aboutsummaryrefslogtreecommitdiff
path: root/vcpkg/ports/sentencepiece
diff options
context:
space:
mode:
Diffstat (limited to 'vcpkg/ports/sentencepiece')
-rw-r--r--vcpkg/ports/sentencepiece/abseil.diff29
-rw-r--r--vcpkg/ports/sentencepiece/linkage.diff13
-rw-r--r--vcpkg/ports/sentencepiece/portfile.cmake36
-rw-r--r--vcpkg/ports/sentencepiece/vcpkg.json15
4 files changed, 93 insertions, 0 deletions
diff --git a/vcpkg/ports/sentencepiece/abseil.diff b/vcpkg/ports/sentencepiece/abseil.diff
new file mode 100644
index 0000000..781afae
--- /dev/null
+++ b/vcpkg/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 ${CMAKE_CURRENT_SOURCE_DIR}/third_party/absl)
+diff --git a/src/sentencepiece_processor.h b/src/sentencepiece_processor.h
+index dd3f092..c1551a7 100644
+--- a/src/sentencepiece_processor.h
++++ b/src/sentencepiece_processor.h
+@@ -23,9 +23,7 @@
+ #include <vector>
+
+ #ifndef SWIG
+-namespace absl {
+-using std::string_view;
+-} // namespace absl
++#include <absl/strings/string_view.h>
+ #endif // SWIG
+
+ namespace sentencepiece {
diff --git a/vcpkg/ports/sentencepiece/linkage.diff b/vcpkg/ports/sentencepiece/linkage.diff
new file mode 100644
index 0000000..60c81f2
--- /dev/null
+++ b/vcpkg/ports/sentencepiece/linkage.diff
@@ -0,0 +1,13 @@
+diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
+index 8d4a34f..6ef8aa9 100644
+--- a/src/CMakeLists.txt
++++ b/src/CMakeLists.txt
+@@ -250,6 +250,8 @@ if (SPM_ENABLE_SHARED)
+ set_target_properties(sentencepiece PROPERTIES IMPORT_SUFFIX ".dll.a")
+ set_target_properties(sentencepiece_train PROPERTIES IMPORT_SUFFIX ".dll.a")
+ endif()
++ list(REMOVE_ITEM SPM_INSTALLTARGETS sentencepiece-static sentencepiece_train-static)
++ set_target_properties(sentencepiece-static sentencepiece_train-static PROPERTIES EXCLUDE_FROM_ALL 1)
+ else()
+ add_library(sentencepiece ALIAS sentencepiece-static)
+ add_library(sentencepiece_train ALIAS sentencepiece_train-static)
diff --git a/vcpkg/ports/sentencepiece/portfile.cmake b/vcpkg/ports/sentencepiece/portfile.cmake
new file mode 100644
index 0000000..bde53b3
--- /dev/null
+++ b/vcpkg/ports/sentencepiece/portfile.cmake
@@ -0,0 +1,36 @@
+if(VCPKG_TARGET_IS_WINDOWS)
+ vcpkg_check_linkage(ONLY_STATIC_LIBRARY)
+endif()
+
+vcpkg_from_github(
+ OUT_SOURCE_PATH SOURCE_PATH
+ REPO google/sentencepiece
+ REF "v${VERSION}"
+ SHA512 012850b63b2323e16acc5dacc0a494ad3f6375425ee86274f0946032e47c088a3b307758b99d752fcf54acf76c82d7d13d0c14bbf07aa9b612c4f1fbd30cf1cf
+ HEAD_REF master
+ PATCHES
+ abseil.diff
+ linkage.diff
+)
+
+string(COMPARE EQUAL "${VCPKG_LIBRARY_LINKAGE}" "dynamic" SPM_ENABLE_SHARED)
+
+vcpkg_cmake_configure(
+ SOURCE_PATH "${SOURCE_PATH}"
+ OPTIONS
+ -DSPM_ENABLE_SHARED=${SPM_ENABLE_SHARED}
+ -DSPM_ENABLE_TCMALLOC=OFF
+ -DSPM_ABSL_PROVIDER=package
+ -DSPM_PROTOBUF_PROVIDER=package
+ -DPROTOBUF_LITE_LIBRARY=protobuf::libprotobuf-lite
+)
+
+vcpkg_cmake_install()
+vcpkg_copy_pdbs()
+vcpkg_fixup_pkgconfig()
+
+vcpkg_copy_tools(TOOL_NAMES spm_decode spm_encode spm_export_vocab spm_normalize spm_train AUTO_CLEAN)
+
+file(REMOVE_RECURSE "${CURRENT_PACKAGES_DIR}/debug/include")
+
+vcpkg_install_copyright(FILE_LIST "${SOURCE_PATH}/LICENSE")
diff --git a/vcpkg/ports/sentencepiece/vcpkg.json b/vcpkg/ports/sentencepiece/vcpkg.json
new file mode 100644
index 0000000..0750a8c
--- /dev/null
+++ b/vcpkg/ports/sentencepiece/vcpkg.json
@@ -0,0 +1,15 @@
+{
+ "name": "sentencepiece",
+ "version": "0.2.1",
+ "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",
+ "dependencies": [
+ "abseil",
+ "protobuf",
+ {
+ "name": "vcpkg-cmake",
+ "host": true
+ }
+ ]
+}