update files

app.py

@@ -10,12 +10,15 @@ import uvicorn
 app = FastAPI()
 
 # ---------------------------------------------------------------------------------------------------------------------
+import gradio as gr
 import os
 import torch
+import json
 import numpy as np
 import requests
 from rank_bm25 import BM25Okapi
 from bs4 import BeautifulSoup
+from datetime import datetime
 
 from transformers import BartTokenizer, BartForConditionalGeneration
 from transformers import BloomTokenizerFast, BloomForCausalLM, BertTokenizer, BertForSequenceClassification
@@ -39,6 +42,29 @@ os.system("python -m spacy download en_core_web_sm")
 nlp = spacy.load("en_core_web_sm")
 
 # ---------------------------------------------------------------------------------------------------------------------
+# ---------------------------------------------------------------------------
+# load .env
+from utils import create_user_id
+user_id = create_user_id()
+
+from azure.storage.fileshare import ShareServiceClient
+try:
+    from dotenv import load_dotenv
+    load_dotenv()
+except Exception as e:
+    pass
+
+account_url = os.environ["AZURE_ACCOUNT_URL"]
+credential = {
+    "account_key":  os.environ['AZURE_ACCOUNT_KEY'],
+    "account_name": os.environ['AZURE_ACCOUNT_NAME']
+}
+
+file_share_name = "averitec"
+azure_service = ShareServiceClient(account_url=account_url, credential=credential)
+azure_share_client = azure_service.get_share_client(file_share_name)
+
+# ---------- Setting ----------
 # ---------- Load Veracity and Justification prediction model ----------
 LABEL = [
     "Supported",
@@ -330,12 +356,18 @@ def greet_json():
     return {"Hello": "World!"}
 
 
+def log_on_azure(file, logs, azure_share_client):
+    logs = json.dumps(logs)
+    file_client = azure_share_client.get_file_client(file)
+    file_client.upload_file(logs)
+
+
 @app.post("/predict/")
 def fact_checking(item: Item):
-    # claim = item['claim']
-    # source = item['source']
-    claim = item.claim
-    source = item.source
+    claim = item['claim']
+    source = item['source']
+    # claim = item.claim
+    # source = item.source
 
     # Step1: Evidence Retrieval
     if source == "Wikipedia":
@@ -347,9 +379,35 @@ def fact_checking(item: Item):
     verdict_label = veracity_prediction(claim, evidence)
     justification_label = justification_generation(claim, evidence, verdict_label)
 
+    ############################################################
     evidence_list = []
     for evi in evidence:
-        evidence_list.append(evi.metadata["query"])
+        title_str = evi.metadata['title']
+        evi_str = evi.metadata['evidence']
+        url_str = evi.metadata['url']
+        evidence_list.append([title_str, evi_str, url_str])
+
+    try:
+        # Log answer on Azure Blob Storage
+        # IF AZURE_ISSAVE=TRUE, save the logs into the Azure share client.
+        if os.environ["AZURE_ISSAVE"] == "TRUE":
+            timestamp = str(datetime.now().timestamp())
+            # timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+            file = timestamp + ".json"
+            logs = {
+                "user_id": str(user_id),
+                "claim": claim,
+                "sources": source,
+                "evidence": evidence_list,
+                "answer": [verdict_label, justification_label],
+                "time": timestamp,
+            }
+            log_on_azure(file, logs, azure_share_client)
+    except Exception as e:
+        print(f"Error logging on Azure Blob Storage: {e}")
+        raise gr.Error(
+            f"AVeriTeC Error: {str(e)[:100]} - The error has been noted, try another question and if the error remains, you can contact us :)")
+    ##########
 
     return  {"Verdict": verdict_label, "Justification": justification_label, "Evidence": evidence_list}
 

utils.py

@@ -0,0 +1,14 @@
+import numpy as np
+import random
+import string
+import uuid
+from datetime import datetime
+
+
+def create_user_id():
+    """Create user_id
+        str: String to id user
+    """
+    current_date = datetime.now().strftime("%Y-%m-%d-%H-%M-%S")
+    user_id = str(uuid.uuid4())
+    return current_date + '_' +user_id