log hash of ip with hash of query so can track number of users and number of queries anonymously

video_semantic_search/app.py

@@ -9,7 +9,9 @@ import faiss
 import numpy as np
 import pandas as pd
 import streamlit as st
+from streamlit import runtime
 from streamlit.logger import get_logger
+from streamlit.runtime.scriptrunner import get_script_run_ctx
 
 from pipeline import clip_wrapper
 from pipeline.process_videos import DATAFRAME_PATH
@@ -117,6 +119,20 @@ def display_search_results(results: list[SearchResult]) -> None:
             col_num = 0
 
 
+def get_remote_ip() -> str:
+    """Get remote ip."""
+    try:
+        ctx = get_script_run_ctx()
+        if ctx is None:
+            return None
+        session_info = runtime.get_instance().get_client(ctx.session_id)
+        if session_info is None:
+            return None
+    except Exception as e:
+        return None
+    return session_info.request.remote_ip
+
+
 def main():
     st.set_page_config(page_title="video-semantic-search", layout="wide")
     st.header("Visual content search over music videos")
@@ -128,7 +144,9 @@ def main():
     )
     query = st.session_state["query"]
     if query:
-        logger.info(f"Recieved query... {hashlib.md5(query.encode()).hexdigest()}")
+        query_sha256 = hashlib.sha256(query.encode()).hexdigest()[:10]
+        ip_sha256 = hashlib.sha256(get_remote_ip().encode()).hexdigest()[:10]
+        logger.info(f"sha256(ip)={ip_sha256} sha256(query)={query_sha256}")
         st.text("Click image to open video")
         display_search_results(searcher.search(query))
     if get_git_hash():