b4709575c19cbba634e44273e9ed80c9359d91bc
[librarian.git] / 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, *args, **kwargs):
42         return cls.from_file(StringIO(xml), *args, **kwargs)
43
44     @classmethod
45     def from_file(cls, xmlfile, swap_endlines=False, parse_dublincore=True, preserve_lines=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             sub = u'<br/>'
62             if preserve_lines:
63                 sub += u'\n'
64             data = cls.LINE_SWAP_EXPR.sub(sub, data)
65
66         try:
67             parser = etree.XMLParser(remove_blank_text=False)
68             return cls(etree.parse(StringIO(data), parser), parse_dublincore=parse_dublincore)
69         except (ExpatError, XMLSyntaxError, XSLTApplyError), e:
70             raise ParseError(e)
71
72     def chunk(self, path):
73         # convert the path to XPath
74         expr = self.path_to_xpath(path)
75         elems = self.edoc.xpath(expr)
76
77         if len(elems) == 0:
78             return None
79         else:
80             return elems[0]
81
82     def path_to_xpath(self, path):
83         parts = []
84
85         for part in path.split('/'):
86             match = re.match(r'([^\[]+)\[(\d+)\]', part)
87             if not match:
88                 parts.append(part)
89             else:
90                 tag, n = match.groups()
91                 parts.append("*[%d][name() = '%s']" % (int(n)+1, tag) )
92
93         if parts[0] == '.':
94             parts[0] = ''
95
96         return '/'.join(parts)
97
98     def transform(self, stylesheet, **options):
99         return self.edoc.xslt(stylesheet, **options)
100
101     def update_dc(self):
102         if self.book_info:
103             parent = self.rdf_elem.getparent()
104             parent.replace( self.rdf_elem, self.book_info.to_etree(parent) )
105
106     def serialize(self):
107         self.update_dc()
108         return etree.tostring(self.edoc, encoding=unicode, pretty_print=True)
109
110     def merge_chunks(self, chunk_dict):
111         unmerged = []
112
113         for key, data in chunk_dict.iteritems():
114             try:
115                 xpath = self.path_to_xpath(key)
116                 node = self.edoc.xpath(xpath)[0]
117                 repl = etree.fromstring(u"<%s>%s</%s>" %(node.tag, data, node.tag) )
118                 node.getparent().replace(node, repl);
119             except Exception, e:
120                 unmerged.append( repr( (key, xpath, e) ) )
121
122         return unmerged