Add UAT file.

[librarian.git] / librarian / html.py

diff --git a/librarian/html.py b/librarian/html.py
index 1c141f3..096e399 100644 (file)
--- a/librarian/html.py
+++ b/librarian/html.py
@@ -3,17 +3,19 @@
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #

+from __future__ import print_function, unicode_literals
+

 import os
 import os

-import cStringIO

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

-from librarian.parser import WLDocument
-from librarian import XHTMLNS, ParseError
+from librarian import XHTMLNS, ParseError, OutputFile

 from librarian import functions
 
 from lxml.etree import XMLSyntaxError, XSLTApplyError
 from librarian import functions
 
 from lxml.etree import XMLSyntaxError, XSLTApplyError

+import six
+

 
 functions.reg_substitute_entities()
 functions.reg_person_name()
 
 functions.reg_substitute_entities()
 functions.reg_person_name()


@@ -24,15 +26,27 @@ STYLESHEETS = {
     'partial': 'xslt/wl2html_partial.xslt'
 }
 
     'partial': 'xslt/wl2html_partial.xslt'
 }
 

+

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

+

 def html_has_content(text):
     return etree.ETXPath('//p|//{%(ns)s}p|//h1|//{%(ns)s}h1' % {'ns': str(XHTMLNS)})(text)
 
 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={}, flags=None):
-    """Transforms file input_filename in XML to output_filename in XHTML.
+
+def transform_abstrakt(abstrakt_element):
+    style_filename = get_stylesheet('legacy')
+    style = etree.parse(style_filename)
+    xml = etree.tostring(abstrakt_element, encoding='unicode')
+    document = etree.parse(six.StringIO(xml.replace('abstrakt', 'dlugi_cytat')))  # HACK
+    result = document.xslt(style)
+    html = re.sub('<a name="sec[0-9]*"/>', '', etree.tostring(result, encoding='unicode'))
+    return re.sub('</?blockquote[^>]*>', '', html)
+
+
+def transform(wldoc, stylesheet='legacy', options=None, flags=None, css=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.
 
     If output_filename is None, returns an XML,
     otherwise returns True if file has been written,False if it hasn't.


@@ -43,41 +57,42 @@ def transform(input, output_filename=None, is_file=True, \
         style_filename = get_stylesheet(stylesheet)
         style = etree.parse(style_filename)
 
         style_filename = get_stylesheet(stylesheet)
         style = etree.parse(style_filename)
 

-        if is_file:
-            document = WLDocument.from_file(input, True, \
-                parse_dublincore=parse_dublincore)
-        else:
-            document = WLDocument.from_string(input, True, \
-                parse_dublincore=parse_dublincore)
+        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 flags:
             for flag in flags:
                 document.edoc.getroot().set(flag, 'yes')
 
         document.clean_ed_note()

+        document.clean_ed_note('abstrakt')
+
+        if not options:
+            options = {}
+        options.setdefault('gallery', "''")

 
 

-        result = document.transform(style, **options)
-        del document # no longer needed large object :)
+        css = css or 'https://static.wolnelektury.pl/css/compressed/book_text.css'
+        css = "'%s'" % css
+        result = document.transform(style, css=css, **options)
+        del document  # no longer needed large object :)

 
         if html_has_content(result):
             add_anchors(result.getroot())
 
         if html_has_content(result):
             add_anchors(result.getroot())

+            add_table_of_themes(result.getroot())

             add_table_of_contents(result.getroot())
 
             add_table_of_contents(result.getroot())
 

-            if output_filename is not None:
-                result.write(output_filename, method='html', xml_declaration=False, pretty_print=True, encoding='utf-8')
-            else:
-                return result
-            return True
+            return OutputFile.from_bytes(etree.tostring(
+                result, method='html', xml_declaration=False, pretty_print=True, encoding='utf-8'))

         else:
         else:

-            if output_filename is not None:
-                return False
-            else:
-                return "<empty />"
+            return None

     except KeyError:
         raise ValueError("'%s' is not a valid stylesheet.")
     except KeyError:
         raise ValueError("'%s' is not a valid stylesheet.")

-    except (XMLSyntaxError, XSLTApplyError), e:
+    except (XMLSyntaxError, XSLTApplyError) as e:

         raise ParseError(e)
 
         raise ParseError(e)
 

+
+@six.python_2_unicode_compatible

 class Fragment(object):
     def __init__(self, id, themes):
         super(Fragment, self).__init__()
 class Fragment(object):
     def __init__(self, id, themes):
         super(Fragment, self).__init__()


@@ -97,7 +112,7 @@ class Fragment(object):
                 try:
                     stack.pop()
                 except IndexError:
                 try:
                     stack.pop()
                 except IndexError:

-                    print 'CLOSED NON-OPEN TAG:', element
+                    print('CLOSED NON-OPEN TAG:', element)

 
         stack.reverse()
         return self.events + stack
 
         stack.reverse()
         return self.events + stack


@@ -106,7 +121,8 @@ 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':


@@ -118,7 +134,7 @@ class Fragment(object):
 
         return ''.join(result)
 
 
         return ''.join(result)
 

-    def __unicode__(self):
+    def __str__(self):

         return self.to_string()
 
 
         return self.to_string()
 
 


@@ -129,28 +145,32 @@ def extract_fragments(input_filename):
 
     # iterparse would die on a HTML document
     parser = etree.HTMLParser(encoding='utf-8')
 
     # iterparse would die on a HTML document
     parser = etree.HTMLParser(encoding='utf-8')

-    buf = cStringIO.StringIO()
+    buf = six.BytesIO()

     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'):
     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
+            if not event == 'end':
+                continue  # Process elements only once, on end event

 
             # Open new fragment
             if element.get('class', '') == 'theme-begin':
                 fragment = Fragment(id=element.get('fid'), themes=element.text)
 
                 # Append parents
 
             # Open new fragment
             if element.get('class', '') == 'theme-begin':
                 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
 
 
                 open_fragments[fragment.id] = fragment
 


@@ -159,7 +179,7 @@ def extract_fragments(input_filename):
                 try:
                     fragment = open_fragments[element.get('fid')]
                 except KeyError:
                 try:
                     fragment = open_fragments[element.get('fid')]
                 except KeyError:

-                    print '%s:closed not open fragment #%s' % (input_filename, element.get('fid'))
+                    print('%s:closed not open fragment #%s' % (input_filename, element.get('fid')))

                 else:
                     closed_fragments[fragment.id] = fragment
                     del open_fragments[fragment.id]
                 else:
                     closed_fragments[fragment.id] = fragment
                     del open_fragments[fragment.id]


@@ -169,11 +189,11 @@ 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)


@@ -185,25 +205,22 @@ def extract_fragments(input_filename):
 
 
 def add_anchor(element, prefix, with_link=True, with_target=True, link_text=None):
 
 
 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')
     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)
+        anchor.text = six.text_type(link_text)
+        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 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):
 
 
 def any_ancestor(element, test):


@@ -216,9 +233,10 @@ def any_ancestor(element, test):
 def add_anchors(root):
     counter = 1
     for element in root.iterdescendants():
 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'):
+        def f(e):
+            return e.get('class') in ('note', 'motto', 'motto_podpis', 'dedication', 'frame') or \
+                e.get('id') == 'nota_red' or e.tag == 'blockquote'
+        if any_ancestor(element, f):

             continue
 
         if element.tag == 'p' and 'verse' in element.get('class', ''):
             continue
 
         if element.tag == 'p' and 'verse' in element.get('class', ''):


@@ -230,18 +248,28 @@ 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') 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'):
 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
 
                 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
 


@@ -257,21 +285,78 @@ def add_table_of_contents(root):
 
         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, subtext, _ 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=subtext)

 
     root.insert(0, toc)
 
 
     root.insert(0, toc)
 

+    
+def add_table_of_themes(root):
+    try:
+        from sortify import sortify
+    except ImportError:
+        def sortify(x):
+            return 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 = list(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):
 
 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)
     parser = etree.HTMLParser(encoding='utf-8')
     tree = etree.parse(html_path, parser)

-    for footnote in tree.find('//*[@id="footnotes"]').findall('div'):
-        anchor = footnote.find('a[@href]').get('href')
-        del footnote[:2]
-        text_str = etree.tostring(footnote, method='text', encoding='utf-8')
-        html_str = etree.tostring(footnote, method='html', encoding='utf-8')
-        yield anchor, text_str, html_str
+    footnotes = tree.find('//*[@id="footnotes"]')
+    re_qualifier = re.compile(r'[^\u2014]+\s+\(([^\)]+)\)\s+\u2014')
+    if footnotes is not None:
+        for footnote in footnotes.findall('div'):
+            fn_type = footnote.get('class').split('-')[1]
+            anchor = footnote.find('a[@class="annotation"]').get('href')[1:]
+            del footnote[:2]
+            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