Update app.py

app.py

@@ -16,22 +16,32 @@ import socket
 # from send_email_user import send_user_email
 from huggingface_hub import HfApi
 import smtplib
-# api = HfApi()
-# api.upload_folder(
-#     folder_path="/media/pragnakalpl20/Projects/Pragnakalp_projects/gradio_demo/images",
-#     path_in_repo="my-dataset/images",
-#     repo_id="pragnakalp/OCR-image-to-text",
-#     repo_type="dataset",
-#     ignore_patterns="**/logs/*.txt",
-# )
-
-# if not os.path.isdir('images'):
-# os.mkdir('images')
-# print("create folder--->")
-
-HF_TOKEN = os.getenv('HF_TOKEN')
-print("$$$$$$$$$$$$$$$$$$$$$$$$$$$$",type(HF_TOKEN))
-hf_writer = gr.HuggingFaceDatasetSaver(HF_TOKEN,'OCR-image-to-text',True,True)
+
+# HF_TOKEN = os.getenv('HF_TOKEN')
+# print("$$$$$$$$$$$$$$$$$$$$$$$$$$$$",type(HF_TOKEN))
+# hf_writer = gr.HuggingFaceDatasetSaver(HF_TOKEN,'OCR-image-to-text',True,True)
+
+
+DATASET_REPO_URL = "https://huggingface.co/datasets/pragnakalp/OCR-img-to-text"
+DATA_FILENAME = "ocr_data.csv"
+DATA_FILE = os.path.join("ocr_data", DATA_FILENAME)
+HF_TOKEN = os.environ.get("HF_TOKEN")
+DATASET_REPO_ID = "pragnakalp/OCR-img-to-text"
+print("is none?", HF_TOKEN is None)
+try:
+    hf_hub_download(
+        repo_id=DATASET_REPO_ID,
+        filename=DATA_FILENAME,
+        cache_dir=DATA_DIRNAME,
+        force_filename=DATA_FILENAME
+    )
+except:
+    print("file not found")
+
+repo = Repository(
+    local_dir="ocr_data", clone_from=DATASET_REPO_URL, use_auth_token=HF_TOKEN
+)
+
 def get_device_ip_address():
 
     if os.name == "nt":
@@ -55,7 +65,6 @@ def get_device_ip_address():
         return result
 
             
-   
 """
 Paddle OCR
 """
@@ -115,6 +124,15 @@ def generate_ocr(Method,img):
             text_output = ocr_with_keras(img)
         if Method == 'PaddleOCR':
             text_output = ocr_with_paddle(img)
+
+        with open(DATA_FILE, "a") as csvfile:
+            writer = csv.DictWriter(csvfile, fieldnames=["method", "image", "generated_text"])
+            writer.writerow(
+                {"method": Method, "image": img, "generated_text": text_output}
+            )
+        commit_url = repo.push_to_hub()
+        print(commit_url)
+        
         # save_details(Method,text_output,img)
         # sender="[email protected]"
         # password="httscgatatbbxxur"
@@ -197,11 +215,10 @@ demo = gr.Interface(
     [method,image],
     output,
     title="Optical Character Recognition",
-    description="Try OCR with different methods", 
-    theme="darkpeach",
+    description="Try OCR with different methods",
     css=".gradio-container {background-color: lightgray} #radio_div {background-color: #FFD8B4; font-size: 40px;}",
     allow_flagging = "manual",
-    flagging_dir = "flagged",
-	flagging_callback=hf_writer
+ #    flagging_dir = "flagged",
+	# flagging_callback=hf_writer
 )
 demo.launch(enable_queue = False)