mmdeploy/csrc/model/zip_model_impl.cpp

152 lines
4.3 KiB
C++
Raw Normal View History

Merge sdk (#251) * check in cmake * move backend_ops to csrc/backend_ops * check in preprocess, model, some codebase and their c-apis * check in CMakeLists.txt * check in parts of test_csrc * commit everything else * add readme * update core's BUILD_INTERFACE directory * skip codespell on third_party * update trt_net and ort_net's CMakeLists * ignore clion's build directory * check in pybind11 * add onnx.proto. Remove MMDeploy's dependency on ncnn's source code * export MMDeployTargets only when MMDEPLOY_BUILD_SDK is ON * remove useless message * target include directory is wrong * change target name from mmdeploy_ppl_net to mmdeploy_pplnn_net * skip install directory * update project's cmake * remove useless code * set CMAKE_BUILD_TYPE to Release by force if it isn't set by user * update custom ops CMakeLists * pass object target's source lists * fix lint end-of-file * fix lint: trailing whitespace * fix codespell hook * remove bicubic_interpolate to csrc/backend_ops/ * set MMDEPLOY_BUILD_SDK OFF * change custom ops build command * add spdlog installation command * update docs on how to checkout pybind11 * move bicubic_interpolate to backend_ops/tensorrt directory * remove useless code * correct cmake * fix typo * fix typo * fix install directory * correct sdk's readme * set cub dir when cuda version < 11.0 * change directory where clang-format will apply to * fix build command * add .clang-format * change clang-format style from google to file * reformat csrc/backend_ops * format sdk's code * turn off clang-format for some files * add -Xcompiler=-fno-gnu-unique * fix trt topk initialize * check in config for sdk demo * update cmake script and csrc's readme * correct config's path * add cuda include directory, otherwise compile failed in case of tensorrt8.2 * clang-format onnx2ncnn.cpp Co-authored-by: zhangli <lzhang329@gmail.com> Co-authored-by: grimoire <yaoqian@sensetime.com>
2021-12-07 10:57:55 +08:00
// Copyright (c) OpenMMLab. All rights reserved.
#include <fstream>
#include <map>
#include "archive/json_archive.h"
#include "core/logger.h"
#include "core/model.h"
#include "core/model_impl.h"
#include "zip.h"
#if __GNUC__ >= 8
#include <filesystem>
namespace fs = std::filesystem;
#else
#include <experimental/filesystem>
namespace fs = std::experimental::filesystem;
#endif
Merge sdk (#251) * check in cmake * move backend_ops to csrc/backend_ops * check in preprocess, model, some codebase and their c-apis * check in CMakeLists.txt * check in parts of test_csrc * commit everything else * add readme * update core's BUILD_INTERFACE directory * skip codespell on third_party * update trt_net and ort_net's CMakeLists * ignore clion's build directory * check in pybind11 * add onnx.proto. Remove MMDeploy's dependency on ncnn's source code * export MMDeployTargets only when MMDEPLOY_BUILD_SDK is ON * remove useless message * target include directory is wrong * change target name from mmdeploy_ppl_net to mmdeploy_pplnn_net * skip install directory * update project's cmake * remove useless code * set CMAKE_BUILD_TYPE to Release by force if it isn't set by user * update custom ops CMakeLists * pass object target's source lists * fix lint end-of-file * fix lint: trailing whitespace * fix codespell hook * remove bicubic_interpolate to csrc/backend_ops/ * set MMDEPLOY_BUILD_SDK OFF * change custom ops build command * add spdlog installation command * update docs on how to checkout pybind11 * move bicubic_interpolate to backend_ops/tensorrt directory * remove useless code * correct cmake * fix typo * fix typo * fix install directory * correct sdk's readme * set cub dir when cuda version < 11.0 * change directory where clang-format will apply to * fix build command * add .clang-format * change clang-format style from google to file * reformat csrc/backend_ops * format sdk's code * turn off clang-format for some files * add -Xcompiler=-fno-gnu-unique * fix trt topk initialize * check in config for sdk demo * update cmake script and csrc's readme * correct config's path * add cuda include directory, otherwise compile failed in case of tensorrt8.2 * clang-format onnx2ncnn.cpp Co-authored-by: zhangli <lzhang329@gmail.com> Co-authored-by: grimoire <yaoqian@sensetime.com>
2021-12-07 10:57:55 +08:00
using nlohmann::json;
namespace mmdeploy {
class ZipModelImpl : public ModelImpl {
public:
~ZipModelImpl() override {
if (zip_ != nullptr) {
zip_close(zip_);
}
#if LIBZIP_VERSION_MAJOR >= 1
if (source_) {
zip_source_close(source_);
}
#endif
}
// @brief load an sdk model, which HAS TO BE a zip file.
// Meta file (i.e. deploy.json) will be extracted and parsed from the zip file
// @param sdk_model_path path of sdk model file, in zip format
Result<void> Init(const std::string& model_path) override {
int ret = 0;
zip_ = zip_open(model_path.c_str(), 0, &ret);
if (ret != 0) {
INFO("open zip file {} failed, ret {}", model_path.c_str(), ret);
return Status(eInvalidArgument);
}
INFO("open sdk model file {} successfully", model_path.c_str());
return InitZip();
}
Result<void> Init(const void* buffer, size_t size) override {
#if LIBZIP_VERSION_MAJOR >= 1
zip_error_t error{};
source_ = zip_source_buffer_create(buffer, size, 0, &error);
if (zip_error_code_zip(&error) != ZIP_ER_OK) {
return Status(eFail);
}
zip_ = zip_open_from_source(source_, ZIP_RDONLY, &error);
if (zip_error_code_zip(&error) != ZIP_ER_OK) {
return Status(eFail);
}
return InitZip();
#else
return Status(eNotSupported);
#endif
}
Result<std::string> ReadFile(const std::string& file_path) const override {
int ret = 0;
int index = -1;
Merge sdk (#251) * check in cmake * move backend_ops to csrc/backend_ops * check in preprocess, model, some codebase and their c-apis * check in CMakeLists.txt * check in parts of test_csrc * commit everything else * add readme * update core's BUILD_INTERFACE directory * skip codespell on third_party * update trt_net and ort_net's CMakeLists * ignore clion's build directory * check in pybind11 * add onnx.proto. Remove MMDeploy's dependency on ncnn's source code * export MMDeployTargets only when MMDEPLOY_BUILD_SDK is ON * remove useless message * target include directory is wrong * change target name from mmdeploy_ppl_net to mmdeploy_pplnn_net * skip install directory * update project's cmake * remove useless code * set CMAKE_BUILD_TYPE to Release by force if it isn't set by user * update custom ops CMakeLists * pass object target's source lists * fix lint end-of-file * fix lint: trailing whitespace * fix codespell hook * remove bicubic_interpolate to csrc/backend_ops/ * set MMDEPLOY_BUILD_SDK OFF * change custom ops build command * add spdlog installation command * update docs on how to checkout pybind11 * move bicubic_interpolate to backend_ops/tensorrt directory * remove useless code * correct cmake * fix typo * fix typo * fix install directory * correct sdk's readme * set cub dir when cuda version < 11.0 * change directory where clang-format will apply to * fix build command * add .clang-format * change clang-format style from google to file * reformat csrc/backend_ops * format sdk's code * turn off clang-format for some files * add -Xcompiler=-fno-gnu-unique * fix trt topk initialize * check in config for sdk demo * update cmake script and csrc's readme * correct config's path * add cuda include directory, otherwise compile failed in case of tensorrt8.2 * clang-format onnx2ncnn.cpp Co-authored-by: zhangli <lzhang329@gmail.com> Co-authored-by: grimoire <yaoqian@sensetime.com>
2021-12-07 10:57:55 +08:00
auto iter = file_index_.find(file_path);
if (iter == file_index_.end()) {
ERROR("cannot find file {} under dir {}", file_path.c_str(), root_dir_.c_str());
return Status(eFail);
}
index = iter->second;
struct zip_file* pzip = zip_fopen_index(zip_, index, 0);
if (nullptr == pzip) {
ERROR("read file {} in zip file failed, whose index is {}", file_path.c_str(), index);
return Status(eFail);
}
struct zip_stat stat {};
if ((ret = zip_stat_index(zip_, index, 0, &stat)) < 0) {
ERROR("get stat of file {} error, ret {}", file_path.c_str(), ret);
return Status(eFail);
}
DEBUG("file size {}", (int)stat.size);
std::vector<char> buf(stat.size);
if ((ret = zip_fread(pzip, buf.data(), stat.size)) < 0) {
ERROR("read data of file {} error, ret {}", file_path.c_str(), ret);
return Status(eFail);
}
return std::string(buf.begin(), buf.end());
}
Result<deploy_meta_info_t> ReadMeta() const override {
OUTCOME_TRY(auto deploy_json, ReadFile("deploy.json"));
try {
deploy_meta_info_t meta;
from_json(json::parse(deploy_json), meta);
return meta;
} catch (std::exception& e) {
ERROR("exception happened: {}", e.what());
return Status(eFail);
}
}
private:
Result<void> InitZip() {
int files = zip_get_num_files(zip_);
INFO("there are {} files in sdk model file", files);
if (files == 0) {
return Status(eFail);
}
Merge sdk (#251) * check in cmake * move backend_ops to csrc/backend_ops * check in preprocess, model, some codebase and their c-apis * check in CMakeLists.txt * check in parts of test_csrc * commit everything else * add readme * update core's BUILD_INTERFACE directory * skip codespell on third_party * update trt_net and ort_net's CMakeLists * ignore clion's build directory * check in pybind11 * add onnx.proto. Remove MMDeploy's dependency on ncnn's source code * export MMDeployTargets only when MMDEPLOY_BUILD_SDK is ON * remove useless message * target include directory is wrong * change target name from mmdeploy_ppl_net to mmdeploy_pplnn_net * skip install directory * update project's cmake * remove useless code * set CMAKE_BUILD_TYPE to Release by force if it isn't set by user * update custom ops CMakeLists * pass object target's source lists * fix lint end-of-file * fix lint: trailing whitespace * fix codespell hook * remove bicubic_interpolate to csrc/backend_ops/ * set MMDEPLOY_BUILD_SDK OFF * change custom ops build command * add spdlog installation command * update docs on how to checkout pybind11 * move bicubic_interpolate to backend_ops/tensorrt directory * remove useless code * correct cmake * fix typo * fix typo * fix install directory * correct sdk's readme * set cub dir when cuda version < 11.0 * change directory where clang-format will apply to * fix build command * add .clang-format * change clang-format style from google to file * reformat csrc/backend_ops * format sdk's code * turn off clang-format for some files * add -Xcompiler=-fno-gnu-unique * fix trt topk initialize * check in config for sdk demo * update cmake script and csrc's readme * correct config's path * add cuda include directory, otherwise compile failed in case of tensorrt8.2 * clang-format onnx2ncnn.cpp Co-authored-by: zhangli <lzhang329@gmail.com> Co-authored-by: grimoire <yaoqian@sensetime.com>
2021-12-07 10:57:55 +08:00
for (int i = 0; i < files; ++i) {
struct zip_stat stat;
zip_stat_init(&stat);
zip_stat_index(zip_, i, 0, &stat);
fs::path path(stat.name);
auto file_name = path.filename().string();
if (file_name == ".") {
Merge sdk (#251) * check in cmake * move backend_ops to csrc/backend_ops * check in preprocess, model, some codebase and their c-apis * check in CMakeLists.txt * check in parts of test_csrc * commit everything else * add readme * update core's BUILD_INTERFACE directory * skip codespell on third_party * update trt_net and ort_net's CMakeLists * ignore clion's build directory * check in pybind11 * add onnx.proto. Remove MMDeploy's dependency on ncnn's source code * export MMDeployTargets only when MMDEPLOY_BUILD_SDK is ON * remove useless message * target include directory is wrong * change target name from mmdeploy_ppl_net to mmdeploy_pplnn_net * skip install directory * update project's cmake * remove useless code * set CMAKE_BUILD_TYPE to Release by force if it isn't set by user * update custom ops CMakeLists * pass object target's source lists * fix lint end-of-file * fix lint: trailing whitespace * fix codespell hook * remove bicubic_interpolate to csrc/backend_ops/ * set MMDEPLOY_BUILD_SDK OFF * change custom ops build command * add spdlog installation command * update docs on how to checkout pybind11 * move bicubic_interpolate to backend_ops/tensorrt directory * remove useless code * correct cmake * fix typo * fix typo * fix install directory * correct sdk's readme * set cub dir when cuda version < 11.0 * change directory where clang-format will apply to * fix build command * add .clang-format * change clang-format style from google to file * reformat csrc/backend_ops * format sdk's code * turn off clang-format for some files * add -Xcompiler=-fno-gnu-unique * fix trt topk initialize * check in config for sdk demo * update cmake script and csrc's readme * correct config's path * add cuda include directory, otherwise compile failed in case of tensorrt8.2 * clang-format onnx2ncnn.cpp Co-authored-by: zhangli <lzhang329@gmail.com> Co-authored-by: grimoire <yaoqian@sensetime.com>
2021-12-07 10:57:55 +08:00
DEBUG("{}-th file name is: {} which is a directory", i, stat.name);
} else {
DEBUG("{}-th file name is: {} which is a file", i, stat.name);
file_index_[file_name] = i;
Merge sdk (#251) * check in cmake * move backend_ops to csrc/backend_ops * check in preprocess, model, some codebase and their c-apis * check in CMakeLists.txt * check in parts of test_csrc * commit everything else * add readme * update core's BUILD_INTERFACE directory * skip codespell on third_party * update trt_net and ort_net's CMakeLists * ignore clion's build directory * check in pybind11 * add onnx.proto. Remove MMDeploy's dependency on ncnn's source code * export MMDeployTargets only when MMDEPLOY_BUILD_SDK is ON * remove useless message * target include directory is wrong * change target name from mmdeploy_ppl_net to mmdeploy_pplnn_net * skip install directory * update project's cmake * remove useless code * set CMAKE_BUILD_TYPE to Release by force if it isn't set by user * update custom ops CMakeLists * pass object target's source lists * fix lint end-of-file * fix lint: trailing whitespace * fix codespell hook * remove bicubic_interpolate to csrc/backend_ops/ * set MMDEPLOY_BUILD_SDK OFF * change custom ops build command * add spdlog installation command * update docs on how to checkout pybind11 * move bicubic_interpolate to backend_ops/tensorrt directory * remove useless code * correct cmake * fix typo * fix typo * fix install directory * correct sdk's readme * set cub dir when cuda version < 11.0 * change directory where clang-format will apply to * fix build command * add .clang-format * change clang-format style from google to file * reformat csrc/backend_ops * format sdk's code * turn off clang-format for some files * add -Xcompiler=-fno-gnu-unique * fix trt topk initialize * check in config for sdk demo * update cmake script and csrc's readme * correct config's path * add cuda include directory, otherwise compile failed in case of tensorrt8.2 * clang-format onnx2ncnn.cpp Co-authored-by: zhangli <lzhang329@gmail.com> Co-authored-by: grimoire <yaoqian@sensetime.com>
2021-12-07 10:57:55 +08:00
}
}
return success();
}
#if LIBZIP_VERSION_MAJOR >= 1
struct zip_source* source_{};
#endif
struct zip* zip_{};
// root directory in zip file
std::string root_dir_;
// a map between file path and its index in zip file
std::map<std::string, int> file_index_;
};
class ZipModelImplRegister {
public:
ZipModelImplRegister() {
(void)ModelRegistry::Get().Register("ZipModel", []() -> std::unique_ptr<ModelImpl> {
return std::make_unique<ZipModelImpl>();
});
}
};
static ZipModelImplRegister folder_model_register;
} // namespace mmdeploy