added some changes

app.py

@@ -16,7 +16,7 @@ from fastapi import FastAPI, HTTPException, Request
 from transformers import LayoutLMv3Processor, LayoutLMv3ForTokenClassification
 from dotenv import load_dotenv
 import urllib.parse
-from utils import doc_processing
+from utils import doc_processing, extract_document_number_from_file
 
 # Load .env file
 load_dotenv()
@@ -156,6 +156,7 @@ def perform_inference(file_paths: Dict[str, str], upload_to_s3: bool):
         "gst_file": gst_model,
         "cheque_file": cheque_model,
     }
+
     try:
         inference_results = {}
 

folder.py

@@ -0,0 +1,21 @@
+
+import boto3
+# AWS credentials (if not set in environment variables or AWS CLI config)
+from dotenv import load_dotenv
+import os
+
+# Load .env file
+load_dotenv()
+# Access variables
+AWS_ACCESS_KEY_ID = os.getenv("AWS_ACCESS_KEY_ID")
+AWS_SECRET_ACCESS_KEY = os.getenv("AWS_SECRET_ACCESS_KEY")
+print("AWS_ACCESS_KEY_ID",AWS_ACCESS_KEY_ID)
+print("AWS_SECRET_ACCESS_KEY",AWS_SECRET_ACCESS_KEY)
+# Initialize S3 client
+
+s3 = boto3.client('s3')
+bucket_name = "edgekycdocs"
+folder_name = "cinllpindocs"
+
+s3.put_object(Bucket=bucket_name, Key=folder_name)
+print(f"Folder '{folder_name}' created in bucket '{bucket_name}'")

s3_setup.py

@@ -1,44 +1,95 @@
 import boto3
+
 # AWS credentials (if not set in environment variables or AWS CLI config)
 from dotenv import load_dotenv
 import os
 import sys
-
 from utils import doc_processing
+import mimetypes
+from pathlib import Path
+from django.conf import settings
+
+BASE_DIR = Path(__file__).resolve().parent.parent.parent
 
 # Load .env file
 load_dotenv()
 # Access variables
 AWS_ACCESS_KEY_ID = os.getenv("AWS_ACCESS_KEY_ID")
 AWS_SECRET_ACCESS_KEY = os.getenv("AWS_SECRET_ACCESS_KEY")
-print("AWS_ACCESS_KEY_ID",AWS_ACCESS_KEY_ID)
-print("AWS_SECRET_ACCESS_KEY",AWS_SECRET_ACCESS_KEY)
+print("AWS_ACCESS_KEY_ID", AWS_ACCESS_KEY_ID)
+print("AWS_SECRET_ACCESS_KEY", AWS_SECRET_ACCESS_KEY)
 # Initialize S3 client
 
-class s3_client:
+# class s3_client:
+#     def __init__(self):
+#         self.aws_access_key_id = AWS_ACCESS_KEY_ID
+#         self.aws_secret_access_key = AWS_SECRET_ACCESS_KEY
+
+#     def initialize(self):
+#         return boto3.client(
+#                 's3',
+#                 aws_access_key_id=self.aws_access_key_id,
+#                 aws_secret_access_key=self.aws_secret_access_key
+#                             )
+
+#     def upload_file(self,local_file_path, bucket_name,folder_name,file_name):
+#         try:
+#             client = self.initialize()
+#             client.upload_file(local_file_path, bucket_name, f"{folder_name}/{file_name}")
+#             print(f"File uploaded successfully to {bucket_name}/{folder_name}{file_name}")
+#             url = f"https://edgekycdocs.s3.eu-north-1.amazonaws.com/{folder_name}/{file_name}"
+#             print("file url",url)
+#             return {"status": 200, "message":"file uploaded successfully" , "url" : url}
+#         except Exception as e:
+#             print("Error uploading file:", e)
+#             return {"status": 400, "message":e}
+
+
+class S3Client:
     def __init__(self):
-        self.aws_access_key_id = AWS_ACCESS_KEY_ID
-        self.aws_secret_access_key = AWS_SECRET_ACCESS_KEY
-
-    def initialize(self):
-        return boto3.client(
-                's3',
-                aws_access_key_id=self.aws_access_key_id,
-                aws_secret_access_key=self.aws_secret_access_key
-                            )
-    
-    def upload_file(self,local_file_path, bucket_name,folder_name,file_name):
+        self.aws_access_key_id = settings.AWS_ACCESS_KEY_ID
+        self.aws_secret_access_key = settings.AWS_SECRET_ACCESS_KEY
+        self.s3_client = boto3.client(
+            "s3",
+            aws_access_key_id=self.aws_access_key_id,
+            aws_secret_access_key=self.aws_secret_access_key,
+        )
+
+    def upload_file(self, local_file_path, bucket_name, folder_name, file_name):
         try:
-            client = self.initialize()
-            client.upload_file(local_file_path, bucket_name, f"{folder_name}/{file_name}")
-            print(f"File uploaded successfully to {bucket_name}/{folder_name}{file_name}")
-            url = f"https://edgekycdocs.s3.eu-north-1.amazonaws.com/{folder_name}/{file_name}"
-            print("file url",url)
-            return {"status": 200, "message":"file uploaded successfully" , "url" : url}
-        except Exception as e:
-            print("Error uploading file:", e)
-            return {"status": 400, "message":e}
+            file_key = f"{folder_name}/{file_name}"
 
+            # Determine the MIME type dynamically
+            # Determine the correct MIME type
+            content_type, _ = mimetypes.guess_type(local_file_path)
+            if content_type is None:
+                content_type = "application/octet-stream"  # Default fallback
 
+            # Explicitly handle common file types
+            if file_name.lower().endswith(
+                (".pdf", ".jpg", ".jpeg", ".png", ".gif", ".webp")
+            ):
+                content_type = (
+                    mimetypes.guess_type(file_name)[0] or "application/octet-stream"
+                )
 
+            # Upload with correct Content-Type
+            self.s3_client.upload_file(
+                local_file_path,
+                bucket_name,
+                file_key,
+                ExtraArgs={
+                    "ContentDisposition": "inline",  # Ensure inline display
+                    "ContentType": content_type,
+                },
+            )
 
+            file_url = f"https://{bucket_name}.s3.amazonaws.com/{file_key}"
+            print(f"Uploading {file_name} with ContentType: {content_type}")
+            return {
+                "status": 200,
+                "message": "File uploaded successfully",
+                "url": file_url,
+            }
+        except Exception as e:
+            return {"status": 400, "message": str(e)}

utils.py

@@ -7,8 +7,7 @@ import logging
 import shutil
 from fastapi import FastAPI, UploadFile, File, HTTPException
 from google.cloud import vision
-
-# from pdf2image import convert_from_path
+from pdf2image import convert_from_path
 
 
 class doc_processing: