[vcpkg baseline][sentencepiece] Fix flags, linkage, deps (#42226)

ports/sentencepiece/abseil.diff

@@ -0,0 +1,29 @@
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index 9fa37fa..0e220e6 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -173,6 +173,9 @@ elseif (SPM_ABSL_PROVIDER STREQUAL "package")
+   find_package(absl REQUIRED)
+   get_target_property(ABSL_INCLUDE_DIRS absl::base INTERFACE_INCLUDE_DIRECTORIES)
+   if (NOT EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/third_party/absl.org)
++    file(REMOVE_RECURSE ${CMAKE_CURRENT_SOURCE_DIR}/third_party/absl)
++    file(COPY "${ABSL_INCLUDE_DIRS}/absl" DESTINATION "${PROJECT_BINARY_DIR}/third_party")
++  elseif(0)
+     file(RENAME ${CMAKE_CURRENT_SOURCE_DIR}/third_party/absl ${CMAKE_CURRENT_SOURCE_DIR}/third_party/absl.org)
+     execute_process(COMMAND ${CMAKE_COMMAND} -E create_symlink
+         ${ABSL_INCLUDE_DIRS}/absl ${CMAKE_CURRENT_SOURCE_DIR}/third_party/absl)
+diff --git a/src/sentencepiece_processor.h b/src/sentencepiece_processor.h
+index dd3f092..c1551a7 100644
+--- a/src/sentencepiece_processor.h
++++ b/src/sentencepiece_processor.h
+@@ -23,9 +23,7 @@
+ #include <vector>
+ 
+ #ifndef SWIG
+-namespace absl {
+-using std::string_view;
+-}  // namespace absl
++#include <absl/strings/string_view.h>
+ #endif  // SWIG
+ 
+ namespace sentencepiece {

ports/sentencepiece/cxxflags.diff

@@ -0,0 +1,12 @@
+diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
+index d983089..19736d8 100644
+--- a/src/CMakeLists.txt
++++ b/src/CMakeLists.txt
+@@ -264,7 +264,6 @@ if (NOT MSVC)
+   if (SPM_COVERAGE)
+     set(CMAKE_CXX_FLAGS "-O0 -Wall -fPIC -coverage ${CMAKE_CXX_FLAGS}")
+   else()
+-    set(CMAKE_CXX_FLAGS "-O3 -Wall -fPIC ${CMAKE_CXX_FLAGS}")
+   endif()
+   if (SPM_ENABLE_TENSORFLOW_SHARED)
+     add_definitions(-D_GLIBCXX_USE_CXX11_ABI=0)

ports/sentencepiece/linkage.diff

@@ -0,0 +1,13 @@
+diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
+index 8d4a34f..6ef8aa9 100644
+--- a/src/CMakeLists.txt
++++ b/src/CMakeLists.txt
+@@ -250,6 +250,8 @@ if (SPM_ENABLE_SHARED)
+     set_target_properties(sentencepiece PROPERTIES IMPORT_SUFFIX ".dll.a")
+     set_target_properties(sentencepiece_train PROPERTIES IMPORT_SUFFIX ".dll.a")
+   endif()
++  list(REMOVE_ITEM SPM_INSTALLTARGETS sentencepiece-static sentencepiece_train-static)
++  set_target_properties(sentencepiece-static sentencepiece_train-static PROPERTIES EXCLUDE_FROM_ALL 1)
+ else()
+   add_library(sentencepiece ALIAS sentencepiece-static)
+   add_library(sentencepiece_train ALIAS sentencepiece_train-static)

ports/sentencepiece/portfile.cmake

@@ -1,5 +1,5 @@
-if(NOT VCPKG_CMAKE_SYSTEM_NAME)
-    vcpkg_check_linkage(ONLY_STATIC_LIBRARY ONLY_STATIC_CRT)
+if(VCPKG_TARGET_IS_WINDOWS)
+    vcpkg_check_linkage(ONLY_STATIC_LIBRARY)
 endif()
 
 vcpkg_from_github(
@@ -8,24 +8,30 @@ vcpkg_from_github(
     REF "v${VERSION}"
     SHA512 b4214f5bfbe2a0757794c792e87e7c53fda7e65b2511b37fc757f280bf9287ba59b5d630801e17de6058f8292a3c6433211917324cb3446a212a51735402e614
     HEAD_REF master
+    PATCHES
+        abseil.diff
+        cxxflags.diff
+        linkage.diff
+        protobuf.diff
 )
 
+string(COMPARE EQUAL "${VCPKG_LIBRARY_LINKAGE}" "dynamic" SPM_ENABLE_SHARED)
+
 vcpkg_cmake_configure(
     SOURCE_PATH "${SOURCE_PATH}"
     OPTIONS
-        -DSPM_ENABLE_SHARED=OFF
-        -DSPM_USE_BUILTIN_PROTOBUF=ON
-        -DSPM_USE_EXTERNAL_ABSL=OFF
+        -DSPM_ENABLE_SHARED=${SPM_ENABLE_SHARED}
+        -DSPM_ENABLE_TCMALLOC=OFF
+        -DSPM_ABSL_PROVIDER=package
+        -DSPM_PROTOBUF_PROVIDER=package
 )
 
 vcpkg_cmake_install()
+vcpkg_copy_pdbs()
+vcpkg_fixup_pkgconfig()
+
+vcpkg_copy_tools(TOOL_NAMES spm_decode spm_encode spm_export_vocab spm_normalize spm_train AUTO_CLEAN)
 
 file(REMOVE_RECURSE "${CURRENT_PACKAGES_DIR}/debug/include")
 
 vcpkg_install_copyright(FILE_LIST "${SOURCE_PATH}/LICENSE")
-
-vcpkg_copy_tools(TOOL_NAMES spm_decode spm_encode spm_export_vocab spm_normalize spm_train AUTO_CLEAN)
-
-vcpkg_copy_pdbs()
-
-vcpkg_fixup_pkgconfig()

ports/sentencepiece/protobuf.diff

@@ -0,0 +1,19 @@
+diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
+index 6ef8aa9..d983089 100644
+--- a/src/CMakeLists.txt
++++ b/src/CMakeLists.txt
+@@ -70,12 +70,11 @@ if (SPM_PROTOBUF_PROVIDER STREQUAL "internal")
+   include_directories(${CMAKE_CURRENT_SOURCE_DIR}/../third_party/protobuf-lite)
+   include_directories(builtin_pb)
+ elseif (SPM_PROTOBUF_PROVIDER STREQUAL "package")
+-  find_package(Protobuf REQUIRED)
+-  include_directories(${Protobuf_INCLUDE_DIRS})
++  find_package(protobuf CONFIG REQUIRED)
++  set(PROTOBUF_LITE_LIBRARY protobuf::libprotobuf-lite)
+   protobuf_generate_cpp(SPM_PROTO_SRCS SPM_PROTO_HDRS sentencepiece.proto)
+   protobuf_generate_cpp(SPM_MODEL_PROTO_SRCS SPM_MODEL_PROTO_HDRS sentencepiece_model.proto)
+   set(PROTOBUF_LITE_SRCS "")
+-  include_directories(${PROTOBUF_INCLUDE_DIR})
+   if (MSVC)
+     add_definitions("/D_USE_EXTERNAL_PROTOBUF")
+   else()

ports/sentencepiece/vcpkg.json

@@ -1,10 +1,13 @@
 {
   "name": "sentencepiece",
   "version": "0.2.0",
+  "port-version": 1,
   "description": "SentencePiece is an unsupervised text tokenizer and detokenizer mainly for Neural Network-based text generation systems where the vocabulary size is predetermined prior to the neural model training",
+  "homepage": "https://github.com/google/sentencepiece",
   "license": "Apache-2.0",
-  "supports": "static",
   "dependencies": [
+    "abseil",
+    "protobuf",
     {
       "name": "vcpkg-cmake",
       "host": true

versions/baseline.json

@@ -8266,7 +8266,7 @@
     },
     "sentencepiece": {
       "baseline": "0.2.0",
-      "port-version": 0
+      "port-version": 1
     },
     "sentry-native": {
       "baseline": "0.7.12",

versions/s-/sentencepiece.json

@@ -1,5 +1,10 @@
 {
   "versions": [
+    {
+      "git-tree": "626e5f5d9e302b9095128148ff4ccd046486c13e",
+      "version": "0.2.0",
+      "port-version": 1
+    },
     {
       "git-tree": "60678651110bfbd252017ee568f2cb6484aaaf46",
       "version": "0.2.0",