Add OCR on ingest

LICENSE

@@ -66,7 +66,7 @@
    2. Grant of Copyright License. Subject to the terms and conditions of
       this License, each Contributor hereby grants to You a perpetual,
       worldwide, non-exclusive, no-charge, royalty-free, irrevocable
-      copyright license to reproduce, prepare Derivative Works of,
+      copyright license to reproduce, prep Derivative Works of,
       publicly display, publicly perform, sublicense, and distribute the
       Work and such Derivative Works in Source or Object form.
 

MANIFEST.in

@@ -1,4 +1,5 @@
 include LICENSE
 include README.md
 recursive-include readux_ingest_ecds/templates *
+recursive-include readux_ingest_ecds/services *
 prune test*

README.md

@@ -37,6 +37,7 @@ python manage.py migrate readux_ingest_ecds
 | IIIF_RELATED_LINK_MODEL | Model reference, eg. 'iiif.RelatedLink' |
 | IIIF_CANVAS_MODEL | Model reference, eg. 'iiif.Canvas' |
 | IIIF_COLLECTION_MODEL | Model reference, eg. 'iiif.Collection' |
+| IIIF_OCR_MODEL | Model reference, eg. 'iiif.OCR' |
 | INGEST_TMP_DIR | Absolute path where files will be temporarily stored. |
 | INGEST_PROCESSING_DIR | Absolute path where Lambda will look for images. |
 | INGEST_OCR_DIR | Absolute path where OCR files will be preserved. |

readux_ingest_ecds/admin.py

@@ -1,10 +1,7 @@
 import os
 import logging
 from django.contrib import admin
-from django.urls import reverse
-from django.utils.html import format_html
 from django.shortcuts import redirect
-from django_celery_results.models import TaskResult
 from .models import Local
 from .tasks import local_ingest_task_ecds
 
@@ -18,15 +15,15 @@ class LocalAdmin(admin.ModelAdmin):
     def save_model(self, request, obj, form, change):
         LOGGER.info(f'INGEST: Local ingest started by {request.user.username}')
         obj.creator = request.user
-        obj.process()
+        obj.prep()
         super().save_model(request, obj, form, change)
-
-    def response_add(self, request, obj, post_url_continue=None):
-        obj.refresh_from_db()
         if os.environ["DJANGO_ENV"] != 'test': # pragma: no cover
             local_ingest_task_ecds.apply_async(args=[obj.id])
         else:
             local_ingest_task_ecds(obj.id)
+
+    def response_add(self, request, obj, post_url_continue=None):
+        obj.refresh_from_db()
         LOGGER.info(f'INGEST: Local ingest - {obj.id} - added for {obj.manifest.pid}')
         return redirect('/admin/manifests/manifest/{m}/change/'.format(m=obj.manifest.pk))
 

readux_ingest_ecds/helpers.py

@@ -10,6 +10,7 @@ def get_iiif_models():
          'RelatedLink': apps.get_model(settings.IIIF_RELATED_LINK_MODEL),
          'Canvas': apps.get_model(settings.IIIF_CANVAS_MODEL),
          'Collection': apps.get_model(settings.IIIF_COLLECTION_MODEL),
+         'OCR': apps.get_model(settings.IIIF_OCR_MODEL),
       }
    except AppRegistryNotReady:
       return {
@@ -18,4 +19,5 @@ def get_iiif_models():
          'RelatedLink': settings.IIIF_RELATED_LINK_MODEL,
          'Canvas': settings.IIIF_CANVAS_MODEL,
          'Collection': settings.IIIF_COLLECTION_MODEL,
+         'OCR': settings.IIIF_OCR_MODEL,
       }

readux_ingest_ecds/models.py

@@ -4,7 +4,9 @@
 from django.core.files.storage import FileSystemStorage
 from django.db import models
 from django.conf import settings
-from .services import is_image, is_ocr, is_junk, metadata_from_file, create_manifest, move_image_file, move_ocr_file, canvas_dimensions, upload_trigger_file
+from .services.file_services import is_image, is_ocr, is_junk, move_image_file, move_ocr_file, canvas_dimensions, upload_trigger_file
+from .services.iiif_services import create_manifest
+from .services.metadata_services import metadata_from_file
 from .helpers import get_iiif_models
 
 Manifest = get_iiif_models()['Manifest']
@@ -74,12 +76,13 @@ def ocr_directory(self):
     def trigger_file(self):
         return os.path.join(settings.INGEST_TMP_DIR, f'{self.manifest.pid}.txt')
 
-    def process(self):
+    def prep(self):
         """
         Open metadata
         Create manifest
         Unzip bundle
         """
+        LOGGER.info(f'INGEST: Local ingest - preparing new local ingest')
         os.makedirs(settings.INGEST_TMP_DIR, exist_ok=True)
         os.makedirs(settings.INGEST_PROCESSING_DIR, exist_ok=True)
         os.makedirs(settings.INGEST_OCR_DIR, exist_ok=True)