fix slug RE and handle Wybor with no <pytanie> tag

[librarian.git] / librarian / html.py

diff --git a/librarian/html.py b/librarian/html.py
index 997f904..985970a 100644 (file)
--- a/librarian/html.py
+++ b/librarian/html.py
@@ -8,7 +8,7 @@ import cStringIO
 import copy
 
 from lxml import etree
-from librarian import XHTMLNS, ParseError, OutputFile
+from librarian import XHTMLNS, ParseError, IOFile
 from librarian import functions
 
 from lxml.etree import XMLSyntaxError, XSLTApplyError
@@ -59,7 +59,7 @@ def transform(wldoc, stylesheet='legacy', options=None, flags=None):
             add_anchors(result.getroot())
             add_table_of_contents(result.getroot())
 
-            return OutputFile.from_string(etree.tostring(result, method='html',
+            return IOFile.from_string(etree.tostring(result, method='html',
                 xml_declaration=False, pretty_print=True, encoding='utf-8'))
         else:
             return None
@@ -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)
@@ -220,18 +221,27 @@ 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
     for element in root.iterdescendants():
         if element.tag in ('h2', 'h3'):
-            if any_ancestor(element, lambda e: e.get('id') in ('footnotes',) or e.get('class') in ('person-list',)):
+            if any_ancestor(element, lambda e: e.get('id') in ('footnotes', 'nota_red') or e.get('class') in ('person-list',)):
                 continue
 
+            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, ''.join(element.xpath('text()')), []))
+                sections[-1][3].append((counter, element.tag, element_text, []))
             else:
-                sections.append((counter, element.tag, ''.join(element.xpath('text()')), []))
+                sections.append((counter, element.tag, element_text, []))
             add_anchor(element, "s%d" % counter, with_link=False)
             counter += 1