Merge pull request #26 from kalessin/emptyregion

avoid exception when instantiating htmlregion if parent htmlpage has empty body. Added test.

Author: pablohoffman

scrapely/htmlpage.py

@@ -133,9 +133,11 @@ class HtmlPageParsedRegion(HtmlPageRegion):
     fragments contained within this region
     """
     def __new__(cls, htmlpage, start_index, end_index):
-        text_start = htmlpage.parsed_body[start_index].start
-        text_end = htmlpage.parsed_body[end_index or -1].end
-        text = htmlpage.body[text_start:text_end]
+        text = htmlpage.body
+        if text:
+            text_start = htmlpage.parsed_body[start_index].start
+            text_end = htmlpage.parsed_body[end_index or -1].end
+            text = htmlpage.body[text_start:text_end]
         return HtmlPageRegion.__new__(cls, htmlpage, text)
 
     def __init__(self, htmlpage, start_index, end_index):

scrapely/tests/test_htmlpage.py

@@ -5,7 +5,7 @@
 from unittest import TestCase
 
 from scrapely.tests import iter_samples
-from scrapely.htmlpage import parse_html, HtmlTag, HtmlDataFragment
+from scrapely.htmlpage import parse_html, HtmlTag, HtmlDataFragment, HtmlPage
 from scrapely.tests.test_htmlpage_data import *
 
 def _encode_element(el):
@@ -135,3 +135,6 @@ def test_malformed2(self):
         parsed = [_decode_element(d) for d in PARSED9]
         self._test_sample(PAGE9, parsed)
 
+    def test_empty_subregion(self):
+        htmlpage = HtmlPage(body=u"")
+        self.assertEqual(htmlpage.subregion(), u"")