X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/ccd5654568bb362a15185bcf0c1dfa5cb9f3ace7..0859d20e39ee940904f2bfdc36572000109b2214:/librarian/epub.py

diff --git a/librarian/epub.py b/librarian/epub.py
index ef07c3f..a5607f7 100644
--- a/librarian/epub.py
+++ b/librarian/epub.py
@@ -82,7 +82,8 @@ def replace_characters(node):
     def replace_chars(text):
         if text is None:
             return None
-        return text.replace("---", u"\u2014")\
+        return text.replace(u"\ufeff", u"")\
+                   .replace("---", u"\u2014")\
                    .replace("--", u"\u2013")\
                    .replace(",,", u"\u201E")\
                    .replace('"', u"\u201D")\
@@ -258,7 +259,7 @@ def transform_chunk(chunk_xml, chunk_no, annotations):
     replace_by_verse(chunk_xml)
     html_tree = xslt(chunk_xml, res('xsltScheme.xsl'))
     chars = used_chars(html_tree.getroot())
-    output_html = etree.tostring(html_tree, pretty_print=True)
+    output_html = etree.tostring(html_tree, method="html", pretty_print=True)
     return output_html, toc, chars
 
 
@@ -275,6 +276,8 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir=
     def transform_file(input_xml, chunk_counter=1, first=True):
         """ processes one input file and proceeds to its children """
 
+        replace_characters(input_xml.getroot())
+
         children = [child.text for child in input_xml.findall('.//'+DCNS('relation.hasPart'))]
 
         # every input file will have a TOC entry,
@@ -286,13 +289,13 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir=
             html_tree = xslt(input_xml, res('xsltTitle.xsl'))
             chars = used_chars(html_tree.getroot())
             zip.writestr('OPS/title.html',
-                 etree.tostring(html_tree, pretty_print=True))
+                 etree.tostring(html_tree, method="html", pretty_print=True))
         elif children:
             # write title page for every parent
             html_tree = xslt(input_xml, res('xsltChunkTitle.xsl'))
             chars = used_chars(html_tree.getroot())
             zip.writestr('OPS/part%d.html' % chunk_counter, 
-                etree.tostring(html_tree, pretty_print=True))
+                etree.tostring(html_tree, method="html", pretty_print=True))
             add_to_manifest(manifest, chunk_counter)
             add_to_spine(spine, chunk_counter)
             chunk_counter += 1
@@ -307,8 +310,6 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir=
                 main_text = None
 
         if main_text is not None:
-            replace_characters(main_text)
-
             for chunk_xml in chop(main_text):
                 chunk_html, chunk_toc, chunk_chars = transform_chunk(chunk_xml, chunk_counter, annotations)
                 toc.extend(chunk_toc)
@@ -409,7 +410,7 @@ def transform(provider, slug=None, file_path=None, output_file=None, output_dir=
         html_tree = xslt(annotations, res("xsltAnnotations.xsl"))
         chars = chars.union(used_chars(html_tree.getroot()))
         zip.writestr('OPS/annotations.html', etree.tostring(
-                            html_tree, pretty_print=True))
+                            html_tree, method="html", pretty_print=True))
 
     # strip fonts
     tmpdir = mkdtemp('-librarian-epub')