dev-python/lxml: Fix building with cython-3.1*

Signed-off-by: Michał Górny <mgorny@gentoo.org>

dev-python/lxml/files/lxml-5.4.0-cython-3.1.patch

@@ -0,0 +1,91 @@
+From 6d5d6aed2e38e1abc625f29c0b3e97fc8c60ae3b Mon Sep 17 00:00:00 2001
+From: Stefan Behnel <stefan_ml@behnel.de>
+Date: Wed, 28 Aug 2024 11:16:55 +0200
+Subject: [PATCH] Avoid custom "tp_new()" call and add a safe-guard that
+ element lookups actually return a type.
+
+---
+ src/lxml/etree.pyx             | 12 +++++-------
+ src/lxml/includes/etree_defs.h |  8 ++------
+ src/lxml/python.pxd            |  1 -
+ 3 files changed, 7 insertions(+), 14 deletions(-)
+
+diff --git a/src/lxml/etree.pyx b/src/lxml/etree.pyx
+index c21d1343..90579af9 100644
+--- a/src/lxml/etree.pyx
++++ b/src/lxml/etree.pyx
+@@ -1636,11 +1636,6 @@ cdef public class _Element [ type LxmlElementType, object LxmlElement ]:
+         return CSSSelector(expr, translator=translator)(self)
+ 
+ 
+-cdef extern from "includes/etree_defs.h":
+-    # macro call to 't->tp_new()' for fast instantiation
+-    cdef object NEW_ELEMENT "PY_NEW" (object t)
+-
+-
+ @cython.linetrace(False)
+ cdef _Element _elementFactory(_Document doc, xmlNode* c_node):
+     cdef _Element result
+@@ -1650,12 +1645,15 @@ cdef _Element _elementFactory(_Document doc, xmlNode* c_node):
+     if c_node is NULL:
+         return None
+ 
+-    element_class = LOOKUP_ELEMENT_CLASS(
++    element_class = <type> LOOKUP_ELEMENT_CLASS(
+         ELEMENT_CLASS_LOOKUP_STATE, doc, c_node)
++    if type(element_class) is not type:
++        if not isinstance(element_class, type):
++            raise TypeError(f"Element class is not a type, got {type(element_class)}")
+     if hasProxy(c_node):
+         # prevent re-entry race condition - we just called into Python
+         return getProxy(c_node)
+-    result = NEW_ELEMENT(element_class)
++    result = element_class.__new__(element_class)
+     if hasProxy(c_node):
+         # prevent re-entry race condition - we just called into Python
+         result._c_node = NULL
+diff --git a/src/lxml/includes/etree_defs.h b/src/lxml/includes/etree_defs.h
+index 17d470d0..8645869f 100644
+--- a/src/lxml/includes/etree_defs.h
++++ b/src/lxml/includes/etree_defs.h
+@@ -177,7 +177,7 @@ long _ftol2( double dblSource ) { return _ftol( dblSource ); }
+ 
+ #ifdef __GNUC__
+ /* Test for GCC > 2.95 */
+-#if __GNUC__ > 2 || (__GNUC__ == 2 && (__GNUC_MINOR__ > 95)) 
++#if __GNUC__ > 2 || (__GNUC__ == 2 && (__GNUC_MINOR__ > 95))
+ #define unlikely_condition(x) __builtin_expect((x), 0)
+ #else /* __GNUC__ > 2 ... */
+ #define unlikely_condition(x) (x)
+@@ -190,10 +190,6 @@ long _ftol2( double dblSource ) { return _ftol( dblSource ); }
+   #define Py_TYPE(ob)   (((PyObject*)(ob))->ob_type)
+ #endif
+ 
+-#define PY_NEW(T) \
+-     (((PyTypeObject*)(T))->tp_new( \
+-             (PyTypeObject*)(T), __pyx_empty_tuple, NULL))
+-
+ #define _fqtypename(o)  ((Py_TYPE(o))->tp_name)
+ 
+ #define lxml_malloc(count, item_size) \
+@@ -268,7 +264,7 @@ static void* lxml_unpack_xmldoc_capsule(PyObject* capsule, int* is_owned) {
+  * 'inclusive' is 1).  The _ELEMENT_ variants will only stop on nodes
+  * that match _isElement(), the normal variant will stop on every node
+  * except text nodes.
+- * 
++ *
+  * To traverse the node and all of its children and siblings in Pyrex, call
+  *    cdef xmlNode* some_node
+  *    BEGIN_FOR_EACH_ELEMENT_FROM(some_node.parent, some_node, 1)
+diff --git a/src/lxml/python.pxd b/src/lxml/python.pxd
+index d0877355..e0ec762e 100644
+--- a/src/lxml/python.pxd
++++ b/src/lxml/python.pxd
+@@ -131,7 +131,6 @@ cdef extern from "includes/etree_defs.h": # redefines some functions as macros
+     cdef void* lxml_unpack_xmldoc_capsule(object capsule, bint* is_owned) except? NULL
+     cdef bint _isString(object obj)
+     cdef const_char* _fqtypename(object t)
+-    cdef object PY_NEW(object t)
+     cdef bint IS_PYPY
+     cdef object PyOS_FSPath(object obj)
+ 

dev-python/lxml/lxml-5.4.0-r1.ebuild

@@ -0,0 +1,122 @@
+# Copyright 1999-2025 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..13} pypy3 pypy3_11 )
+
+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 ~loong ~m68k ~mips ~ppc ~ppc64 ~riscv ~s390 ~sparc ~x86 ~amd64-linux ~x86-linux ~arm64-macos ~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/libxml2-2.10.3
+	>=dev-libs/libxslt-1.1.38
+"
+RDEPEND="
+	${DEPEND}
+"
+BDEPEND="
+	virtual/pkgconfig
+	>=dev-python/cython-3.0.10[${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}-5.3.0-pypy.patch"
+	# https://github.com/lxml/lxml/commit/6d5d6aed2e38e1abc625f29c0b3e97fc8c60ae3b
+	"${FILESDIR}/${PN}-5.4.0-cython-3.1.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() {
+	# 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() {
+	local DISTUTILS_ARGS=(
+		# by default it adds -w to CFLAGS
+		--warnings
+	)
+	tc-export PKG_CONFIG
+	distutils-r1_python_compile
+}
+
+python_compile_all() {
+	# disable automagic dep on coverage
+	use doc && emake CYTHON_WITH_COVERAGE= 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 "${dir}/html/" || die
+	mkdir "${dir}"/../../doc || die
+	# this one needs to be copied, because upstream uses doc/../../../doc
+	cp -r "${S}"/doc "${dir}"/../../ || die
+	ln -s "${S}"/doc "${dir}"/../../../../ || die
+
+	"${EPYTHON}" test.py --no-src -vv --all-levels -p ||
+		die "Tests fail on ${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
+	optfeature "Support for lxml.html.clean sanitizer" dev-python/lxml-html-clean
+}