tests for handling "2 poł XIX w." dates + utf8 check

[librarian.git] / librarian / html.py

diff --git a/librarian/html.py b/librarian/html.py
index 997f904..70fc6e5 100644 (file)
--- a/librarian/html.py
+++ b/librarian/html.py
@@ -57,6 +57,7 @@ def transform(wldoc, stylesheet='legacy', options=None, flags=None):
 
         if html_has_content(result):
             add_anchors(result.getroot())
+            add_table_of_themes(result.getroot())
             add_table_of_contents(result.getroot())
 
             return OutputFile.from_string(etree.tostring(result, method='html',
@@ -163,7 +164,8 @@ 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 +222,27 @@ 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
 
@@ -253,6 +264,34 @@ def add_table_of_contents(root):
 
     root.insert(0, toc)
 
+    
+def add_table_of_themes(root):
+    try:
+        from sortify import sortify
+    except ImportError:
+        sortify = lambda x: x
+
+    book_themes = {}
+    for fragment in root.findall('.//a[@class="theme-begin"]'):
+        if not fragment.text:
+            continue
+        theme_names = [s.strip() for s in fragment.text.split(',')]
+        for theme_name in theme_names:
+            book_themes.setdefault(theme_name, []).append(fragment.get('name'))
+    book_themes = book_themes.items()
+    book_themes.sort(key=lambda s: sortify(s[0]))
+    themes_div = etree.Element('div', id="themes")
+    themes_ol = etree.SubElement(themes_div, 'ol')
+    for theme_name, fragments in book_themes:
+        themes_li = etree.SubElement(themes_ol, 'li')
+        themes_li.text = "%s: " % theme_name
+        for i, fragment in enumerate(fragments):
+            item = etree.SubElement(themes_li, 'a', href="#%s" % fragment)
+            item.text = str(i + 1)
+            item.tail = ' '
+    root.insert(0, themes_div)
+
+
 
 def extract_annotations(html_path):
     """For each annotation, yields a tuple: anchor, text, html."""