X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/f8e5f031c04122d65d1066077be3920ae95518ae..f8442984a021793164661a54ec84ae9caaf9862b:/librarian/html.py?ds=sidebyside diff --git a/librarian/html.py b/librarian/html.py index 6551995..5974d93 100644 --- a/librarian/html.py +++ b/librarian/html.py @@ -1,4 +1,8 @@ # -*- coding: utf-8 -*- +# +# This file is part of Librarian, licensed under GNU Affero GPLv3 or later. +# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. +# import os import cStringIO import re @@ -7,41 +11,33 @@ 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': 'book2html.xslt', - 'full': 'wl2html_full.xslt', - 'partial': 'wl2html_partial.xslt' + 'legacy': 'xslt/book2html.xslt', + 'full': 'xslt/wl2html_full.xslt', + 'partial': 'xslt/wl2html_partial.xslt' } 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={}): - """Transforms file input_filename in XML to output_filename in XHTML.""" + 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, + otherwise returns True if file has been written,False if it hasn't. + File won't be written if it has no content. + """ # Parse XSLT try: style_filename = get_stylesheet(stylesheet) @@ -54,21 +50,29 @@ 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 :) - - if etree.ETXPath('//p|//{%s}p' % str(XHTMLNS))(result) is not None: + del document # no longer needed large object :) + + if html_has_content(result): add_anchors(result.getroot()) 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 else: - print "[Librarian] didn't find any paragraphs" - return "" + if output_filename is not None: + return False + else: + return "" except KeyError: raise ValueError("'%s' is not a valid stylesheet.") except (XMLSyntaxError, XSLTApplyError), e: @@ -123,7 +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 @@ -185,7 +195,7 @@ def add_anchor(element, prefix, with_link=True, with_target=True, link_text=None anchor.tail = element.text element.text = u'' element.insert(0, anchor) - + if with_target: anchor_target = etree.Element('a', name='%s' % prefix) anchor_target.set('class', 'target') @@ -207,9 +217,10 @@ 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 - + if element.tag == 'p' and 'verse' in element.get('class', ''): if counter == 1 or counter % 5 == 0: add_anchor(element, "f%d" % counter, link_text=counter) @@ -226,14 +237,14 @@ def add_table_of_contents(root): if element.tag in ('h2', 'h3'): if any_ancestor(element, lambda e: e.get('id') in ('footnotes',) or e.get('class') in ('person-list',)): continue - + if element.tag == 'h3' and len(sections) and sections[-1][1] == 'h2': sections[-1][3].append((counter, element.tag, ''.join(element.xpath('text()')), [])) else: sections.append((counter, element.tag, ''.join(element.xpath('text()')), [])) add_anchor(element, "s%d" % counter, with_link=False) counter += 1 - + toc = etree.Element('div') toc.set('id', 'toc') toc_header = etree.SubElement(toc, 'h2') @@ -243,12 +254,26 @@ def add_table_of_contents(root): for n, section, text, subsections in sections: section_element = etree.SubElement(toc_list, 'li') add_anchor(section_element, "s%d" % n, with_target=False, link_text=text) - + if len(subsections): subsection_list = etree.SubElement(section_element, 'ol') for n, subsection, text, _ in subsections: subsection_element = etree.SubElement(subsection_list, 'li') add_anchor(subsection_element, "s%d" % n, with_target=False, link_text=text) - + root.insert(0, toc) + +def extract_annotations(html_path): + """For each annotation, yields a tuple: anchor, text, html.""" + parser = etree.HTMLParser(encoding='utf-8') + tree = etree.parse(html_path, parser) + footnotes = tree.find('//*[@id="footnotes"]') + if footnotes is not None: + for footnote in footnotes.findall('div'): + anchor = footnote.find('a[@name]').get('name') + 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 +