From 37bf997dc0220a30605249655056e90f04e33366 Mon Sep 17 00:00:00 2001 From: Bryan Newbold Date: Thu, 26 Sep 2019 12:00:01 -0700 Subject: lots of grobid tool implementation (still WIP) --- python/tests/test_misc.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'python/tests/test_misc.py') diff --git a/python/tests/test_misc.py b/python/tests/test_misc.py index 02deec9..420bc07 100644 --- a/python/tests/test_misc.py +++ b/python/tests/test_misc.py @@ -50,9 +50,9 @@ def test_parse_cdx_line(): 'surt': "edu,upenn,ldc)/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf", 'url': "https://www.ldc.upenn.edu/sites/www.ldc.upenn.edu/files/medar2009-large-arabic-broadcast-collection.pdf", 'datetime': "20170828233154", - 'warc': "SEMSCHOLAR-PDF-CRAWL-2017-08-04-20170828231135742-00000-00009-wbgrp-svc284/SEMSCHOLAR-PDF-CRAWL-2017-08-04-20170828232253025-00005-3480~wbgrp-svc284.us.archive.org~8443.warc.gz", - 'offset': 931661233, - 'c_size': 210251, + 'warc_path': "SEMSCHOLAR-PDF-CRAWL-2017-08-04-20170828231135742-00000-00009-wbgrp-svc284/SEMSCHOLAR-PDF-CRAWL-2017-08-04-20170828232253025-00005-3480~wbgrp-svc284.us.archive.org~8443.warc.gz", + 'warc_offset': 931661233, + 'warc_csize': 210251, 'http_status': 200, } -- cgit v1.2.3