import os
import requests
import zipfile
import io
import minsearch
import urllib3
# Disable SSL warnings for the environment
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
def download_and_index():
zip_url = "https://github.com/jlowin/fastmcp/archive/refs/heads/main.zip"
zip_filename = "fastmcp.zip"
# 1. Download if not already downloaded
if not os.path.exists(zip_filename):
print(f"Downloading {zip_url}...")
response = requests.get(zip_url, verify=False)
response.raise_for_status()
with open(zip_filename, 'wb') as f:
f.write(response.content)
else:
print(f"{zip_filename} already exists.")
# 2. Iterate over zip and read md/mdx
documents = []
with zipfile.ZipFile(zip_filename, 'r') as z:
for member in z.infolist():
if not member.is_dir() and (member.filename.endswith('.md') or member.filename.endswith('.mdx')):
# 3. Remove first part of the path
# Example: "fastmcp-main/docs/welcome.mdx" -> "docs/welcome.mdx"
parts = member.filename.split('/')
clean_filename = '/'.join(parts[1:])
if not clean_filename: # Skip the top-level directory if it was somehow included
continue
with z.open(member) as f:
content = f.read().decode('utf-8')
documents.append({
"filename": clean_filename,
"content": content
})
# 4. Index with minsearch
index = minsearch.Index(
text_fields=["content"],
keyword_fields=["filename"]
)
index.fit(documents)
return index
def search(index, query):
# 5. Search function retrieves 5 most relevant
results = index.search(
query=query,
filter_dict={},
boost_dict={"content": 1.0},
num_results=5
)
return results
if __name__ == "__main__":
idx = download_and_index()
query = "demo"
results = search(idx, query)
if results:
print(f"Top result for '{query}': {results[0]['filename']}")
# Printing top 5 for clarity
for i, res in enumerate(results):
print(f"{i+1}. {res['filename']}")
else:
print("No results found.")
