Function to check internet access before network calls

retriever/lib/rdatasets.py

@@ -7,13 +7,16 @@
 from retriever.lib.defaults import RDATASETS_URL, RDATASET_PATH, RDATASET_SCRIPT_WRITE_PATH
 from retriever.lib.create_scripts import create_package
 from retriever.lib.scripts import reload_scripts
+from retriever.lib.tools import check_network
 
 
 def update_rdataset_catalog(test=False):
     '''Updates the datasets_url.json from the github repo'''
     if not os.path.exists(RDATASET_SCRIPT_WRITE_PATH):
         os.makedirs(RDATASET_SCRIPT_WRITE_PATH)
 
+    check_network()
+
     df = pd.read_csv(RDATASETS_URL)
     dataset_url = {}
 

retriever/lib/scripts.py

@@ -18,6 +18,8 @@
                                     RETRIEVER_DATASETS)
 from retriever.lib.load_json import read_json
 from retriever.lib.provenance_tools import get_script_provenance
+from retriever.lib.tools import check_network
+
 
 global_script_list = None
 
@@ -322,6 +324,7 @@ def get_dataset_names_upstream(keywords=None, licenses=None, repo=REPOSITORY):
     in the repositories. Else, the version.txt file is read and the script
     names are then returned.
     """
+    check_network()
     if not keywords and not licenses:
         version_file_request = get_data_upstream(repo + "version.txt")
         if not version_file_request:

retriever/lib/tools.py

@@ -4,6 +4,7 @@
 import sys
 
 import xlrd
+import socket
 import pandas as pd
 
 from retriever.lib.defaults import ENCODING
@@ -78,3 +79,10 @@ def walk_relative_path(dir_name):
         for dir_, _, files in os.walk(dir_name, topdown=False)
         for file_name in files
     ]
+
+def check_network():
+    try:
+        socket.create_connection(("1.1.1.1", 53))
+    except OSError:
+        print("Network Error.")
+        sys.exit(1)