[Feat][Java] Fill two incompatible gaps between C++ and Java (#279)

.github/workflows/java.yml

@@ -7,7 +7,6 @@ on:
     branches:
       - main
     paths:
-      - 'cpp/include/**'
       - 'java/**'
       - '.github/workflows/java.yml'
   pull_request:
@@ -37,10 +36,10 @@ jobs:
           sudo apt-get install -y /tmp/apache-arrow-apt-source-latest-"$(lsb_release --codename --short)".deb
           sudo apt-get update -y
           # TODO: ISSUE-241
-          sudo apt install -y libarrow-dev=13.0.0-1 \
-                              libarrow-dataset-dev=13.0.0-1 \
-                              libarrow-acero-dev=13.0.0-1 \
-                              libparquet-dev=13.0.0-1
+          sudo apt install -y libarrow-dev=14.0.1-1 \
+                            libarrow-dataset-dev=14.0.1-1 \
+                            libarrow-acero-dev=14.0.1-1 \
+                            libparquet-dev=14.0.1-1
           sudo apt-get install libcurl4-openssl-dev -y
           sudo apt-get install llvm-11 clang-11 lld-11 libclang-11-dev libz-dev -y
           
@@ -53,7 +52,8 @@ jobs:
 
       - name: Run test
         run: |
-          export JAVA_HOME=${JAVA_HOME_11_X64}
+          # Temporarily using Java 8, related issue: https://github.com/alibaba/GraphAr/issues/277
+          export JAVA_HOME=${JAVA_HOME_8_X64}
           export LLVM11_HOME=/usr/lib/llvm-11
           pushd java
           mvn test -Dspotless.check.skip=true

cpp/include/gar/graph.h

@@ -1121,7 +1121,7 @@ ConstructVerticesCollection(const GraphInfo& graph_info,
  * @param edge_label The edge label.
  * @param dst_label The destination vertex label.
  * @param adj_list_type The adjList type.
- * @param vertex_chunk_begin The index of the begin vertex chunk.
+ * @param vertex_chunk_begin The index of the beginning vertex chunk.
  * @param vertex_chunk_end The index of the end vertex chunk (not included).
  * @return The constructed collection or error.
  */

java/CMakeLists.txt

@@ -1,11 +1,13 @@
 cmake_minimum_required(VERSION 3.6)
 project(gar-java)
 
+option(BUILD_GAR_CPP "Build GraphAr C++ library with specific version" ON)
 set(CMAKE_MODULE_PATH ${CMAKE_CURRENT_SOURCE_DIR}/cmake)
-SET(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -O3 -g -std=c++17 -Wall")
+set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -O3 -g -std=c++17 -Wall")
 
 # set auto-generated JNI code and handwriting JNI code as source files
-file(GLOB SOURCES "${CMAKE_CURRENT_SOURCE_DIR}/target/generated-sources/annotations/*.cc" "${CMAKE_CURRENT_SOURCE_DIR}/target/generated-test-sources/test-annotations/*.cc"
+file(GLOB SOURCES "${CMAKE_CURRENT_SOURCE_DIR}/target/generated-sources/annotations/*.cc"
+                  "${CMAKE_CURRENT_SOURCE_DIR}/target/generated-test-sources/test-annotations/*.cc"
                   "${CMAKE_CURRENT_SOURCE_DIR}/src/main/cpp/ffi/*.cc")
 # remove auto-generated JNI code for specific method cause we have handwriting JNI code for it
 list(REMOVE_ITEM SOURCES "${CMAKE_CURRENT_SOURCE_DIR}/target/generated-sources/annotations/jni_com_alibaba_graphar_arrow_ArrowTable_Static_cxx_0x58c7409.cc")
@@ -19,19 +21,23 @@ find_package(JNI REQUIRED)
 include_directories(SYSTEM ${JAVA_INCLUDE_PATH})
 include_directories(SYSTEM ${JAVA_INCLUDE_PATH2})
 
-# some JNI code depends on arrow
+# some JNI code depends on arrow directly
 find_package(Arrow REQUIRED)
-# build graphar-cpp in specific version
-include(graphar-cpp)
-build_graphar_cpp()
 
 # build the bridge JNI library
 add_library(${LIBNAME} SHARED ${SOURCES})
-# include graphar-cpp headers
-target_include_directories(${LIBNAME} SYSTEM BEFORE PRIVATE ${GAR_INCLUDE_DIR})
-# link graphar-cpp and arrow
-target_link_libraries(${LIBNAME} ${CMAKE_JNI_LINKER_FLAGS} gar_shared)
-target_link_libraries(${LIBNAME} ${CMAKE_JNI_LINKER_FLAGS} Arrow::arrow_static)
+# build GraphAr C++ library or link system installed GraphAr C++ library
+if(BUILD_GAR_CPP)
+    include(graphar-cpp)
+    build_graphar_cpp()
+    target_link_libraries(${LIBNAME} ${CMAKE_JNI_LINKER_FLAGS} gar_shared)
+    # include graphar-cpp headers
+    target_include_directories(${LIBNAME} SYSTEM BEFORE PRIVATE ${GAR_INCLUDE_DIR})
+else()
+    find_package(gar REQUIRED)
+    target_link_libraries(${LIBNAME} ${CMAKE_JNI_LINKER_FLAGS} gar)
+endif()
+target_link_libraries(${LIBNAME} ${CMAKE_JNI_LINKER_FLAGS} Arrow::arrow_shared)
 
 set_target_properties(${LIBNAME} PROPERTIES LINKER_LANGUAGE CXX)
 
@@ -43,4 +49,4 @@ add_custom_command(TARGET ${LIBNAME}
 add_custom_command(TARGET ${LIBNAME}
                    POST_BUILD
                    COMMAND ${CMAKE_COMMAND} -E make_directory ${CMAKE_CURRENT_SOURCE_DIR}/target/native/bitcode
-                   COMMAND ${CMAKE_COMMAND} -E copy_if_different $<TARGET_OBJECTS:${LIBNAME}> ${CMAKE_CURRENT_SOURCE_DIR}/target/native/bitcode COMMAND_EXPAND_LISTS)
+                   COMMAND ${CMAKE_COMMAND} -E copy_if_different $<TARGET_OBJECTS:${LIBNAME}> ${CMAKE_CURRENT_SOURCE_DIR}/target/native/bitcode COMMAND_EXPAND_LISTS)

java/build.xml

@@ -18,7 +18,7 @@
         <echo message="Native Library Name: ${native.library.name}" level="info"/>
 
         <exec executable="cmake" dir="${project.build.directory}/native" failonerror="true">
-            <arg line="-DCMAKE_AR=&quot;${env.LLVM11_HOME}/bin/llvm-ar&quot; -DCMAKE_RANLIB=&quot;${env.LLVM11_HOME}/bin/llvm-ranlib&quot; -DCMAKE_C_COMPILER=&quot;${env.LLVM11_HOME}/bin/clang&quot; -DCMAKE_CXX_COMPILER=&quot;${env.LLVM11_HOME}/bin/clang++&quot; -DCMAKE_CXX_FLAGS=&quot;-flto -fforce-emit-vtables&quot; ${basedir}"/>
+            <arg line="-DBUILD_GAR_CPP=${build_gar_cpp}; -DCMAKE_AR=&quot;${env.LLVM11_HOME}/bin/llvm-ar&quot; -DCMAKE_RANLIB=&quot;${env.LLVM11_HOME}/bin/llvm-ranlib&quot; -DCMAKE_C_COMPILER=&quot;${env.LLVM11_HOME}/bin/clang&quot; -DCMAKE_CXX_COMPILER=&quot;${env.LLVM11_HOME}/bin/clang++&quot; -DCMAKE_CXX_FLAGS=&quot;-flto -fforce-emit-vtables&quot; ${basedir}"/>
         </exec>
 
         <exec executable="make" dir="${project.build.directory}/native" failonerror="true">

java/cmake/graphar-cpp.cmake

@@ -48,7 +48,7 @@ function(build_graphar_cpp)
     set(GAR_INCLUDE_DIR "${GAR_PREFIX}/include" CACHE INTERNAL "graphar cpp include directory")
     set(GAR_BUILD_BYPRODUCTS "${GAR_SHARED_LIB}")
 
-    set(GAR_VERSION_TO_BUILD "v0.9.0")
+    set(GAR_VERSION_TO_BUILD "v0.10.0")
 
     include(ExternalProject)
     ExternalProject_Add(graphar_ep

java/pom.xml

@@ -129,7 +129,7 @@
                         <configuration>
                             <target>
                                 <property name="compile_classpath" refid="maven.compile.classpath"/>
-
+                                <property name="build_gar_cpp" value="${buildGarCPP}"/>
                                 <ant antfile="${basedir}/build.xml">
                                     <target name="run-llvm4jni" />
                                 </ant>

java/src/main/java/com/alibaba/graphar/edges/EdgesCollection.java

@@ -14,96 +14,29 @@
 
 package com.alibaba.graphar.edges;
 
+import static com.alibaba.graphar.util.CppClassName.GAR_EDGES_COLLECTION;
+import static com.alibaba.graphar.util.CppHeaderName.GAR_GRAPH_H;
+
+import com.alibaba.fastffi.CXXHead;
 import com.alibaba.fastffi.CXXPointer;
+import com.alibaba.fastffi.CXXReference;
+import com.alibaba.fastffi.CXXValue;
+import com.alibaba.fastffi.FFIGen;
 import com.alibaba.fastffi.FFINameAlias;
-import com.alibaba.graphar.graphinfo.EdgeInfo;
-import com.alibaba.graphar.graphinfo.GraphInfo;
-import com.alibaba.graphar.stdcxx.StdString;
-import com.alibaba.graphar.types.AdjListType;
-import com.alibaba.graphar.util.Result;
+import com.alibaba.fastffi.FFITypeAlias;
 import java.util.Iterator;
 
 /** EdgesCollection is designed for reading a collection of edges. */
+@FFIGen
+@FFITypeAlias(GAR_EDGES_COLLECTION)
+@CXXHead(GAR_GRAPH_H)
 public interface EdgesCollection extends CXXPointer, Iterable<Edge> {
-    /**
-     * Construct the collection for a range of edges.
-     *
-     * @param graphInfo The GraphInfo for the graph.
-     * @param srcLabel The source vertex label.
-     * @param edgeLabel The edge label.
-     * @param dstLabel The destination vertex label.
-     * @param adjListType The adjList type.
-     * @param vertexChunkBegin The index of the beginning vertex chunk.
-     * @param vertexChunkEnd The index of the end vertex chunk (not included).
-     * @return The constructed collection or error.
-     */
-    static EdgesCollection create(
-            final GraphInfo graphInfo,
-            String srcLabel,
-            String edgeLabel,
-            String dstLabel,
-            AdjListType adjListType,
-            long vertexChunkBegin,
-            long vertexChunkEnd) {
-        Result<EdgeInfo> maybeEdgeInfo =
-                graphInfo.getEdgeInfo(
-                        StdString.create(srcLabel),
-                        StdString.create(edgeLabel),
-                        StdString.create(dstLabel));
-        if (maybeEdgeInfo.hasError()) {
-            throw new RuntimeException(
-                    "graphInfo to create EdgesCollection has error: "
-                            + maybeEdgeInfo.status().message().toJavaString());
-        }
-        EdgeInfo edgeInfo = maybeEdgeInfo.value();
-        if (!edgeInfo.containAdjList(adjListType)) {
-            throw new RuntimeException(
-                    "The edge "
-                            + edgeLabel
-                            + " of adj list type "
-                            + adjListType
-                            + " doesn't exist.");
-        }
-        switch (adjListType) {
-            case ordered_by_source:
-                return EdgesCollectionOrderedBySource.factory.create(
-                        edgeInfo, graphInfo.getPrefix(), vertexChunkBegin, vertexChunkEnd);
-            case ordered_by_dest:
-                return EdgesCollectionOrderedByDest.factory.create(
-                        edgeInfo, graphInfo.getPrefix(), vertexChunkBegin, vertexChunkEnd);
-            case unordered_by_source:
-                return EdgesCollectionUnorderedBySource.factory.create(
-                        edgeInfo, graphInfo.getPrefix(), vertexChunkBegin, vertexChunkEnd);
-            case unordered_by_dest:
-                return EdgesCollectionUnorderedByDest.factory.create(
-                        edgeInfo, graphInfo.getPrefix(), vertexChunkBegin, vertexChunkEnd);
-        }
-        throw new RuntimeException("Unknown adj list type: " + adjListType);
-    }
-
-    /**
-     * Construct the collection for a range of edges.
-     *
-     * @param graphInfo The GraphInfo for the graph.
-     * @param srcLabel The source vertex label.
-     * @param edgeLabel The edge label.
-     * @param dstLabel The destination vertex label.
-     * @param adjListType The adjList type.
-     * @return The constructed collection or error.
-     */
-    static EdgesCollection create(
-            final GraphInfo graphInfo,
-            String srcLabel,
-            String edgeLabel,
-            String dstLabel,
-            AdjListType adjListType) {
-        return create(graphInfo, srcLabel, edgeLabel, dstLabel, adjListType, 0, Long.MAX_VALUE);
-    }
-
     /** The iterator pointing to the first edge. */
+    @CXXValue
     EdgeIter begin();
 
     /** The iterator pointing to the past-the-end element. */
+    @CXXValue
     EdgeIter end();
 
     /**
@@ -115,7 +48,8 @@ static EdgesCollection create(
      * @return The new constructed iterator.
      */
     @FFINameAlias("find_src")
-    EdgeIter findSrc(long id, EdgeIter from);
+    @CXXValue
+    EdgeIter findSrc(long id, @CXXReference EdgeIter from);
 
     /**
      * Construct and return the iterator pointing to the first incoming edge of the vertex with
@@ -126,7 +60,8 @@ static EdgesCollection create(
      * @return The new constructed iterator.
      */
     @FFINameAlias("find_dst")
-    EdgeIter findDst(long id, EdgeIter from);
+    @CXXValue
+    EdgeIter findDst(long id, @CXXReference EdgeIter from);
 
     /** Get the number of edges in the collection. */
     long size();