Showing only in books with no parents in tagged_object_list.

[wolnelektury.git] / lib / librarian / html.py

diff --git a/lib/librarian/html.py b/lib/librarian/html.py
index c75dd6b..9763428 100644 (file)
--- a/lib/librarian/html.py
+++ b/lib/librarian/html.py
@@ -54,7 +54,8 @@ def transform(input_filename, output_filename):
 
     result = doc.xslt(style)
     add_anchors(result.getroot())
-    result.write(output_filename, xml_declaration=True, pretty_print=True, encoding='utf-8')
+    add_table_of_contents(result.getroot())
+    result.write(output_filename, xml_declaration=False, pretty_print=True, encoding='utf-8')
 
 
 class Fragment(object):
@@ -108,7 +109,7 @@ def extract_fragments(input_filename):
 
     for event, element in etree.iterparse(input_filename, events=('start', 'end')):
         # Process begin and end elements
-        if element.tag == 'span' and element.get('class', '') in ('theme-begin', 'theme-end'):
+        if element.get('class', '') in ('theme-begin', 'theme-end'):
             if not event == 'end': continue # Process elements only once, on end event
 
             # Open new fragment
@@ -116,9 +117,9 @@ def extract_fragments(input_filename):
                 fragment = Fragment(id=element.get('fid'), themes=element.text)
 
                 # Append parents
-                if element.getparent().tag != 'body':
+                if element.getparent().get('id', None) != 'book-text':
                     parents = [element.getparent()]
-                    while parents[-1].getparent().tag != 'body':
+                    while parents[-1].getparent().get('id', None) != 'book-text':
                         parents.append(parents[-1].getparent())
 
                     parents.reverse()
@@ -170,12 +171,18 @@ def add_anchor(element, number):
     element.insert(0, anchor_target)
 
 
+def any_ancestor(element, test):
+    for ancestor in element.iterancestors():
+        if test(ancestor):
+            return True
+    return False
+
+
 def add_anchors(root):
     counter = 1
     for element in root.iterdescendants():
-        if element.getparent().tag in 'div' and 'note' in element.getparent().get('class', ''):
-            continue
-        if element.getparent().tag in 'blockquote':
+        if any_ancestor(element, lambda e: e.get('class') in ('note', 'motto', 'motto_podpis', 'dedication')
+        or e.tag == 'blockquote'):
             continue
         
         if element.tag == 'p' and 'verse' in element.get('class', ''):
@@ -187,3 +194,29 @@ def add_anchors(root):
             counter += 1
 
 
+def add_table_of_contents(root):
+    sections = []
+
+    for element in root.iterdescendants():
+        if element.tag in ('h2', 'h3'):
+            if element.tag == 'h3' and len(sections) and sections[-1][0] == 'h2':
+                sections[-1][2].append((element.tag, ''.join(element.xpath('descendant-or-self::text()')), []))
+            else:
+                sections.append((element.tag, ''.join(element.xpath('descendant-or-self::text()')), []))
+            
+    toc = etree.Element('div')
+    toc.set('id', 'toc')
+    toc_header = etree.SubElement(toc, 'h2')
+    toc_header.text = u'Spis treści'
+    toc_list = etree.SubElement(toc, 'ol')
+
+    for section, text, subsections in sections:
+        section_element = etree.SubElement(toc_list, 'li')
+        section_element.text = text
+        if len(subsections):
+            subsection_list = etree.SubElement(section_element, 'ol')
+            for subsection, text, _ in subsections:
+                subsection_element = etree.SubElement(subsection_list, 'li')
+                subsection_element.text = text
+    root.insert(0, toc)
+