more updates

Author: zingale

content/12-extensions/extensions-example.md

@@ -250,7 +250,7 @@ Here's the implementation of our Mandelbrot generator:
 We build the shared library as:
 
 ```bash
-g++ -O3  -Wall -Wextra -shared -std=c++17 -fPIC $(python3 -m pybind11 --includes) mandel.cpp -o mandel$(python3-config --extension-suffix)
+g++ -DNDEBUG -O3  -Wall -Wextra -shared -std=c++17 -fPIC $(python3 -m pybind11 --includes) mandel.cpp -o mandel$(python3-config --extension-suffix)
 ```
 
 Our driver is essentially the same as the Fortran one.

content/12-extensions/extensions-overview.md

@@ -19,6 +19,9 @@ and weaknesses.
     header-only library that allows you to call C++ functions directly
     from python.
 
+    A related library is [nanobind](https://github.com/wjakob/nanobind),
+    which has similar syntax but may be more efficient.
+
 * C
 
   * [C-API](https://docs.python.org/3/c-api/index.html) : the
@@ -33,9 +36,12 @@ and weaknesses.
     represent the arrays we use.  This means writing a lot of
     boilerplate code just to deal with some simple operations.
 
-    This underlies most of the techniques that we'll see here.
+    This underlies many of the techniques that we'll see here.
+
+    .. note::
 
-    These days, there are better methods for most applications.
+       These days, there are better methods for most applications,
+       and you should probably not use the C-API directly.
 
   * [ctypes](https://docs.python.org/3/library/ctypes.html) : this
     is a module that allows you to call functions in shared libraries.

examples/extensions/pybind11/contiguous/mandel.cpp

@@ -1,7 +1,6 @@
-#include <iostream>
+#include <cstddef>
 #include <cmath>
 #include <complex>
-#include <vector>
 
 #include <pybind11/pybind11.h>
 #include <pybind11/numpy.h>
@@ -33,9 +32,9 @@ py::array_t<int> mandelbrot(int N,
     // construct the numpy array we will return
     // we need to specify the strides manually
 
-    constexpr size_t elsize = sizeof(int);
-    size_t shape[2]{N, N};
-    size_t strides[2]{N * elsize, elsize};
+    constexpr std::size_t elsize = sizeof(int);
+    std::size_t shape[2]{N, N};
+    std::size_t strides[2]{N * elsize, elsize};
     auto m = py::array_t<int>(shape, strides);
     auto m_view = m.mutable_unchecked<2>();
 

examples/extensions/pybind11/mandel.cpp

@@ -21,9 +21,9 @@ py::array_t<int> mandelbrot(int N,
     // construct the numpy array we will return
     // we need to specify the strides manually
 
-    constexpr size_t elsize = sizeof(int);
-    size_t shape[2]{N, N};
-    size_t strides[2]{N * elsize, elsize};
+    constexpr std::size_t elsize = sizeof(int);
+    std::size_t shape[2]{N, N};
+    std::size_t strides[2]{N * elsize, elsize};
     auto m = py::array_t<int>(shape, strides);
     auto m_view = m.mutable_unchecked<2>();