Spaces:
Running
Running
File size: 10,142 Bytes
17f9660 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 |
import streamlit as st
import os
import tempfile
from qdrant_client import QdrantClient, models
from langchain.vectorstores import Qdrant
from langchain.embeddings import FakeEmbeddings
from pathlib import Path
import shutil
st.set_page_config(
page_title="FormPilot - Qdrant Dependency Test",
page_icon="🧪",
layout="wide",
)
st.title("🧪 FormPilot - Qdrant Dependency Test UI")
# Create a sidebar with instructions
with st.sidebar:
st.header("Test Steps")
st.info("""
This app helps verify your Qdrant setup for FormPilot.
1. Check package versions
2. Test basic Qdrant functionality
3. Test LangChain integration
4. Test form ingestion process simulation
""")
st.markdown("---")
st.subheader("Package Versions")
try:
import pkg_resources
packages = [
'streamlit', 'pandas', 'qdrant-client', 'langchain',
'langchain-openai', 'openai', 'tqdm', 'python-dotenv'
]
for package in packages:
try:
version = pkg_resources.get_distribution(package).version
st.success(f"{package}: {version}")
except pkg_resources.DistributionNotFound:
st.error(f"{package}: Not installed")
except ImportError:
st.error("Could not import pkg_resources")
# Main content area
st.header("1. Create Test Environment")
# Use temporary directory or let user specify path
use_temp_dir = st.checkbox("Use temporary directory for tests", value=True)
if use_temp_dir:
temp_dir = tempfile.mkdtemp()
qdrant_path = os.path.join(temp_dir, "qdrant_test_data")
st.info(f"Using temporary directory: {qdrant_path}")
else:
qdrant_path = st.text_input("Enter path for Qdrant data:", value="./qdrant_test_data")
os.makedirs(qdrant_path, exist_ok=True)
# Create a test collection
st.header("2. Test Basic Qdrant Functionality")
test_collection_name = "test_collection"
if st.button("Create Test Collection"):
with st.spinner("Creating collection..."):
try:
client = QdrantClient(path=qdrant_path)
try:
# Check if collection exists using get_collection API
client.get_collection(collection_name=test_collection_name)
st.info(f"Collection '{test_collection_name}' already exists")
except Exception:
# Collection doesn't exist, create it
client.create_collection(
collection_name=test_collection_name,
vectors_config=models.VectorParams(size=2, distance=models.Distance.COSINE),
)
st.success(f"Created collection '{test_collection_name}'")
# Insert test vectors
vectors = [[1.0, 0.0], [0.0, 1.0]]
payloads = [{"text": "test1"}, {"text": "test2"}]
ids = [1, 2]
client.upload_collection(
collection_name=test_collection_name,
vectors=vectors,
payload=payloads,
ids=ids
)
# Check count
count = client.count(test_collection_name).count
st.success(f"Number of vectors in collection: {count}")
# Search for similar vectors
search_result = client.search(
collection_name=test_collection_name,
query_vector=[1.0, 0.0],
limit=1
)
st.json({"Search Result": [{"id": res.id, "score": res.score, "payload": res.payload} for res in search_result]})
except Exception as e:
st.error(f"Error: {e}")
# Test LangChain integration
st.header("3. Test LangChain Integration")
langchain_collection = "langchain_test"
if st.button("Test LangChain + Qdrant"):
with st.spinner("Testing LangChain integration..."):
try:
# Create a fake embeddings model for testing
class TestEmbeddings(FakeEmbeddings):
def embed_documents(self, texts):
# Return 1536-dim vectors (like OpenAI)
return [[1.0] * 1536 for _ in texts]
def embed_query(self, text):
# Return 1536-dim vector
return [1.0] * 1536
embeddings = TestEmbeddings(size=1536)
# Create a new directory for this test
langchain_path = os.path.join(Path(qdrant_path).parent, "langchain_test")
os.makedirs(langchain_path, exist_ok=True)
st.info(f"LangChain test path: {langchain_path}")
# Initialize Qdrant with LangChain
try:
from langchain.schema.document import Document
docs = [Document(page_content="This is a test document", metadata={"source": "test"})]
vectordb = Qdrant.from_documents(
documents=docs,
embedding=embeddings,
path=langchain_path,
collection_name=langchain_collection,
)
st.success("Successfully created Qdrant vector store with LangChain")
# Test search
results = vectordb.similarity_search("test query")
st.json({"Search Results": [{"content": doc.page_content, "metadata": doc.metadata} for doc in results]})
except Exception as e:
st.error(f"LangChain integration error: {e}")
st.error("This may indicate version incompatibility between LangChain and Qdrant")
except Exception as e:
st.error(f"Error: {e}")
# Simulate form ingestion
st.header("4. Test Form Ingestion Simulation")
if st.button("Simulate Form Ingestion"):
with st.spinner("Simulating form ingestion process..."):
try:
# Create path for this test
ingest_path = os.path.join(Path(qdrant_path).parent, "ingest_test")
os.makedirs(ingest_path, exist_ok=True)
# Create collection for ingestion test
collection_name = "formpilot_test"
client = QdrantClient(path=ingest_path)
try:
# Check if collection exists
client.get_collection(collection_name=collection_name)
st.info(f"Collection '{collection_name}' already exists")
except Exception:
# Collection doesn't exist, create it
client.create_collection(
collection_name=collection_name,
vectors_config=models.VectorParams(size=1536, distance=models.Distance.COSINE),
)
st.success(f"Created collection '{collection_name}'")
# Simulate adding vectors similar to ingest process
test_vectors = [
[0.1] * 1536, # Simplified vector
[0.2] * 1536
]
test_payloads = [
{"text": "I-485 form instructions page 1", "source": "I-485-instr.pdf:page-1", "form": "I-485"},
{"text": "I-485 form instructions page 2", "source": "I-485-instr.pdf:page-2", "form": "I-485"}
]
test_ids = [1, 2]
client.upload_collection(
collection_name=collection_name,
vectors=test_vectors,
payload=test_payloads,
ids=test_ids,
batch_size=64
)
# Verify count
count = client.count(collection_name).count
st.success(f"Number of vectors in test collection: {count}")
# Retrieve a point to confirm structure
point = client.retrieve(collection_name, ids=[1])
if point:
st.json({"Retrieved Point": {"id": point[0].id, "payload": point[0].payload}})
else:
st.error("Could not retrieve point")
except Exception as e:
st.error(f"Error: {e}")
# Cleanup
st.header("5. Cleanup")
if st.button("Clean Up Test Directories"):
with st.spinner("Cleaning up..."):
try:
if use_temp_dir:
shutil.rmtree(Path(qdrant_path).parent)
st.success(f"Removed temp directory: {Path(qdrant_path).parent}")
else:
if os.path.exists(qdrant_path):
shutil.rmtree(qdrant_path)
st.success(f"Removed directory: {qdrant_path}")
langchain_path = os.path.join(Path(qdrant_path).parent, "langchain_test")
if os.path.exists(langchain_path):
shutil.rmtree(langchain_path)
st.success(f"Removed directory: {langchain_path}")
ingest_path = os.path.join(Path(qdrant_path).parent, "ingest_test")
if os.path.exists(ingest_path):
shutil.rmtree(ingest_path)
st.success(f"Removed directory: {ingest_path}")
except Exception as e:
st.error(f"Error during cleanup: {e}")
# Summary
st.header("6. Summary")
st.markdown("""
If all tests passed successfully, your Qdrant and LangChain dependencies are correctly configured for FormPilot.
### Common Issues:
1. **Version Incompatibilities**: Ensure you're using compatible versions of Qdrant client and LangChain.
2. **API Changes**: The LangChain API has changed significantly in recent versions. Your code may need updates.
3. **Missing Dependencies**: Make sure all required packages are installed.
### Recommendations:
Based on your code analysis, consider using these pinned versions:
```
qdrant-client==1.7.3
langchain==0.1.12
langchain-openai==0.1.0
```
Check the notebook for a more detailed dependency test if you need it.
""") |