librarian/parser.py

   1 # -*- coding: utf-8 -*-
   2 #
   3 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
   4 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
   5 #
   6 from librarian import ValidationError, NoDublinCore,  ParseError
   7 from librarian import RDFNS
   8 from librarian import dcparser
   9
  10 from xml.parsers.expat import ExpatError
  11 from lxml import etree
  12 from lxml.etree import XMLSyntaxError, XSLTApplyError
  13
  14 import re
  15 from StringIO import StringIO
  16
  17 class WLDocument(object):
  18     LINE_SWAP_EXPR = re.compile(r'/\s', re.MULTILINE | re.UNICODE);
  19
  20     def __init__(self, edoc, parse_dublincore=True):
  21         self.edoc = edoc
  22
  23         root_elem = edoc.getroot()
  24
  25         dc_path = './/' + RDFNS('RDF')
  26
  27         if root_elem.tag != 'utwor':
  28             raise ValidationError("Invalid root element. Found '%s', should be 'utwor'" % root_elem.tag)
  29
  30         if parse_dublincore:
  31             self.rdf_elem = root_elem.find(dc_path)
  32
  33             if self.rdf_elem is None:
  34                 raise NoDublinCore('Document has no DublinCore - which is required.')
  35
  36             self.book_info = dcparser.BookInfo.from_element(self.rdf_elem)
  37         else:
  38             self.book_info = None
  39
  40     @classmethod
  41     def from_string(cls, xml, swap_endlines=False, parse_dublincore=True):
  42         return cls.from_file(StringIO(xml), swap_endlines, parse_dublincore=parse_dublincore)
  43
  44     @classmethod
  45     def from_file(cls, xmlfile, swap_endlines=False, parse_dublincore=True):
  46
  47         # first, prepare for parsing
  48         if isinstance(xmlfile, basestring):
  49             file = open(xmlfile, 'rb')
  50             try:
  51                 data = file.read()
  52             finally:
  53                 file.close()
  54         else:
  55             data = xmlfile.read()
  56
  57         if not isinstance(data, unicode):
  58             data = data.decode('utf-8')
  59
  60         if swap_endlines:
  61             data = cls.LINE_SWAP_EXPR.sub(u'<br />\n', data)
  62
  63         try:
  64             parser = etree.XMLParser(remove_blank_text=False)
  65             return cls(etree.parse(StringIO(data), parser), parse_dublincore=parse_dublincore)
  66         except (ExpatError, XMLSyntaxError, XSLTApplyError), e:
  67             raise ParseError(e)
  68
  69     def chunk(self, path):
  70         # convert the path to XPath
  71         expr = self.path_to_xpath(path)
  72         elems = self.edoc.xpath(expr)
  73
  74         if len(elems) == 0:
  75             return None
  76         else:
  77             return elems[0]
  78
  79     def path_to_xpath(self, path):
  80         parts = []
  81
  82         for part in path.split('/'):
  83             match = re.match(r'([^\[]+)\[(\d+)\]', part)
  84             if not match:
  85                 parts.append(part)
  86             else:
  87                 tag, n = match.groups()
  88                 parts.append("*[%d][name() = '%s']" % (int(n)+1, tag) )
  89
  90         if parts[0] == '.':
  91             parts[0] = ''
  92
  93         return '/'.join(parts)
  94
  95     def transform(self, stylesheet, **options):
  96         return self.edoc.xslt(stylesheet, **options)
  97
  98     def update_dc(self):
  99         if self.book_info:
 100             parent = self.rdf_elem.getparent()
 101             parent.replace( self.rdf_elem, self.book_info.to_etree(parent) )
 102
 103     def serialize(self):
 104         self.update_dc()
 105         return etree.tostring(self.edoc, encoding=unicode, pretty_print=True)
 106
 107     def merge_chunks(self, chunk_dict):
 108         unmerged = []
 109
 110         for key, data in chunk_dict.iteritems():
 111             try:
 112                 xpath = self.path_to_xpath(key)
 113                 node = self.edoc.xpath(xpath)[0]
 114                 repl = etree.fromstring(u"<%s>%s</%s>" %(node.tag, data, node.tag) )
 115                 node.getparent().replace(node, repl);
 116             except Exception, e:
 117                 unmerged.append( repr( (key, xpath, e) ) )
 118
 119         return unmerged