diff options
| author | Gertjan van den Burg <gertjanvandenburg@gmail.com> | 2020-08-17 20:56:12 +0100 |
|---|---|---|
| committer | Gertjan van den Burg <gertjanvandenburg@gmail.com> | 2020-08-17 20:56:12 +0100 |
| commit | 38800a8e4ce2328548a10ea31062089900385075 (patch) | |
| tree | 53259b8ed39a6e9d9b39cb894a2e45780c666f42 /datasets/iceland_tourism/get_iceland_tourism.py | |
| parent | Use read_table instead of read_csv (diff) | |
| parent | Add retries to all download scripts (diff) | |
| download | TCPD-38800a8e4ce2328548a10ea31062089900385075.tar.gz TCPD-38800a8e4ce2328548a10ea31062089900385075.zip | |
Merge branch 'bugfix/measles_download' into master
Diffstat (limited to 'datasets/iceland_tourism/get_iceland_tourism.py')
| -rw-r--r-- | datasets/iceland_tourism/get_iceland_tourism.py | 18 |
1 files changed, 17 insertions, 1 deletions
diff --git a/datasets/iceland_tourism/get_iceland_tourism.py b/datasets/iceland_tourism/get_iceland_tourism.py index 752f07d..b9c8347 100644 --- a/datasets/iceland_tourism/get_iceland_tourism.py +++ b/datasets/iceland_tourism/get_iceland_tourism.py @@ -17,9 +17,12 @@ import hashlib import json import os import xlrd +import sys +import time from functools import wraps from urllib.request import urlretrieve +from urllib.error import URLError XLSX_URL = "https://web.archive.org/web/20191121170223if_/https://www.ferdamalastofa.is/static/files/ferdamalastofa/Frettamyndir/2019/nov/visitors-to-iceland-2002-2019-oct.xlsx" @@ -84,7 +87,20 @@ def validate(checksum): @validate(MD5_XLSX) def download_xlsx(target_path=None): - urlretrieve(XLSX_URL, target_path) + count = 0 + while count < 5: + count += 1 + try: + urlretrieve(XLSX_URL, target_path) + return + except URLError as err: + print( + "Error occurred (%r) when trying to download xlsx. Retrying in 5 seconds" + % err, + sys.stderr, + ) + time.sleep(5) + def format_ym(year, month): |