Increase version to 2.0.0 and update docs

- move examples to dedicated folder and make sure they compile
- remove old docs about laziness

Author: Thoemi09

CMakeLists.txt

@@ -19,7 +19,7 @@ cmake_policy(VERSION ${CMAKE_VERSION})
 
 # ############
 # Define Project
-project(mpi VERSION 1.3.0 LANGUAGES CXX)
+project(mpi VERSION 2.0.0 LANGUAGES CXX)
 get_directory_property(IS_SUBPROJECT PARENT_DIRECTORY)
 
 # Get the git hash & print status

doc/CMakeLists.txt

@@ -25,6 +25,9 @@ add_custom_target(doxygen ALL
   VERBATIM
 )
 
+# Compile example codes
+add_subdirectory(examples)
+
 # ############
 # Install
 install(DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}/html/ COMPONENT documentation DESTINATION share/doc/${PROJECT_NAME}

doc/documentation.md

@@ -28,12 +28,14 @@ If you are looking for a specific function, class, etc., try using the search ba
   Besides storing the `MPI_Group` object, it also provides some convient functions for getting the size of the
   group, the rank of the current process or for splitting the group based on include rules.
 
-It further contains the convenient function mpi::is_initialized and the static boolean mpi::has_env.
+It further contains the convenient functions mpi::is_initialized and mpi::is_finalized and the static boolean 
+mpi::has_env.
 
 ## MPI datatypes and operations
 
 @ref mpi_types_ops map various C++ datatypes to MPI datatypes and help the user with registering their own datatypes to
 be used in MPI communications.
+
 Furthermore, it offers tools to simplify the creation of custom MPI operations usually required in `MPI_Reduce` or
 `MPI_Accumulate` functions.
 
@@ -71,13 +73,6 @@ Another use-case of @ref mpi_osc_shm is the shared memory aspect by which
 MPI applications can reduce their memory requirements through the deduplication
 of replicated data between MPI ranks that are executed on the same SMP node.
 
-## Lazy MPI communication
-
-@ref mpi_lazy can be used to provied collective MPI communication for lazy expression types.
-Most users probably won't need to use this functionality directly.
-
-We refer the interested reader to [TRIQS/nda](https://triqs.github.io/nda/latest/group__av__mpi.html) for more details.
-
 ## Event handling
 
 @ref event_handling provides the mpi::monitor class which can be used to communicate and handle events across multiple

doc/ex1.md

@@ -4,20 +4,7 @@
 
 In this example, we show how to implement the standard `Hello world` program using **mpi**.
 
-```cpp
-#include <mpi/mpi.hpp>
-#include <iostream>
-
-int main(int argc, char *argv[]) {
-  // initialize MPI environment and communicator
-  mpi::environment env(argc, argv);
-  mpi::communicator world;
-
-  // get rank and greet world
-  int rank = world.rank();
-  std::cout << "Hello from processor " << rank << "\n";
-}
-```
+@include ex1.cpp
 
 Output (depends on the number of processes and the order is arbitrary):
 

doc/ex2.md

@@ -4,52 +4,7 @@
 
 In this example, we show how to use the mpi::monitor class to communicate and process errors across a communicator.
 
-```cpp
-#include <mpi/mpi.hpp>
-#include <mpi/monitor.hpp>
-#include <iostream>
-
-int main(int argc, char *argv[]) {
-  // initialize MPI environment
-  mpi::environment env(argc, argv);
-  mpi::communicator world;
-
-  // initialize monitor
-  mpi::monitor monitor(world);
-
-  // in case an event has occurred, print some info and return true
-  auto stop = [&monitor, world](int i) {
-    bool res = false;
-    if (monitor.event_on_any_rank()) {
-      std::cerr << "Processor " << world.rank() << ": After " << i << " steps an event has been communicated.\n";
-      res = true;
-    }
-    return res;
-  };
-
-  // loop as long as no event has occurred
-  int event_rank = 3;
-  for (int i = 0; i < 1000000; ++i) {
-    // report a local event on the event_rank
-    if (world.rank() == event_rank) {
-      std::cerr << "Processor " << event_rank << ": Local event reported.\n";
-      monitor.report_local_event();
-    }
-
-    // should we stop the loop?
-    if (stop(i)) break;
-  }
-
-  // check if all processes finished the loop
-  if (world.rank() == 0) {
-    if (monitor.event_on_any_rank()) {
-      std::cout << "Oh no! An event occurred somewhere and the loop has not been finished on all processes.\n";
-    } else {
-      std::cout << "No worries, all processes have finished the loop.\n";
-    }
-  }
-}
-```
+@include ex2.cpp
 
 Output (running with `-n 12`):
 

doc/ex3.md

@@ -2,55 +2,11 @@
 
 [TOC]
 
-In this example, we show how to register a new MPI datatype and how to use mpi::map_C_function and mpi::map_add to
-define MPI operations for it.
+In this example, we show 
+- how to register a new MPI datatype by providing a `tie_data` function for our C++ type and 
+- how to use mpi::map_C_function and mpi::map_add to define MPI operations for it.
 
-```cpp
-#include <mpi/mpi.hpp>
-#include <iostream>
-
-// define a custom complex type
-struct my_complex {
-  double real;
-  double imag;
-};
-
-// define an addition for my_complex
-inline my_complex operator+(const my_complex& z1, const my_complex& z2) {
-  return { z1.real + z2.real, z1.imag + z2.imag };
-}
-
-// define a tie_data function for my_complex to make it MPI compatible
-inline auto tie_data(const my_complex& z) {
-  return std::tie(z.real, z.imag);
-}
-
-int main(int argc, char *argv[]) {
-  // initialize MPI environment
-  mpi::environment env(argc, argv);
-  mpi::communicator world;
-
-  // create complex number z
-  my_complex z = { world.rank() + 1.0, static_cast<double>(world.rank()) };
-
-  // sum z over all processes
-  auto sum = mpi::reduce(z, world, 0, false, mpi::map_add<my_complex>());
-
-  // define a product for my_complex
-  auto my_product = [](const my_complex& z1, const my_complex& z2) {
-    return my_complex { z1.real * z2.real - z1.imag * z2.imag, z1.real * z2.imag + z1.imag * z2.real };
-  };
-
-  // multiply z over all processes
-  auto product = mpi::reduce(z, world, 0, false, mpi::map_C_function<my_complex, my_product>());
-
-  // print result
-  if (world.rank() == 0) {
-    std::cout << "sum = (" << sum.real << ", " << sum.imag << ")\n";
-    std::cout << "product = (" << product.real << ", " << product.imag << ")\n";
-  }
-}
-```
+@include ex3.cpp
 
 Output (running with `-n 5`):
 

doc/ex4.md

@@ -4,48 +4,7 @@
 
 In this example, we show how to write a specialized `mpi_reduce_into` for a custom type.
 
-```cpp
-#include <mpi/mpi.hpp>
-#include <iostream>
-#include <vector>
-
-// Custom type.
-class foo {
-  public:
-  // Constructor.
-  foo(int x = 5) : x_(x) {}
-
-  // Get the value stored in the class.
-  int x() const { return x_; }
-
-  // Specialization of mpi_reduce_into for the custom type.
-  friend void mpi_reduce_into(foo const &f_in, foo &f_out, mpi::communicator c = {}, int root = 0, bool all = false, MPI_Op op = MPI_SUM) {
-    mpi::reduce_into(f_in.x_, f_out.x_, c, root, all, op);
-  }
-
-  private:
-  int x_;
-};
-
-int main(int argc, char *argv[]) {
-  // initialize MPI environment
-  mpi::environment env(argc, argv);
-  mpi::communicator world;
-
-  // create a vector of foo objects
-  std::vector<foo> vec {foo{1}, foo{2}, foo{3}, foo{4}, foo{5}};
-
-  // reduce the vector of foo objects
-  auto result = mpi::reduce(vec, world);
-
-  // print the result on rank 0
-  if (world.rank() == 0) {
-    std::cout << "Reduced vector: ";
-    for (auto const &f : result) std::cout << f.x() << " ";
-    std::cout << "\n";
-  }
-}
-```
+@include ex4.cpp
 
 Output (running with `-n 4`):
 

doc/examples.md

@@ -11,7 +11,7 @@
 
 All examples have been compiled on a MacBook Pro with an Apple M2 Max chip and [open-mpi](https://www.open-mpi.org/)
 5.0.1.
-We further used clang 19.1.7 together with cmake 3.31.5.
+We further used clang 21.1.2 together with cmake 4.1.2.
 
 Assuming that the actual example code is in a file `main.cpp`, the following generic `CMakeLists.txt` should work for
 all examples:
@@ -31,7 +31,7 @@ include (FetchContent)
 FetchContent_Declare(
   mpi
   GIT_REPOSITORY https://github.com/TRIQS/mpi.git
-  GIT_TAG        1.3.x
+  GIT_TAG        2.0.x
 )
 FetchContent_MakeAvailable(mpi)
 

doc/examples/CMakeLists.txt

@@ -0,0 +1,8 @@
+# List of all examples
+file(GLOB_RECURSE all_examples RELATIVE ${CMAKE_CURRENT_SOURCE_DIR} *.cpp)
+
+foreach(ex_src ${all_examples})
+    get_filename_component(ex_name ${ex_src} NAME_WE)
+    add_executable(${ex_name} ${ex_src})
+    target_link_libraries(${ex_name} mpi::mpi_c)
+endforeach()

doc/examples/ex1.cpp

@@ -0,0 +1,12 @@
+#include <mpi/mpi.hpp>
+#include <iostream>
+
+int main(int argc, char *argv[]) {
+  // initialize MPI environment and communicator
+  mpi::environment env(argc, argv);
+  mpi::communicator world;
+
+  // get rank and greet world
+  int rank = world.rank();
+  std::cout << "Hello from processor " << rank << "\n";
+}