1 # -*- coding: utf-8 -*-
3 # Copyright © 2008,2009,2010 Fundacja Nowoczesna Polska
5 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
6 # For full license text see COPYING or <http://www.gnu.org/licenses/agpl.html>
13 from lxml import etree
14 from librarian.parser import WLDocument
15 from librarian import XHTMLNS, ParseError
17 from lxml.etree import XMLSyntaxError, XSLTApplyError
19 ENTITY_SUBSTITUTIONS = [
28 'legacy': 'xslt/book2html.xslt',
29 'full': 'xslt/wl2html_full.xslt',
30 'partial': 'xslt/wl2html_partial.xslt'
33 def get_stylesheet(name):
34 return os.path.join(os.path.dirname(__file__), STYLESHEETS[name])
36 def substitute_entities(context, text):
37 """XPath extension function converting all entites in passed text."""
38 if isinstance(text, list):
40 for entity, substitutution in ENTITY_SUBSTITUTIONS:
41 text = text.replace(entity, substitutution)
44 # Register substitute_entities function with lxml
45 ns = etree.FunctionNamespace('http://wolnelektury.pl/functions')
46 ns['substitute_entities'] = substitute_entities
48 def transform(input, output_filename=None, is_file=True, \
49 parse_dublincore=True, stylesheet='legacy', options={}):
50 """Transforms file input_filename in XML to output_filename in XHTML."""
53 style_filename = get_stylesheet(stylesheet)
54 style = etree.parse(style_filename)
57 document = WLDocument.from_file(input, True, \
58 parse_dublincore=parse_dublincore)
60 document = WLDocument.from_string(input, True, \
61 parse_dublincore=parse_dublincore)
63 result = document.transform(style, **options)
64 del document # no longer needed large object :)
66 if etree.ETXPath('//p|//{%s}p' % str(XHTMLNS))(result) is not None:
67 add_anchors(result.getroot())
68 add_table_of_contents(result.getroot())
70 if output_filename is not None:
71 result.write(output_filename, xml_declaration=False, pretty_print=True, encoding='utf-8')
78 raise ValueError("'%s' is not a valid stylesheet.")
79 except (XMLSyntaxError, XSLTApplyError), e:
82 class Fragment(object):
83 def __init__(self, id, themes):
84 super(Fragment, self).__init__()
89 def append(self, event, element):
90 self.events.append((event, element))
92 def closed_events(self):
94 for event, element in self.events:
96 stack.append(('end', element))
101 print 'CLOSED NON-OPEN TAG:', element
104 return self.events + stack
108 for event, element in self.closed_events():
110 result.append(u'<%s %s>' % (element.tag, ' '.join('%s="%s"' % (k, v) for k, v in element.attrib.items())))
112 result.append(element.text)
114 result.append(u'</%s>' % element.tag)
116 result.append(element.tail)
118 result.append(element)
120 return ''.join(result)
122 def __unicode__(self):
123 return self.to_string()
126 def extract_fragments(input_filename):
127 """Extracts theme fragments from input_filename."""
129 closed_fragments = {}
131 for event, element in etree.iterparse(input_filename, events=('start', 'end')):
132 # Process begin and end elements
133 if element.get('class', '') in ('theme-begin', 'theme-end'):
134 if not event == 'end': continue # Process elements only once, on end event
137 if element.get('class', '') == 'theme-begin':
138 fragment = Fragment(id=element.get('fid'), themes=element.text)
141 if element.getparent().get('id', None) != 'book-text':
142 parents = [element.getparent()]
143 while parents[-1].getparent().get('id', None) != 'book-text':
144 parents.append(parents[-1].getparent())
147 for parent in parents:
148 fragment.append('start', parent)
150 open_fragments[fragment.id] = fragment
152 # Close existing fragment
155 fragment = open_fragments[element.get('fid')]
157 print '%s:closed not open fragment #%s' % (input_filename, element.get('fid'))
159 closed_fragments[fragment.id] = fragment
160 del open_fragments[fragment.id]
162 # Append element tail to lost_text (we don't want to lose any text)
164 for fragment_id in open_fragments:
165 open_fragments[fragment_id].append('text', element.tail)
168 # Process all elements except begin and end
170 # Omit annotation tags
171 if len(element.get('name', '')) or element.get('class', '') == 'annotation':
172 if event == 'end' and element.tail:
173 for fragment_id in open_fragments:
174 open_fragments[fragment_id].append('text', element.tail)
176 for fragment_id in open_fragments:
177 open_fragments[fragment_id].append(event, copy.copy(element))
179 return closed_fragments, open_fragments
182 def add_anchor(element, prefix, with_link=True, with_target=True, link_text=None):
184 if link_text is None:
186 anchor = etree.Element('a', href='#%s' % prefix)
187 anchor.set('class', 'anchor')
188 anchor.text = unicode(link_text)
190 anchor.tail = element.text
192 element.insert(0, anchor)
195 anchor_target = etree.Element('a', name='%s' % prefix)
196 anchor_target.set('class', 'target')
197 anchor_target.text = u' '
199 anchor_target.tail = element.text
201 element.insert(0, anchor_target)
204 def any_ancestor(element, test):
205 for ancestor in element.iterancestors():
211 def add_anchors(root):
213 for element in root.iterdescendants():
214 if any_ancestor(element, lambda e: e.get('class') in ('note', 'motto', 'motto_podpis', 'dedication')
215 or e.tag == 'blockquote'):
218 if element.tag == 'p' and 'verse' in element.get('class', ''):
219 if counter == 1 or counter % 5 == 0:
220 add_anchor(element, "f%d" % counter, link_text=counter)
222 elif 'paragraph' in element.get('class', ''):
223 add_anchor(element, "f%d" % counter, link_text=counter)
227 def add_table_of_contents(root):
230 for element in root.iterdescendants():
231 if element.tag in ('h2', 'h3'):
232 if any_ancestor(element, lambda e: e.get('id') in ('footnotes',) or e.get('class') in ('person-list',)):
235 if element.tag == 'h3' and len(sections) and sections[-1][1] == 'h2':
236 sections[-1][3].append((counter, element.tag, ''.join(element.xpath('text()')), []))
238 sections.append((counter, element.tag, ''.join(element.xpath('text()')), []))
239 add_anchor(element, "s%d" % counter, with_link=False)
242 toc = etree.Element('div')
244 toc_header = etree.SubElement(toc, 'h2')
245 toc_header.text = u'Spis treści'
246 toc_list = etree.SubElement(toc, 'ol')
248 for n, section, text, subsections in sections:
249 section_element = etree.SubElement(toc_list, 'li')
250 add_anchor(section_element, "s%d" % n, with_target=False, link_text=text)
253 subsection_list = etree.SubElement(section_element, 'ol')
254 for n, subsection, text, _ in subsections:
255 subsection_element = etree.SubElement(subsection_list, 'li')
256 add_anchor(subsection_element, "s%d" % n, with_target=False, link_text=text)