accept missing dc tags

[librarian.git] / librarian / html.py

diff --git a/librarian/html.py b/librarian/html.py
index 4edbf33..848935a 100644 (file)
--- a/librarian/html.py
+++ b/librarian/html.py
@@ -1,61 +1,75 @@
 # -*- coding: utf-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 os
 import cStringIO

-import re

 import copy
 
 from lxml import etree
 import copy
 
 from lxml import etree

-from librarian.parser import WLDocument
+from librarian import XHTMLNS, ParseError, IOFile
+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': 'xslt/book2html.xslt',
+    'full': 'xslt/wl2html_full.xslt',
+    'partial': 'xslt/wl2html_partial.xslt'
+}

 
 

-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

 
 

+def get_stylesheet(name):
+    return os.path.join(os.path.dirname(__file__), STYLESHEETS[name])

 
 

-# 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):
-    """Transforms file input_filename in XML to output_filename in XHTML."""
+
+def transform(wldoc, stylesheet='legacy', options=None, flags=None):
+    """Transforms the WL document to 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
     # Parse XSLT

-    style_filename = os.path.join(os.path.dirname(__file__), 'book2html.xslt')
-    style = etree.parse(style_filename)
-
-    if is_file:
-        document = WLDocument.from_file(input, True)
-    else:
-        document = WLDocument.from_string(input, True)
-
-    result = document.transform(style)
-    del document # no longer needed large object :)
-
-    if result.find('//p') is not None:
-        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')
+    try:
+        style_filename = get_stylesheet(stylesheet)
+        style = etree.parse(style_filename)
+
+        document = copy.deepcopy(wldoc)
+        del wldoc
+        document.swap_endlines()
+
+        if flags:
+            for flag in flags:
+                document.edoc.getroot().set(flag, 'yes')
+
+        document.clean_ed_note()
+
+        if not options:
+            options = {}
+        result = document.transform(style, **options)
+        del document  # no longer needed large object :)
+
+        if html_has_content(result):
+            add_anchors(result.getroot())
+            add_table_of_contents(result.getroot())
+
+            return IOFile.from_string(
+                etree.tostring(result, method='html', xml_declaration=False, pretty_print=True, encoding='utf-8'))

         else:
         else:

-            return result
-        return True
-    else:
-        return False
+            return None
+    except KeyError:
+        raise ValueError("'%s' is not a valid stylesheet.")
+    except (XMLSyntaxError, XSLTApplyError), e:
+        raise ParseError(e)

 
 
 class Fragment(object):
 
 
 class Fragment(object):


@@ -86,7 +100,8 @@ class Fragment(object):
         result = []
         for event, element in self.closed_events():
             if event == 'start':
         result = []
         for event, element in self.closed_events():
             if event == 'start':

-                result.append(u'<%s %s>' % (element.tag, ' '.join('%s="%s"' % (k, v) for k, v in element.attrib.items())))
+                result.append(u'<%s %s>' % (
+                    element.tag, ' '.join('%s="%s"' % (k, v) for k, v in element.attrib.items())))

                 if element.text:
                     result.append(element.text)
             elif event == 'end':
                 if element.text:
                     result.append(element.text)
             elif event == 'end':


@@ -107,10 +122,17 @@ def extract_fragments(input_filename):
     open_fragments = {}
     closed_fragments = {}
 
     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'):
         # 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
+            if not event == 'end':
+                continue  # Process elements only once, on end event

 
             # Open new fragment
             if element.get('class', '') == 'theme-begin':
 
             # Open new fragment
             if element.get('class', '') == 'theme-begin':


@@ -143,11 +165,11 @@ def extract_fragments(input_filename):
                 for fragment_id in open_fragments:
                     open_fragments[fragment_id].append('text', element.tail)
 
                 for fragment_id in open_fragments:
                     open_fragments[fragment_id].append('text', element.tail)
 

-

         # Process all elements except begin and end
         else:
             # Omit annotation tags
         # Process all elements except begin and end
         else:
             # Omit annotation tags

-            if len(element.get('name', '')) or element.get('class', '') == 'annotation':
+            if (len(element.get('name', '')) or
+                    element.get('class', '') in ('annotation', 'anchor')):

                 if event == 'end' and element.tail:
                     for fragment_id in open_fragments:
                         open_fragments[fragment_id].append('text', element.tail)
                 if event == 'end' and element.tail:
                     for fragment_id in open_fragments:
                         open_fragments[fragment_id].append('text', element.tail)


@@ -169,7 +191,7 @@ 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)
             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')
     if with_target:
         anchor_target = etree.Element('a', name='%s' % prefix)
         anchor_target.set('class', 'target')


@@ -189,11 +211,15 @@ def any_ancestor(element, test):
 
 def add_anchors(root):
     counter = 1
 
 def add_anchors(root):
     counter = 1

+
+    def is_side_text(e):
+        side_classes = ('note', 'motto', 'motto_podpis', 'dedication')
+        return e.get('class') in side_classes or e.get('id') == 'nota_red' or e.tag == 'blockquote'
+

     for element in root.iterdescendants():
     for element in root.iterdescendants():

-        if any_ancestor(element, lambda e: e.get('class') in ('note', 'motto', 'motto_podpis', 'dedication')
-        or e.tag == 'blockquote'):
+        if any_ancestor(element, is_side_text):

             continue
             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)
         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)


@@ -203,21 +229,34 @@ def add_anchors(root):
             counter += 1
 
 
             counter += 1
 
 

+def raw_printable_text(element):
+    working = copy.deepcopy(element)
+    for e in working.findall('a'):
+        if e.get('class') == 'annotation':
+            e.text = ''
+    return etree.tostring(working, method='text', encoding=unicode).strip()
+
+

 def add_table_of_contents(root):
     sections = []
     counter = 1
 def add_table_of_contents(root):
     sections = []
     counter = 1

+
+    def is_side_text(e):
+        return e.get('id') in ('footnotes', 'nota_red') or e.get('class') == 'person-list'
+

     for element in root.iterdescendants():
         if element.tag in ('h2', 'h3'):
     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, is_side_text):

                 continue
                 continue

-            
+
+            element_text = raw_printable_text(element)

             if element.tag == 'h3' and len(sections) and sections[-1][1] == 'h2':
             if element.tag == 'h3' and len(sections) and sections[-1][1] == 'h2':

-                sections[-1][3].append((counter, element.tag, ''.join(element.xpath('text()')), []))
+                sections[-1][3].append((counter, element.tag, element_text, []))

             else:
             else:

-                sections.append((counter, element.tag, ''.join(element.xpath('text()')), []))
+                sections.append((counter, element.tag, element_text, []))

             add_anchor(element, "s%d" % counter, with_link=False)
             counter += 1
             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')
     toc = etree.Element('div')
     toc.set('id', 'toc')
     toc_header = etree.SubElement(toc, 'h2')


@@ -227,12 +266,25 @@ 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)
     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')
         if len(subsections):
             subsection_list = etree.SubElement(section_element, 'ol')

-            for n, subsection, text, _ in subsections:
+            for n1, subsection, text1, _ in subsections:

                 subsection_element = etree.SubElement(subsection_list, 'li')
                 subsection_element = etree.SubElement(subsection_list, 'li')

-                add_anchor(subsection_element, "s%d" % n, with_target=False, link_text=text)
-    
+                add_anchor(subsection_element, "s%d" % n1, with_target=False, link_text=text1)
+

     root.insert(0, toc)
 
     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