chore: prepare the repository for v0.3.0

.asf.yaml

@@ -32,7 +32,7 @@ github:
   enabled_merge_buttons:
     merge: false
     squash: true
-    rebase: true
+    rebase: false
   rulesets:
     - name: "Branch Protection"
       type: branch
@@ -52,17 +52,19 @@ github:
       required_status_checks:
         - name: "pre-commit"
           app_slug: -1
-        - name: "gcc-test"
-          app_slug: -1
         - name: "rat-license-check"
           app_slug: -1
-        - name: "test-with-sanitizer"
+        - name: "asan-ubsan"
+          app_slug: -1
+        - name: "tsan"
+          app_slug: -1
+        - name: "clang-debug"
           app_slug: -1
         - name: "clang-release"
           app_slug: -1
-        - name: "gcc-release"
+        - name: "gcc-debug"
           app_slug: -1
-        - name: "clang-test"
+        - name: "gcc-release"
           app_slug: -1
         - name: "gcc8-test"
           app_slug: -1

.github/workflows/build_and_test.yaml

@@ -51,10 +51,10 @@ jobs:
             cc: gcc-14
             cxx: g++-14
           - name: clang-debug
-            # fetch_depth: '0' # fetch the PR target branch history for clang-tidy
-            # build_args: >-
-            #   --check_clang_tidy
-            #   --lint_git_target_commit "origin/${{ github.base_ref || github.event.repository.default_branch }}"
+            fetch_depth: '0' # fetch the PR target branch history for clang-tidy
+            build_args: >-
+              --check_clang_tidy
+              --lint_git_target_commit "origin/${{ github.base_ref || github.event.repository.default_branch }}"
           - name: asan-ubsan
             build_args: --enable_asan --enable_ubsan
           - name: tsan

.github/workflows/release_candidate.yaml

@@ -122,6 +122,15 @@ jobs:
           name: source-archive
           path: release/ci
 
+      - name: Install Rust toolchain (tantivy-fts)
+        shell: bash
+        run: ci/scripts/setup_rust.sh
+
+      - name: Install HTTP and TLS development dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y libcurl4-openssl-dev libssl-dev
+
       - name: Verify, build, test, and install
         shell: bash
         env:

CMakeLists.txt

@@ -64,6 +64,7 @@ option(PAIMON_ENABLE_JINDO "Whether to enable jindo file system" OFF)
 option(PAIMON_ENABLE_S3 "Whether to enable S3 file system" OFF)
 option(PAIMON_ENABLE_NETWORK_TESTS
        "Whether to enable tests that access real remote services over the network" OFF)
+option(PAIMON_ENABLE_LUMINA "Whether to enable lumina vector index" OFF)
 option(PAIMON_ENABLE_LUCENE "Whether to enable lucene index" OFF)
 option(PAIMON_ENABLE_TANTIVY
        "Whether to enable tantivy-fulltext global index (Rust FFI, experimental)" OFF)
@@ -485,7 +486,7 @@ include(CMakePackageConfigHelpers)
 write_basic_package_version_file(
     "${CMAKE_CURRENT_BINARY_DIR}/PaimonConfigVersion.cmake"
     VERSION ${PROJECT_VERSION}
-    COMPATIBILITY AnyNewerVersion)
+    COMPATIBILITY SameMinorVersion)
 
 configure_package_config_file("${CMAKE_CURRENT_SOURCE_DIR}/PaimonConfig.cmake.in"
                               "${CMAKE_CURRENT_BINARY_DIR}/PaimonConfig.cmake"

CONTRIBUTING.md

@@ -80,7 +80,7 @@ Before submitting, please verify:
 ### Prerequisites
 
 - **C++17** compatible compiler (GCC recommended)
-- **CMake** ≥ 3.16
+- **CMake** ≥ 3.22
 - **Python 3** (for linting scripts and pre-commit)
 
 ---

LICENSE

@@ -429,16 +429,6 @@ License: https://www.apache.org/licenses/LICENSE-2.0
 
 --------------------------------------------------------------------------------
 
-This product includes code from LucenePlusPlus.
-
-* LucenePlusPlus utility in src/paimon/global_index/lucene/ directory
-
-Copyright: 2009-2014 Alan Wright.
-Home page: https://github.com/luceneplusplus/LucenePlusPlus
-License: https://www.apache.org/licenses/LICENSE-2.0
-
---------------------------------------------------------------------------------
-
 This product includes code derived from PyTorch TH simd.h.
 
 * SIMD detection code in third_party/roaring_bitmap/roaring.cpp

NOTICE

@@ -50,10 +50,5 @@ PalDB
 
 --------------------------------------------------------------------------------
 
-This product includes software from LucenePlusPlus project (Apache 2.0)
-Copyright 2009-2014 Alan Wright.
-
---------------------------------------------------------------------------------
-
 JindoSDK NextArch C++
 Copyright 2024-present Alibaba Cloud.

README.md

@@ -26,10 +26,6 @@ It provides native, high-performance, and extensible access to the Paimon lake f
 
 Background and documentation are available at [paimon.apache.org](https://paimon.apache.org).
 
-## Status
-
-Paimon-cpp is currently undergoing repository migration. The original repository is hosted at [github.com/alibaba/paimon-cpp](https://github.com/alibaba/paimon-cpp/), and the codebase is being migrated incrementally to the Apache Paimon community repository.
-
 ## Features
 
 Paimon-cpp currently provides:
@@ -53,7 +49,7 @@ Paimon-cpp currently provides:
 > Paimon C++ therefore currently treats the `bitmap` global index type as unsupported. The legacy
 > implementation remains in the codebase pending migration to the Java-compatible format.
 
-Note: Linux x86_64 and macOS arm64 builds are currently verified.
+Note: Only Linux x86_64 builds are currently supported and verified.
 
 ## Building
 

docs/source/build_system.rst

@@ -44,7 +44,7 @@ for data format and file system.
 
 .. code-block:: cmake
 
-   cmake_minimum_required(VERSION 3.16)
+   cmake_minimum_required(VERSION 3.22)
 
    project(MyExample)
 

docs/source/building.rst

@@ -32,9 +32,8 @@ out-of-source. For example, you could create ``paimon-cpp/build`` and invoke
 
 Building requires:
 
-* A C++17-enabled compiler. On Linux, gcc 8 and higher should be
-  sufficient. On macOS, use AppleClang from Xcode Command Line Tools or
-  LLVM clang from Homebrew. Windows is not supported for now.
+* A C++17-enabled compiler. On Linux, gcc 8 and higher should be sufficient.
+  macOS and Windows are not supported for now.
 * At least 2GB of RAM for a minimal build, 8GB for a minimal
   debug build with tests and 16GB for a full build.
 
@@ -46,29 +45,6 @@ On Ubuntu/Debian you can install the requirements with:
         build-essential \
         cmake
 
-On macOS you can install the requirements with:
-
-.. code-block:: shell
-
-   xcode-select --install
-   brew install cmake
-
-The same CMake build options apply on Linux and macOS. If you prefer upstream
-LLVM clang instead of AppleClang on macOS, install LLVM and pass the Homebrew
-compiler paths when configuring:
-
-.. code-block:: shell
-
-   brew install llvm
-   cmake -B build \
-       -DCMAKE_C_COMPILER="$(brew --prefix llvm)/bin/clang" \
-       -DCMAKE_CXX_COMPILER="$(brew --prefix llvm)/bin/clang++"
-
-When building with upstream Clang on macOS, Paimon uses Apple SDK libc++
-headers by default to avoid incompatibilities in bundled third-party
-dependencies. Pass ``-DPAIMON_USE_APPLE_LIBCXX_WITH_CLANG=OFF`` to disable
-this behavior.
-
 We also provide a docker template to help you get started quickly. See in
 ``.devcontainer`` folder for more details.
 
@@ -149,6 +125,7 @@ boolean flags to ``cmake``.
 * ``-DPAIMON_ENABLE_JINDO=ON``: Support for Alibaba Jindo filesystems
 * ``-DPAIMON_ENABLE_LUMINA=ON``: Support for the Lumina vector index.
 * ``-DPAIMON_ENABLE_LUCENE=ON``: Support for Lucene full-text search indexes
+* ``-DPAIMON_ENABLE_TANTIVY=ON``: Enable the experimental Tantivy full-text index Rust FFI.
 
 Third-party dependency source
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -241,7 +218,7 @@ several times with different options if you want to exercise all of them.
 CMake version requirements
 ~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-We support CMake 3.16 and higher.
+We support CMake 3.22 and higher.
 
 LLVM and Clang Tools
 ~~~~~~~~~~~~~~~~~~~~

docs/source/conf.py

@@ -32,8 +32,8 @@
 import os
 import datetime
 
-project = u'C++ Paimon'
-copyright = f'2024-{datetime.datetime.now().year} Apache Software Foundation'
+project = u'Apache Paimon C++'
+copyright = f'2026-{datetime.datetime.now().year} Apache Software Foundation'
 author = u'Apache Software Foundation'
 
 # -- General configuration ---------------------------------------------------
@@ -131,7 +131,7 @@
         },
     ],
     "logo": {
-        "text": "Paimon C++",
+        "text": "Apache Paimon C++",
     },
     "show_version_warning_banner": True,
 }
@@ -143,6 +143,6 @@
     "doc_path": "docs/source",
 }
 
-html_title = f"C++ Paimon"
+html_title = "Apache Paimon C++"
 
 html_show_sourcelink = False

docs/source/examples/benchmark.rst

@@ -54,9 +54,11 @@ specific options below:
 
 ``--paimon_option=<key1>:<value1>;<key2>:<value2>``
    Repeatable table options passed through to Paimon. The default table file
-   format is ``parquet``; use ``--paimon_option file.format:<format>`` to
-   override it. For ``BM_PK_Write`` and ``BM_MOR_Read``, ``bucket`` is forced to
-   ``1``.
+   format is ``parquet``. Benchmark output supports ``parquet`` and, when
+   built with ``PAIMON_ENABLE_ORC=ON``, ``orc``; use
+   ``--paimon_option file.format:parquet`` or
+   ``--paimon_option file.format:orc`` to select one. For ``BM_PK_Write`` and
+   ``BM_MOR_Read``, ``bucket`` is forced to ``1``.
 
 Examples
 ========

docs/source/index.rst

@@ -17,8 +17,8 @@
 
 .. _implementations:
 
-C++ Paimon Documentation
-========================
+Apache Paimon C++
+=================
 
 Paimon C++ is a high-performance C++ implementation of Apache Paimon. We aim to
 provide a native, high-performance and extensible implementation that allows

docs/source/user_guide/global_index.rst

@@ -20,12 +20,12 @@ Global Index
 
 Global Index is a powerful indexing mechanism for append-only tables.
 It enables efficient row-level lookups and filtering without full-table scans.
-Paimon supports multiple global index types:
+Paimon C++ supports the following global index types:
 
-- **Bitmap Index**: A bitmap-based index. Each distinct value is mapped to a compressed bitmap (RoaringBitmap) that records which rows contain that value, enabling extremely fast set membership tests.
 - **BTree Index**: An efficient index based on multi-level SST files for scalar column lookups.
 - **Range Bitmap Index**: A range bitmap index optimized for range predicates on ordered scalar columns. Extends the bitmap approach by encoding value ordering, enabling efficient less-than, greater-than, and range conditions.
 - **Lucene Index**: A full-text search index powered by Lucene++. Supports tokenized text search with multiple modes including match-all, match-any, phrase, prefix, and wildcard queries.
+- **Tantivy Full-Text Index (experimental)**: A Rust FFI-based full-text search index.
 - **Vector Index (Lumina)**: An approximate nearest neighbor (ANN) index powered by Lumina for vector similarity search with configurable distance metrics.
 
 Global indexes work on top of Data Evolution tables. To use global indexes, your table must have:
@@ -34,12 +34,10 @@ Global indexes work on top of Data Evolution tables. To use global indexes, your
 - ``'row-tracking.enabled' = 'true'``
 - ``'data-evolution.enabled' = 'true'``
 
-Bitmap Index
-------------
+Bitmap Index Compatibility
+--------------------------
 
-A bitmap-based index for Equal and In predicates. Each distinct value in the indexed column
-is mapped to a compressed bitmap (RoaringBitmap) that records which rows contain that value.
-This allows extremely fast set membership tests.
+The current Paimon C++ version does not support bitmap global indexes.
 
 BTree Index
 -----------
@@ -94,6 +92,14 @@ search modes including match-all, match-any, phrase, prefix, and wildcard querie
 
   - **Description**: Specifies the directory containing Jieba dictionary files for Chinese text tokenization. At runtime, the system first checks this environment variable; if not set, it falls back to the compile-time ``JIEBA_TEST_DICT_DIR`` macro (only available in test builds). If neither is available, will fail with an error.
 
+Tantivy Full-Text Index (Experimental)
+---------------------------------------
+
+The Tantivy full-text index is an experimental Rust FFI-based backend. Enable it at build time
+with ``-DPAIMON_ENABLE_TANTIVY=ON``. Its global index identifier is ``tantivy-fulltext``.
+Because this backend is experimental, validate its behavior and compatibility for your workload
+before production use.
+
 Vector Index (Lumina)
 ---------------------
 

docs/source/user_guide/manifest_cache.rst

@@ -22,10 +22,9 @@ Overview
 --------
 
 paimon-cpp caches raw manifest file bytes at the ``ObjectsFile<T>::Read()``
-layer. The cache uses the public ``Cache`` abstraction and is injected through
-``ScanContextBuilder`` or ``ReadContextBuilder``. The cache covers data
-manifests, manifest lists, and index manifests because they all read through
-``ObjectsFile<T>``.
+layer. The cache uses the public ``Cache`` abstraction and is enabled through
+``ScanContextBuilder::WithCache()``. The cache covers data manifests, manifest
+lists, and index manifests because they all read through ``ObjectsFile<T>``.
 
 For repeated ``get``, ``scan``, or batch ``get/scan -f`` requests in the same
 process, the same snapshot often reads the same manifest files repeatedly. On a
@@ -84,10 +83,8 @@ Example:
    paimon::ScanContextBuilder scan_builder(table_path);
    scan_builder.WithCache(cache);
 
-   paimon::ReadContextBuilder read_builder(table_path);
-   read_builder.WithCache(cache);
-
-Passing ``nullptr`` or omitting ``WithCache()`` leaves manifest caching disabled.
+Passing ``nullptr`` or omitting ``ScanContextBuilder::WithCache()`` leaves
+manifest caching disabled.
 
 Future Optimizations
 --------------------

include/paimon/utils/special_field_ids.h

@@ -28,20 +28,20 @@ namespace paimon {
 class SpecialFieldIds {
  protected:
     /// System defined constant for field id boundary. Value: INT32_MAX - 10000
-    static const int32_t CPP_FIELD_ID_END = std::numeric_limits<int32_t>::max() - 10000;
+    inline static constexpr int32_t CPP_FIELD_ID_END = std::numeric_limits<int32_t>::max() - 10000;
 
  public:
     /// Special field ID reserved for sequence number. Value: INT32_MAX - 1
-    static const int32_t SEQUENCE_NUMBER = std::numeric_limits<int32_t>::max() - 1;
+    inline static constexpr int32_t SEQUENCE_NUMBER = std::numeric_limits<int32_t>::max() - 1;
     /// Special field ID reserved for value kind. Value: INT32_MAX - 2
-    static const int32_t VALUE_KIND = std::numeric_limits<int32_t>::max() - 2;
+    inline static constexpr int32_t VALUE_KIND = std::numeric_limits<int32_t>::max() - 2;
     /// Special field ID reserved for row kind. Value: INT32_MAX - 3
-    static const int32_t ROW_KIND = std::numeric_limits<int32_t>::max() - 3;
+    inline static constexpr int32_t ROW_KIND = std::numeric_limits<int32_t>::max() - 3;
     /// Special field ID reserved for row ID. Value: INT32_MAX - 5
-    static const int32_t ROW_ID = std::numeric_limits<int32_t>::max() - 5;
+    inline static constexpr int32_t ROW_ID = std::numeric_limits<int32_t>::max() - 5;
 
     /// Special field ID reserved for index score. Value: CPP_FIELD_ID_END - 1
-    static const int32_t INDEX_SCORE = CPP_FIELD_ID_END - 1;
+    inline static constexpr int32_t INDEX_SCORE = CPP_FIELD_ID_END - 1;
 };
 
 }  // namespace paimon