Strip breadcrumbs from the documentation

Change-Id: I2a1469844e5a20eecdd735f38d68e571fce3f3fa
diff --git a/scripts/make_docs.py b/scripts/make_docs.py
index efbbf88..5d5656a 100644
--- a/scripts/make_docs.py
+++ b/scripts/make_docs.py
@@ -1,7 +1,8 @@
 #!/usr/bin/python
+# encoding: utf-8
 #
 # Ceres Solver - A fast non-linear least squares minimizer
-# Copyright 2013 Google Inc. All rights reserved.
+# Copyright 2014 Google Inc. All rights reserved.
 # http://code.google.com/p/ceres-solver/
 #
 # Redistribution and use in source and binary forms, with or without
@@ -40,42 +41,76 @@
 N = len(sys.argv)
 
 if N < 3:
-  print "make_docs.py src_root destination_root"
+  print 'make_docs.py src_root destination_root'
   sys.exit(1)
 
-src_dir    = sys.argv[1] + "/docs/source"
+src_dir    = sys.argv[1] + '/docs/source'
 build_root = sys.argv[2]
-cache_dir  = build_root + "/doctrees"
-html_dir   = build_root + "/html"
+cache_dir  = build_root + '/doctrees'
+html_dir   = build_root + '/html'
 
 # Called from Command Line
 if N == 3:
-  sphinx_exe = "sphinx-build"
+  sphinx_exe = 'sphinx-build'
 
 # Called from CMake (using the SPHINX_EXECUTABLE found)
 elif N == 4:
   sphinx_exe = sys.argv[3]
 
 # Run Sphinx to build the documentation.
-os.system("%s -b html -d %s %s %s" %(sphinx_exe, cache_dir, src_dir, html_dir))
+os.system('%s -b html -d %s %s %s' %(sphinx_exe, cache_dir, src_dir, html_dir))
 
-input_pattern = """config=TeX-AMS-MML_HTMLorMML"></script>"""
-output_pattern = """config=TeX-AMS_HTML">
-  MathJax.Hub.Config({
-    "HTML-CSS": {
-      availableFonts: ["TeX"]
-    }
-  });
-</script>"""
+replacements = [
+  # By default MathJax uses does not use TeX fonts. This simple search
+  # and replace fixes that.
+  ('''config=TeX-AMS-MML_HTMLorMML"></script>''',
+   '''config=TeX-AMS_HTML">
+      MathJax.Hub.Config({
+          "HTML-CSS": {
+            availableFonts: ["TeX"]
+          }
+        });
+      </script>'''),
 
-# By default MathJax uses does not use TeX fonts. This simple search
-# and replace fixes that.
-for name in glob.glob("%s/*.html" % html_dir):
-  print "Postprocessing: ", name
-  fptr = open(name)
-  out = fptr.read().replace(input_pattern, output_pattern)
-  fptr.close()
+  # The title for the homepage is not ideal, so change it.
+  ('<title>Ceres Solver &mdash; Ceres Solver</title>',
+   '<title>Ceres Solver &mdash; A Nonlinear Least Squares Minimizer</title>')
+]
 
-  fptr = open(name, "w")
-  fptr.write(out)
-  fptr.close()
+# This is a nasty hack to strip the breadcrumb navigation. A better strategy is
+# to fork the upstream template, but that is no fun either. Whitespace matters!
+# This doesn't use regular expressions since the escaping makes it untenable.
+breadcrumb_start = \
+'''<div role="navigation" aria-label="breadcrumbs navigation">
+  <ul class="wy-breadcrumbs">
+    <li><a href="index.html">Docs</a> &raquo;</li>
+      
+    <li>'''
+
+breadcrumb_end = \
+'''</li>
+      <li class="wy-breadcrumbs-aside">
+        
+      </li>
+  </ul>
+  <hr/>
+</div>'''
+
+for name in glob.glob('%s/*.html' % html_dir):
+  print 'Postprocessing: ', name
+  with open(name) as fptr:
+    out = fptr.read()
+
+  for input_pattern, output_pattern in replacements:
+    out = out.replace(input_pattern, output_pattern)
+
+  try:
+    pre_breadcrumb_start, post_breadcrumb_start = out.split(breadcrumb_start)
+    title, post_breadcrumb_end = post_breadcrumb_start.split(breadcrumb_end)
+    print 'Stripping breadcrumb for -', title
+    out = pre_breadcrumb_start + post_breadcrumb_end
+  except ValueError:
+    print 'Skipping breadcrumb strip for', name
+
+  with open(name, 'w') as fptr:
+    fptr.write(out)