encode unicode for lxml

[librarian.git] / librarian / html.py

diff --git a/librarian/html.py b/librarian/html.py
index 9401a01..5f832e3 100644 (file)
--- a/librarian/html.py
+++ b/librarian/html.py
@@ -16,6 +16,7 @@ from librarian import functions
 from lxml.etree import XMLSyntaxError, XSLTApplyError
 
 functions.reg_substitute_entities()
+functions.reg_person_name()
 
 STYLESHEETS = {
     'legacy': 'xslt/book2html.xslt',
@@ -30,7 +31,7 @@ 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,
@@ -49,6 +50,10 @@ 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)
@@ -122,7 +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