Fixing ugly mpi bug in tensorproductspace.py and bumping version

Author: mikaem

docs/Makefile

@@ -57,13 +57,14 @@ help:
 
 %.ipynb2:
 	doconce format ipynb demos/$(basename $@).do.txt
+	python add_metadata.py demos/$(basename $@).ipynb
 	jupyter nbconvert --inplace --clear-output demos/${basename $@}.ipynb
 	#jupyter nbconvert --inplace --execute demos/${basename $@}.ipynb
 	doconce subst 'XXX' '   ' demos/$(basename $@).ipynb
 	doconce subst "<div" "<a" demos/$(basename $@).ipynb
 	doconce subst "div>" "a>" demos/$(basename $@).ipynb
 	#doconce subst '<!--.*-->' '' demos/$(basename $@).ipynb
-	python add_metadata.py demos/$(basename $@).ipynb
+	#python add_metadata.py demos/$(basename $@).ipynb
 	cp demos/$(basename $@).ipynb ../../shenfun-demos/content/
 	cp demos/$(basename $@).ipynb notebooks/
 

docs/source/conf.py

@@ -100,8 +100,8 @@
 #html_theme = 'bootstrap'
 #html_theme_path = sphinx_bootstrap_theme.get_html_theme_path()
 
-# For some reason mathjax 2.7.5 renders bold type ugly. Use 2.7.1
-mathjax_path = 'https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-svg.js'
+#mathjax_path = 'https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-svg.js'
+mathjax_path = 'https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js'
 #mathjax_path = 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.1/MathJax.js?config=TeX-AMS-MML_HTMLorMML'
 
 intersphinx_mapping = {'mpi4py_fft': ('https://mpi4py-fft.readthedocs.io/en/latest/', None)}

shenfun/__init__.py

@@ -29,7 +29,7 @@
 """
 #pylint: disable=wildcard-import,no-name-in-module
 
-__version__ = '4.1.0'
+__version__ = '4.1.1'
 __author__ = 'Mikael Mortensen'
 
 import numpy as np

shenfun/tensorproductspace.py

@@ -926,8 +926,8 @@ def local_mesh(self, bcast=False, map_true_domain=True, kind='quadrature'):
         """
         mesh = self.mesh(map_true_domain=map_true_domain, kind=kind)
         lm = []
-        ss = [slice(None)]*len(mesh)
         for axis, (n, s) in enumerate(zip(mesh, self.local_slice(False))):
+            ss = [slice(None)]*len(mesh)
             ss[axis] = s
             lm.append(n[tuple(ss)])
         if bcast is True: