aboutsummaryrefslogtreecommitdiffstats
path: root/mapreduce/tests/test_grobid2json.py
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2018-04-06 14:17:30 -0700
committerBryan Newbold <bnewbold@archive.org>2018-04-06 14:17:30 -0700
commit82cbabc249e5c0375dbb82106451969467f8da00 (patch)
tree0369d2a2f834b18a051da0d7f41c1625543894ab /mapreduce/tests/test_grobid2json.py
parent159d049fe072d38c0297dea6b4ed82a37eff36b7 (diff)
downloadsandcrawler-82cbabc249e5c0375dbb82106451969467f8da00.tar.gz
sandcrawler-82cbabc249e5c0375dbb82106451969467f8da00.zip
add test for grobid2json
Diffstat (limited to 'mapreduce/tests/test_grobid2json.py')
-rw-r--r--mapreduce/tests/test_grobid2json.py14
1 files changed, 14 insertions, 0 deletions
diff --git a/mapreduce/tests/test_grobid2json.py b/mapreduce/tests/test_grobid2json.py
new file mode 100644
index 0000000..40853dd
--- /dev/null
+++ b/mapreduce/tests/test_grobid2json.py
@@ -0,0 +1,14 @@
+
+import json
+import pytest
+from grobid2json import *
+
+
+def test_small_xml():
+
+ with open('tests/files/small.xml', 'r') as f:
+ tei_xml = f.read()
+ with open('tests/files/small.json', 'r') as f:
+ json_form = json.loads(f.read())
+
+ assert teixml2json(tei_xml) == json_form