From 1b4136d415cbe549c735d459cd0ab1345bb37011 Mon Sep 17 00:00:00 2001 From: Martin Czygan Date: Fri, 15 Oct 2021 21:56:52 +0200 Subject: pubmed: switch default http site to retrieve update files Proxy started to throw: "dial tcp: lookup ftp.ncbi.nlm.nih.gov on [::1]:53: read udp [::1]:45178->[::1]:53: read: connection refused" NIH has a http version on it's own, try to use that. --- python/fatcat_tools/harvest/pubmed.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'python/fatcat_tools/harvest') diff --git a/python/fatcat_tools/harvest/pubmed.py b/python/fatcat_tools/harvest/pubmed.py index 579dd505..92798a99 100644 --- a/python/fatcat_tools/harvest/pubmed.py +++ b/python/fatcat_tools/harvest/pubmed.py @@ -263,10 +263,12 @@ def ftpretr(url, max_retries=10, retry_delay=1, proxy_hostport=None): return f.name -def ftpretr_via_http_proxy(url, proxy_hostport="159.69.240.245:15201", max_retries=10, retry_delay=1): +def ftpretr_via_http_proxy(url, proxy_hostport="ftp.ncbi.nlm.nih.gov", max_retries=10, retry_delay=1): """ Fetch file from FTP via external HTTP proxy, e.g. ftp.host.com:/a/b/c would - be retrievable via proxy.com/a/b/c. + be retrievable via proxy.com/a/b/c; (in 09/2021 we used + "159.69.240.245:15201" as proxy_hostport but that started to fail + 2021-10-15; just switch to NIH's http version). """ parsed = urlparse(url) server, path = parsed.netloc, parsed.path -- cgit v1.2.3