6 from .parser import parser
7 from . import dcparser, DCNS, DirDocProvider
8 from .functions import lang_code_3to2
12 def __init__(self, filename=None, url=None, provider=None):
13 source = filename or six.moves.urllib.request.urlopen(url)
14 tree = etree.parse(source, parser=parser)
16 tree.getroot().document = self
17 self.base_meta = dcparser.BookInfo({}, {
18 DCNS('language'): ["pol"],
19 }, validate_required=False)
21 self.provider = provider if provider is not None else DirDocProvider('.')
23 self.tree.getroot().validate()
27 # Allow metadata of the master element as document meta.
28 #master = self.tree.getroot()[-1]
29 return self.tree.getroot().meta
34 for part_uri in self.meta.parts or []:
36 filename=self.provider.by_slug(part_uri.slug),
37 provider=self.provider
40 def build(self, builder, base_url=None, **kwargs):
41 return builder(base_url=base_url).build(self, **kwargs)
43 def assign_ids(self, existing=None):
44 # Find all existing IDs.
45 existing = existing or set()
46 que = [self.tree.getroot()]
50 item.normalize_insides()
51 except AttributeError:
53 existing.add(item.attrib.get('id'))
57 que = [self.tree.getroot()]
61 if item.attrib.get('id'):
63 if not getattr(item, 'SHOULD_HAVE_ID', False):
65 while f'e{i}' in existing:
67 item.attrib['id'] = f'e{i}'
70 def _compat_assign_ordered_ids(self):
72 Compatibility: ids in document order, to be roughly compatible with legacy
73 footnote ids. Just for testing consistency, change to some sane identifiers
76 EXPR = re.compile(r'/\s', re.MULTILINE | re.UNICODE)
77 def _compat_assign_ordered_ids_in_elem(elem, i):
78 elem.attrib['_compat_ordered_id'] = str(i)
80 if getattr(elem, 'HTML_CLASS', None) == 'stanza':
82 i += len(EXPR.split(elem.text)) - 1
84 i = _compat_assign_ordered_ids_in_elem(sub, i)
86 i += len(EXPR.split(sub.tail)) - 1
88 if elem.tag in ('uwaga', 'extra'):
91 i = _compat_assign_ordered_ids_in_elem(sub, i)
94 _compat_assign_ordered_ids_in_elem(self.tree.getroot(), 4)
96 def _compat_assign_section_ids(self):
98 Ids in master-section order. These need to be compatible with the
99 #secN anchors used by WL search results page to link to fragments.
101 def _compat_assigns_section_ids_in_elem(elem, prefix='sec'):
102 for i, child in enumerate(elem):
103 idfier = '{}{}'.format(prefix, i + 1)
105 child.attrib['_compat_section_id'] = idfier
108 _compat_assigns_section_ids_in_elem(child, idfier + '-')
109 _compat_assigns_section_ids_in_elem(self.tree.getroot().master)
113 persons = set(self.meta.editors
114 + self.meta.technical_editors)
115 for child in self.children:
116 persons.update(child.editors())
121 def references(self):
122 return self.tree.findall('.//ref')