ocr: import tesseract package from upstream
Upgraded app-text/tesseract to version 4.1.1 on amd64, arm
BUG=b:159229207
TEST=build successfully with emerge and test on hatch device:
tesseract phototest.tif phototest -l eng --psm 1 --oem 1
--tessdata-dir /usr/share/tessdata txt pdf hocr
Cq-Depend: chromium:2264287
Change-Id: Ibc913c4d31b394399a21c0328f9a2f6e53144a79
Reviewed-on: https://chromium-review.googlesource.com/c/chromiumos/overlays/portage-stable/+/2258057
Reviewed-by: Mike Frysinger <vapier@chromium.org>
Commit-Queue: Efi Mavroudi <emavroudi@google.com>
Tested-by: Efi Mavroudi <emavroudi@google.com>
diff --git a/app-text/tesseract/Manifest b/app-text/tesseract/Manifest
new file mode 100644
index 0000000..f3e80c5
--- /dev/null
+++ b/app-text/tesseract/Manifest
@@ -0,0 +1 @@
+DIST tesseract-4.1.1.tar.gz 1974988 BLAKE2B f729beb017ec81704b84f5e6539706ff0cd6a51a4d1b7b730c394af58ce2ea2a659203bf239417e8d500cfbde845a73d6e6e181ca946257c2f22525820899602 SHA512 017723a2268be789fe98978eed02fd294968cc8050dde376dee026f56f2b99df42db935049ae5e72c4519a920e263b40af1a6a40d9942e66608145b3131a71a2
diff --git a/app-text/tesseract/metadata.xml b/app-text/tesseract/metadata.xml
new file mode 100644
index 0000000..c3f2037
--- /dev/null
+++ b/app-text/tesseract/metadata.xml
@@ -0,0 +1,25 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE pkgmetadata SYSTEM "http://www.gentoo.org/dtd/metadata.dtd">
+<pkgmetadata>
+ <maintainer type="person">
+ <email>voyageur@gentoo.org</email>
+ <name>Bernard Cafarelli</name>
+ </maintainer>
+ <maintainer type="person">
+ <email>chutzpah@gentoo.org</email>
+ <name>Patrick McLean</name>
+ </maintainer>
+ <longdescription lang="en">A commercial quality OCR engine originally developed at HP between 1985 and
+ 1995. In 1995, this engine was among the top 3 evaluated by UNLV. It was
+ open-sourced by HP and UNLV in 2005.</longdescription>
+ <upstream>
+ <remote-id type="github">tesseract-ocr/tesseract</remote-id>
+ </upstream>
+ <use>
+ <flag name="math"> Enable support for recognition of equations.</flag>
+ <flag name="scrollview"> Install viewer to debug recognition (ScrollView).</flag>
+ <flag name="osd"> Enable support orientation and script detection.</flag>
+ <flag name="opencl">Enable opencl support for speedup using GPU computation.</flag>
+ <flag name="training">Install training applications to add support for new languages.</flag>
+ </use>
+</pkgmetadata>
diff --git a/app-text/tesseract/tesseract-4.1.1.ebuild b/app-text/tesseract/tesseract-4.1.1.ebuild
new file mode 100644
index 0000000..1b16d83
--- /dev/null
+++ b/app-text/tesseract/tesseract-4.1.1.ebuild
@@ -0,0 +1,91 @@
+# Copyright 1999-2020 Gentoo Authors
+# Distributed under the terms of the GNU General Public License v2
+
+EAPI=7
+
+inherit autotools multilib-minimal toolchain-funcs
+
+DESCRIPTION="An OCR Engine, originally developed at HP, now open source."
+HOMEPAGE="https://github.com/tesseract-ocr"
+SRC_URI="https://github.com/tesseract-ocr/${PN}/archive/${PV}.tar.gz -> ${P}.tar.gz"
+
+LICENSE="Apache-2.0"
+SLOT="0"
+KEYWORDS="*"
+IUSE="doc jpeg opencl openmp png static-libs tiff training webp"
+
+COMMON_DEPEND=">=media-libs/leptonica-1.74:=[${MULTILIB_USEDEP},zlib,tiff?,jpeg?,png?,webp?]
+ opencl? (
+ virtual/opencl[${MULTILIB_USEDEP}]
+ media-libs/tiff:0=[${MULTILIB_USEDEP}]
+ media-libs/leptonica:=[tiff]
+ )
+ training? (
+ dev-libs/icu:=
+ x11-libs/pango:=
+ x11-libs/cairo:=
+ )"
+RDEPEND="${COMMON_DEPEND}
+ || (
+ >=app-text/tessdata_fast-4.0.0
+ >=app-text/tessdata_best-4.0.0
+ >=app-text/tessdata_legacy-4.0.0
+ )"
+DEPEND="${COMMON_DEPEND}
+ app-text/asciidoc
+ app-text/docbook-xsl-stylesheets
+ dev-libs/libxslt
+ doc? ( app-doc/doxygen )"
+
+pkg_pretend() {
+ [[ ${MERGE_TYPE} != binary ]] && use openmp && tc-check-openmp
+}
+
+pkg_setup() {
+ [[ ${MERGE_TYPE} != binary ]] && use openmp && tc-check-openmp
+}
+
+src_prepare() {
+ default
+ eautoreconf
+}
+
+multilib_src_configure() {
+ # scrollview disabled for now, see bug #686944
+ local myeconfargs=(
+ --enable-shared
+ --disable-graphics
+ $(use_enable opencl)
+ $(use_enable openmp)
+ $(use_enable static-libs static)
+ )
+
+ ECONF_SOURCE="${S}" econf "${myeconfargs[@]}"
+}
+
+multilib_src_compile() {
+ default
+ if multilib_is_native_abi; then
+ use doc && emake doc
+ use training && emake training
+ fi
+}
+
+multilib_src_install() {
+ if multilib_is_native_abi; then
+ DOCS=( "${S}"/AUTHORS "${S}"/ChangeLog "${S}"/README.md )
+ if use doc; then
+ HTML_DOCS=( doc/html/. )
+ fi
+ einstalldocs
+
+ if use training; then
+ emake DESTDIR="${D}" training-install
+ fi
+ fi
+ emake DESTDIR="${D}" install
+}
+
+multilib_src_install_all() {
+ find "${D}" -name '*.la' -type f -delete || die
+}
diff --git a/metadata/md5-cache/app-text/tesseract-4.1.1 b/metadata/md5-cache/app-text/tesseract-4.1.1
new file mode 100644
index 0000000..f63dc13
--- /dev/null
+++ b/metadata/md5-cache/app-text/tesseract-4.1.1
@@ -0,0 +1,14 @@
+BDEPEND=>=app-portage/elt-patches-20170815
+DEFINED_PHASES=compile configure install prepare pretend setup test
+DEPEND=>=media-libs/leptonica-1.74:=[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?,abi_ppc_32(-)?,abi_ppc_64(-)?,abi_s390_32(-)?,abi_s390_64(-)?,abi_arm_32(-)?,abi_arm_64(-)?,zlib,tiff?,jpeg?,png?,webp?] opencl? ( virtual/opencl[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?,abi_ppc_32(-)?,abi_ppc_64(-)?,abi_s390_32(-)?,abi_s390_64(-)?,abi_arm_32(-)?,abi_arm_64(-)?] media-libs/tiff:0=[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?,abi_ppc_32(-)?,abi_ppc_64(-)?,abi_s390_32(-)?,abi_s390_64(-)?,abi_arm_32(-)?,abi_arm_64(-)?] media-libs/leptonica:=[tiff] ) training? ( dev-libs/icu:= x11-libs/pango:= x11-libs/cairo:= ) app-text/asciidoc app-text/docbook-xsl-stylesheets dev-libs/libxslt doc? ( app-doc/doxygen ) !<sys-devel/gettext-0.18.1.1-r3 || ( >=sys-devel/automake-1.16.1:1.16 >=sys-devel/automake-1.15.1:1.15 ) >=sys-devel/autoconf-2.69 >=sys-devel/libtool-2.4
+DESCRIPTION=An OCR Engine, originally developed at HP, now open source.
+EAPI=7
+HOMEPAGE=https://github.com/tesseract-ocr
+IUSE=doc jpeg opencl openmp png static-libs tiff training webp abi_x86_32 abi_x86_64 abi_x86_x32 abi_mips_n32 abi_mips_n64 abi_mips_o32 abi_ppc_32 abi_ppc_64 abi_s390_32 abi_s390_64 abi_arm_32 abi_arm_64
+KEYWORDS=*
+LICENSE=Apache-2.0
+RDEPEND=>=media-libs/leptonica-1.74:=[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?,abi_ppc_32(-)?,abi_ppc_64(-)?,abi_s390_32(-)?,abi_s390_64(-)?,abi_arm_32(-)?,abi_arm_64(-)?,zlib,tiff?,jpeg?,png?,webp?] opencl? ( virtual/opencl[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?,abi_ppc_32(-)?,abi_ppc_64(-)?,abi_s390_32(-)?,abi_s390_64(-)?,abi_arm_32(-)?,abi_arm_64(-)?] media-libs/tiff:0=[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?,abi_ppc_32(-)?,abi_ppc_64(-)?,abi_s390_32(-)?,abi_s390_64(-)?,abi_arm_32(-)?,abi_arm_64(-)?] media-libs/leptonica:=[tiff] ) training? ( dev-libs/icu:= x11-libs/pango:= x11-libs/cairo:= ) || ( >=app-text/tessdata_fast-4.0.0 >=app-text/tessdata_best-4.0.0 >=app-text/tessdata_legacy-4.0.0 )
+SLOT=0
+SRC_URI=https://github.com/tesseract-ocr/tesseract/archive/4.1.1.tar.gz -> tesseract-4.1.1.tar.gz
+_eclasses_=autotools d0e5375d47f4c809f406eb892e531513 libtool f143db5a74ccd9ca28c1234deffede96 multibuild 40fe59465edacd730c644ec2bc197809 multilib b2f01ad412baf81650c23fcf0975fa33 multilib-build 1979aa0ff4d356d32507ca4650d9f37d multilib-minimal 8bddda43703ba94d8341f4e247f97566 toolchain-funcs 209edad4a5c4812e7b2f8021650974f0
+_md5_=6dfbe5ea57150f91db4a350dcf3468f2