X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/b24b166cc4de6ba7e9b1559717bb5ff6e27bdacd..141733d3db8c11f1eb69a9c0195f07c3c2ed3f8f:/librarian/html.py?ds=inline diff --git a/librarian/html.py b/librarian/html.py index b126638..6115b31 100644 --- a/librarian/html.py +++ b/librarian/html.py @@ -4,6 +4,7 @@ # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. # import os +import re import cStringIO import copy @@ -57,6 +58,7 @@ 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', @@ -133,14 +135,17 @@ def extract_fragments(input_filename): fragment = Fragment(id=element.get('fid'), themes=element.text) # Append parents - if element.getparent().get('id', None) != 'book-text': - parents = [element.getparent()] - while parents[-1].getparent().get('id', None) != 'book-text': - parents.append(parents[-1].getparent()) - - parents.reverse() - for parent in parents: - fragment.append('start', parent) + parent = element.getparent() + parents = [] + while parent.get('id', None) != 'book-text': + cparent = copy.deepcopy(parent) + cparent.text = None + parents.append(cparent) + parent = parent.getparent() + + parents.reverse() + for parent in parents: + fragment.append('start', parent) open_fragments[fragment.id] = fragment @@ -176,25 +181,22 @@ def extract_fragments(input_filename): def add_anchor(element, prefix, with_link=True, with_target=True, link_text=None): + parent = element.getparent() + index = parent.index(element) + if with_link: if link_text is None: link_text = prefix anchor = etree.Element('a', href='#%s' % prefix) anchor.set('class', 'anchor') anchor.text = unicode(link_text) - if element.text: - anchor.tail = element.text - element.text = u'' - element.insert(0, anchor) + parent.insert(index, anchor) if with_target: anchor_target = etree.Element('a', name='%s' % prefix) anchor_target.set('class', 'target') anchor_target.text = u' ' - if element.text: - anchor_target.tail = element.text - element.text = u'' - element.insert(0, anchor_target) + parent.insert(index, anchor_target) def any_ancestor(element, test): @@ -221,16 +223,23 @@ 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') in ('annotation', 'theme-begin'): + 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 = etree.tostring(element, method='text', - encoding=unicode).strip() + 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, element_text, [])) else: @@ -256,17 +265,72 @@ 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.""" + """Extracts annotations from HTML for annotations dictionary. + + For each annotation, yields a tuple of: + anchor, footnote type, valid qualifiers, text, html. + + """ + from .fn_qualifiers import FN_QUALIFIERS + parser = etree.HTMLParser(encoding='utf-8') tree = etree.parse(html_path, parser) footnotes = tree.find('//*[@id="footnotes"]') + re_qualifier = re.compile(ur'[^\u2014]+\s+\(([^\)]+)\)\s+\u2014') if footnotes is not None: for footnote in footnotes.findall('div'): - anchor = footnote.find('a[@name]').get('name') + fn_type = footnote.get('class').split('-')[1] + anchor = footnote.find('a[@class="annotation"]').get('href')[1:] del footnote[:2] - text_str = etree.tostring(footnote, method='text', encoding='utf-8').strip() - html_str = etree.tostring(footnote, method='html', encoding='utf-8') - yield anchor, text_str, html_str + footnote.text = None + if len(footnote) and footnote[-1].tail == '\n': + footnote[-1].tail = None + text_str = etree.tostring(footnote, method='text', encoding=unicode).strip() + html_str = etree.tostring(footnote, method='html', encoding=unicode).strip() + + match = re_qualifier.match(text_str) + if match: + qualifier_str = match.group(1) + qualifiers = [] + for candidate in re.split('[;,]', qualifier_str): + candidate = candidate.strip() + if candidate in FN_QUALIFIERS: + qualifiers.append(candidate) + elif candidate.startswith('z '): + subcandidate = candidate.split()[1] + if subcandidate in FN_QUALIFIERS: + qualifiers.append(subcandidate) + else: + qualifiers = [] + + yield anchor, fn_type, qualifiers, text_str, html_str