#2044: handle weird whitespace for covers

[librarian.git] / librarian / html.py

diff --git a/librarian/html.py b/librarian/html.py
index 39e5a01..e04790f 100644 (file)
--- a/librarian/html.py
+++ b/librarian/html.py
@@ -163,7 +163,8 @@ def extract_fragments(input_filename):
         # Process all elements except begin and end
         else:
             # Omit annotation tags
-            if len(element.get('name', '')) or element.get('class', '') == 'annotation':
+            if (len(element.get('name', '')) or 
+                    element.get('class', '') in ('annotation', 'anchor')):
                 if event == 'end' and element.tail:
                     for fragment_id in open_fragments:
                         open_fragments[fragment_id].append('text', element.tail)
@@ -220,6 +221,14 @@ def add_anchors(root):
             counter += 1
 
 
+def raw_printable_text(element):
+    working = copy.deepcopy(element)
+    for e in working.findall('a'):
+        if e.get('class') == 'annotation':
+            e.text = ''
+    return etree.tostring(working, method='text', encoding=unicode).strip()
+
+
 def add_table_of_contents(root):
     sections = []
     counter = 1
@@ -228,8 +237,7 @@ def add_table_of_contents(root):
             if any_ancestor(element, lambda e: e.get('id') in ('footnotes',) or e.get('class') in ('person-list',)):
                 continue
 
-            element_text = etree.tostring(element, method='text',
-                    encoding=unicode).strip()
+            element_text = raw_printable_text(element)
             if element.tag == 'h3' and len(sections) and sections[-1][1] == 'h2':
                 sections[-1][3].append((counter, element.tag, element_text, []))
             else: