extract annotations: return anchor to annotation, not the word in text

[librarian.git] / librarian / parser.py

diff --git a/librarian/parser.py b/librarian/parser.py
index 727f02e..afc4f1a 100644 (file)
--- a/librarian/parser.py
+++ b/librarian/parser.py
@@ -1,9 +1,7 @@
 # -*- coding: utf-8 -*-
 #
 # -*- coding: utf-8 -*-
 #

-# Copyright © 2008,2009,2010 Fundacja Nowoczesna Polska  
-#

 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.

-# For full license text see COPYING or <http://www.gnu.org/licenses/agpl.html>
+# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.

 #
 from librarian import ValidationError, NoDublinCore,  ParseError
 from librarian import RDFNS
 #
 from librarian import ValidationError, NoDublinCore,  ParseError
 from librarian import RDFNS


@@ -23,9 +21,9 @@ class WLDocument(object):
         self.edoc = edoc
 
         root_elem = edoc.getroot()
         self.edoc = edoc
 
         root_elem = edoc.getroot()

-       
+

         dc_path = './/' + RDFNS('RDF')
         dc_path = './/' + RDFNS('RDF')

-        
+

         if root_elem.tag != 'utwor':
             raise ValidationError("Invalid root element. Found '%s', should be 'utwor'" % root_elem.tag)
 
         if root_elem.tag != 'utwor':
             raise ValidationError("Invalid root element. Found '%s', should be 'utwor'" % root_elem.tag)
 


@@ -34,14 +32,14 @@ class WLDocument(object):
 
             if self.rdf_elem is None:
                 raise NoDublinCore('Document has no DublinCore - which is required.')
 
             if self.rdf_elem is None:
                 raise NoDublinCore('Document has no DublinCore - which is required.')

-            
+

             self.book_info = dcparser.BookInfo.from_element(self.rdf_elem)
         else:
             self.book_info = None
             self.book_info = dcparser.BookInfo.from_element(self.rdf_elem)
         else:
             self.book_info = None

-    
+

     @classmethod
     @classmethod

-    def from_string(cls, xml, swap_endlines=False, parse_dublincore=True):
-        return cls.from_file(StringIO(xml), swap_endlines, parse_dublincore=parse_dublincore)
+    def from_string(cls, xml, *args, **kwargs):
+        return cls.from_file(StringIO(xml), *args, **kwargs)

 
     @classmethod
     def from_file(cls, xmlfile, swap_endlines=False, parse_dublincore=True):
 
     @classmethod
     def from_file(cls, xmlfile, swap_endlines=False, parse_dublincore=True):


@@ -59,23 +57,48 @@ class WLDocument(object):
         if not isinstance(data, unicode):
             data = data.decode('utf-8')
 
         if not isinstance(data, unicode):
             data = data.decode('utf-8')
 

-        if swap_endlines:
-            data = cls.LINE_SWAP_EXPR.sub(u'<br />\n', data)
-    
+        data = data.replace(u'\ufeff', '')
+

         try:
             parser = etree.XMLParser(remove_blank_text=False)
         try:
             parser = etree.XMLParser(remove_blank_text=False)

-            return cls(etree.parse(StringIO(data), parser), parse_dublincore=parse_dublincore)
+            tree = etree.parse(StringIO(data.encode('utf-8')), parser)
+
+            if swap_endlines:
+                cls.swap_endlines(tree)
+
+            return cls(tree, parse_dublincore=parse_dublincore)

         except (ExpatError, XMLSyntaxError, XSLTApplyError), e:
         except (ExpatError, XMLSyntaxError, XSLTApplyError), e:

-            raise ParseError(e)                  
+            raise ParseError(e)
+
+    @classmethod
+    def swap_endlines(cls, tree):
+        # only swap inside stanzas
+        for elem in tree.iter('strofa'):
+            for child in list(elem):
+                if child.tail:
+                    chunks = cls.LINE_SWAP_EXPR.split(child.tail)
+                    ins_index = elem.index(child) + 1
+                    while len(chunks) > 1:
+                        ins = etree.Element('br')
+                        ins.tail = chunks.pop()
+                        elem.insert(ins_index, ins)
+                    child.tail = chunks.pop(0)
+            if elem.text:
+                chunks = cls.LINE_SWAP_EXPR.split(elem.text)
+                while len(chunks) > 1:
+                    ins = etree.Element('br')
+                    ins.tail = chunks.pop()
+                    elem.insert(0, ins)
+                elem.text = chunks.pop(0)

 
     def chunk(self, path):
 
     def chunk(self, path):

-        # convert the path to XPath        
+        # convert the path to XPath

         expr = self.path_to_xpath(path)
         elems = self.edoc.xpath(expr)
 
         if len(elems) == 0:
             return None
         expr = self.path_to_xpath(path)
         elems = self.edoc.xpath(expr)
 
         if len(elems) == 0:
             return None

-        else:        
+        else:

             return elems[0]
 
     def path_to_xpath(self, path):
             return elems[0]
 
     def path_to_xpath(self, path):


@@ -118,4 +141,14 @@ class WLDocument(object):
             except Exception, e:
                 unmerged.append( repr( (key, xpath, e) ) )
 
             except Exception, e:
                 unmerged.append( repr( (key, xpath, e) ) )
 

-        return unmerged
\ No newline at end of file
+        return unmerged
+
+    def clean_ed_note(self):
+        """ deletes forbidden tags from nota_red """
+
+        for node in self.edoc.xpath('|'.join('//nota_red//%s' % tag for tag in
+                    ('pa', 'pe', 'pr', 'pt', 'begin', 'end', 'motyw'))):
+            tail = node.tail
+            node.clear()
+            node.tag = 'span'
+            node.tail = tail