Upload 2 files

app.py

@@ -45,8 +45,6 @@ os.environ['SERPAPI_API_KEY']='2e3312eee2ce5e3d9d97989c87999ae121765134a5f713b91
 from langchain_community.utilities import SerpAPIWrapper
 # search = SerpAPIWrapper()
 
-import trafilatura
-
 # from google.colab import userdata
 # groq_api_key = userdata.get('GROQ_API_KEY')
 # api_key = userdata.get('twelvelabs')
@@ -235,67 +233,31 @@ from langchain_community.document_loaders import WebBaseLoader
 #     return docs
 
 
-# def get_docs(destination):
-    
-#     tool = TavilySearchResults(max_results=5)
-#     resp =tool.invoke({"query": f"Tell me about tourist destination place called: {destination}"})
-#     urls = [result['url'] for result in resp]
-#     print("urls ",urls)
-    
-
-#     essay = llamaver.invoke(f"You are a great travel advisor. Write an insightful and helpful essay for a traveller who plans to visit {destination}")
-#     essaydoc = Document(page_content=essay.content)
-
-#     cont = []
-#     cont.append(essaydoc)
-    
-#     for url in urls:
-#       try:
-#         downloaded = trafilatura.fetch_url(url)
-#         text = trafilatura.extract(downloaded) #, config=trafilatura_config)
-#       except Exception:
-#         pass
-#       txt = Document(page_content = text)
-#       cont.append(txt)
-    
-#     # docs=[]
-#     # docs.append(essaydoc)
-#     # for url in urls:
-#     #   if "youtube.com" not in url:
-#     #     loader = WebBaseLoader(url)
-#     #     try:
-#     #         docs1 = loader.load()
-#     #         docs.extend(docs1)
-#     #     except Exception:
-#     #         pass
-    
-#     print("docs created")
-#     return cont
-
-
-def get_docs(destination,prompt):
+def get_docs(destination):
     
     tool = TavilySearchResults(max_results=5)
-    resp =tool.invoke({"query": f"{prompt}"})
+    resp =tool.invoke({"query": f"Tell me about tourist destination place called: {destination}"})
     urls = [result['url'] for result in resp]
     print("urls ",urls)
     
 
-    # essay = llamaver.invoke(f"You are a great travel advisor. Write an insightful and helpful essay for a traveller who plans to visit {destination}")
-    # essaydoc = Document(page_content=essay.content)
+    essay = llamaver.invoke(f"You are a great travel advisor. Write an insightful and helpful essay for a traveller who plans to visit {destination}")
+    essaydoc = Document(page_content=essay.content)
 
-    cont = []
-    # cont.append(essaydoc)
     
+    docs=[]
+    docs.append(essaydoc)
     for url in urls:
-      try:
-        downloaded = trafilatura.fetch_url(url)
-        text = trafilatura.extract(downloaded) #, config=trafilatura_config)
-        txt = Document(page_content = text)
-        cont.append(txt)
-      except Exception:
-        pass 
-    return cont
+      if "youtube.com" not in url:
+        loader = WebBaseLoader(url)
+        try:
+            docs1 = loader.load()
+            docs.extend(docs1)
+        except Exception:
+            pass
+    print("docs created")
+    return docs
+
 
 from langchain import hub
 from langchain.chains.combine_documents.stuff import StuffDocumentsChain
@@ -307,7 +269,7 @@ from langchain_text_splitters import CharacterTextSplitter
 
 from duckduckgo_search import DDGS
 
-def get_blogsummary(docs,llm):
+def get_blogsummary(docs):
 
     # Map
     map_template = """The following is a set of documents. \
@@ -442,17 +404,16 @@ def get_transcriptsummary(docs):
 
 
     
-def get_summary(destination,vlogsummary,prompt):      
-     
-    # print("blog docs obtained")
+def get_summary(destination,vlogsummary): 
+    docs = get_docs(destination)
+
+    print("blog docs obtained")
 
-    prompt = f"Tell me about the tourist destination place called {destination}"
-    docs = get_docs(destination, prompt)
     if vlogsummary:
         vlog_doc = Document(page_content=vlogsummary)
         docs.append(vlog_doc)  
         
-    summary = get_blogsummary(docs,llm)  
+    summary = get_blogsummary(docs)  
     
     return summary   
 
@@ -639,12 +600,6 @@ def get_traveladvisory(origin, destination,prompt,llm):
 ##########################################################################################
 ##########################################################################################
 
-def get_info(destination,prompt,llm):   
-    docs = get_docs(destination,prompt)
-    print("get info docs obtained") 
-    info = get_blogsummary(destination,llm)
-    return info
-
 # Functions called by gradio
 
 def travel_advisory(origin, selection, custom_input,days, purpose):      
@@ -708,19 +663,13 @@ def travel_advisory(origin, selection, custom_input,days, purpose):
     image = get_image_url(destination)    
     # print("image extracted")
 
-    # general = get_traveladvisory(origin,destination,general_pr,llm) 
-    # print("general adv obtained") 
+    general = get_traveladvisory(origin,destination,general_pr,llm)
+    print("general adv obtained") 
 
-    general = get_info(destination,general_pr,llm)
-    print("general info obtained")
-    
     places = get_traveladvisory(origin,destination,places_pr,llama70)
     print("interesting places obtained") 
     
-    # weather = get_traveladvisory(origin,destination,weather_pr,llm)
-    # print("weather obtained")
-
-    weather = get_info(destination,weather_pr,llm)
+    weather = get_traveladvisory(origin,destination,weather_pr,llm)
     print("weather obtained")
     
     politics = get_traveladvisory(origin,destination,politics_pr,llmgemma)

requirements.txt

@@ -1,4 +1,3 @@
-trafilatura
 langchain 
 langchain-groq 
 langchainhub