Early days, but useful for some testing still. src_test fails dismally still.

(Portage version: 2.2.0_alpha142/cvs/Linux x86_64, unsigned Manifest commit)

2 files changed, 225 insertions, 1 deletions

diff --git a/sys-libs/db/ChangeLog b/sys-libs/db/ChangeLog
index d44b8ea61055..c7d9bd596b68 100644
--- a/sys-libs/db/ChangeLog
+++ b/sys-libs/db/ChangeLog
@@ -1,6 +1,11 @@
 # ChangeLog for sys-libs/db
 # Copyright 1999-2012 Gentoo Foundation; Distributed under the GPL v2
-# $Header: /var/cvsroot/gentoo-x86/sys-libs/db/ChangeLog,v 1.332 2012/10/21 00:03:09 blueness Exp $
+# $Header: /var/cvsroot/gentoo-x86/sys-libs/db/ChangeLog,v 1.333 2012/11/12 23:09:59 robbat2 Exp $
+
+*db-5.3.21 (12 Nov 2012)
+
+  12 Nov 2012; Robin H. Johnson <robbat2@gentoo.org> +db-5.3.21.ebuild:
+  Early days, but useful for some testing still. src_test fails dismally still.
 
   21 Oct 2012; Anthony G. Basile <blueness@gentoo.org> db-4.8.30.ebuild:
   stable arm, bug #429188
diff --git a/sys-libs/db/db-5.3.21.ebuild b/sys-libs/db/db-5.3.21.ebuild
new file mode 100644
index 000000000000..0b9d68780a1c
--- /dev/null
+++ b/sys-libs/db/db-5.3.21.ebuild
@@ -0,0 +1,219 @@
+# Copyright 1999-2012 Gentoo Foundation
+# Distributed under the terms of the GNU General Public License v2
+# $Header: /var/cvsroot/gentoo-x86/sys-libs/db/db-5.3.21.ebuild,v 1.1 2012/11/12 23:09:59 robbat2 Exp $
+
+EAPI=2
+inherit eutils db flag-o-matic java-pkg-opt-2 autotools multilib
+
+#Number of official patches
+#PATCHNO=`echo ${PV}|sed -e "s,\(.*_p\)\([0-9]*\),\2,"`
+PATCHNO=${PV/*.*.*_p}
+if [[ ${PATCHNO} == "${PV}" ]] ; then
+	MY_PV=${PV}
+	MY_P=${P}
+	PATCHNO=0
+else
+	MY_PV=${PV/_p${PATCHNO}}
+	MY_P=${PN}-${MY_PV}
+fi
+
+S_BASE="${WORKDIR}/${MY_P}"
+S="${S_BASE}/build_unix"
+DESCRIPTION="Oracle Berkeley DB"
+HOMEPAGE="http://www.oracle.com/technology/software/products/berkeley-db/index.html"
+SRC_URI="http://download.oracle.com/berkeley-db/${MY_P}.tar.gz"
+for (( i=1 ; i<=${PATCHNO} ; i++ )) ; do
+	export SRC_URI="${SRC_URI} http://www.oracle.com/technology/products/berkeley-db/db/update/${MY_PV}/patch.${MY_PV}.${i}"
+done
+
+LICENSE="OracleDB"
+SLOT="5.3"
+KEYWORDS="~alpha ~amd64 ~arm ~hppa ~ia64 ~m68k ~ppc ~ppc64 ~s390 ~sh ~sparc ~x86 ~sparc-fbsd ~x86-fbsd"
+IUSE="doc java cxx tcl test"
+
+# the entire testsuite needs the TCL functionality
+DEPEND="tcl? ( >=dev-lang/tcl-8.4 )
+	test? ( >=dev-lang/tcl-8.4 )
+	java? ( >=virtual/jdk-1.5 )
+	>=sys-devel/binutils-2.16.1"
+RDEPEND="tcl? ( dev-lang/tcl )
+	java? ( >=virtual/jre-1.5 )"
+
+src_unpack() {
+	unpack "${MY_P}".tar.gz
+}
+
+src_prepare() {
+	cd "${WORKDIR}"/"${MY_P}"
+	for (( i=1 ; i<=${PATCHNO} ; i++ ))
+	do
+		epatch "${DISTDIR}"/patch."${MY_PV}"."${i}"
+	done
+	#epatch "${FILESDIR}"/${PN}-4.8-libtool.patch
+	# upstreamed:5.2.36
+	#epatch "${FILESDIR}"/${PN}-4.8.24-java-manifest-location.patch
+
+	# use the includes from the prefix
+	epatch "${FILESDIR}"/${PN}-4.6-jni-check-prefix-first.patch
+	epatch "${FILESDIR}"/${PN}-4.3-listen-to-java-options.patch
+
+	# upstream autoconf fails to build DBM when it's supposed to
+	# merged upstream in 5.0.26
+	#epatch "${FILESDIR}"/${PN}-5.0.21-enable-dbm-autoconf.patch
+
+	# sqlite configure call has an extra leading ..
+	# upstreamed:5.2.36, missing in 5.3.x
+	epatch "${FILESDIR}"/${PN}-5.2.28-sqlite-configure-path.patch
+
+	# Upstream release script grabs the dates when the script was run, so lets
+	# end-run them to keep the date the same.
+	export REAL_DB_RELEASE_DATE="$(awk \
+		'/^DB_VERSION_STRING=/{ gsub(".*\\(|\\).*","",$0); print $0; }' \
+		"${S_BASE}"/dist/configure)"
+	sed -r -i \
+		-e "/^DB_RELEASE_DATE=/s~=.*~='${REAL_DB_RELEASE_DATE}'~g" \
+		"${S_BASE}"/dist/RELEASE
+
+	# Include the SLOT for Java JAR files
+	# This supersedes the unused jarlocation patches.
+	sed -r -i \
+		-e '/jarfile=.*\.jar$/s,(.jar$),-$(LIBVERSION)\1,g' \
+		"${S_BASE}"/dist/Makefile.in
+
+	cd "${S_BASE}"/dist
+	rm -f aclocal/libtool.m4
+	sed -i \
+		-e '/AC_PROG_LIBTOOL$/aLT_OUTPUT' \
+		configure.ac
+	sed -i \
+		-e '/^AC_PATH_TOOL/s/ sh, none/ bash, none/' \
+		aclocal/programs.m4
+	AT_M4DIR="aclocal aclocal_java" eautoreconf
+	# Upstream sucks - they do autoconf and THEN replace the version variables.
+	. ./RELEASE
+	for v in \
+		DB_VERSION_{FAMILY,LETTER,RELEASE,MAJOR,MINOR} \
+		DB_VERSION_{PATCH,FULL,UNIQUE_NAME,STRING,FULL_STRING} \
+		DB_VERSION \
+		DB_RELEASE_DATE ; do
+		local ev="__EDIT_${v}__"
+		sed -i -e "s/${ev}/${!v}/g" configure
+	done
+}
+
+src_configure() {
+	local myconf=''
+
+	# compilation with -O0 fails on amd64, see bug #171231
+	if use amd64; then
+		replace-flags -O0 -O2
+		is-flagq -O[s123] || append-flags -O2
+	fi
+
+	# use `set` here since the java opts will contain whitespace
+	set --
+	if use java ; then
+		set -- "$@" \
+			--with-java-prefix="${JAVA_HOME}" \
+			--with-javac-flags="$(java-pkg_javac-args)"
+	fi
+
+	# Add linker versions to the symbols. Easier to do, and safer than header file
+	# mumbo jumbo.
+	if use userland_GNU ; then
+		append-ldflags -Wl,--default-symver
+	fi
+
+	# Bug #270851: test needs TCL support
+	if use tcl || use test ; then
+		myconf="${myconf} --enable-tcl"
+		myconf="${myconf} --with-tcl=/usr/$(get_libdir)"
+	else
+		myconf="${myconf} --disable-tcl"
+	fi
+
+	# sql_compat will cause a collision with sqlite3
+	# --enable-sql_compat
+	cd "${S}"
+	ECONF_SOURCE="${S_BASE}"/dist \
+	STRIP="true" \
+	econf \
+		--enable-compat185 \
+		--enable-dbm \
+		--enable-o_direct \
+		--without-uniquename \
+		--enable-sql \
+		--enable-sql_codegen \
+		--disable-sql_compat \
+		$(use arm && echo --with-mutex=ARM/gcc-assembly) \
+		$(use amd64 && echo --with-mutex=x86/gcc-assembly) \
+		$(use_enable cxx) \
+		$(use_enable cxx stl) \
+		$(use_enable java) \
+		${myconf} \
+		$(use_enable test) \
+		"$@"
+}
+
+src_compile() {
+	emake || die "make failed"
+}
+
+src_install() {
+	emake install DESTDIR="${D}" || die
+
+	db_src_install_usrbinslot
+
+	db_src_install_headerslot
+
+	db_src_install_doc
+
+	db_src_install_usrlibcleanup
+
+	dodir /usr/sbin
+	# This file is not always built, and no longer exists as of db-4.8
+	[[ -f "${D}"/usr/bin/berkeley_db_svc ]] && \
+	mv "${D}"/usr/bin/berkeley_db_svc "${D}"/usr/sbin/berkeley_db"${SLOT/./}"_svc
+
+	if use java; then
+		java-pkg_regso "${D}"/usr/"$(get_libdir)"/libdb_java*.so
+		java-pkg_dojar "${D}"/usr/"$(get_libdir)"/*.jar
+		rm -f "${D}"/usr/"$(get_libdir)"/*.jar
+	fi
+}
+
+pkg_postinst() {
+	db_fix_so
+}
+
+pkg_postrm() {
+	db_fix_so
+}
+
+src_test() {
+	# db_repsite is impossible to build, as upstream strips those sources.
+	# db_repsite is used directly in the setup_site_prog,
+	# setup_site_prog is called from open_site_prog
+	# which is called only from tests in the multi_repmgr group.
+	#sed -ri \
+	#	-e '/set subs/s,multi_repmgr,,g' \
+	#	"${S_BASE}/test/testparams.tcl"
+	sed -ri \
+		-e '/multi_repmgr/d' \
+		"${S_BASE}/test/tcl/test.tcl"
+
+	# This is the only failure in 5.2.28 so far, and looks like a false positive.
+	# Repmgr018 (btree): Test of repmgr stats.
+	#     Repmgr018.a: Start a master.
+	#     Repmgr018.b: Start a client.
+	#     Repmgr018.c: Run some transactions at master.
+	#         Rep_test: btree 20 key/data pairs starting at 0
+	#         Rep_test.a: put/get loop
+	# FAIL:07:05:59 (00:00:00) perm_no_failed_stat: expected 0, got 1
+	sed -ri \
+		-e '/set parms.*repmgr018/d' \
+		-e 's/repmgr018//g' \
+		"${S_BASE}/test/tcl/test.tcl"
+
+	db_src_test
+}