Feature/dev

lazygitgpt/agents/cli_agent.py

@@ -1,30 +1,43 @@
-import json
-from langchain.prompts import ChatPromptTemplate
-from langchain.output_parsers import ResponseSchema, StructuredOutputParser
-import re
-
+from langchain.chains import ConversationalRetrievalChain
+from langchain.agents import Tool
+from langchain.tools import DuckDuckGoSearchRun
+from langchain.agents import initialize_agent
 from lazygitgpt.llms import chat_model
 from lazygitgpt.datasources.repos import read_repository_contents
 from lazygitgpt.git.operations import update_files
+from lazygitgpt.retrievers.retrievalqa import retriever
+from lazygitgpt.memory.memory import memory
+
+search = DuckDuckGoSearchRun()
+
+def generate_response(prompt):
+    inputs = {'chat_history': '', 'question': prompt}
+    qa = ConversationalRetrievalChain.from_llm(chat_model, retriever=retriever, memory=memory)
+    result = qa(inputs)
+    return result["answer"]
+
+# tools = [
+#     Tool(
+#         name='DuckDuckGo Search',
+#         func= search.run,
+#         description="Useful for when you need to do a search on the internet to find information that another tool can't find. be specific with your input."
+#     ),
+#     Tool(
+#         name='Conversational Retrieval',
+#         func=generate_response,
+#         description="This is Conversational Retrieval chain which has content of the entire repository."
+#     )
+# ]
 
-output_schema = ResponseSchema(name='filename', description='contents', type='string')
-output_parser = StructuredOutputParser(response_schemas=[output_schema])
-format_instructions = output_parser.get_format_instructions()
-template_string = """You are an expert programmer. 
-You are reviewing a code repository.
-Read the code and make changes to the code as per the user requirements.
-user requirements: {user_requirements}
-code repository: {code_repository}
-Output the contents of the file that you changed as per the format instructions : {format_instructions}
-"""
+# zero_shot_agent = initialize_agent(
+#     agent="zero-shot-react-description",
+#     tools=tools,
+#     llm=chat_model,
+#     verbose=True,
+#     max_iterations=30,
+#     retriever=retriever
+# )
 
-def generate_response(prompt, sources=read_repository_contents()):
-    sources_str = json.dumps(sources, indent=4)
-    prompt_template = ChatPromptTemplate.from_template(template_string)
-    messages = prompt_template.format_messages(user_requirements = prompt,  
-                            code_repository = sources_str,
-                            format_instructions=format_instructions)
-    response = chat_model(messages)
-    response_json = response.to_json()
-    data = response_json['kwargs']['content']
-    return data
+# def run(prompt):
+#     reponse = zero_shot_agent.run(prompt)
+#     return reponse

lazygitgpt/cli.py

@@ -1,8 +1,7 @@
 #!/usr/bin/env python3
-
 import click
 from .git.operations import clone_repository, checkout_branch, create_branch
-from .agents.cli_agent import generate_response
+from .agents.cli_agent import generate_response#, run
 
 @click.group()
 def cli():

lazygitgpt/datasources/repos.py

@@ -1,26 +1,30 @@
 import os
 import glob
 import json
+from git import Repo
+from langchain.document_loaders import GitLoader
+from langchain.document_loaders.generic import GenericLoader
+from langchain.document_loaders.parsers import LanguageParser
+from langchain.text_splitter import Language
+from langchain.text_splitter import RecursiveCharacterTextSplitter
 
-def read_repository_contents(directory_path=os.getcwd(), file_pattern="*"):
-    """
-    Reads all files in the specified directory matching the file pattern,
-    and creates a JSON object with file names and their contents.
+def read_repository_contents():
+    repo_path = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+    Repo.git_dir=repo_path
+    repo = Repo(Repo.git_dir)
+    branch = repo.head.reference
 
-    Args:
-    directory_path (str): Path to the directory containing the files.
-    file_pattern (str): Pattern to match files. Defaults to '*' (all files).
-
-    Returns:
-    str: A JSON string containing the file names and their contents.
-    """
-    data = {}
-    for file_path in glob.glob(f"{directory_path}/{file_pattern}"):
-        if os.path.isfile(file_path):
-            try:
-                with open(file_path, 'r', encoding='utf-8') as file:
-                    data[file_path] = file.read()
-            except Exception as e:
-                print(f"Error reading file: {file_path} - {e}")
-
-    return json.dumps(data, indent=4)
+    loader = GitLoader(repo_path, branch=branch)
+    docs = loader.load()
+    loader = GenericLoader.from_filesystem(
+        repo_path,
+        glob="**/*",
+        suffixes=[".py"],
+        parser=LanguageParser(language=Language.PYTHON, parser_threshold=500),
+    )
+    documents = loader.load()
+    python_splitter = RecursiveCharacterTextSplitter.from_language(
+        language=Language.PYTHON, chunk_size=2000, chunk_overlap=200
+    )
+    texts = python_splitter.split_documents(documents)
+    return texts

lazygitgpt/memory/__init__.py

@@ -0,0 +1 @@
+from .memory import memory

lazygitgpt/memory/memory.py

@@ -0,0 +1,4 @@
+from langchain.memory import ConversationBufferMemory
+
+memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
+

lazygitgpt/retrievers/retrievalqa.py

@@ -0,0 +1,9 @@
+from langchain.embeddings.openai import OpenAIEmbeddings
+from langchain.vectorstores import Chroma
+from lazygitgpt.datasources.repos import read_repository_contents
+
+db = Chroma.from_documents(read_repository_contents(), OpenAIEmbeddings(disallowed_special=()))
+retriever = db.as_retriever(
+    search_type="mmr",  # Also test "similarity"
+    search_kwargs={"k": 1000},
+)