Merge pull request #3140 from t20100/update-doc

Update documentation following drop of python<=3.4

CHANGELOG.rst

@@ -60,10 +60,10 @@ This version drops the support of Python 2.7 and Python <= 3.4.
     * Limit `silxGlobalThreadPool` function to use 4 threads maximum (PR #3072)
 
   * `silx.gui.utils.glutils`: Added `isOpenGLAvailable` to check the availability of OpenGL (PR #2878)
-  * `silx.gui.widgets:
+  * `silx.gui.widgets`:
 
     * Added `ElidedLabel` widget (PR #3110, #3111)
-    * Fixed `LegendIconWidget (PR #3112)
+    * Fixed `LegendIconWidget` (PR #3112)
 
 * `silx.io`:
 
@@ -85,7 +85,7 @@ This version drops the support of Python 2.7 and Python <= 3.4.
 
 * Build, documentation and tests:
 
-  * Dropped Python2 support (PR #3119) and removed Python 2 tests and packaging (PR #2838, #2917)
+  * Dropped Python2 support (PR #3119, #3140) and removed Python 2 tests and packaging (PR #2838, #2917)
   * Added debian 11/Ubuntu 20.04 packaging (PR #2875)
   * Improved test environment (PR #2870, #2949, #2995, #3009, #3061, #3086, #3087, #3122), documentation (PR #2872, #2894, #2937, #2987, #3042, #3053, #3068, #3091, #3103, #3115) and sample code (PR #2978, #3130, #3138)
   * Fixed Windows "fat binary" build (PR #2971)

doc/source/install.rst

@@ -10,11 +10,11 @@ This table summarizes the support matrix of silx:
 +------------+--------------+---------------------+
 | System     | Python vers. | Qt and its bindings |
 +------------+--------------+---------------------+
-| `Windows`_ | 3.5, 3.6-3.7 | PyQt5.6+, PySide2   |
+| `Windows`_ | 3.5-3.8      | PyQt5.6+, PySide2   |
 +------------+--------------+---------------------+
-| `MacOS`_   | 2.7, 3.5-3.7 | PyQt5.6+, PySide2   |
+| `MacOS`_   | 3.5-3.8      | PyQt5.6+, PySide2   |
 +------------+--------------+---------------------+
-| `Linux`_   | 2.7, 3.4-3.7 | PyQt5.3+, PySide2   |
+| `Linux`_   | 3.5-3.8      | PyQt5.3+, PySide2   |
 +------------+--------------+---------------------+
 
 For the description of *silx* dependencies, see the Dependencies_ section.
@@ -211,8 +211,8 @@ prompt.
 MacOS
 -----
 
-While Apple ships Python 2.7 by default on their operating systems, we recommend
-using Python 3.5 or newer to ease the installation of the Qt library.
+Apple ships Python 2.7 by default on their operating systems.
+You must install Python 3.5 or newer to install silx.
 
 Then, install *silx* with ``pip``, see `Installing with pip`_::
 

doc/source/modules/gui/plot/getting_started.rst

@@ -20,7 +20,7 @@ For a complete description of the API, see :mod:`silx.gui.plot`.
 Use :mod:`silx.gui.plot` from (I)Python console
 -----------------------------------------------
 
-We recommend to use (I)Python 3.x and PyQt5.
+We recommend to use (I)Python >=3.5 and PyQt5.
 
 From a Python or IPython interpreter, the simplest way is to import the :mod:`silx.sx` module:
 
@@ -48,10 +48,6 @@ the way silx loads Qt and the way IPython is doing it through the ``--gui`` opti
 `%pylab <http://ipython.org/ipython-doc/stable/interactive/magics.html#magic-pylab>`_ magics.
 In this case, IPython magics that initialize Qt might not work after importing modules from silx.gui.
 
-When using Python2.7 and PyQt4, there is another incompatibility to deal with as
-silx requires PyQt4 API version 2 (See note below for explanation).
-In this case, start IPython with the ``QT_API`` environment variable set to ``pyqt``.
-
 On Linux and MacOS X, run from the command line::
 
   QT_API=pyqt ipython
@@ -61,16 +57,6 @@ On Windows, run from the command line::
   set QT_API=pyqt&&ipython
 
 
-.. note:: PyQt4 used from Python 2.x provides 2 incompatible versions of QString and QVariant:
-
-   - version 1, the legacy version which is also the default, and
-   - version 2, a more pythonic one, which is the only one supported by *silx*.
-
-   All other configurations (i.e., PyQt4 on Python 3.x, PySide2, PyQt5, IPython QtConsole widget) uses version 2.
-
-   For more information, see `IPython, PyQt and PySide <http://ipython.org/ipython-doc/stable/interactive/reference.html#pyqt-and-pyside>`_.
-
-
 Plot functions
 ++++++++++++++
 

doc/source/modules/gui/plot/tools/index.rst

@@ -104,13 +104,6 @@ Other tools API
 .. autoclass:: RegionOfInterestTableWidget
    :members:
 
-:mod:`~silx.gui.plot.tools.profile`: Profile Tools
---------------------------------------------------
-
-.. automodule:: silx.gui.plot.tools.profile
-
-.. currentmodule:: silx.gui.plot.tools.profile
-
 :mod:`ColorBar`: ColorBar Widget
 ================================
 

doc/source/virtualenv.rst

@@ -12,30 +12,12 @@ Prerequisites
 
 This guide assumes that your system meets the following requirements:
 
-   - a version of python compatible with *silx* is installed (python 2.7 or python >= 3.5)
+   - a version of python compatible with *silx* is installed (Python >= 3.5)
    - the *pip* installer for python packages is installed
-   - the Qt and PyQt libraries are installed (optional, required for using ``silx.gui``)
 
 Installation procedure
 ----------------------
 
-
-Install vitrualenv
-******************
-
-.. code-block:: bash
-
-    pip install virtualenv --user
-
-.. note::
-
-    This step is not required for recent version of Python 3.
-    Virtual environments are created using a builtin standard library,
-    ``venv``.
-    On Debian platforms, you might need to install the ``python3-venv``
-    package.
-
-
 Create a virtualenv
 *******************
 
@@ -51,32 +33,33 @@ a virtual environment named ``silx_venv``
     cd
     mkdir -p venvs
     cd venvs
-    virtualenv silx_venv
-
+    python -m venv silx_venv
 
 A virtualenv contains a copy of your default python interpreter with a few tools
 to install packages (pip, setuptools).
 
-To use a different python interpreter, you can specify it on the command line.
-For example, to use python 3.4:
-
-.. code-block:: bash
-
-    virtualenv -p /usr/bin/python3.4 silx_venv
-
-But for python 3  you should use the builtin ``venv`` module:
+Virtual environments are created using a builtin standard library, ``venv`` (Python3 only):
 
 .. code-block:: bash
 
     python3 -m venv /path/to/new/virtual/environment
 
 .. note::
 
+    On Debian platforms, you might need to install the ``python3-venv`` package.
+
     If you don't need to start with a clean environment and you don't want
     to install each required library one by one, you can use a command line
     option to create a virtualenv with access to all system packages:
     ``--system-site-packages``
 
+To use a different python interpreter, use it to create the virtual environment.
+For example, to use python 3.5:
+
+.. code-block:: bash
+
+    /usr/bin/python3.5 -m venv silx_venv
+
 
 Activate a virtualenv
 *********************
@@ -128,64 +111,28 @@ install *silx*:
 
 .. since 0.5, numpy is now automatically installed when doing `pip install silx`
 
-Install optional dependencies
-*****************************
-
-The following command installs libraries that are needed by various modules
-of *silx*:
-
-.. code-block:: bash
-
-    pip install matplotlib fabio h5py
-
-The next command installs libraries that are used by the python modules
-handling parallel computing:
-
-.. code-block:: bash
-
-    pip install pyopencl mako
-
-
-Install pyqt
+Install silx
 ************
 
-If your python version is 3.5 or newer, installing PyQt5 and all required packages
-is as simple as typing:
-
-.. code-block:: bash
-
-    pip install PyQt5
-
-For previous versions of python, there are no PyQt wheels available, so the installation
-is not as simple.
-
-The simplest way, assuming that PyQt is installed on your system, is to use that
-system package directly. For this, you need to add a symbolic link to your virtualenv.
-
-If you want to use PyQt5 installed in ``/usr/lib/python2.7/dist-packages/``, type:
+To install silx with minimal dependencies, run:
 
 .. code-block:: bash
 
-    ln -s /usr/lib/python2.7/dist-packages/PyQt5 silx_venv/lib/python2.7/site-packages/
-    ln -s /usr/lib/python2.7/dist-packages/sip.so silx_venv/lib/python2.7/site-packages/
-
+    pip install silx
 
-Install silx
-************
+To install silx with all dependencies, run:
 
 .. code-block:: bash
 
-    pip install silx
-
+    pip install silx[full]
 
-To test *silx*, open an interactive python console. If you managed to install PyQt5 or PySide2
-in your virtualenv, type:
+To test *silx*, open an interactive python console:
 
 .. code-block:: bash
 
     python
 
-If you don't have PyQt, use:
+If you don't have PyQt5 or PySide2, run:
 
 .. code-block:: bash
 
@@ -195,9 +142,3 @@ Run the test suite using:
 
     >>> import silx.test
     >>> silx.test.run_tests()
-
-
-
-
-
-

package/debian10/control

@@ -8,6 +8,7 @@ Priority: optional
 Build-Depends: cython3 (>= 0.23.2),
                debhelper (>= 10),
                dh-python,
+               graphviz,
                help2man,
                ipython3,
                ipython3-qtconsole,

package/debian11/control

@@ -9,6 +9,7 @@ Build-Depends: cython3 (>= 0.23.2),
                cython3-dbg (>= 0.23.2),
                debhelper-compat (= 12),
                dh-python,
+               graphviz,
                help2man,
                ipython3,
                pandoc <!nodoc>,

package/debian9/control

@@ -8,6 +8,7 @@ Build-Depends: cython3,
                libstdc++-4.9-dev|libstdc++6,
                debhelper (>=9.20150101+deb8u2),
                dh-python,
+               graphviz,
                python3-all-dev,
                python3-numpy,
                python3-fabio,

silx/gui/plot/backends/BackendMatplotlib.py

@@ -450,6 +450,7 @@ def contains(self, mouseevent):
         return inside, info
 
     def set_data(self, A):
+        """Overridden to add a fast path for RGBA unit8 images"""
         A = numpy.array(A, copy=False)
         if A.ndim != 3 or A.shape[2] != 4 or A.dtype != numpy.uint8:
             super(Image, self).set_data(A)