whitespace, line breaks etc.

[librarian.git] / librarian / html.py

diff --git a/librarian/html.py b/librarian/html.py
index 92fc134..5f832e3 100644 (file)
--- a/librarian/html.py
+++ b/librarian/html.py
@@ -11,16 +11,12 @@ import copy
 from lxml import etree
 from librarian.parser import WLDocument
 from librarian import XHTMLNS, ParseError
+from librarian import functions
 
 from lxml.etree import XMLSyntaxError, XSLTApplyError
 
-ENTITY_SUBSTITUTIONS = [
-    (u'---', u'—'),
-    (u'--', u'–'),
-    (u'...', u'…'),
-    (u',,', u'„'),
-    (u'"', u'”'),
-]
+functions.reg_substitute_entities()
+functions.reg_person_name()
 
 STYLESHEETS = {
     'legacy': 'xslt/book2html.xslt',
@@ -31,23 +27,11 @@ STYLESHEETS = {
 def get_stylesheet(name):
     return os.path.join(os.path.dirname(__file__), STYLESHEETS[name])
 
-def substitute_entities(context, text):
-    """XPath extension function converting all entites in passed text."""
-    if isinstance(text, list):
-        text = ''.join(text)
-    for entity, substitutution in ENTITY_SUBSTITUTIONS:
-        text = text.replace(entity, substitutution)
-    return text
-
-# Register substitute_entities function with lxml
-ns = etree.FunctionNamespace('http://wolnelektury.pl/functions')
-ns['substitute_entities'] = substitute_entities
-
 def html_has_content(text):
     return etree.ETXPath('//p|//{%(ns)s}p|//h1|//{%(ns)s}h1' % {'ns': str(XHTMLNS)})(text)
 
 def transform(input, output_filename=None, is_file=True, \
-    parse_dublincore=True, stylesheet='legacy', options={}):
+    parse_dublincore=True, stylesheet='legacy', options={}, flags=None):
     """Transforms file input_filename in XML to output_filename in XHTML.
 
     If output_filename is None, returns an XML,
@@ -66,6 +50,12 @@ def transform(input, output_filename=None, is_file=True, \
             document = WLDocument.from_string(input, True, \
                 parse_dublincore=parse_dublincore)
 
+        if flags:
+            for flag in flags:
+                document.edoc.getroot().set(flag, 'yes')
+
+        document.clean_ed_note()
+
         result = document.transform(style, **options)
         del document # no longer needed large object :)
 
@@ -74,7 +64,7 @@ def transform(input, output_filename=None, is_file=True, \
             add_table_of_contents(result.getroot())
 
             if output_filename is not None:
-                result.write(output_filename, xml_declaration=False, pretty_print=True, encoding='utf-8')
+                result.write(output_filename, method='html', xml_declaration=False, pretty_print=True, encoding='utf-8')
             else:
                 return result
             return True
@@ -137,7 +127,13 @@ def extract_fragments(input_filename):
     open_fragments = {}
     closed_fragments = {}
 
-    for event, element in etree.iterparse(input_filename, events=('start', 'end')):
+    # iterparse would die on a HTML document
+    parser = etree.HTMLParser(encoding='utf-8')
+    buf = cStringIO.StringIO()
+    buf.write(etree.tostring(etree.parse(input_filename, parser).getroot()[0][0], encoding='utf-8'))
+    buf.seek(0)
+
+    for event, element in etree.iterparse(buf, events=('start', 'end')):
         # Process begin and end elements
         if element.get('class', '') in ('theme-begin', 'theme-end'):
             if not event == 'end': continue # Process elements only once, on end event
@@ -221,6 +217,7 @@ def add_anchors(root):
     counter = 1
     for element in root.iterdescendants():
         if any_ancestor(element, lambda e: e.get('class') in ('note', 'motto', 'motto_podpis', 'dedication')
+        or e.get('id') == 'nota_red'
         or e.tag == 'blockquote'):
             continue