Add strip_math_delims option to sphinxify so that its current default

behavior can be disabled.

If I understand correctly, the SageNB instead uses a MathJax extension
which applies math rendering to any <span> or <div> with the "math"
class.  However, this does not work by default with the MathJax settings
in the Jupyter Notebook, so it's better to keep the math delimiters.

Also updates the code that removes them to use a regular expression
instead.

src/sage/misc/sphinxify.py

@@ -29,7 +29,7 @@
 from sphinx.application import Sphinx
 
 
-def sphinxify(docstring, format='html'):
+def sphinxify(docstring, format='html', strip_math_delims=True):
     r"""
     Runs Sphinx on a ``docstring``, and outputs the processed
     documentation.
@@ -41,6 +41,13 @@ def sphinxify(docstring, format='html'):
     - ``format`` -- string (optional, default 'html') -- either 'html' or
       'text'
 
+    - ``strip_math_delims`` -- string (optional, default True) -- if True,
+      remove content like ``\( .. \)`` and ``\[ .. \]`` that are used by
+      default to delimit math content processed by MathJax.  This should be
+      set to False when generating output to be displayed in the Jupyter
+      Notebook, as its MathJax is configured to look for these.  The default
+      behavior is for legacy support of SageNB.
+
     OUTPUT:
 
     - string -- Sphinx-processed documentation, in either HTML or
@@ -55,6 +62,8 @@ def sphinxify(docstring, format='html'):
         '<div class="docstring"...<strong>Testing</strong>\n<span class="math...</p>\n\n\n</div>'
         sage: sphinxify('`x=y`')
         '<div class="docstring">\n    \n  <p><span class="math notranslate nohighlight">x=y</span></p>\n\n\n</div>'
+        sage: sphinxify(':math:`x=y`', strip_math_delims=False)
+        '<div class="docstring">\n    \n  <p><span class="math notranslate nohighlight">\\(x=y\\)</span></p>\n\n\n</div>'
         sage: sphinxify('`x=y`', format='text')
         'x=y\n'
         sage: sphinxify(':math:`x=y`', format='text')
@@ -141,7 +150,8 @@ def sphinxify(docstring, format='html'):
                         'src="/doc/static/reference/media/\\2"',
                         output)
         # Remove spurious \(, \), \[, \].
-        output = output.replace(r'\(', '').replace(r'\)', '').replace(r'\[', '').replace(r'\]', '')
+        if strip_math_delims:
+            output = re.sub(r'\\[()[\]]', '', output)
     else:
         from warnings import warn
         warn("Sphinx did not produce any output", Warning)