[GLUTEN-8194][VL] Add icu port and enable test component on vcpkg

dev/vcpkg/ports/icu/disable-escapestr-tool.patch

@@ -0,0 +1,17 @@
+diff --git a/source/tools/Makefile.in b/source/tools/Makefile.in
+index e0896f1..5ead980 100644
+--- a/source/tools/Makefile.in
++++ b/source/tools/Makefile.in
+@@ -19,9 +19,9 @@ SUBDIRS = toolutil ctestfw makeconv genrb genbrk \
+ gencnval gensprep icuinfo genccode gencmn icupkg pkgdata \
+ gentest gennorm2 gencfu gendict icuexportdata
+
+-ifneq (@platform_make_fragment_name@,mh-cygwin-msvc)
+-SUBDIRS += escapesrc
+-endif
++#ifneq (@platform_make_fragment_name@,mh-cygwin-msvc)
++#SUBDIRS += escapesrc
++#endif
+
+ ## List of phony targets
+ .PHONY : all all-local all-recursive install install-local	\

dev/vcpkg/ports/icu/disable-static-prefix.patch

@@ -0,0 +1,20 @@
+diff --git a/source/icudefs.mk.in b/source/icudefs.mk.in
+index 24bd97a..33169de 100644
+--- a/source/icudefs.mk.in
++++ b/source/icudefs.mk.in
+@@ -213,13 +213,13 @@ LIBICU = $(LIBPREFIX)$(ICUPREFIX)
+
+ ## If we can't use the shared libraries, use the static libraries
+ ifneq ($(ENABLE_SHARED),YES)
+-STATIC_PREFIX_WHEN_USED = s
++STATIC_PREFIX_WHEN_USED =
+ else
+ STATIC_PREFIX_WHEN_USED = 
+ endif
+
+ # Static library prefix and file extension
+-STATIC_PREFIX = s
++STATIC_PREFIX =
+ LIBSICU = $(LIBPREFIX)$(STATIC_PREFIX)$(ICUPREFIX)
+ A = a
+ SOBJ = $(SO)

dev/vcpkg/ports/icu/fix-extra.patch

@@ -0,0 +1,13 @@
+diff --urN a/source/extra/Makefile.in b/source/extra/Makefile.in
+--- a/source/extra/Makefile.in
++++ b/source/extra/Makefile.in 
+@@ -23,7 +23,7 @@
+ ## Files to remove for 'make clean'
+ CLEANFILES = *~
+
+-SUBDIRS = scrptrun uconv
++SUBDIRS = uconv
+
+ ## List of phony targets
+ .PHONY : all all-local all-recursive install install-local		\
+

dev/vcpkg/ports/icu/fix-win-build.patch

@@ -0,0 +1,35 @@
+diff --git a/source/configure.ac b/source/configure.ac
+index 425fdc7..97210f8 100644
+--- a/source/configure.ac
++++ b/source/configure.ac
+@@ -114,6 +114,15 @@ AC_ARG_ENABLE([icu-config],
+         esac], [enable_icu_config=true])
+ AC_SUBST(INSTALL_ICU_CONFIG, [$enable_icu_config])
+
++AC_ARG_ENABLE([icu-build-win],
++    AS_HELP_STRING([--enable-icu-build-win], [install icu-build-win]),
++        [case "${enableval}" in
++            yes) enable_icu_build_win=true ;;
++            no) enable_icu_build_win=false ;;
++            *) AC_MSG_ERROR([bad value '${enableval}' for --enable-icu-build-win]) ;;
++        esac], [enable_icu_build_win=true])
++AC_SUBST(INSTALL_ICU_BUILD_WIN, [$enable_icu_build_win])
++
+ # Check whether to build debug libraries
+ AC_MSG_CHECKING([whether to build debug libraries])
+ enabled=no
+@@ -263,12 +272,14 @@ ICU_CHECK_MH_FRAG
+ # Checks for libraries and other host specific stuff
+ # On HP/UX, don't link to -lm from a shared lib because it isn't
+ #  PIC (at least on 10.2)
++if test "$enable_icu_build_win" = no; then
+ case "${host}" in
+ 	*-*-hpux*) 	AC_CHECK_LIB(m, floor, LIB_M="-lm") ;;
+
+ 	*) 		AC_CHECK_LIB(m, floor)
+ 			LIB_M="" ;;
+ esac
++fi
+ AC_SUBST(LIB_M)
+
+ # Check whether to build shared libraries

dev/vcpkg/ports/icu/fix_parallel_build_on_windows.patch

@@ -0,0 +1,25 @@
+diff --urN a/source/data/Makefile.in b/source/data/Makefile.in
+--- a/source/data/Makefile.in
++++ b/source/data/Makefile.in
+@@ -236,11 +236,12 @@
+ ## Include the main build rules for data files
+ include $(top_builddir)/$(subdir)/rules.mk
+
++PKGDATA_LIST = $(TMP_DIR)/icudata.lst
+
+ ifeq ($(ENABLE_SO_VERSION_DATA),1)
+ ifeq ($(PKGDATA_MODE),dll)
+ SO_VERSION_DATA = $(OUTTMPDIR)/icudata.res
+-$(SO_VERSION_DATA) : $(MISCSRCDIR)/icudata.rc | $(TMP_DIR)/dirs.timestamp
++$(SO_VERSION_DATA) : $(MISCSRCDIR)/icudata.rc $(PKGDATA_LIST)
+ ifeq ($(MSYS_RC_MODE),1)
+ 	rc.exe -i$(srcdir)/../common -i$(top_builddir)/common -fo$@ $(CPPFLAGS) $<
+ else
+@@ -249,7 +250,6 @@
+ endif
+ endif
+
+-PKGDATA_LIST = $(TMP_DIR)/icudata.lst
+
+
+ #####################################################

dev/vcpkg/ports/icu/mingw-dll-install.patch

@@ -0,0 +1,26 @@
+diff --git a/source/config/mh-mingw b/source/config/mh-mingw
+index 30f6e5be81..b6364551ea 100644
+--- a/source/config/mh-mingw
++++ b/source/config/mh-mingw
+@@ -13,7 +13,7 @@
+ # On Windows we generally have the DLLs in the bin directory rather than the lib directory.
+ # This setting moves the ICU DLLs into the bin folder for MinGW/MSYS2 when "make install" is run.
+ # If you prefer to have the DLLs in the lib folder, then set this to NO instead.
+-MINGW_MOVEDLLSTOBINDIR = YES
++MINGW_MOVEDLLSTOBINDIR = NO
+
+ # We install sbin tools into the same bin directory because
+ # pkgdata needs some of the tools in sbin, and we can't always depend on
+diff --git a/source/config/mh-mingw64 b/source/config/mh-mingw64
+index fb64c56260..a43cc4dd71 100644
+--- a/source/config/mh-mingw64
++++ b/source/config/mh-mingw64
+@@ -10,7 +10,7 @@
+ # On Windows we generally have the DLLs in the bin directory rather than the lib directory.
+ # This setting moves the ICU DLLs into the bin folder for MinGW/MSYS2 when "make install" is run.
+ # If you prefer to have the DLLs in the lib folder, then set this to NO instead.
+-MINGW_MOVEDLLSTOBINDIR = YES
++MINGW_MOVEDLLSTOBINDIR = NO
+
+ # This file is similar to mh-mingw
+ # Any changes made here may also need to be made in mh-mingw

dev/vcpkg/ports/icu/portfile.cmake

@@ -0,0 +1,200 @@
+string(REGEX MATCH "^[0-9]*" ICU_VERSION_MAJOR "${VERSION}")
+string(REPLACE "." "_" VERSION2 "${VERSION}")
+string(REPLACE "." "-" VERSION3 "${VERSION}")
+
+vcpkg_download_distfile(
+    ARCHIVE
+    URLS "https://github.com/unicode-org/icu/releases/download/release-${VERSION3}/icu4c-${VERSION2}-src.tgz"
+    FILENAME "icu4c-${VERSION2}-src.tgz"
+    SHA512 e788e372716eecebc39b56bbc88f3a458e21c3ef20631c2a3d7ef05794a678fe8dad482a03a40fdb9717109a613978c7146682e98ee16fade5668d641d5c48f8
+)
+
+vcpkg_extract_source_archive(SOURCE_PATH
+    ARCHIVE "${ARCHIVE}"
+    PATCHES
+        disable-escapestr-tool.patch
+        remove-MD-from-configure.patch
+        fix_parallel_build_on_windows.patch
+        fix-extra.patch
+        mingw-dll-install.patch
+        disable-static-prefix.patch # https://gitlab.kitware.com/cmake/cmake/-/issues/16617; also mingw.
+        fix-win-build.patch
+)
+
+vcpkg_find_acquire_program(PYTHON3)
+set(ENV{PYTHON} "${PYTHON3}")
+
+if(VCPKG_TARGET_IS_WINDOWS)
+    list(APPEND CONFIGURE_OPTIONS --enable-icu-build-win)
+endif()
+
+list(APPEND CONFIGURE_OPTIONS --disable-samples --disable-layoutex)
+
+list(APPEND CONFIGURE_OPTIONS_RELEASE --disable-debug --enable-release)
+list(APPEND CONFIGURE_OPTIONS_DEBUG  --enable-debug --disable-release)
+
+set(RELEASE_TRIPLET ${TARGET_TRIPLET}-rel)
+set(DEBUG_TRIPLET ${TARGET_TRIPLET}-dbg)
+
+if("tools" IN_LIST FEATURES)
+  list(APPEND CONFIGURE_OPTIONS --enable-tools)
+else()
+  list(APPEND CONFIGURE_OPTIONS --disable-tools)
+endif()
+if(CMAKE_HOST_WIN32 AND VCPKG_TARGET_IS_MINGW AND NOT HOST_TRIPLET MATCHES "mingw")
+    # Assuming no cross compiling because the host (windows) pkgdata tool doesn't
+    # use the '/' path separator when creating compiler commands for mingw bash.
+elseif(VCPKG_CROSSCOMPILING)
+    set(TOOL_PATH "${CURRENT_HOST_INSTALLED_DIR}/tools/${PORT}")
+    # convert to unix path
+    string(REGEX REPLACE "^([a-zA-Z]):/" "/\\1/" _VCPKG_TOOL_PATH "${TOOL_PATH}")
+    list(APPEND CONFIGURE_OPTIONS "--with-cross-build=${_VCPKG_TOOL_PATH}")
+endif()
+
+vcpkg_configure_make(
+    SOURCE_PATH "${SOURCE_PATH}"
+    AUTOCONFIG
+    PROJECT_SUBPATH source
+    ADDITIONAL_MSYS_PACKAGES autoconf-archive
+    OPTIONS ${CONFIGURE_OPTIONS}
+    OPTIONS_RELEASE ${CONFIGURE_OPTIONS_RELEASE}
+    OPTIONS_DEBUG ${CONFIGURE_OPTIONS_DEBUG}
+    DETERMINE_BUILD_TRIPLET
+)
+
+if(VCPKG_TARGET_IS_OSX AND VCPKG_LIBRARY_LINKAGE STREQUAL "dynamic")
+
+    vcpkg_build_make()
+    # remove this block if https://unicode-org.atlassian.net/browse/ICU-21458
+    # is resolved and use the configure script instead
+    if(DEFINED CMAKE_INSTALL_NAME_DIR)
+        set(ID_PREFIX "${CMAKE_INSTALL_NAME_DIR}")
+    else()
+        set(ID_PREFIX "@rpath")
+    endif()
+
+    # install_name_tool may be missing if cross-compiling
+    find_program(
+        INSTALL_NAME_TOOL
+        install_name_tool
+        HINTS /usr/bin /Library/Developer/CommandLineTools/usr/bin/
+        DOC "Absolute path of install_name_tool"
+        REQUIRED
+    )
+
+    message(STATUS "setting rpath prefix for macOS dynamic libraries")
+
+    if("tools" IN_LIST FEATURES)
+        set(LIBICUTU_RPATH "libicutu")
+    endif()
+
+    #31680: Fix @rpath in both debug and release build
+    foreach(CONFIG_TRIPLE IN ITEMS ${DEBUG_TRIPLET} ${RELEASE_TRIPLET})
+        # add ID_PREFIX to libicudata libicui18n libicuio libicutu libicuuc
+        foreach(LIB_NAME IN ITEMS libicudata libicui18n libicuio ${LIBICUTU_RPATH} libicuuc)
+            vcpkg_execute_build_process(
+                COMMAND "${INSTALL_NAME_TOOL}" -id "${ID_PREFIX}/${LIB_NAME}.${ICU_VERSION_MAJOR}.dylib"
+                "${LIB_NAME}.${VERSION}.dylib"
+                WORKING_DIRECTORY "${CURRENT_BUILDTREES_DIR}/${CONFIG_TRIPLE}/lib"
+                LOGNAME "make-build-fix-rpath-${CONFIG_TRIPLE}"
+            )
+        endforeach()
+
+        # add ID_PREFIX to libicui18n libicuio libicutu dependencies
+        foreach(LIB_NAME IN ITEMS libicui18n libicuio)
+            vcpkg_execute_build_process(
+                COMMAND "${INSTALL_NAME_TOOL}" -change "libicuuc.${ICU_VERSION_MAJOR}.dylib"
+                                                    "${ID_PREFIX}/libicuuc.${ICU_VERSION_MAJOR}.dylib"
+                                                    "${LIB_NAME}.${VERSION}.dylib"
+                WORKING_DIRECTORY "${CURRENT_BUILDTREES_DIR}/${CONFIG_TRIPLE}/lib"
+                LOGNAME "make-build-fix-rpath-${CONFIG_TRIPLE}"
+            )
+            vcpkg_execute_build_process(
+                COMMAND "${INSTALL_NAME_TOOL}" -change "libicudata.${ICU_VERSION_MAJOR}.dylib"
+                                                    "${ID_PREFIX}/libicudata.${ICU_VERSION_MAJOR}.dylib"
+                                                    "${LIB_NAME}.${VERSION}.dylib"
+                WORKING_DIRECTORY "${CURRENT_BUILDTREES_DIR}/${CONFIG_TRIPLE}/lib"
+                LOGNAME "make-build-fix-rpath-${CONFIG_TRIPLE}"
+            )
+        endforeach()
+
+        # add ID_PREFIX to remaining libicuio libicutu dependencies
+        foreach(LIB_NAME libicuio libicutu)
+            vcpkg_execute_build_process(
+                COMMAND "${INSTALL_NAME_TOOL}" -change "libicui18n.${ICU_VERSION_MAJOR}.dylib"
+                                                    "${ID_PREFIX}/libicui18n.${ICU_VERSION_MAJOR}.dylib"
+                                                    "${LIB_NAME}.${VERSION}.dylib"
+                WORKING_DIRECTORY "${CURRENT_BUILDTREES_DIR}/${CONFIG_TRIPLE}/lib"
+                LOGNAME "make-build-fix-rpath-${CONFIG_TRIPLE}"
+            )
+        endforeach()
+
+        # add ID_PREFIX to libicuuc dependencies
+        vcpkg_execute_build_process(
+            COMMAND "${INSTALL_NAME_TOOL}" -change "libicudata.${ICU_VERSION_MAJOR}.dylib"
+                                                "${ID_PREFIX}/libicudata.${ICU_VERSION_MAJOR}.dylib"
+                                                "libicuuc.${VERSION}.dylib"
+            WORKING_DIRECTORY "${CURRENT_BUILDTREES_DIR}/${CONFIG_TRIPLE}/lib"
+            LOGNAME "make-build-fix-rpath-${CONFIG_TRIPLE}"
+        )
+    endforeach()
+
+endif()
+
+vcpkg_install_make()
+
+file(REMOVE_RECURSE
+    "${CURRENT_PACKAGES_DIR}/share"
+    "${CURRENT_PACKAGES_DIR}/debug/share"
+    "${CURRENT_PACKAGES_DIR}/lib/icu"
+    "${CURRENT_PACKAGES_DIR}/debug/lib/icu"
+    "${CURRENT_PACKAGES_DIR}/debug/lib/icud")
+
+if(VCPKG_LIBRARY_LINKAGE STREQUAL "static")
+    # force U_STATIC_IMPLEMENTATION macro
+    foreach(HEADER utypes.h utf_old.h platform.h)
+        vcpkg_replace_string("${CURRENT_PACKAGES_DIR}/include/unicode/${HEADER}" "defined(U_STATIC_IMPLEMENTATION)" "1")
+    endforeach()
+endif()
+
+# Install executables from /tools/icu/sbin to /tools/icu/bin on unix (/bin because icu require this for cross compiling)
+if(VCPKG_TARGET_IS_LINUX OR VCPKG_TARGET_IS_OSX AND "tools" IN_LIST FEATURES)
+    vcpkg_copy_tools(
+        TOOL_NAMES icupkg gennorm2 gencmn genccode gensprep
+        SEARCH_DIR "${CURRENT_PACKAGES_DIR}/tools/icu/sbin"
+        DESTINATION "${CURRENT_PACKAGES_DIR}/tools/${PORT}/bin"
+    )
+endif()
+
+file(REMOVE_RECURSE
+    "${CURRENT_PACKAGES_DIR}/tools/icu/sbin"
+    "${CURRENT_PACKAGES_DIR}/tools/icu/debug")
+
+# To cross compile, we need some files at specific positions. So lets copy them
+file(GLOB CROSS_COMPILE_DEFS "${CURRENT_BUILDTREES_DIR}/${RELEASE_TRIPLET}/config/icucross.*")
+file(INSTALL ${CROSS_COMPILE_DEFS} DESTINATION "${CURRENT_PACKAGES_DIR}/tools/${PORT}/config")
+
+file(GLOB RELEASE_DLLS "${CURRENT_PACKAGES_DIR}/lib/*icu*${ICU_VERSION_MAJOR}.dll")
+file(COPY ${RELEASE_DLLS} DESTINATION "${CURRENT_PACKAGES_DIR}/tools/${PORT}/bin")
+
+# copy dlls
+file(GLOB RELEASE_DLLS "${CURRENT_PACKAGES_DIR}/lib/*icu*${ICU_VERSION_MAJOR}.dll")
+file(COPY ${RELEASE_DLLS} DESTINATION "${CURRENT_PACKAGES_DIR}/bin")
+if(NOT VCPKG_BUILD_TYPE)
+    file(GLOB DEBUG_DLLS "${CURRENT_PACKAGES_DIR}/debug/lib/*icu*${ICU_VERSION_MAJOR}.dll")
+    file(COPY ${DEBUG_DLLS} DESTINATION "${CURRENT_PACKAGES_DIR}/debug/bin")
+endif()
+
+# remove any remaining dlls in /lib
+file(GLOB DUMMY_DLLS "${CURRENT_PACKAGES_DIR}/lib/*.dll" "${CURRENT_PACKAGES_DIR}/debug/lib/*.dll")
+if(DUMMY_DLLS)
+    file(REMOVE ${DUMMY_DLLS})
+endif()
+
+vcpkg_copy_pdbs()
+vcpkg_fixup_pkgconfig()
+
+vcpkg_replace_string("${CURRENT_PACKAGES_DIR}/tools/icu/bin/icu-config" "${CURRENT_INSTALLED_DIR}" "`dirname $0`/../../../")
+
+file(INSTALL "${CMAKE_CURRENT_LIST_DIR}/vcpkg-cmake-wrapper.cmake" DESTINATION "${CURRENT_PACKAGES_DIR}/share/${PORT}")
+vcpkg_install_copyright(FILE_LIST "${SOURCE_PATH}/LICENSE")

dev/vcpkg/ports/icu/remove-MD-from-configure.patch

@@ -0,0 +1,18 @@
+diff -urN a/source/runConfigureICU b/source/runConfigureICU
+--- a/source/runConfigureICU	2018-03-26 21:38:44.000000000 +0800
++++ b/source/runConfigureICU	2018-08-26 09:04:53.197454400 +0800
+@@ -322,10 +322,10 @@
+         THE_COMP="Microsoft Visual C++"
+         CC=cl; export CC
+         CXX=cl; export CXX
+-        RELEASE_CFLAGS='-Gy -MD'
+-        RELEASE_CXXFLAGS='-Gy -MD'
+-        DEBUG_CFLAGS='-FS -Zi -MDd'
+-        DEBUG_CXXFLAGS='-FS -Zi -MDd'
++        RELEASE_CFLAGS='-Gy'
++        RELEASE_CXXFLAGS='-Gy'
++        DEBUG_CFLAGS='-FS -Zi'
++        DEBUG_CXXFLAGS='-FS -Zi'
+         DEBUG_LDFLAGS='-DEBUG'
+         ;;
+     *BSD)

dev/vcpkg/ports/icu/vcpkg-cmake-wrapper.cmake

@@ -0,0 +1,8 @@
+cmake_policy(PUSH)
+cmake_policy(SET CMP0057 NEW)
+if(NOT "COMPONENTS" IN_LIST ARGS)
+    _find_package(${ARGS} COMPONENTS data)
+else()
+    _find_package(${ARGS})
+endif()
+cmake_policy(POP)

dev/vcpkg/ports/icu/vcpkg.json

@@ -0,0 +1,22 @@
+{
+  "name": "icu",
+  "version": "73.1",
+  "description": "Mature and widely used Unicode and localization library.",
+  "homepage": "https://icu.unicode.org/home",
+  "license": "ICU",
+  "supports": "!uwp",
+  "dependencies": [
+    {
+      "name": "icu",
+      "host": true,
+      "features": [
+        "tools"
+      ]
+    }
+  ],
+  "features": {
+    "tools": {
+      "description": "Build tools"
+    }
+  }
+}