+import os
+import subprocess
+import uuid
+from django.apps import apps
+from django.conf import settings
+from django.db import models
+from django.urls import reverse
+from django.utils.timezone import now
+from django.utils.translation import gettext_lazy as _
+from . import conversion
+from . import document
+from . import utils
+
+
+class Source(models.Model):
+ name = models.CharField(_('name'), max_length=1024)
+ notes = models.TextField(blank=True, help_text=_('private'))
+ wikisource = models.CharField(max_length=1024, blank=True)
+ modified_at = models.DateTimeField(null=True, blank=True)
+ processed_at = models.DateTimeField(null=True, blank=True)
+
+ def __str__(self):
+ return self.name
+
+ def get_absolute_url(self):
+ return reverse('source', args=[self.pk])
+
+ def touch(self):
+ self.modified_at = now()
+ self.save(update_fields=['modified_at'])
+
+ def get_upload_directory(self):
+ return f'sources/upload/{self.pk}/'
+
+ def get_view_directory(self):
+ return f'sources/view/{self.pk}/'
+
+ def get_ocr_directory(self):
+ return f'sources/ocr/{self.pk}/'
+
+ def get_view_files(self):
+ d = self.get_view_directory()
+ return [
+ d + name
+ for name in sorted(os.listdir(
+ os.path.join(settings.MEDIA_ROOT, d)
+ ))
+ ]
+
+ def get_ocr_files(self):
+ d = os.path.join(settings.MEDIA_ROOT, self.get_ocr_directory())
+ return [
+ d + name
+ for name in sorted(os.listdir(d))
+ ]
+
+ def process(self):
+ updir = os.path.join(
+ settings.MEDIA_ROOT,
+ self.get_upload_directory()
+ )
+ view_dir = os.path.join(
+ settings.MEDIA_ROOT,
+ self.get_view_directory()
+ )
+ ocr_dir = os.path.join(
+ settings.MEDIA_ROOT,
+ self.get_ocr_directory()
+ )
+ with utils.replace_dir(view_dir) as d:
+ self.build_view_directory(updir, d)
+ with utils.replace_dir(ocr_dir) as d:
+ self.build_ocr_directory(updir, d)
+ self.processed_at = now()
+ self.save(update_fields=['processed_at'])
+
+ def build_view_directory(self, srcpath, targetpath):
+ for source_file_name in os.listdir(srcpath):
+ print(source_file_name)
+ src = os.path.join(srcpath, source_file_name)
+ ext = source_file_name.rsplit('.', 1)[-1].lower()
+ if ext in ('png', 'jpg', 'jpeg'):
+ conversion.resize_image(src, targetpath)
+ # cp?
+ # maybe resize
+ elif ext in ('tiff', 'tif'):
+ conversion.convert_image(src, targetpath)
+ elif ext == 'pdf':
+ conversion.convert_pdf(src, targetpath)
+ elif ext == 'djvu':
+ conversion.convert_djvu(src, targetpath)
+ else:
+ pass
+
+ def build_ocr_directory(self, srcpath, targetpath):
+ for source_file_name in os.listdir(srcpath):
+ print(source_file_name)
+ subprocess.run([
+ 'tesseract',
+ os.path.join(srcpath, source_file_name),
+ os.path.join(targetpath, source_file_name),
+ '-l', 'pol'
+ ])
+
+
+class BookSource(models.Model):
+ book = models.ForeignKey('catalogue.Book', models.CASCADE)
+ source = models.ForeignKey(Source, models.CASCADE)
+ page_start = models.IntegerField(null=True, blank=True)
+ page_end = models.IntegerField(null=True, blank=True)
+
+ class Meta:
+ ordering = ('page_start',)
+
+ def __str__(self):
+ return f'{self.source} -> {self.book}'
+
+ def get_absolute_url(self):
+ return reverse('source_book_prepare', args=[self.pk])
+
+ def get_view_files(self):
+ # TODO: won't work for PDFs.
+ files = self.source.get_view_files()
+ if self.page_end:
+ files = files[:self.page_end]
+ if self.page_start:
+ files = files[self.page_start - 1:]
+ return files
+
+ def get_ocr_files(self):
+ # TODO: won't work for PDFs.
+ files = self.source.get_ocr_files()
+ if self.page_end:
+ files = files[:self.page_end]
+ if self.page_start:
+ files = files[self.page_start - 1:]
+ return files
+
+ def get_document(self):
+ return self.book.document_books.first()
+
+ def prepare_document(self, user=None):
+ DBook = apps.get_model('documents', 'Book')
+ texts = document.build_document_texts(self)
+
+ dbook = self.get_document()
+ if dbook is None:
+ dbook = DBook.create(
+ user, texts[0],
+ title=self.book.title,
+ slug=str(uuid.uuid4()),
+ )
+ else:
+ dbook[0].commit(text=texts[0], description='OCR', author=user)
+ for text in texts[1:]:
+ dbook[0].commit(text=text, description='OCR', author=user)
+
+ dbook[0].head.set_publishable(True)
+ return dbook
+
fnp / redakcja.git / commitdiff