Refactor the OCR app API

This refactor adds two new endpoints to view the OCR
content of versions and documents.

Signed-off-by: Roberto Rosario <Roberto.Rosario.Gonzalez@mayan-edms.com>
This commit is contained in:
Roberto Rosario
2019-02-19 03:21:36 -04:00
parent e03f017e7f
commit c2dd01d51e
7 changed files with 327 additions and 175 deletions

View File

@@ -2,95 +2,104 @@ from __future__ import absolute_import, unicode_literals
from django.shortcuts import get_object_or_404
from rest_framework import generics, status
from rest_framework import status
from rest_framework.decorators import action
from rest_framework.response import Response
from mayan.apps.documents.models import Document, DocumentVersion
from mayan.apps.rest_api.permissions import MayanPermission
from mayan.apps.rest_api.viewsets import MayanAPIViewSet
from .models import DocumentPageOCRContent
from .permissions import permission_ocr_content_view, permission_ocr_document
from .serializers import DocumentPageOCRContentSerializer
from .serializers import (
DocumentOCRSerializer, DocumentPageOCRContentSerializer,
DocumentVersionOCRSerializer
)
class APIDocumentOCRView(generics.GenericAPIView):
"""
post: Submit a document for OCR.
"""
mayan_object_permissions = {
'POST': (permission_ocr_document,)
class DocumentOCRAPIViewSet(MayanAPIViewSet):
lookup_url_kwarg = 'document_id'
object_permission_map = {
'ocr_content': permission_ocr_content_view,
'ocr_submit': permission_ocr_document,
}
permission_classes = (MayanPermission,)
queryset = Document.objects.all()
serializer_class = DocumentOCRSerializer
def get_serializer(self, *args, **kwargs):
return None
def get_serializer_class(self):
return None
def post(self, request, *args, **kwargs):
self.get_object().submit_for_ocr()
return Response(status=status.HTTP_202_ACCEPTED)
class APIDocumentVersionOCRView(generics.GenericAPIView):
"""
post: Submit a document version for OCR.
"""
lookup_url_kwarg = 'document_version_pk'
mayan_object_permissions = {
'POST': (permission_ocr_document,)
}
permission_classes = (MayanPermission,)
queryset = DocumentVersion.objects.all()
def get_document(self):
return get_object_or_404(klass=Document, pk=self.kwargs['document_pk'])
def get_queryset(self):
return self.get_document().versions.all()
def get_serializer(self, *args, **kwargs):
return None
def get_serializer_class(self):
return None
def post(self, request, *args, **kwargs):
self.get_object().submit_for_ocr()
return Response(status=status.HTTP_202_ACCEPTED)
class APIDocumentPageOCRContentView(generics.RetrieveAPIView):
"""
get: Returns the OCR content of the selected document page.
"""
lookup_url_kwarg = 'document_page_pk'
mayan_object_permissions = {
'GET': (permission_ocr_content_view,),
}
permission_classes = (MayanPermission,)
serializer_class = DocumentPageOCRContentSerializer
def get_document(self):
return get_object_or_404(klass=Document, pk=self.kwargs['document_pk'])
def get_document_version(self):
return get_object_or_404(
klass=self.get_document().versions.all(), pk=self.kwargs['document_version_pk']
@action(
detail=True, url_name='ocr-content', url_path='ocr'
)
def ocr_content(self, request, *args, **kwargs):
instance = self.get_object()
serializer = self.get_serializer(instance)
headers = self.get_success_headers(data=serializer.data)
return Response(
serializer.data, status=status.HTTP_200_OK, headers=headers
)
def get_queryset(self):
return self.get_document_version().pages.all()
def retrieve(self, request, *args, **kwargs):
@action(
detail=True, methods=('post',), url_name='ocr-submit',
url_path='ocr/submit'
)
def ocr_submit(self, request, *args, **kwargs):
instance = self.get_object()
instance.submit_for_ocr(_user=request.user)
return Response(
data=None, status=status.HTTP_202_ACCEPTED
)
try:
ocr_content = instance.ocr_content
except DocumentPageOCRContent.DoesNotExist:
ocr_content = DocumentPageOCRContent.objects.none()
serializer = self.get_serializer(ocr_content)
return Response(serializer.data)
class DocumentVersionOCRAPIViewSet(MayanAPIViewSet):
lookup_url_kwarg = 'document_version_id'
object_permission_map = {
'ocr_content': permission_ocr_content_view,
'ocr_submit': permission_ocr_document,
}
queryset = DocumentVersion.objects.all()
serializer_class = DocumentVersionOCRSerializer
@action(
detail=True, url_name='ocr-content', url_path='ocr'
)
def ocr_content(self, request, *args, **kwargs):
instance = self.get_object()
serializer = self.get_serializer(instance)
headers = self.get_success_headers(data=serializer.data)
return Response(
serializer.data, status=status.HTTP_200_OK, headers=headers
)
@action(
detail=True, methods=('post',), url_name='ocr-submit',
url_path='ocr/submit'
)
def ocr_submit(self, request, *args, **kwargs):
instance = self.get_object()
instance.submit_for_ocr(_user=request.user)
return Response(
data=None, status=status.HTTP_202_ACCEPTED
)
class DocumentPageOCRAPIViewSet(MayanAPIViewSet):
lookup_url_kwarg = 'document_page_id'
object_permission_map = {
'ocr_content': permission_ocr_content_view,
}
serializer_class = DocumentPageOCRContentSerializer
def get_queryset(self):
return get_object_or_404(
klass=DocumentVersion, document_id=self.kwargs['document_id'],
pk=self.kwargs['document_version_id']
).pages.all()
@action(
detail=True, url_name='content', url_path='ocr'
)
def ocr_content(self, request, *args, **kwargs):
instance = self.get_object()
serializer = self.get_serializer(instance)
headers = self.get_success_headers(data=serializer.data)
return Response(
serializer.data, status=status.HTTP_200_OK, headers=headers
)