src/librarian/document.py

   1 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
   2 # Copyright © Fundacja Wolne Lektury. See NOTICE for more information.
   3 #
   4 import gettext
   5 import os
   6 import re
   7 import urllib.request
   8 from lxml import etree
   9 from .parser import parser
  10 from . import dcparser, DCNS, DirDocProvider
  11 from .functions import lang_code_3to2
  12
  13
  14 class WLDocument:
  15     def __init__(self, filename=None, url=None, provider=None):
  16         source = filename or urllib.request.urlopen(url)
  17         tree = etree.parse(source, parser=parser)
  18         self.tree = tree
  19         tree.getroot().document = self
  20         self.base_meta = dcparser.BookInfo({}, {
  21             DCNS('language'): ["pol"],
  22         }, validate_required=False)
  23
  24         self.provider = provider if provider is not None else DirDocProvider('.')
  25
  26         self.tree.getroot().validate()
  27
  28     @property
  29     def meta(self):
  30         # Allow metadata of the master element as document meta.
  31         #master = self.tree.getroot()[-1]
  32         return self.tree.getroot().meta
  33         return master.meta
  34
  35     @property
  36     def children(self):
  37         for part_uri in self.meta.parts or []:
  38             with self.provider.by_slug(part_uri.slug) as f:
  39                 yield type(self)(filename=f, provider=self.provider)
  40
  41     def build(self, builder, base_url=None, **kwargs):
  42         return builder(base_url=base_url).build(self, **kwargs)
  43
  44     def assign_ids(self, existing=None):
  45         # Find all existing IDs.
  46         existing = existing or set()
  47         que = [self.tree.getroot()]
  48         while que:
  49             item = que.pop(0)
  50             try:
  51                 item.normalize_insides()
  52             except AttributeError:
  53                 pass
  54             existing.add(item.attrib.get('id'))
  55             que.extend(item)
  56
  57         i = 1
  58         que = [self.tree.getroot()]
  59         while que:
  60             item = que.pop(0)
  61             que.extend(item)
  62             if item.attrib.get('id'):
  63                 continue
  64             if not getattr(item, 'SHOULD_HAVE_ID', False):
  65                 continue
  66             while f'e{i}' in existing:
  67                 i += 1
  68             item.attrib['id'] = f'e{i}'
  69             i += 1
  70
  71     def _compat_assign_ordered_ids(self):
  72         """
  73         Compatibility: ids in document order, to be roughly compatible with legacy
  74         footnote ids. Just for testing consistency, change to some sane identifiers
  75         at convenience.
  76         """
  77         EXPR = re.compile(r'/\s', re.MULTILINE | re.UNICODE)
  78         def _compat_assign_ordered_ids_in_elem(elem, i):
  79             elem.attrib['_compat_ordered_id'] = str(i)
  80             i += 1
  81             if getattr(elem, 'HTML_CLASS', None) == 'stanza':
  82                 if elem.text:
  83                     i += len(EXPR.split(elem.text)) - 1
  84                 for sub in elem:
  85                     i = _compat_assign_ordered_ids_in_elem(sub, i)
  86                     if sub.tail:
  87                         i += len(EXPR.split(sub.tail)) - 1
  88             else:
  89                 if elem.tag in ('uwaga', 'extra'):
  90                     return i
  91                 for sub in elem:
  92                     i = _compat_assign_ordered_ids_in_elem(sub, i)
  93             return i
  94
  95         _compat_assign_ordered_ids_in_elem(self.tree.getroot(), 4)
  96
  97     def _compat_assign_section_ids(self):
  98         """
  99         Ids in master-section order. These need to be compatible with the
 100         #secN anchors used by WL search results page to link to fragments.
 101         """
 102         def _compat_assigns_section_ids_in_elem(elem, prefix='sec'):
 103             for i, child in enumerate(elem):
 104                 idfier = '{}{}'.format(prefix, i + 1)
 105                 try:
 106                     child.attrib['_compat_section_id'] = idfier
 107                 except:
 108                     pass
 109                 _compat_assigns_section_ids_in_elem(child, idfier + '-')
 110         _compat_assigns_section_ids_in_elem(self.tree.getroot().master)
 111
 112
 113     def editors(self):
 114         persons = set(self.meta.editors
 115                       + self.meta.technical_editors)
 116         for child in self.children:
 117             persons.update(child.editors())
 118         if None in persons:
 119             persons.remove(None)
 120         return persons
 121
 122     def references(self):
 123         return self.tree.findall('.//ref')