src/dictionary/models.py

   1 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
   2 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
   3 #
   4 from django.db import models, transaction
   5 from celery import shared_task
   6 from sortify import sortify
   7 from celery.utils.log import get_task_logger
   8
   9 from catalogue.models import Book
  10
  11 task_logger = get_task_logger(__name__)
  12
  13
  14 class Qualifier(models.Model):
  15     qualifier = models.CharField(max_length=128, db_index=True, unique=True)
  16     name = models.CharField(max_length=255)
  17
  18     class Meta:
  19         ordering = ['qualifier']
  20
  21     def __str__(self):
  22         return self.name or self.qualifier
  23
  24
  25 class Note(models.Model):
  26     """Represents a single annotation from a book."""
  27     html = models.TextField()
  28     sort_key = models.CharField(max_length=128, db_index=True)
  29     fn_type = models.CharField(max_length=10, db_index=True)
  30     qualifiers = models.ManyToManyField(Qualifier)
  31     language = models.CharField(max_length=10, db_index=True)
  32
  33     class Meta:
  34         ordering = ['sort_key']
  35
  36
  37 class NoteSource(models.Model):
  38     """Represents a single annotation from a book."""
  39     note = models.ForeignKey(Note, models.CASCADE)
  40     book = models.ForeignKey(Book, models.CASCADE)
  41     anchor = models.CharField(max_length=64)
  42
  43     class Meta:
  44         ordering = ['book']
  45
  46
  47 @shared_task(ignore_result=True)
  48 def build_notes(book):
  49     if not book.findable:
  50         return
  51     task_logger.info(book.slug)
  52     with transaction.atomic():
  53         book.notesource_set.all().delete()
  54         if book.html_file:
  55             from librarian import html
  56             from librarian.fn_qualifiers import FN_QUALIFIERS
  57
  58             for anchor, fn_type, qualifiers, text_str, html_str in \
  59                     html.extract_annotations(book.html_file.path):
  60                 sort_key = sortify(text_str).strip()[:128]
  61
  62                 language = book.language
  63                 notes = Note.objects.filter(sort_key=sort_key, fn_type=fn_type, language=language, html=html_str)
  64                 if notes:
  65                     note = notes[0]
  66                 else:
  67                     note = Note.objects.create(
  68                         sort_key=sort_key,
  69                         html=html_str,
  70                         fn_type=fn_type,
  71                         language=language
  72                         )
  73                 qualifier_objects = []
  74                 for qualifier in qualifiers:
  75                     obj, created = Qualifier.objects.get_or_create(
  76                         qualifier=qualifier, defaults={
  77                             'name': FN_QUALIFIERS.get(qualifier, '')
  78                         })
  79                     qualifier_objects.append(obj)
  80                 note.qualifiers.set(qualifier_objects)
  81                 note.notesource_set.create(book=book, anchor=anchor)
  82
  83         Note.objects.filter(notesource=None).delete()
  84
  85
  86 def notes_from_book(sender, instance, **kwargs):
  87     if instance.findable:
  88         build_notes.delay(instance)
  89 Book.html_built.connect(notes_from_book)