Add regression tests for html nbconvert

IPython/nbconvert/exporters/tests/files/notebook2.ipynb

@@ -169,6 +169,15 b''
      "language": "python",
      "metadata": {},
      "outputs": []
+    },
+    {
+     "cell_type": "code",
+     "collapsed": false,
+     "input": [],
+     "language": "python",
+     "metadata": {},
+     "outputs": [],
+     "prompt_number": null
     }
    ],
    "metadata": {}

IPython/nbconvert/exporters/tests/test_html.py

@@ -15,6 +15,7 b''
 from .base import ExportersTestsBase
 from ..html import HTMLExporter
 from IPython.testing.decorators import onlyif_any_cmd_exists
+import re
 
 #-----------------------------------------------------------------------------
 # Class
@@ -59,3 +60,17 b' class TestHTMLExporter(ExportersTestsBase):'
         (output, resources) = HTMLExporter(template_file='full').from_filename(self._get_notebook())
         assert len(output) > 0
 
+    @onlyif_any_cmd_exists('nodejs', 'node', 'pandoc')
+    def test_prompt_number(self):
+        """
+        Does HTMLExporter properly format input and output prompts?
+        """
+        (output, resources) = HTMLExporter(template_file='full').from_filename(self._get_notebook())
+        in_regex = r"In \[(.*)\]:"
+        out_regex = r"Out\[(.*)\]:"
+
+        ins = ["1", "2", "6", "7", "8", "10", "14", " ", " "]
+        outs = ["7", "10", "14"]
+
+        assert re.findall(in_regex, output) == ins
+        assert re.findall(out_regex, output) == outs