From c4981cd87ebae1c31e9de1c130f85aa6c0e5b808 Mon Sep 17 00:00:00 2001
From: Rosen Penev <rosenp@gmail.com>
Date: Sun, 14 Mar 2021 19:21:58 -0700
Subject: [PATCH] tesseract: update to 4.1.1

Switch to AUTORELEASE for simplicity.

Add libarchive dependency since it's now used.

Build with ninja for faster compilation.

Refresh SIMD CMake patch and add musl patch.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
---
 utils/tesseract/Makefile                | 11 +++--
 utils/tesseract/patches/010-cmake.patch | 56 +++++++------------------
 utils/tesseract/patches/020-musl.patch  | 10 +++++
 3 files changed, 30 insertions(+), 47 deletions(-)
 create mode 100644 utils/tesseract/patches/020-musl.patch

diff --git a/utils/tesseract/Makefile b/utils/tesseract/Makefile
index b3aebe5a0..8911bf65c 100644
--- a/utils/tesseract/Makefile
+++ b/utils/tesseract/Makefile
@@ -7,22 +7,21 @@
 include $(TOPDIR)/rules.mk
 
 PKG_NAME:=tesseract
-PKG_VERSION:=4.0.0
-PKG_RELEASE:=2
+PKG_VERSION:=4.1.1
+PKG_RELEASE:=$(AUTORELEASE)
 
 PKG_SOURCE:=$(PKG_NAME)-$(PKG_VERSION).tar.gz
 PKG_SOURCE_URL:=https://codeload.github.com/tesseract-ocr/tesseract/tar.gz/$(PKG_VERSION)?
-PKG_HASH:=a1f5422ca49a32e5f35c54dee5112b11b99928fc9f4ee6695cdc6768d69f61dd
+PKG_HASH:=2a66ff0d8595bff8f04032165e6c936389b1e5727c3ce5a27b3e059d218db1cb
 
 PKG_MAINTAINER:=Valentin Kivachuk <vk18496@gmail.com>
 PKG_LICENSE:=Apache-2.0
 PKG_LICENSE_FILES:=LICENSE
 
-PKG_BUILD_PARALLEL:=1
 CMAKE_BINARY_SUBDIR:=openwrt-build
 
 include $(INCLUDE_DIR)/package.mk
-include $(INCLUDE_DIR)/cmake.mk
+include ../../devel/ninja/ninja-cmake.mk
 
 define Package/tesseract
   MENU:=1
@@ -30,7 +29,7 @@ define Package/tesseract
   CATEGORY:=Utilities
   TITLE:=Tesseract Open Source OCR Engine
   URL:=https://github.com/tesseract-ocr/tesseract
-  DEPENDS:=+libleptonica +libpthread +libstdcpp
+  DEPENDS:=+libleptonica +libarchive +libpthread +libstdcpp
 endef
 
 TARGET_CFLAGS:=$(filter-out -O%,$(TARGET_CFLAGS)) -O3
diff --git a/utils/tesseract/patches/010-cmake.patch b/utils/tesseract/patches/010-cmake.patch
index 5b77e57bf..2a9f88c79 100644
--- a/utils/tesseract/patches/010-cmake.patch
+++ b/utils/tesseract/patches/010-cmake.patch
@@ -1,44 +1,18 @@
 --- a/CMakeLists.txt
 +++ b/CMakeLists.txt
-@@ -224,41 +224,6 @@ set(tesseract_src ${tesseract_src}
-     src/api/pdfrenderer.cpp
- )
- 
--if (WIN32)
--    if (MSVC)
--        include_directories(src/vs2010/tesseract)
--        set(tesseract_hdr
--            ${tesseract_hdr}
--            ${CMAKE_CURRENT_SOURCE_DIR}/src/vs2010/tesseract/resource.h)
--        set(tesseract_rsc ${CMAKE_BINARY_DIR}/vs2010/tesseract/libtesseract.rc)
--        set_source_files_properties(
--            ${CMAKE_CURRENT_SOURCE_DIR}/src/arch/dotproductsse.cpp
--            PROPERTIES COMPILE_DEFINITIONS __SSE4_1__)
--        set_source_files_properties(
--            ${CMAKE_CURRENT_SOURCE_DIR}/src/arch/intsimdmatrixsse.cpp
--            PROPERTIES COMPILE_DEFINITIONS __SSE4_1__)
--        set_source_files_properties(
--            ${CMAKE_CURRENT_SOURCE_DIR}/src/arch/dotproductavx.cpp
--            PROPERTIES COMPILE_FLAGS "/arch:AVX")
--        set_source_files_properties(
--            ${CMAKE_CURRENT_SOURCE_DIR}/src/arch/intsimdmatrixavx2.cpp
--            PROPERTIES COMPILE_FLAGS "/arch:AVX2")
+@@ -252,15 +252,6 @@ set(AVX2_OPT OFF)
+ set(FMA_OPT OFF)
+ set(SSE41_OPT OFF)
+ set(MARCH_NATIVE_OPT OFF)
+-foreach(flag ${_enable_vector_unit_list})  # from OptimizeForArchitecture()
+-    string(TOUPPER "${flag}" flag)
+-    string(REPLACE "\." "_" flag "${flag}")
+-    set(simd_flags "${simd_flags} -D${flag}")
+-    string(REPLACE "_" "" flag "${flag}")
+-    if("${flag}" MATCHES "AVX|AVX2|FMA|SSE41")
+-        set("${flag}_OPT" ON)
 -    endif()
--else()
--    set_source_files_properties(
--            ${CMAKE_CURRENT_SOURCE_DIR}/src/arch/dotproductsse.cpp
--            PROPERTIES COMPILE_FLAGS "-msse4.1")
--    set_source_files_properties(
--            ${CMAKE_CURRENT_SOURCE_DIR}/src/arch/intsimdmatrixsse.cpp
--            PROPERTIES COMPILE_FLAGS "-msse4.1")
--    set_source_files_properties(
--            ${CMAKE_CURRENT_SOURCE_DIR}/src/arch/dotproductavx.cpp
--            PROPERTIES COMPILE_FLAGS "-mavx")
--    set_source_files_properties(
--            ${CMAKE_CURRENT_SOURCE_DIR}/src/arch/intsimdmatrixavx2.cpp
--            PROPERTIES COMPILE_FLAGS "-mavx2")
--endif()
--
- add_library                     (libtesseract ${LIBRARY_TYPE} ${tesseract_src} ${tesseract_hdr}
-     ${tesseract_rsc}
-     )
+-endforeach(flag)
+ if (NOT MSVC)
+    set(MARCH_NATIVE_FLAGS "${MARCH_NATIVE_FLAGS} -O3 -ffast-math")
+ endif()
diff --git a/utils/tesseract/patches/020-musl.patch b/utils/tesseract/patches/020-musl.patch
new file mode 100644
index 000000000..821a51d06
--- /dev/null
+++ b/utils/tesseract/patches/020-musl.patch
@@ -0,0 +1,10 @@
+--- a/src/ccutil/ocrclass.h
++++ b/src/ccutil/ocrclass.h
+@@ -28,6 +28,7 @@
+ 
+ #include <chrono>
+ #include <ctime>
++#include <sys/time.h>
+ #ifdef _WIN32
+ #include <winsock2.h> // for timeval
+ #endif