feat: add audio transcription, youtube search and fileRead

tools.py

@@ -16,7 +16,10 @@ from langchain_tavily import TavilySearch, TavilyExtract
 from langchain_google_genai import ChatGoogleGenerativeAI
 from langchain_community.utilities.wikipedia import WikipediaAPIWrapper
 from langchain_community.tools.wikipedia.tool import WikipediaQueryRun
+from langchain_google_community import SpeechToTextLoader
+from langchain_community.tools import YouTubeSearchTool
 from youtube_transcript_api import YouTubeTranscriptApi
+from langchain_community.tools.file_management.read import ReadFileTool
 
 from basic_agent import print_conversation
 
@@ -111,6 +114,22 @@ def search_and_extract(query: str) -> list[dict]:
     return structured_results
 
 
+youtube_search_api = YouTubeSearchTool()
+
+@tool
+def youtube_search_tool(query: str, number_of_results:int=3) -> list:
+    """Search YouTube for a query and return the top number_of_results."""
+    if CUSTOM_DEBUG:
+        print_tool_call(
+            youtube_search_tool,
+            tool_name='youtube_search_tool',
+            args={'query': query, number_of_results: number_of_results},
+        )
+    response = youtube_search_api.run(f"{query},{number_of_results}")
+    if CUSTOM_DEBUG:
+        print_tool_response(response)
+    return response
+
 
 def extract_video_id(url: str) -> str:
     parsed = urlparse(url)
@@ -198,4 +217,48 @@ def search_and_extract_from_wikipedia(query: str) -> list:
     response = wiki_tool.invoke(query)
     if CUSTOM_DEBUG:
         print_tool_response(response)
+    return response
+
+
+@tool
+def transcribe_audio(file_path: str) -> list:
+    """Transcribe audio from a file using Google Speech-to-Text."""
+    if CUSTOM_DEBUG:
+        print_tool_call(
+            transcribe_audio,
+            tool_name='transcribe_audio',
+            args={'file_path': file_path},
+        )
+    project_id = os.getenv("GOOGLE_CLOUD_PROJECT_ID")
+    loader = SpeechToTextLoader(
+        project_id=project_id,
+        file_path=file_path,
+        is_long = False,  # Set to True for long audio files
+    )
+
+    docs = loader.load()
+    docs_content = [doc.page_content for doc in docs]
+
+    if CUSTOM_DEBUG:
+        print_tool_response(docs_content)
+    return docs_content
+
+
+read_tool = ReadFileTool()
+
+
+@tool
+def read_file_tool(file_path: str) -> str:
+    """Read the content of a file. Use this tool to read .py, .csv, .md, text files, PDFs, etc."""
+    if CUSTOM_DEBUG:
+        print_tool_call(
+            read_file_tool,
+            tool_name='read_file_tool',
+            args={'file_path': file_path},
+        )
+    response = read_tool.invoke({"file_path": file_path})
+    if not os.path.exists(file_path):
+        response = f"File not found: {file_path}"
+        print_tool_response(response)
+    print_tool_response(response)
     return response