Updated docker settings, removed unnecessary package.

Author: vladd-bit

config.py

@@ -9,7 +9,7 @@
 TMP_FILE_DIR = os.path.join(ROOT_DIR, "tmp")
 
 # Should we actually ocr or just extract text from PDFs ? NOTE: OCR IS STILL APPLIED TO IMAGES if detected | possible vals : "OCR", "NO_OCR"
-OPERATION_MODE= os.environ.get("OCR_SERVICE_OPERATION_MODE", "NO_OCR")
+OPERATION_MODE= os.environ.get("OCR_SERVICE_OPERATION_MODE", "OCR")
 
 # basic app settings
 OCR_SERVICE_PORT = os.environ.get("OCR_SERVICE_PORT", 8090)

docker-compose.yml

@@ -22,10 +22,10 @@ services:
       - OMP_THREAD_LIMIT=1
 
       # ocr service variables
-      - OCR_SERVICE_THREADS=1
       - OCR_SERVICE_LOG_LEVEL=20
-      - OCR_SERVICE_CPU_THREADS=1
-      - OCR_SERVICE_CONVERTER_THREADS=1
+      - OCR_SERVICE_CPU_THREADS=4
+      - OCR_SERVICE_CONVERTER_THREADS=4
+      - OCR_SERVICE_OPERATION_MODE=OCR
     ulimits:
       memlock:
         soft: -1

ocr_service/processor/processor.py

@@ -7,7 +7,6 @@
 import time
 import traceback
 import uuid
-import pikepdf
 
 from subprocess import PIPE, Popen
 from io import BytesIO