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         self.tree.getroot().validate()
  24
  25     @property
  26     def meta(self):
  27         # Allow metadata of the master element as document meta.
  28         #master = self.tree.getroot()[-1]
  29         return self.tree.getroot().meta
  30         return master.meta
  31
  32     @property
  33     def children(self):
  34         for part_uri in self.meta.parts or []:
  35             yield type(self)(
  36                 filename=self.provider.by_uri(part_uri),
  37                 provider=self.provider
  38             )
  39
  40
  41     def build(self, builder, base_url=None, **kwargs):
  42         return builder(base_url=base_url).build(self, **kwargs)
  43
  44     def _compat_assign_ordered_ids(self):
  45         """
  46         Compatibility: ids in document order, to be roughly compatible with legacy
  47         footnote ids. Just for testing consistency, change to some sane identifiers
  48         at convenience.
  49         """
  50         EXPR = re.compile(r'/\s', re.MULTILINE | re.UNICODE)
  51         def _compat_assign_ordered_ids_in_elem(elem, i):
  52             elem.attrib['_compat_ordered_id'] = str(i)
  53             i += 1
  54             if getattr(elem, 'HTML_CLASS', None) == 'stanza':
  55                 if elem.text:
  56                     i += len(EXPR.split(elem.text)) - 1
  57                 for sub in elem:
  58                     i = _compat_assign_ordered_ids_in_elem(sub, i)
  59                     if sub.tail:
  60                         i += len(EXPR.split(sub.tail)) - 1
  61             else:
  62                 if elem.tag in ('uwaga', 'extra'):
  63                     return i
  64                 for sub in elem:
  65                     i = _compat_assign_ordered_ids_in_elem(sub, i)
  66             return i
  67
  68         _compat_assign_ordered_ids_in_elem(self.tree.getroot(), 4)
  69
  70     def _compat_assign_section_ids(self):
  71         """
  72         Ids in master-section order. These need to be compatible with the
  73         #secN anchors used by WL search results page to link to fragments.
  74         """
  75         def _compat_assigns_section_ids_in_elem(elem, prefix='sec'):
  76             for i, child in enumerate(elem):
  77                 idfier = '{}{}'.format(prefix, i + 1)
  78                 child.attrib['_compat_section_id'] = idfier
  79                 _compat_assigns_section_ids_in_elem(child, idfier + '-')
  80         _compat_assigns_section_ids_in_elem(self.tree.getroot().master)
  81
  82
  83     def editors(self):
  84         persons = set(self.meta.editors
  85                       + self.meta.technical_editors)
  86         for child in self.children:
  87             persons.update(child.editors())
  88         if None in persons:
  89             persons.remove(None)
  90         return persons
  91