app-doc/doxygen: bump to 1.9.1

* USE=doc now generates fine with the
  added dependency of dev-texlive/texlive-plaingeneric
  for listofitems.

* Add a patch to ignore files we can't decode
  when generating docs. Ideally those files
  shouldn't exist anyway when we're generating docs,
  or be ignored earlier, but this works.

Closes: https://bugs.gentoo.org/688658
Closes: https://bugs.gentoo.org/749630
Package-Manager: Portage-3.0.12, Repoman-3.0.2
Signed-off-by: Sam James <sam@gentoo.org>

3 files changed, 159 insertions, 0 deletions

diff --git a/app-doc/doxygen/Manifest b/app-doc/doxygen/Manifest
index 0f3605334495..4903673007fb 100644
--- a/app-doc/doxygen/Manifest
+++ b/app-doc/doxygen/Manifest
@@ -1,2 +1,3 @@
 DIST doxygen-1.8.17.src.tar.gz 5138888 BLAKE2B 79d7d64fe2104f3fc802e9ee8f5fcd13ed4f0e23fe3e8d8013d1724d917d5f4e7d8c2df672e253aa2c8d67e839a6169632ebd87bc281b8ee70d217f486e402f7 SHA512 2fd087d127e301ea48355ea52c9af4f2091df06551cf64da80df81f0758194b296efb1e8d3962867a6a6d2da5a3fc323842f7766a445748005b30097ded30a75
 DIST doxygen-1.8.20.src.tar.gz 5132703 BLAKE2B 2a9e6df833c64d7052331b52647943f7eaad8214d66962970ad55628cc2ab26883655908588bee2d4872bb2d6d85f8d18cbb03e7dabe0280f337ab53d8538ec9 SHA512 65d104d25061ee59199c74c0328f59fbeaf14f0dade755187ebd43f59008adfef243d4da448b71ae04dc325b848f9bdd109eb20e6f6092f3ed19862426d060cf
+DIST doxygen-1.9.1.src.tar.gz 5117435 BLAKE2B 28c1a1cb8390e17e2b399cea765f837782ee268a999606eef4031a5f65943c793c97bdbd2e2a25a39358ecac5df2c7a62aa4cca72753a924e4e718c512b8a9c7 SHA512 637496c549a4a150cfaeb5d4913de512262145ecd7d455d7b7f3dd68f9416e47d931a6c1efd8a17d931e4baf4a8a9f2ed21124664003b123b6f89ca4abf263ed
diff --git a/app-doc/doxygen/doxygen-1.9.1.ebuild b/app-doc/doxygen/doxygen-1.9.1.ebuild
new file mode 100644
index 000000000000..949ef011d174
--- /dev/null
+++ b/app-doc/doxygen/doxygen-1.9.1.ebuild
@@ -0,0 +1,135 @@
+# Copyright 1999-2021 Gentoo Authors
+# Distributed under the terms of the GNU General Public License v2
+
+EAPI=7
+
+PYTHON_COMPAT=( python3_{7,8,9} )
+
+inherit cmake flag-o-matic llvm python-any-r1
+if [[ ${PV} = *9999* ]]; then
+	inherit git-r3
+	EGIT_REPO_URI="https://github.com/doxygen/doxygen.git"
+	SRC_URI=""
+else
+	SRC_URI="http://doxygen.nl/files/${P}.src.tar.gz"
+	KEYWORDS="~alpha ~amd64 ~arm ~arm64 ~hppa ~ia64 ~mips ~ppc ~ppc64 ~riscv ~s390 ~sparc ~x86 ~amd64-linux ~x86-linux ~ppc-macos ~x64-macos ~sparc-solaris ~sparc64-solaris ~x64-solaris ~x86-solaris"
+fi
+
+DESCRIPTION="Documentation system for most programming languages"
+HOMEPAGE="http://www.doxygen.org"
+
+LICENSE="GPL-2"
+SLOT="0"
+IUSE="clang debug doc dot doxysearch qt5 sqlite userland_GNU"
+
+BDEPEND="sys-devel/bison
+	sys-devel/flex
+	doc? ( ${PYTHON_DEPS} )
+"
+RDEPEND="app-text/ghostscript-gpl
+	dev-lang/perl
+	media-libs/libpng:0=
+	virtual/libiconv
+	clang? ( >=sys-devel/clang-10:= )
+	dot? (
+		media-gfx/graphviz
+		media-libs/freetype
+	)
+	doc? (
+		dev-texlive/texlive-bibtexextra
+		dev-texlive/texlive-fontsextra
+		dev-texlive/texlive-fontutils
+		dev-texlive/texlive-latex
+		dev-texlive/texlive-latexextra
+		dev-texlive/texlive-plaingeneric
+	)
+	doxysearch? ( dev-libs/xapian:= )
+	qt5? (
+		dev-qt/qtgui:5
+		dev-qt/qtwidgets:5
+		dev-qt/qtxml:5
+	)
+	sqlite? ( dev-db/sqlite:3 )
+"
+DEPEND="${RDEPEND}"
+
+PATCHES=(
+	"${FILESDIR}/${PN}-1.8.9.1-empty-line-sigsegv.patch" #454348
+	"${FILESDIR}/${PN}-1.8.16-link_with_pthread.patch"
+	"${FILESDIR}/${PN}-1.8.17-ensure_static_support_libraries.patch"
+	"${FILESDIR}/${PN}-1.9.1-ignore-bad-encoding.patch"
+)
+
+DOCS=( LANGUAGE.HOWTO README.md )
+
+pkg_setup() {
+	use clang && llvm_pkg_setup
+	use doc && python-any-r1_pkg_setup
+}
+
+src_prepare() {
+	cmake_src_prepare
+
+	# Ensure we link to -liconv
+	if use elibc_FreeBSD && has_version dev-libs/libiconv || use elibc_uclibc; then
+		local pro
+		for pro in */*.pro.in */*/*.pro.in; do
+			echo "unix:LIBS += -liconv" >> "${pro}" || die
+		done
+	fi
+
+	# Call dot with -Teps instead of -Tps for EPS generation - bug #282150
+	sed -i -e '/addJob("ps"/ s/"ps"/"eps"/g' src/dot.cpp || die
+
+	# fix pdf doc
+	sed -i.orig -e "s:g_kowal:g kowal:" \
+		doc/maintainers.txt || die
+
+	if is-flagq "-O3" ; then
+		ewarn
+		ewarn "Compiling with -O3 is known to produce incorrectly"
+		ewarn "optimized code which breaks doxygen."
+		ewarn
+		elog
+		elog "Continuing with -O2 instead ..."
+		elog
+		replace-flags "-O3" "-O2"
+	fi
+}
+
+src_configure() {
+	local mycmakeargs=(
+		-Duse_libclang=$(usex clang)
+		-Dbuild_doc=$(usex doc)
+		-Dbuild_search=$(usex doxysearch)
+		-Dbuild_wizard=$(usex qt5)
+		-Duse_sqlite3=$(usex sqlite)
+		-DGIT_EXECUTABLE="false"
+	)
+
+	use doc && mycmakeargs+=(
+		-DDOC_INSTALL_DIR="share/doc/${P}"
+	)
+
+	cmake_src_configure
+}
+
+src_compile() {
+	if use doc; then
+		export VARTEXFONTS="${T}/fonts" # bug #564944
+
+		if ! use dot; then
+			sed -i -e "s/HAVE_DOT               = YES/HAVE_DOT    = NO/" \
+				{Doxyfile,doc/Doxyfile} \
+				|| die "disabling dot failed"
+		fi
+
+		cmake_src_compile docs
+	fi
+
+	cmake_src_compile
+}
+
+src_install() {
+	cmake_src_install
+}
diff --git a/app-doc/doxygen/files/doxygen-1.9.1-ignore-bad-encoding.patch b/app-doc/doxygen/files/doxygen-1.9.1-ignore-bad-encoding.patch
new file mode 100644
index 000000000000..5dd7578f24ec
--- /dev/null
+++ b/app-doc/doxygen/files/doxygen-1.9.1-ignore-bad-encoding.patch
@@ -0,0 +1,23 @@
+Hack to avoid "UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 45: invalid start byte"
+in src/._xmlgen.cpp which is binary. This doesn't seem like it should be present at the time the docs
+are generated, but this is simpler for now.
+--- a/doc/translator.py
++++ b/doc/translator.py
+@@ -1454,10 +1454,13 @@ class TrManager:
+ 
+         # Read content of the file as one string.

+         assert os.path.isfile(fname)

+-        f = xopen(fname)

+-        cont = f.read()

+-        cont = ''.join(cont.split('\n')) # otherwise the 'match' function won't work.

+-        f.close()

++        try:

++            with xopen(fname) as f:

++                cont = f.read()

++                cont = ''.join(cont.split('\n')) # otherwise the 'match' function won't work.

++        except UnicodeDecodeError:

++            print("Skipping {0} because of decoding errors".format(fname))

++            return

+ 

+         # Remove the items for identifiers that were found in the file.

+         while lst_in: