initial, incomplete pdf support

[librarian.git] / librarian / html.py

diff --git a/librarian/html.py b/librarian/html.py
index 6551995..3912fbc 100644 (file)
--- 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 +11,32 @@ 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()
 
 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."""
+    """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)
@@ -55,20 +50,22 @@ def transform(input, output_filename=None, is_file=True, \
                 parse_dublincore=parse_dublincore)
 
         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')
             else:
                 return result
             return True
         else:
-            print "[Librarian] didn't find any paragraphs"
-            return "<empty />"
+            if output_filename is not None:
+                return False
+            else:
+                return "<empty />"
     except KeyError:
         raise ValueError("'%s' is not a valid stylesheet.")
     except (XMLSyntaxError, XSLTApplyError), e:
@@ -185,7 +182,7 @@ 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')
@@ -209,7 +206,7 @@ def add_anchors(root):
         if any_ancestor(element, lambda e: e.get('class') in ('note', 'motto', 'motto_podpis', 'dedication')
         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 +223,14 @@ 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 +240,12 @@ 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)