summaryrefslogtreecommitdiffstats
path: root/python/tests/matched_importer.py
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@robocracy.org>2018-09-14 15:44:55 -0700
committerBryan Newbold <bnewbold@robocracy.org>2018-09-14 15:44:55 -0700
commit8b2c46c4d578ba4129316a07dd1e3981520b8921 (patch)
treeb8b7d6e6c85041dca17873bf7338f2ac9587777a /python/tests/matched_importer.py
parentba3adecfaf626b92fb5231cc445ce9e07fb5f877 (diff)
downloadfatcat-8b2c46c4d578ba4129316a07dd1e3981520b8921.tar.gz
fatcat-8b2c46c4d578ba4129316a07dd1e3981520b8921.zip
fixes to matched importer (and a test)
Diffstat (limited to 'python/tests/matched_importer.py')
-rw-r--r--python/tests/matched_importer.py32
1 files changed, 32 insertions, 0 deletions
diff --git a/python/tests/matched_importer.py b/python/tests/matched_importer.py
new file mode 100644
index 00000000..9cc6aa32
--- /dev/null
+++ b/python/tests/matched_importer.py
@@ -0,0 +1,32 @@
+
+import json
+import pytest
+from fatcat.matched_importer import FatcatMatchedImporter
+
+
+@pytest.fixture(scope="function")
+def matched_importer():
+ yield FatcatMatchedImporter("http://localhost:9411/v0")
+
+# TODO: use API to check that entities actually created...
+def test_matched_importer_batch(matched_importer):
+ with open('tests/files/example_matched.json', 'r') as f:
+ matched_importer.process_batch(f)
+
+def test_matched_importer(matched_importer):
+ with open('tests/files/example_matched.json', 'r') as f:
+ matched_importer.process_source(f)
+
+def test_matched_dict_parse(matched_importer):
+ with open('tests/files/example_matched.json', 'r') as f:
+ raw = json.loads(f.readline())
+ f = matched_importer.parse_matched_dict(raw)
+ assert f.sha1 == "00242a192acc258bdfdb151943419437f440c313"
+ assert f.md5 == "f4de91152c7ab9fdc2a128f962faebff"
+ assert f.mimetype == "application/pdf"
+ assert f.size == 255629
+ assert f.urls[1].url.startswith("http://journals.plos.org")
+ assert f.urls[1].rel == "web"
+ assert f.urls[0].url.startswith("https://web.archive.org/")
+ assert f.urls[0].rel == "webarchive"
+ assert len(f.releases) == 1