src/librarian/document.py

   1 import gettext
   2 import os
   3 import re
   4 from lxml import etree
   5 import six
   6 from .parser import parser
   7 from . import dcparser, DCNS, DirDocProvider
   8 from .functions import lang_code_3to2
   9
  10
  11 class WLDocument:
  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)
  15         self.tree = tree
  16         tree.getroot().document = self
  17         self.base_meta = dcparser.BookInfo({}, {
  18             DCNS('language'): ["pol"],
  19         }, validate_required=False)
  20
  21         self.provider = provider if provider is not None else DirDocProvider('.')
  22
  23     @property
  24     def meta(self):
  25         # Allow metadata of the master element as document meta.
  26         #master = self.tree.getroot()[-1]
  27         return self.tree.getroot().meta
  28         return master.meta
  29
  30     @property
  31     def children(self):
  32         for part_uri in self.meta.parts or []:
  33             yield type(self)(
  34                 filename=self.provider.by_uri(part_uri),
  35                 provider=self.provider
  36             )
  37
  38
  39     def build(self, builder, base_url=None, **kwargs):
  40         return builder(base_url=base_url).build(self, **kwargs)
  41
  42     def _compat_assign_ordered_ids(self):
  43         """
  44         Compatibility: ids in document order, to be roughly compatible with legacy
  45         footnote ids. Just for testing consistency, change to some sane identifiers
  46         at convenience.
  47         """
  48         EXPR = re.compile(r'/\s', re.MULTILINE | re.UNICODE)
  49         def _compat_assign_ordered_ids_in_elem(elem, i):
  50             elem.attrib['_compat_ordered_id'] = str(i)
  51             i += 1
  52             if getattr(elem, 'HTML_CLASS', None) == 'stanza':
  53                 if elem.text:
  54                     i += len(EXPR.split(elem.text)) - 1
  55                 for sub in elem:
  56                     i = _compat_assign_ordered_ids_in_elem(sub, i)
  57                     if sub.tail:
  58                         i += len(EXPR.split(sub.tail)) - 1
  59             else:
  60                 if elem.tag in ('uwaga', 'extra'):
  61                     return i
  62                 for sub in elem:
  63                     i = _compat_assign_ordered_ids_in_elem(sub, i)
  64             return i
  65
  66         _compat_assign_ordered_ids_in_elem(self.tree.getroot(), 4)
  67
  68     def _compat_assign_section_ids(self):
  69         """
  70         Ids in master-section order. These need to be compatible with the
  71         #secN anchors used by WL search results page to link to fragments.
  72         """
  73         def _compat_assigns_section_ids_in_elem(elem, prefix='sec'):
  74             for i, child in enumerate(elem):
  75                 idfier = '{}{}'.format(prefix, i + 1)
  76                 child.attrib['_compat_section_id'] = idfier
  77                 _compat_assigns_section_ids_in_elem(child, idfier + '-')
  78         _compat_assigns_section_ids_in_elem(self.tree.getroot().master)
  79
  80
  81     def editors(self):
  82         persons = set(self.meta.editors
  83                       + self.meta.technical_editors)
  84         #for child in self.parts():
  85         #    persons.update(child.editors())
  86         #if None in persons:
  87         #    persons.remove(None)
  88         return persons
  89