Support adding an /etc/host entry for the data proxy, if asked.

Author: Seb Boving

patches/sitecustomize.py

@@ -5,12 +5,27 @@
 import torch
 import os
 
+_HOST_FILE = "/etc/hosts"
+
 kaggle_proxy_token = os.getenv("KAGGLE_DATA_PROXY_TOKEN")
 if kaggle_proxy_token:
     from google.auth import credentials
     from google.cloud import bigquery
     from google.cloud.bigquery._http import Connection
 
+    host_entry = os.getenv("KAGGLE_DATA_PROXY_HOST_ENTRY")
+    if host_entry:
+        host_entry = host_entry.strip()
+        with open(_HOST_FILE, "r") as host_file:
+            for line in host_file:
+                if line.strip() == host_entry:
+                    break
+            else:
+                with open(_HOST_FILE, "a") as host_file_append:
+                    host_file_append.write("\n"  # Just in case it wasn't newline terminated.
+                                           + host_entry
+                                           + "\n")
+
     Connection.API_BASE_URL = os.getenv("KAGGLE_DATA_PROXY_URL")
     Connection._EXTRA_HEADERS["X-KAGGLE-PROXY-DATA"] = kaggle_proxy_token