From fb7717ae410f72ff33017c176f64dff556b86f5b Mon Sep 17 00:00:00 2001 From: Bryan Newbold Date: Tue, 21 Jan 2020 10:57:13 -0800 Subject: grobid: fix error_msg typo; set status_code for timeouts --- python/sandcrawler/grobid.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/python/sandcrawler/grobid.py b/python/sandcrawler/grobid.py index 6c8450b..8c3aec1 100644 --- a/python/sandcrawler/grobid.py +++ b/python/sandcrawler/grobid.py @@ -41,7 +41,8 @@ class GrobidClient(object): except requests.Timeout: return { 'status': 'error-timeout', - 'status': 'GROBID request (HTTP POST) timeout', + 'status_code': -4, # heritrix3 "HTTP timeout" code + 'error_msg': 'GROBID request (HTTP POST) timeout', } info = dict( -- cgit v1.2.3