Fix(Robot): Stop splitting and un-splitting URLs

[bookmarks_db.git] / parse_html / bkmk_ph_beautifulsoup4.py

diff --git a/parse_html/bkmk_ph_beautifulsoup4.py b/parse_html/bkmk_ph_beautifulsoup4.py
index 6549683edcb094a4282d89a6846aba81286016a0..060f0786d473c17de36ad073c05632c06f30c35d 100644 (file)
--- a/parse_html/bkmk_ph_beautifulsoup4.py
+++ b/parse_html/bkmk_ph_beautifulsoup4.py
@@ -39,6 +39,8 @@ def _parse_html(html_text, charset):
 
 
 def parse_html(html_text, charset=None, log=None):
+    if not html_text:
+        return None
     root = _parse_html(html_text, charset)
     if root is None:
         return None
@@ -67,9 +69,12 @@ def parse_html(html_text, charset=None, log=None):
         else:
             parts = []
             for part in title:
-                if not isinstance(part, string_type):
-                    part = part.decode()
-                parts.append(part.strip())
+                #if not isinstance(part, string_type):
+                #    part = part.decode()
+                if part.strip:
+                    parts.append(part.strip())
+                else:
+                    parts.append(' ')  # Skip tags, they're usually `<br>`
             title = ''.join(parts)
 
     meta = head.find(_find_contenttype, recursive=False)
@@ -134,6 +139,7 @@ def _find_refresh(Tag):
 
 
 def _find_icon(Tag):
-    return (Tag.name == "link") and \
-       (Tag.get_attribute_list("rel", '')[0].lower()
-        in ('icon', 'shortcut icon'))
+    if Tag.name != "link":
+        return False
+    rel = ' '.join(Tag.get_attribute_list("rel", ''))
+    return rel in ('icon', 'shortcut icon')