Merge pull request #16 from SylvainCorlay/xtensor-master

Update xtensor-python to reflect semantic changes in xtensor

Author: SylvainCorlay

.appveyor.yml

@@ -24,7 +24,7 @@ install:
   - conda info -a
   - conda install pytest -c conda-forge
   - cd test
-  - conda install xtensor==0.2.0 pytest numpy pybind11==1.8.1 -c conda-forge
+  - conda install xtensor==0.2.2 pytest numpy pybind11==1.8.1 -c conda-forge
   - xcopy /S %APPVEYOR_BUILD_FOLDER%\include %MINICONDA%\include
 
 build_script:

.travis.yml

@@ -59,7 +59,7 @@ install:
     # Useful for debugging any issues with conda
     - conda info -a
     - cd test
-    - conda install xtensor==0.2.0 pytest numpy pybind11==1.8.1 -c conda-forge
+    - conda install xtensor==0.2.2 pytest numpy pybind11==1.8.1 -c conda-forge
     - cp -r $TRAVIS_BUILD_DIR/include/* $HOME/miniconda/include/
 
 script:

README.md

@@ -22,6 +22,18 @@ The Python bindings for `xtensor` are based on the [pybind11](https://github.com
 conda install -c conda-forge xtensor-python
 ```
 
+### Dependences on `xtensor` and `pybind11`
+
+`xtensor-python` depends on the `xtensor` and `pybind11` libraries
+
+| `xtensor-python`  | `xtensor`  |  `pybind11` |
+|-------------------|------------|-------------|
+| master            |  master    | ^1.8.1      |
+| 0.2.0             |  ^0.2.1    | ^1.8.1      |
+| 0.1.0             |  ^0.1.1    | ^1.8.1      |
+
+These dependencies are automatically resolved when using the conda package manager.
+
 ## Usage
 
 ### Example 1: Use an algorithm of the C++ library on a numpy array inplace.

include/xtensor-python/pyarray.hpp

@@ -81,18 +81,17 @@ namespace xt
         using size_type = std::size_t;
         using difference_type = std::ptrdiff_t;
 
-        using stepper = xstepper<self_type>;
-        using const_stepper = xstepper<const self_type>;
-
-        using iterator = xiterator<stepper>;
-        using const_iterator = xiterator<const_stepper>;
-        
         using storage_iterator = T*;
         using const_storage_iterator = const T*;
 
         using shape_type = std::vector<size_type>;
         using strides_type = std::vector<size_type>;
         using backstrides_type = pyarray_backstrides<self_type>;
+        
+        using stepper = xstepper<self_type>;
+        using const_stepper = xstepper<const self_type>;
+        using iterator = xiterator<stepper, shape_type>;
+        using const_iterator = xiterator<const_stepper, shape_type>;
 
         using closure_type = const self_type&;
 
@@ -153,13 +152,18 @@ namespace xt
         const_iterator cbegin() const;
         const_iterator cend() const;
 
-        iterator xbegin(const shape_type& shape);
-        iterator xend(const shape_type& shape);
-
-        const_iterator xbegin(const shape_type& shape) const;
-        const_iterator xend(const shape_type& shape) const;
-        const_iterator cxbegin(const shape_type& shape) const;
-        const_iterator cxend(const shape_type& shape) const;
+        template <class S>
+        xiterator<stepper, S> xbegin(const S& shape);
+        template <class S>
+        xiterator<stepper, S> xend(const S& shape);
+        template <class S>
+        xiterator<const_stepper, S> xbegin(const S& shape) const;
+        template <class S>
+        xiterator<const_stepper, S> xend(const S& shape) const;
+        template <class S>
+        xiterator<const_stepper, S> cxbegin(const S& shape) const;
+        template <class S>
+        xiterator<const_stepper, S> cxend(const S& shape) const;
 
         stepper stepper_begin(const shape_type& shape);
         stepper stepper_end(const shape_type& shape);
@@ -169,9 +173,10 @@ namespace xt
 
         storage_iterator storage_begin();
         storage_iterator storage_end();
-
         const_storage_iterator storage_begin() const;
         const_storage_iterator storage_end() const;
+        const_storage_iterator storage_cbegin() const;
+        const_storage_iterator storage_cend() const;
 
         template <class E>
         pyarray(const xexpression<E>& e);
@@ -427,37 +432,43 @@ namespace xt
     }
 
     template <class T, int ExtraFlags>
-    inline auto pyarray<T, ExtraFlags>::xbegin(const shape_type& shape) -> iterator
+    template <class S>
+    inline auto pyarray<T, ExtraFlags>::xbegin(const S& shape) -> xiterator<stepper, S>
     {
-        return iterator(stepper_begin(shape), shape);
+        return xiterator<stepper, S>(stepper_begin(shape), shape);
     }
 
     template <class T, int ExtraFlags>
-    inline auto pyarray<T, ExtraFlags>::xend(const shape_type& shape) -> iterator
+    template <class S>
+    inline auto pyarray<T, ExtraFlags>::xend(const S& shape) -> xiterator<stepper, S>
     {
-        return iterator(stepper_end(shape), shape);
+        return xiterator<stepper, S>(stepper_end(shape), shape);
     }
 
     template <class T, int ExtraFlags>
-    inline auto pyarray<T, ExtraFlags>::xbegin(const shape_type& shape) const -> const_iterator
+    template <class S>
+    inline auto pyarray<T, ExtraFlags>::xbegin(const S& shape) const -> xiterator<const_stepper, S>
     {
-        return const_iterator(stepper_begin(shape), shape);
+        return xiterator<const_stepper, S>(stepper_begin(shape), shape);
     }
 
     template <class T, int ExtraFlags>
-    inline auto pyarray<T, ExtraFlags>::xend(const shape_type& shape) const -> const_iterator
+    template <class S>
+    inline auto pyarray<T, ExtraFlags>::xend(const S& shape) const -> xiterator<const_stepper, S>
     {
-        return const_iterator(stepper_end(shape), shape);
+        return xiterator<const_stepper, S>(stepper_end(shape), shape);
     }
 
     template <class T, int ExtraFlags>
-    inline auto pyarray<T, ExtraFlags>::cxbegin(const shape_type& shape) const -> const_iterator
+    template <class S>
+    inline auto pyarray<T, ExtraFlags>::cxbegin(const S& shape) const -> xiterator<const_stepper, S>
     {
         return xbegin(shape);
     }
 
     template <class T, int ExtraFlags>
-    inline auto pyarray<T, ExtraFlags>::cxend(const shape_type& shape) const -> const_iterator
+    template <class S>
+    inline auto pyarray<T, ExtraFlags>::cxend(const S& shape) const -> xiterator<const_stepper, S>
     {
         return xend(shape);
     }
@@ -514,6 +525,18 @@ namespace xt
         return storage_begin() + pybind_array::size();
     }
 
+    template <class T, int ExtraFlags>
+    inline auto pyarray<T, ExtraFlags>::storage_cbegin() const -> const_storage_iterator
+    {
+        return reinterpret_cast<const_storage_iterator>(pybind11::backport::array_proxy(m_ptr)->data);
+    }
+
+    template <class T, int ExtraFlags>
+    inline auto pyarray<T, ExtraFlags>::storage_cend() const -> const_storage_iterator
+    {
+        return storage_begin() + pybind_array::size();
+    }
+
     template <class T, int ExtraFlags>
     template <class E>
     inline pyarray<T, ExtraFlags>::pyarray(const xexpression<E>& e)