philincloud commited on
Commit
f33e43c
·
verified ·
1 Parent(s): daafe8e

Update langgraph_agent.py

Browse files
Files changed (1) hide show
  1. langgraph_agent.py +43 -58
langgraph_agent.py CHANGED
@@ -4,50 +4,44 @@ import contextlib
4
  import pandas as pd
5
  from typing import Dict, List, Union
6
 
7
- # New imports for image and audio processing
8
- from PIL import Image as PILImage # Used for type checking/potential future local processing
9
  from huggingface_hub import InferenceClient
10
 
11
  from langgraph.graph import START, StateGraph, MessagesState
12
  from langgraph.prebuilt import tools_condition, ToolNode
13
  from langchain_openai import ChatOpenAI
14
  from langchain_huggingface import ChatHuggingFace, HuggingFaceEndpoint
15
- from langchain_community.tools.tavily_search import TavilySearchResults
16
  from langchain_community.document_loaders import WikipediaLoader, ArxivLoader
17
  from langchain_core.messages import SystemMessage, HumanMessage
18
  from langchain_google_genai import ChatGoogleGenerativeAI
19
  from langchain_core.tools import tool
20
 
 
 
21
  @tool
22
  def multiply(a: int, b: int) -> int:
23
- """Multiply two integers."""
24
  return a * b
25
 
26
  @tool
27
  def add(a: int, b: int) -> int:
28
- """Add two integers."""
29
  return a + b
30
 
31
  @tool
32
  def subtract(a: int, b: int) -> int:
33
- """Subtract the second integer from the first."""
34
  return a - b
35
 
36
  @tool
37
  def divide(a: int, b: int) -> float:
38
- """Divide first integer by second; error if divisor is zero."""
39
  if b == 0:
40
  raise ValueError("Cannot divide by zero.")
41
  return a / b
42
 
43
  @tool
44
  def modulus(a: int, b: int) -> int:
45
- """Return the remainder of dividing first integer by second."""
46
  return a % b
47
 
48
  @tool
49
  def wiki_search(query: str) -> dict:
50
- """Search Wikipedia for a query and return up to 2 documents."""
51
  try:
52
  docs = WikipediaLoader(query=query, load_max_docs=2, lang="en").load()
53
  if not docs:
@@ -61,23 +55,21 @@ def wiki_search(query: str) -> dict:
61
  print(f"Error in wiki_search tool: {e}")
62
  return {"wiki_results": f"Error occurred while searching Wikipedia for '{query}'. Details: {str(e)}"}
63
 
 
 
 
 
64
  @tool
65
- def web_search(query: str) -> dict:
66
- """Perform a web search (via Tavily) and return up to 3 results."""
67
  try:
68
- docs = TavilySearchResults(max_results=3).invoke(query=query)
69
- formatted = "\n\n---\n\n".join(
70
- f'<Document source="{d.metadata["source"]}"/>\n{d.page_content}'
71
- for d in docs
72
- )
73
- return {"web_results": formatted}
74
  except Exception as e:
75
- print(f"Error in web_search tool: {e}")
76
- return {"web_results": f"Error occurred while searching the web for '{query}'. Details: {str(e)}"}
77
 
78
  @tool
79
  def arvix_search(query: str) -> dict:
80
- """Search arXiv for a query and return up to 3 paper excerpts."""
81
  docs = ArxivLoader(query=query, load_max_docs=3).load()
82
  formatted = "\n\n---\n\n".join(
83
  f'<Document source="{d.metadata["source"]}"/>\n{d.page_content[:1000]}'
@@ -85,7 +77,6 @@ def arvix_search(query: str) -> dict:
85
  )
86
  return {"arvix_results": formatted}
87
 
88
- # Initialize Hugging Face Inference Client
89
  HF_API_TOKEN = os.getenv("HF_API_TOKEN")
90
  HF_INFERENCE_CLIENT = None
91
  if HF_API_TOKEN:
@@ -95,10 +86,6 @@ else:
95
 
96
  @tool
97
  def read_file_content(file_path: str) -> Dict[str, str]:
98
- """
99
- Reads the content of a file and returns its primary information.
100
- For text/code/excel, returns content. For media, returns a prompt to use specific tools.
101
- """
102
  try:
103
  _, file_extension = os.path.splitext(file_path)
104
  file_extension = file_extension.lower()
@@ -108,14 +95,12 @@ def read_file_content(file_path: str) -> Dict[str, str]:
108
  content = f.read()
109
  return {"file_type": "text/code", "file_name": file_path, "file_content": content}
110
  elif file_extension == ".xlsx":
111
- df = pd.read_excel(file_path)
112
- content = df.to_string()
113
- return {"file_type": "excel", "file_name": file_path, "file_content": content}
114
  elif file_extension in (".jpeg", ".jpg", ".png"):
115
- # Indicate that it's an image and needs to be described by a specific tool
116
  return {"file_type": "image", "file_name": file_path, "file_content": f"Image file '{file_path}' detected. Use 'describe_image' tool to get a textual description."}
117
  elif file_extension == ".mp3":
118
- # Indicate that it's an audio file and needs to be transcribed by a specific tool
119
  return {"file_type": "audio", "file_name": file_path, "file_content": f"Audio file '{file_path}' detected. Use 'transcribe_audio' tool to get the text transcription."}
120
  else:
121
  return {"file_type": "unsupported", "file_name": file_path, "file_content": f"Unsupported file type: {file_extension}. Only .txt, .py, .xlsx, .jpeg, .jpg, .png, .mp3 files are recognized."}
@@ -126,10 +111,6 @@ def read_file_content(file_path: str) -> Dict[str, str]:
126
 
127
  @tool
128
  def python_interpreter(code: str) -> Dict[str, str]:
129
- """
130
- Executes Python code and returns its standard output.
131
- If there's an error during execution, it returns the error message.
132
- """
133
  old_stdout = io.StringIO()
134
  with contextlib.redirect_stdout(old_stdout):
135
  try:
@@ -143,10 +124,6 @@ def python_interpreter(code: str) -> Dict[str, str]:
143
 
144
  @tool
145
  def describe_image(image_path: str) -> Dict[str, str]:
146
- """
147
- Generates a textual description for an image file (JPEG, JPG, PNG) using an image-to-text model
148
- from the Hugging Face Inference API. Requires HF_API_TOKEN environment variable to be set.
149
- """
150
  if not HF_INFERENCE_CLIENT:
151
  return {"error": "Hugging Face API token not configured for image description. Cannot use this tool."}
152
  try:
@@ -159,38 +136,48 @@ def describe_image(image_path: str) -> Dict[str, str]:
159
  except Exception as e:
160
  return {"error": f"Error describing image {image_path}: {str(e)}"}
161
 
162
-
 
 
 
 
 
 
 
 
 
 
 
 
163
 
164
  API_KEY = os.getenv("GEMINI_API_KEY")
165
  HF_API_TOKEN = os.getenv("HF_SPACE_TOKEN")
166
  GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
167
 
168
-
169
  tools = [
170
  multiply, add, subtract, divide, modulus,
171
- wiki_search, web_search, arvix_search,
 
 
172
  read_file_content,
173
  python_interpreter,
174
  describe_image,
 
175
  ]
176
 
177
-
178
  with open("prompt.txt", "r", encoding="utf-8") as f:
179
  system_prompt = f.read()
180
  sys_msg = SystemMessage(content=system_prompt)
181
 
182
-
183
  def build_graph(provider: str = "gemini"):
184
- """Build the LangGraph agent with chosen LLM (default: Gemini)."""
185
  if provider == "gemini":
186
  llm = ChatGoogleGenerativeAI(
187
- model= "gemini-2.5-pro-preview-05-06", # Reverted model to gemini-2.5-pro-preview-05-06
188
- temperature=1.0,
189
- max_retries=2,
190
- api_key=GEMINI_API_KEY,
191
- max_tokens=5000
192
- )
193
-
194
  elif provider == "huggingface":
195
  llm = ChatHuggingFace(
196
  llm=HuggingFaceEndpoint(
@@ -204,10 +191,10 @@ def build_graph(provider: str = "gemini"):
204
  llm_with_tools = llm.bind_tools(tools)
205
 
206
  def assistant(state: MessagesState):
207
- messages_to_send = [sys_msg] + state["messages"]
208
- llm_response = llm_with_tools.invoke(messages_to_send)
209
- print(f"LLM Raw Response: {llm_response}") # Add this line
210
- return {"messages": [llm_response]}
211
 
212
  builder = StateGraph(MessagesState)
213
  builder.add_node("assistant", assistant)
@@ -219,6 +206,4 @@ def build_graph(provider: str = "gemini"):
219
  return builder.compile()
220
 
221
  if __name__ == "__main__":
222
- # This block is intentionally left empty as per user request to remove examples.
223
- # Your agent will interact with the graph by invoking it with messages.
224
- pass
 
4
  import pandas as pd
5
  from typing import Dict, List, Union
6
 
7
+ from PIL import Image as PILImage
 
8
  from huggingface_hub import InferenceClient
9
 
10
  from langgraph.graph import START, StateGraph, MessagesState
11
  from langgraph.prebuilt import tools_condition, ToolNode
12
  from langchain_openai import ChatOpenAI
13
  from langchain_huggingface import ChatHuggingFace, HuggingFaceEndpoint
 
14
  from langchain_community.document_loaders import WikipediaLoader, ArxivLoader
15
  from langchain_core.messages import SystemMessage, HumanMessage
16
  from langchain_google_genai import ChatGoogleGenerativeAI
17
  from langchain_core.tools import tool
18
 
19
+ from langchain_google_community.tools.Google Search import GoogleSearchResults
20
+
21
  @tool
22
  def multiply(a: int, b: int) -> int:
 
23
  return a * b
24
 
25
  @tool
26
  def add(a: int, b: int) -> int:
 
27
  return a + b
28
 
29
  @tool
30
  def subtract(a: int, b: int) -> int:
 
31
  return a - b
32
 
33
  @tool
34
  def divide(a: int, b: int) -> float:
 
35
  if b == 0:
36
  raise ValueError("Cannot divide by zero.")
37
  return a / b
38
 
39
  @tool
40
  def modulus(a: int, b: int) -> int:
 
41
  return a % b
42
 
43
  @tool
44
  def wiki_search(query: str) -> dict:
 
45
  try:
46
  docs = WikipediaLoader(query=query, load_max_docs=2, lang="en").load()
47
  if not docs:
 
55
  print(f"Error in wiki_search tool: {e}")
56
  return {"wiki_results": f"Error occurred while searching Wikipedia for '{query}'. Details: {str(e)}"}
57
 
58
+ Google Search_tool = GoogleSearchResults(
59
+ api_key=os.getenv("GOOGLE_API_KEY"),
60
+ engine_id=os.getenv("GOOGLE_CSE_ID")
61
+ )
62
  @tool
63
+ def google_web_search(query: str) -> dict:
 
64
  try:
65
+ docs = Google Search_tool.invoke(query)
66
+ return {"google_web_results": docs}
 
 
 
 
67
  except Exception as e:
68
+ print(f"Error in google_web_search tool: {e}")
69
+ return {"google_web_results": f"Error occurred while searching the web for '{query}'. Details: {str(e)}"}
70
 
71
  @tool
72
  def arvix_search(query: str) -> dict:
 
73
  docs = ArxivLoader(query=query, load_max_docs=3).load()
74
  formatted = "\n\n---\n\n".join(
75
  f'<Document source="{d.metadata["source"]}"/>\n{d.page_content[:1000]}'
 
77
  )
78
  return {"arvix_results": formatted}
79
 
 
80
  HF_API_TOKEN = os.getenv("HF_API_TOKEN")
81
  HF_INFERENCE_CLIENT = None
82
  if HF_API_TOKEN:
 
86
 
87
  @tool
88
  def read_file_content(file_path: str) -> Dict[str, str]:
 
 
 
 
89
  try:
90
  _, file_extension = os.path.splitext(file_path)
91
  file_extension = file_extension.lower()
 
95
  content = f.read()
96
  return {"file_type": "text/code", "file_name": file_path, "file_content": content}
97
  elif file_extension == ".xlsx":
98
+ df = pd.read_excel(file_path)
99
+ content = df.to_string()
100
+ return {"file_type": "excel", "file_name": file_path, "file_content": content}
101
  elif file_extension in (".jpeg", ".jpg", ".png"):
 
102
  return {"file_type": "image", "file_name": file_path, "file_content": f"Image file '{file_path}' detected. Use 'describe_image' tool to get a textual description."}
103
  elif file_extension == ".mp3":
 
104
  return {"file_type": "audio", "file_name": file_path, "file_content": f"Audio file '{file_path}' detected. Use 'transcribe_audio' tool to get the text transcription."}
105
  else:
106
  return {"file_type": "unsupported", "file_name": file_path, "file_content": f"Unsupported file type: {file_extension}. Only .txt, .py, .xlsx, .jpeg, .jpg, .png, .mp3 files are recognized."}
 
111
 
112
  @tool
113
  def python_interpreter(code: str) -> Dict[str, str]:
 
 
 
 
114
  old_stdout = io.StringIO()
115
  with contextlib.redirect_stdout(old_stdout):
116
  try:
 
124
 
125
  @tool
126
  def describe_image(image_path: str) -> Dict[str, str]:
 
 
 
 
127
  if not HF_INFERENCE_CLIENT:
128
  return {"error": "Hugging Face API token not configured for image description. Cannot use this tool."}
129
  try:
 
136
  except Exception as e:
137
  return {"error": f"Error describing image {image_path}: {str(e)}"}
138
 
139
+ @tool
140
+ def transcribe_audio(audio_path: str) -> Dict[str, str]:
141
+ if not HF_INFERENCE_CLIENT:
142
+ return {"error": "Hugging Face API token not configured for audio transcription. Cannot use this tool."}
143
+ try:
144
+ with open(audio_path, "rb") as f:
145
+ audio_bytes = f.read()
146
+ transcription = HF_INFERENCE_CLIENT.automatic_speech_recognition(audio_bytes)
147
+ return {"audio_transcription": transcription, "audio_path": audio_path}
148
+ except FileNotFoundError:
149
+ return {"error": f"Audio file not found: {audio_path}. Please ensure the file exists."}
150
+ except Exception as e:
151
+ return {"error": f"Error transcribing audio {audio_path}: {str(e)}"}
152
 
153
  API_KEY = os.getenv("GEMINI_API_KEY")
154
  HF_API_TOKEN = os.getenv("HF_SPACE_TOKEN")
155
  GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
156
 
 
157
  tools = [
158
  multiply, add, subtract, divide, modulus,
159
+ wiki_search,
160
+ google_web_search,
161
+ arvix_search,
162
  read_file_content,
163
  python_interpreter,
164
  describe_image,
165
+ transcribe_audio,
166
  ]
167
 
 
168
  with open("prompt.txt", "r", encoding="utf-8") as f:
169
  system_prompt = f.read()
170
  sys_msg = SystemMessage(content=system_prompt)
171
 
 
172
  def build_graph(provider: str = "gemini"):
 
173
  if provider == "gemini":
174
  llm = ChatGoogleGenerativeAI(
175
+ model="gemini-2.5-pro-preview-05-06",
176
+ temperature=1.0,
177
+ max_retries=2,
178
+ api_key=GEMINI_API_KEY,
179
+ max_tokens=5000
180
+ )
 
181
  elif provider == "huggingface":
182
  llm = ChatHuggingFace(
183
  llm=HuggingFaceEndpoint(
 
191
  llm_with_tools = llm.bind_tools(tools)
192
 
193
  def assistant(state: MessagesState):
194
+ messages_to_send = [sys_msg] + state["messages"]
195
+ llm_response = llm_with_tools.invoke(messages_to_send)
196
+ print(f"LLM Raw Response: {llm_response}")
197
+ return {"messages": [llm_response]}
198
 
199
  builder = StateGraph(MessagesState)
200
  builder.add_node("assistant", assistant)
 
206
  return builder.compile()
207
 
208
  if __name__ == "__main__":
209
+ pass