extract snowpark methods

README.md

@@ -1,5 +1,7 @@
 # SnowDev - Snowpark Devops
 
+[![Documentation](https://img.shields.io/badge/documentation-view-blue)](docs/quickstart.md) ![PyPI Downloads](https://img.shields.io/pypi/dm/snowdev)
+
 SnowDev is a command-line utility designed for deploying various components related to Snowflake such as UDFs, stored procedures, and Streamlit applications using **Snowpark**. This tool streamlines tasks like initializing directories, local testing, uploading, and auto create components code using AI.
 
 ## Setup
@@ -76,10 +78,12 @@ snowdev <command> [options]
 - [x] Support for Streamlit
 - [x] AI interactions for embedding and suggestions
 - [ ] Use AI to modify existing code for optimization
+- [ ] AI to add the packages to toml file
 - [ ] Adding more granularity for AI commands
 - [ ] Support for snowflake Tasks
 
 
+
 ## Contributions
 
 Feel free to contribute to SnowDev by submitting pull requests or opening issues on the project's GitHub repository.

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "snowdev"
-version = "0.1.8"
+version = "0.1.9"
 description = "snowdev: DevOps toolkit for Snowflake, facilitating seamless deployment of UDFs, stored procedures, and Streamlit apps using Snowpark's capabilities right from your local environment."
 authors = ["kaarthik <[email protected]>"]
 readme = "README.md"

snowdev/functions/bot.py

@@ -1,4 +1,5 @@
 import os
+import re
 
 import toml
 from langchain.chains import LLMChain, RetrievalQA
@@ -14,7 +15,9 @@
 )
 from snowdev.functions.utils.templates.udf import TEMPLATE as UDF_TEMPLATE
 from snowdev.functions.utils.ingest import DocumentProcessor, Secrets, Config
+from snowdev.functions.utils.snowpark_methods import SnowparkMethods
 
+MODEL = "gpt-4"
 
 class SnowBot:
 
@@ -33,6 +36,7 @@ def ai_embed():
         config = Config()
         doc_processor = DocumentProcessor(secrets, config)
         try:
+            SnowparkMethods.generate_documentation()
             result = doc_processor.process()
             print(colored("✅ Documents have been successfully embedded!", "green"))
             return result
@@ -62,8 +66,8 @@ def get_chain_gpt(db):
         Get a chain for chatting with a vector database.
         """
         llm = ChatOpenAI(
-            model_name="gpt-4",
-            temperature=0.5,
+            model_name=MODEL,
+            temperature=0,
             openai_api_key=os.environ["OPENAI_API_KEY"],
             max_tokens=500,
         )
@@ -126,7 +130,14 @@ def create_new_ai_component(component_name, prompt, template_type):
         vectordb = Chroma(persist_directory="chroma_db", embedding_function=embeddings)
         chain = SnowBot.get_chain_gpt(vectordb)
         response_content = chain(prompt)["result"]
-        response_content = response_content.split("```python\n")[1].split("\n```")[0]
+
+        matches = re.findall(r"```(?:python)?\n(.*?)\n```", response_content, re.DOTALL)
+
+        if matches:
+            # Take the first match, as there might be multiple code blocks
+            response_content = matches[0].strip() 
+        else:
+            print(colored("Unexpected response content format. Expected code block not found. Please try again", "red"))
 
         component_folder = os.path.join("src", template_type, component_name)
         os.makedirs(component_folder, exist_ok=True)

snowdev/functions/utils/ingest.py

@@ -15,8 +15,8 @@ class Secrets(BaseModel):
 class Config(BaseModel):
     chunk_size: int = 1000
     chunk_overlap: int = 0
-    docs_dir: str = "snowdev/snow_functions/utils/knowledge"
-    docs_glob: str = "**/*.py"
+    docs_dir: str = "snowdev/functions/utils/knowledge"
+    docs_glob: str = "**/*"
 
 
 class DocumentProcessor: