxml.de/
+	https://pypi.org/project/lxml/
+	https://github.com/lxml/lxml/
+"
+SRC_URI="
+	https://github.com/lxml/lxml/archive/${P}.tar.gz
+		-> ${P}.gh.tar.gz
+"
+S=${WORKDIR}/lxml-${P}
+
+LICENSE="BSD ElementTree GPL-2 PSF-2"
+SLOT="0"
+KEYWORDS="~alpha ~amd64 ~arm ~arm64 ~hppa ~ia64 ~loong ~m68k ~mips ~ppc ~ppc64 ~riscv ~s390 ~sparc ~x86 ~amd64-linux ~x86-linux ~ppc-macos ~x64-macos ~x64-solaris"
+IUSE="doc examples +threads test"
+RESTRICT="!test? ( test )"
+
+# Note: lib{xml2,xslt} are used as C libraries, not Python modules.
+DEPEND="
+	>=dev-libs/libxml2-2.9.12-r2
+	>=dev-libs/libxslt-1.1.28
+"
+RDEPEND="
+	${DEPEND}
+"
+BDEPEND="
+	virtual/pkgconfig
+	<dev-python/cython-3[${PYTHON_USEDEP}]
+	>=dev-python/cython-0.29.29[${PYTHON_USEDEP}]
+	doc? (
+		$(python_gen_any_dep '
+			dev-python/docutils[${PYTHON_USEDEP}]
+			dev-python/pygments[${PYTHON_USEDEP}]
+			dev-python/sphinx[${PYTHON_USEDEP}]
+			dev-python/sphinx-rtd-theme[${PYTHON_USEDEP}]
+		')
+	)
+	test? (
+		dev-python/cssselect[${PYTHON_USEDEP}]
+	)
+"
+
+PATCHES=(
+	"${FILESDIR}"/${PN}-4.6.0-tests-pypy.patch
+	"${FILESDIR}"/${P}-py3.12-drop-deprecated-imp.patch
+	"${FILESDIR}"/${P}-py3.12-backport-unicode-fix.patch
+)
+
+python_check_deps() {
+	use doc || return 0
+	python_has_version -b "dev-python/docutils[${PYTHON_USEDEP}]" &&
+	python_has_version -b "dev-python/pygments[${PYTHON_USEDEP}]" &&
+	python_has_version -b "dev-python/sphinx[${PYTHON_USEDEP}]" &&
+	python_has_version -b "dev-python/sphinx-rtd-theme[${PYTHON_USEDEP}]"
+}
+
+python_prepare_all() {
+	# avoid replacing PYTHONPATH in tests.
+	sed -i -e '/sys\.path/d' test.py || die
+
+	# don't use some random SDK on Darwin
+	sed -i -e '/_ldflags =/s/=.*isysroot.*darwin.*None/= None/' \
+		setupinfo.py || die
+
+	distutils-r1_python_prepare_all
+}
+
+python_compile() {
+	tc-export PKG_CONFIG
+	distutils-r1_python_compile
+}
+
+python_compile_all() {
+	use doc && emake html
+}
+
+python_test() {
+	local dir=${BUILD_DIR}/test$(python_get_sitedir)/lxml
+	local -x PATH=${BUILD_DIR}/test/usr/bin:${PATH}
+
+	cp -al "${BUILD_DIR}"/{install,test} || die
+	cp -al src/lxml/tests "${dir}/" || die
+	cp -al src/lxml/html/tests "${dir}/html/" || die
+	ln -rs "${S}"/doc "${dir}"/../../ || die
+
+	"${EPYTHON}" test.py -vv --all-levels -p || die "Test ${test} fails with ${EPYTHON}"
+}
+
+python_install_all() {
+	if use doc; then
+		local DOCS=( README.rst *.txt doc/*.txt )
+		local HTML_DOCS=( doc/html/. )
+	fi
+	if use examples; then
+		dodoc -r samples
+	fi
+
+	distutils-r1_python_install_all
+}
+
+pkg_postinst() {
+	optfeature "Support for BeautifulSoup as a parser backend" dev-python/beautifulsoup4
+	optfeature "Translates CSS selectors to XPath 1.0 expressions" dev-python/cssselect
+}