ebook fix: remove stray markup from the text (#172)

Before this commit, chapter 23 of the ebook had some spurious markup
that found its way into the readable text:

    And Harry brought out the original parchment with the hypotheses,
    and began scribbling.

    plus .5minus 1

    Observation:

    Wizardry isn’t as powerful now as it was when Hogwarts was founded.

    plus .5minus 1

You can see the source of this markup in the comment immediately above
the line changed in this commit:

    # \vskip 1\baselineskip plus .5\textheight minus 1\baselineskip

The problem is this regex:

    "\\vskip .*?\\baselineskip"

In the example above, it matches `"\vskip 1\baselineskip"`, stopping at
the first `"\baselineskip"` instead of the one at the end of the line.
This leaves the errant bit of markup in the text:

    plus .5\textheight minus 1\baselineskip

Which, because it does not start with a backslash, ends up inserted into
the content.

The fix is simple: remove the `'?'`, turning the `".*"` into a greedy
match instead of a minimal match. This matches to the end of the last
`"\baselineskip"`, completely removing this bit of markup from the text
as intended.

scripts/ebook/step_3.py

@@ -51,7 +51,7 @@
     cont = re.sub(r"\\clearpage(\{\}|)\n?", "", cont)
 
     # \vskip 1\baselineskip plus .5\textheight minus 1\baselineskip
-    cont = re.sub(r"\\vskip .*?\\baselineskip", "", cont)
+    cont = re.sub(r"\\vskip .*\\baselineskip", "", cont)
 
     # remove \settowidth{\versewidth}... \begin{verse}[\versewidth]
     cont = re.sub(