complete document ingestion pipeline

ocr.py

@@ -0,0 +1,88 @@
+from flask_restx import Namespace, Resource, fields
+from flask import request, jsonify, current_app as app, send_file
+from ...services.ocr import OCRService
+from ...utils.decorators.auth import protected_route
+from .models.errors import error_404, error_500
+from .models.response import response
+import json
+import os
+import numpy as np
+from werkzeug.datastructures import FileStorage
+
+api = Namespace('OCR',
+                description='Description',
+                path='/v2/api/tools/ocr')
+
+upload_parser = api.parser()
+upload_parser.add_argument('file', location='files',
+                           type=FileStorage, required=True)
+# OCR Data Model
+ocr_model = api.model('OCR', {
+    'format': fields.String(required=True),
+    'data': fields.String(required=False),
+
+})
+
+success_response = api.clone('OCR Model Response', response, {
+    'model': fields.Nested(ocr_model)
+})
+
+
+@api.route('')
+@api.doc(security='apikey')
+class OCRResource(Resource):
+    @api.doc('get_text')
+    @api.expect(upload_parser)
+    @protected_route
+    def post(self):
+        output_format = request.get_json().get('format')
+
+        if not output_format:
+            output_format = 'txt'
+
+        args = upload_parser.parse_args()
+
+        try:
+            if 'file' not in args:
+                raise ValueError("Invalid file")
+
+            file = args['file']
+
+            if file.filename == '':
+                raise ValueError("Invalid file")
+
+            if file:
+                # Save the uploaded file to the UPLOAD_FOLDER
+                filename = os.path.join(file.filename)
+                file.save(filename)
+                ocr_service = OCRService(image_directory='',
+                                         export_directory=os.path.join(app.config['UPLOAD_FOLDER']),
+                                         language='en')
+            output_format = 'text'
+            result = ocr_service.read_text(filename, output_format=output_format)
+            if output_format == 'text':
+                txt_file, message = result
+                if txt_file:
+
+                    return send_file(os.path.join("..",'..','..', txt_file), mimetype='text/plain', as_attachment=True, download_name=txt_file)
+                else:
+                    return jsonify(error=message)
+
+            elif output_format == 'json':
+                json_data, message = result
+                if json_data:
+                    result_json = json.dumps(json_data, default=np_encoder)
+                    return {'model': {
+                        'format': output_format,
+                        'data': result_json
+                    }}
+            api.abort(code=500, message="Invalid Format", error=True)
+
+
+        except Exception as e:
+            api.abort(code=500, message="Something went wrong", error=True)
+
+
+def np_encoder(object):
+    if isinstance(object, np.generic):
+        return object.item()