[epub] output improvements - clean xhtml instead of html tag soup

[librarian.git] / librarian / parser.py

diff --git a/librarian/parser.py b/librarian/parser.py
index 55b4e4b..e06c4d7 100644 (file)
--- a/librarian/parser.py
+++ b/librarian/parser.py
@@ -1,25 +1,34 @@
 # -*- coding: utf-8 -*-
 # -*- coding: utf-8 -*-

-from librarian import ValidationError, NoDublinCore,  ParseError
-from librarian import RDFNS, DCNS
+#
+# This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
+# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
+#
+from librarian import ValidationError, NoDublinCore,  ParseError, NoProvider
+from librarian import RDFNS
+from librarian.cover import DefaultEbookCover

 from librarian import dcparser
 
 from xml.parsers.expat import ExpatError
 from lxml import etree
 from lxml.etree import XMLSyntaxError, XSLTApplyError
 
 from librarian import dcparser
 
 from xml.parsers.expat import ExpatError
 from lxml import etree
 from lxml.etree import XMLSyntaxError, XSLTApplyError
 

+import os

 import re
 from StringIO import StringIO
 
 class WLDocument(object):
 import re
 from StringIO import StringIO
 
 class WLDocument(object):

-    LINE_SWAP_EXPR = re.compile(r'/\s', re.MULTILINE | re.UNICODE);
+    LINE_SWAP_EXPR = re.compile(r'/\s', re.MULTILINE | re.UNICODE)
+    provider = None

 
 

-    def __init__(self, edoc, parse_dublincore=True):
+    def __init__(self, edoc, parse_dublincore=True, provider=None, 
+                    strict=False, meta_fallbacks=None):

         self.edoc = edoc
         self.edoc = edoc

+        self.provider = provider

 
         root_elem = edoc.getroot()
 
         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)
 


@@ -28,17 +37,18 @@ 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)
+
+            self.book_info = dcparser.BookInfo.from_element(
+                    self.rdf_elem, fallbacks=meta_fallbacks, strict=strict)

         else:
             self.book_info = None
         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
 
     @classmethod

-    def from_file(cls, xmlfile, swap_endlines=False, parse_dublincore=True):
+    def from_file(cls, xmlfile, *args, **kwargs):

 
         # first, prepare for parsing
         if isinstance(xmlfile, basestring):
 
         # first, prepare for parsing
         if isinstance(xmlfile, basestring):


@@ -53,27 +63,55 @@ 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:
         try:

-            parser = etree.XMLParser(remove_blank_text=True)
-            return cls(etree.parse(StringIO(data), parser), parse_dublincore=parse_dublincore)
-        except (ExpatError, XMLSyntaxError, XSLTApplyError), e:
-            raise ParseError(e)                  
+            parser = etree.XMLParser(remove_blank_text=False)
+            tree = etree.parse(StringIO(data.encode('utf-8')), parser)

 
 

-    def part_as_text(self, path):
-        # convert the path to XPath        
-        print "[L] Retrieving part:", path
+            return cls(tree, *args, **kwargs)
+        except (ExpatError, XMLSyntaxError, XSLTApplyError), e:
+            raise ParseError(e)
+
+    def swap_endlines(self):
+        """Converts line breaks in stanzas into <br/> tags."""
+        # only swap inside stanzas
+        for elem in self.edoc.iter('strofa'):
+            for child in list(elem):
+                if child.tail:
+                    chunks = self.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 = self.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 parts(self):
+        if self.provider is None:
+            raise NoProvider('No document provider supplied.')
+        if self.book_info is None:
+            raise NoDublinCore('No Dublin Core in document.')
+        for part_uri in self.book_info.parts:
+            yield self.from_file(self.provider.by_uri(part_uri),
+                    provider=self.provider)
+
+    def chunk(self, path):
+        # convert the path to XPath
+        expr = self.path_to_xpath(path)
+        elems = self.edoc.xpath(expr)

 
 

-        elems = self.edoc.xpath(self.path_to_xpath(path))
-        print "[L] xpath", elems
-        

         if len(elems) == 0:
         if len(elems) == 0:

-            return None        
-
-        return etree.tostring(elems[0], encoding=unicode, pretty_print=True)
-
+            return None
+        else:
+            return elems[0]

 
     def path_to_xpath(self, path):
         parts = []
 
     def path_to_xpath(self, path):
         parts = []


@@ -84,7 +122,7 @@ class WLDocument(object):
                 parts.append(part)
             else:
                 tag, n = match.groups()
                 parts.append(part)
             else:
                 tag, n = match.groups()

-                parts.append("node()[position() = %d and name() = '%s']" % (int(n), tag) )
+                parts.append("*[%d][name() = '%s']" % (int(n)+1, tag) )

 
         if parts[0] == '.':
             parts[0] = ''
 
         if parts[0] == '.':
             parts[0] = ''


@@ -95,8 +133,9 @@ class WLDocument(object):
         return self.edoc.xslt(stylesheet, **options)
 
     def update_dc(self):
         return self.edoc.xslt(stylesheet, **options)
 
     def update_dc(self):

-        parent = self.rdf_elem.getparent()
-        parent.replace( self.rdf_elem, self.book_info.to_etree(parent) )
+        if self.book_info:
+            parent = self.rdf_elem.getparent()
+            parent.replace( self.rdf_elem, self.book_info.to_etree(parent) )

 
     def serialize(self):
         self.update_dc()
 
     def serialize(self):
         self.update_dc()


@@ -108,10 +147,82 @@ class WLDocument(object):
         for key, data in chunk_dict.iteritems():
             try:
                 xpath = self.path_to_xpath(key)
         for key, data in chunk_dict.iteritems():
             try:
                 xpath = self.path_to_xpath(key)

-                node = self.edoc.xpath(xpath)[0]                
-                repl = etree.fromstring(data)
-                node.getparent().replace(node, repl);
+                node = self.edoc.xpath(xpath)[0]
+                repl = etree.fromstring(u"<%s>%s</%s>" %(node.tag, data, node.tag) )
+                node.getparent().replace(node, repl)

             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
+
+    def editors(self):
+        """Returns a set of all editors for book and its children.
+
+        :returns: set of dcparser.Person objects
+        """
+        if self.book_info is None:
+            raise NoDublinCore('No Dublin Core in document.')
+        persons = set(self.book_info.editors +
+                        self.book_info.technical_editors)
+        for child in self.parts():
+            persons.update(child.editors())
+        if None in persons:
+            persons.remove(None)
+        return persons
+
+    # Converters
+
+    def as_html(self, *args, **kwargs):
+        from librarian import html
+        return html.transform(self, *args, **kwargs)
+
+    def as_text(self, *args, **kwargs):
+        from librarian import text
+        return text.transform(self, *args, **kwargs)
+
+    def as_epub(self, *args, **kwargs):
+        from librarian import epub
+        return epub.transform(self, *args, **kwargs)
+
+    def as_pdf(self, *args, **kwargs):
+        from librarian import pdf
+        return pdf.transform(self, *args, **kwargs)
+
+    def as_mobi(self, *args, **kwargs):
+        from librarian import mobi
+        return mobi.transform(self, *args, **kwargs)
+
+    def as_fb2(self, *args, **kwargs):
+        from librarian import fb2
+        return fb2.transform(self, *args, **kwargs)
+
+    def as_cover(self, cover_class=None, *args, **kwargs):
+        if cover_class is None:
+            cover_class = DefaultEbookCover
+        return cover_class(self.book_info, *args, **kwargs).output_file()
+
+    def save_output_file(self, output_file, output_path=None,
+            output_dir_path=None, make_author_dir=False, ext=None):
+        if output_dir_path:
+            save_path = output_dir_path
+            if make_author_dir:
+                save_path = os.path.join(save_path,
+                        unicode(self.book_info.author).encode('utf-8'))
+            save_path = os.path.join(save_path,
+                                self.book_info.uri.slug)
+            if ext:
+                save_path += '.%s' % ext
+        else:
+            save_path = output_path
+
+        output_file.save_as(save_path)