- Eliminated the default installation prefix condition to streamline the configuration.
- Simplified the test build logic by ensuring tests are enabled only for top-level projects.
- Cleaned up redundant code for better readability and maintainability.
- Added include directories for the current binary and test directories to improve test file accessibility.
- Ensured proper configuration for test paths in the build process.
- Removed unused Jieba test and associated includes from load_test.cpp.
- Simplified main function to focus on essential operations.
- Ensured consistent exit handling by returning EXIT_SUCCESS.
- Added configuration for test paths in CMake to simplify file references.
- Updated load_test.cpp and various unit tests to use defined path macros for dictionary and test data files.
- Introduced test_paths.h.in to manage directory paths consistently across tests.
- Added entries to .gitignore for temporary test files.
- Included a message to display MSVC value during build.
- Added UTF-8 compile option for MSVC in unittest CMakeLists.
This PR fixes the ambigious `partial_sort` in KeywordExtractor.hpp.
We also have a definition for it and the compiler is consufed which
implementation should be used. To fix it, we can use the `std` namespace
explicitly.
Also, use the `std` namespace for the other data structures and include
their headers.
This commit introduces a new GitHub Actions workflow for building and testing CMake projects on Windows ARM64. The workflow includes steps for checking out the repository, configuring CMake with multiple C++ standards, building the project, and running tests. This enhancement supports continuous integration for ARM64 architecture, improving the project's build versatility.
Updated multiple files to replace instances of DecodeRunesInString with DecodeUTF8RunesInString, ensuring proper handling of UTF-8 encoded strings. This change enhances the robustness of string decoding across the cppjieba library, including updates in DictTrie, HMMModel, PosTagger, PreFilter, SegmentBase, and Unicode files. Additionally, corresponding unit tests have been modified to reflect these changes.
