Merge pull request #5 from mogproject/dev

Dev

.clang-format

@@ -0,0 +1,77 @@
+---
+Language:        Cpp
+# BasedOnStyle:  Google
+AccessModifierOffset: -1
+AlignAfterOpenBracket: true
+AlignConsecutiveAssignments: false
+AlignEscapedNewlinesLeft: true
+AlignOperands:   true
+AlignTrailingComments: true
+AllowAllParametersOfDeclarationOnNextLine: true
+AllowShortBlocksOnASingleLine: true
+AllowShortCaseLabelsOnASingleLine: true
+AllowShortFunctionsOnASingleLine: All
+AllowShortIfStatementsOnASingleLine: true
+AllowShortLoopsOnASingleLine: true
+AlwaysBreakAfterDefinitionReturnType: None
+AlwaysBreakBeforeMultilineStrings: true
+AlwaysBreakTemplateDeclarations: true
+BinPackArguments: true
+BinPackParameters: true
+BreakBeforeBinaryOperators: None
+BreakBeforeBraces: Custom
+BraceWrapping:
+  AfterClass: false
+  AfterControlStatement: false
+  AfterEnum: false
+  AfterFunction: false
+  AfterNamespace: false
+  AfterStruct: false
+  AfterUnion: false
+  BeforeCatch: false
+  BeforeElse: false
+BreakBeforeTernaryOperators: true
+BreakConstructorInitializersBeforeComma: false
+ColumnLimit:     120
+CommentPragmas:  '^ IWYU pragma:'
+ConstructorInitializerAllOnOneLineOrOnePerLine: true
+ConstructorInitializerIndentWidth: 4
+ContinuationIndentWidth: 4
+Cpp11BracedListStyle: true
+DerivePointerAlignment: true
+DisableFormat:   false
+ExperimentalAutoDetectBinPacking: false
+ForEachMacros:   [ foreach, Q_FOREACH, BOOST_FOREACH ]
+IndentCaseLabels: true
+IndentWidth:     2
+IndentWrappedFunctionNames: false
+KeepEmptyLinesAtTheStartOfBlocks: false
+MacroBlockBegin: ''
+MacroBlockEnd:   ''
+MaxEmptyLinesToKeep: 1
+NamespaceIndentation: None
+ObjCBlockIndentWidth: 2
+ObjCSpaceAfterProperty: false
+ObjCSpaceBeforeProtocolList: false
+PenaltyBreakBeforeFirstCallParameter: 1
+PenaltyBreakComment: 300
+PenaltyBreakFirstLessLess: 120
+PenaltyBreakString: 1000
+PenaltyExcessCharacter: 1
+PenaltyReturnTypeOnItsOwnLine: 200
+PointerAlignment: Left
+SpaceAfterCStyleCast: false
+SpaceBeforeAssignmentOperators: true
+SpaceBeforeParens: ControlStatements
+SpaceInEmptyParentheses: false
+SpacesBeforeTrailingComments: 2
+SpacesInAngles:  false
+SpacesInContainerLiterals: true
+SpacesInCStyleCastParentheses: false
+SpacesInParentheses: false
+SpacesInSquareBrackets: false
+Standard:        Auto
+TabWidth:        2
+UseTab:          Never
+...
+

Makefile

@@ -1,5 +1,11 @@
+RM = /bin/rm
 PYTHON=python3
 
+SRC_CPP=src/main/cpp
+BUILD_DIR=$(PWD)/build
+TEST_BIN_DIR=$(BUILD_DIR)/test
+TEST_EXEC=$(TEST_BIN_DIR)/modular_test
+
 SRC_PY=src/main/python
 TEST_PY=src/test/python
 STUB_PY=src/main/stubs
@@ -9,11 +15,31 @@ PYTEST_OPTS=""  # --full-trace
 export PYTHONPATH=$(SRC_PY)
 export MYPYPATH=$(STUB_PY)
 
-test:
+PROFILE_ON ?= false
+TRACE_ON ?= false
+
+build:
+	cd $(SRC_CPP) && cmake -S . -B $(BUILD_DIR)/Release -DCMAKE_BUILD_TYPE=Release -DPROFILE_ON=$(PROFILE_ON) -DTRACE_ON=$(TRACE_ON)
+	cd $(SRC_CPP) && cmake --build $(BUILD_DIR)/Release
+
+test: test-cpp test-py
+
+test-py:
 	mypy $(SRC_PY)
 	$(PYTHON) -m pytest -x --cov=$(SRC_PY) --cov-report=lcov:./coverage/lcov.info $(PYTEST_OPTS) $(TEST_PY)
 
+test-cpp:
+	@echo "GTEST_FILTER: $(GTEST_FILTER)"
+	cd $(SRC_CPP) && cmake -DBUILD_TESTS=ON -S . -B $(BUILD_DIR)/Debug
+	cd $(SRC_CPP) && cmake --build $(BUILD_DIR)/Debug
+	$(TEST_EXEC) --output-on-failure $(GTEST_OPTS)
+
+clean:
+	@echo "Cleaning..."
+	@$(RM) -rf build/*
+	@echo "Cleaning done."
+
 lab:
 	jupyter-lab
 
-.PHONY: test lab
+.PHONY: build test test-cpp clean lab

README.md

@@ -8,14 +8,21 @@ The code implements the algorithm described in *Simpler, Linear-Time Modular Dec
 
 ## Dependencies
 
-- NetworkX (`pip install networkx`)
-- NumPy (`pip install numpy`)
+- C++
+  - gcc version 11 or 12
+  - CMake
+- Python
+  - NetworkX (`pip install networkx`)
+  - NumPy (`pip install numpy`)
 
 #### Dependencies for Unit Testing
 
-- pytest (`pip install pytest`)
-- pytest-cov (`pip install pytest-cov`)
-- MyPy (`pip install mypy`)
+- C++
+  - GoogleTest (automatically installed during the build process)
+- Python
+  - pytest (`pip install pytest`)
+  - pytest-cov (`pip install pytest-cov`)
+  - MyPy (`pip install mypy`)
 
 ## Installation
 
@@ -28,6 +35,9 @@ TBD
 | Task | Command | Note |
 | :--- | :--- | :--- |
 | Run all unit tests | `make test` | Coverage info will be created as `coverage/lcov.info`. |
+| Run C++ unit tests | `make test-cpp` ||
+| Run Python unit tests | `make test-py` ||
+| Clean build and binary files | `make clean` ||
 | Open Jupyter Lab   | `make lab`  | Jupyter Notebooks are in the `notebooks` directory.|
 
 ## Special Thanks

src/main/cpp/CMakeLists.txt

@@ -0,0 +1,65 @@
+cmake_minimum_required(VERSION 3.14)
+
+# Avoid warning about DOWNLOAD_EXTRACT_TIMESTAMP in CMake 3.24:
+if (CMAKE_VERSION VERSION_GREATER_EQUAL "3.24.0")
+  cmake_policy(SET CMP0135 NEW)
+endif()
+
+# options
+option(BUILD_TESTS "Build test programs" OFF)
+
+# specify the C++ standard
+set(CMAKE_CXX_STANDARD 14)
+set(CMAKE_CXX_STANDARD_REQUIRED True)
+
+# specify compilers on local machine (Mac)
+if (EXISTS /opt/homebrew/bin/g++-12)
+  set(CMAKE_C_COMPILER /opt/homebrew/bin/gcc-12)
+  set(CMAKE_CXX_COMPILER /opt/homebrew/bin/g++-12)
+elseif(EXISTS /opt/homebrew/bin/gcc-11)
+  set(CMAKE_C_COMPILER /opt/homebrew/bin/gcc-11)
+  set(CMAKE_CXX_COMPILER /opt/homebrew/bin/g++-11)
+endif ()
+
+message("Compiler: ${CMAKE_CXX_COMPILER}")
+
+# source files
+file(GLOB_RECURSE MAIN_SRC
+  ds/*.[ch]pp
+  modular/*.[ch]pp
+  readwrite/*.[ch]pp
+  util/*.[ch]pp
+)
+
+file(GLOB BENCH_SRC modular-bench.cpp)
+
+include_directories(
+  .
+)
+
+if (CMAKE_BUILD_TYPE STREQUAL "Debug")
+  set (CMAKE_CXX_FLAGS_DEBUG "${CMAKE_CXX_FLAGS_DEBUG} -fno-omit-frame-pointer -fsanitize=leak")
+endif ()
+
+# project
+project(modular)
+set(CMAKE_CXX_FLAGS "-Wall -funroll-loops -fno-stack-limit -O3")
+
+# build options
+if (NOT DEFINED PROFILE_ON)
+  set(PROFILE_ON false)
+endif ()
+
+if (NOT DEFINED TRACE_ON)
+  set(TRACE_ON false)
+endif ()
+
+add_compile_definitions(PROFILE_ON=${PROFILE_ON})
+add_compile_definitions(TRACE_ON=${TRACE_ON})
+
+# tests with GoogleTest
+if (BUILD_TESTS)
+  add_subdirectory(../../test/cpp ../test)
+else ()
+  add_executable(modular-bench ${BENCH_SRC} ${MAIN_SRC})
+endif ()

src/main/cpp/ds/graph/Graph.hpp

@@ -0,0 +1,132 @@
+#pragma once
+
+#include "ds/set/ArrayBitset.hpp"
+#include "ds/set/SortedVectorSet.hpp"
+
+namespace ds {
+namespace graph {
+/**
+ * @brief Graph with integer labels.
+ *
+ * @tparam Set set implementation class
+ */
+class Graph {
+ private:
+  /** Number of vertices. */
+  std::size_t n_;
+
+  /** Number of vertices. */
+  std::size_t m_;
+
+  /** True if the adjacency sets are in dense representation. */
+  bool dense_;
+
+  /** Adjacency sets. */
+  std::vector<basic_set<int>*> adj_;
+
+  /** Masks for removed vertices. */
+  basic_set<int>* removed_;
+
+  bool is_valid(int v) const { return 0 <= v && v < static_cast<int>(adj_.size()) && !removed_->get(v); }
+
+  basic_set<int>* create_set(std::size_t n, bool dense) {
+    if (dense) {
+      if (n <= 1 << 6) {
+        return new ArrayBitset6(n);
+      } else if (n <= 1 << 7) {
+        return new ArrayBitset7(n);
+      } else if (n <= 1 << 8) {
+        return new ArrayBitset8(n);
+      } else if (n <= 1 << 9) {
+        return new ArrayBitset9(n);
+      } else if (n <= 1 << 10) {
+        return new ArrayBitset10(n);
+      } else if (n <= 1 << 11) {
+        return new ArrayBitset11(n);
+      } else if (n <= 1 << 12) {
+        return new ArrayBitset12(n);
+      } else if (n <= 1 << 13) {
+        return new ArrayBitset13(n);
+      } else {
+        throw std::invalid_argument("Graph: n too large for dense representation");
+      }
+    } else {
+      return new SortedVectorSet();
+    }
+  }
+
+ public:
+  Graph(std::size_t n = 0, std::vector<std::pair<int, int>> const& edges = {}, bool dense = false)
+      : n_(n), m_(0), dense_(dense) {
+    for (int i = 0; i < static_cast<int>(n); ++i) adj_.push_back(create_set(n, dense));
+    removed_ = create_set(n, dense);
+    for (auto& p : edges) add_edge(p.first, p.second);
+  }
+
+  std::size_t number_of_nodes() const { return n_; }
+  std::size_t number_of_edges() const { return m_; }
+
+  int add_vertex() {
+    int x;
+    if (removed_->empty()) {  // all vertices are in use
+      if (dense_) {
+        throw std::invalid_argument("add_vertex: cannot extend capacity");
+      } else {
+        adj_.push_back(create_set(n_, dense_));
+        // `removed_` should be SortedVectorSet so no need to replace
+        x = n_++;
+      }
+    } else {
+      // reuse one of the removed vertices
+      x = dense_ ? removed_->pop_front() : removed_->pop_back();
+      ++n_;
+    }
+    return x;
+  }
+
+  void add_edge(int u, int v) {
+    if (!is_valid(u)) throw std::invalid_argument("add_edge: invalid u");
+    if (!is_valid(v)) throw std::invalid_argument("add_edge: invalid v");
+    if (u == v) throw std::invalid_argument("add_edge: loop is not allowed");
+
+    if (!has_edge(u, v)) {
+      adj_[u]->set(v);
+      adj_[v]->set(u);
+      ++m_;
+    }
+  }
+
+  void remove_vertex(int v) {
+    if (!is_valid(v)) throw std::invalid_argument("remove_vertex: invalid v");
+
+    --n_;
+    m_ -= adj_[v]->size();
+
+    for (auto u : adj_[v]->to_vector()) adj_[u]->reset(v);
+    adj_[v]->clear();
+    removed_->set(v);
+  }
+
+  void remove_edge(int u, int v) {
+    if (!is_valid(u)) throw std::invalid_argument("remove_edge: invalid u");
+    if (!is_valid(v)) throw std::invalid_argument("remove_edge: invalid v");
+    if (!adj_[u]->get(v)) throw std::invalid_argument("remove_edge: edge does not exist");
+
+    adj_[u]->reset(v);
+    adj_[v]->reset(u);
+    --m_;
+  }
+
+  std::vector<int> neighbors(int v) const { return adj_[v]->to_vector(); }
+
+  int degree(int v) const { return adj_[v]->size(); }
+
+  bool has_vertex(int v) const { return is_valid(v); }
+
+  bool has_edge(int u, int v) const {
+    if (!is_valid(u) || !is_valid(v) || u == v) return false;
+    return degree(u) <= degree(v) ? adj_[u]->get(v) : adj_[v]->get(u);
+  }
+};
+}  // namespace graph
+}  // namespace ds