Welcome to mirror list, hosted at ThFree Co, Russian Federation.

github.com/stanfordnlp/stanza.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat (limited to 'stanza/tests/test_prepare_resources.py')
-rw-r--r--stanza/tests/test_prepare_resources.py30
1 files changed, 30 insertions, 0 deletions
diff --git a/stanza/tests/test_prepare_resources.py b/stanza/tests/test_prepare_resources.py
new file mode 100644
index 00000000..f7f485e4
--- /dev/null
+++ b/stanza/tests/test_prepare_resources.py
@@ -0,0 +1,30 @@
+import pytest
+
+import stanza
+import stanza.resources.prepare_resources as prepare_resources
+
+from stanza.tests import *
+
+pytestmark = [pytest.mark.travis, pytest.mark.pipeline]
+
+def test_split_model_name():
+ # Basic test
+ lang, package, processor = prepare_resources.split_model_name('ro_nonstandard_tagger.pt')
+ assert lang == 'ro'
+ assert package == 'nonstandard'
+ assert processor == 'pos'
+
+ # Check that nertagger is found even though it also ends with tagger
+ # Check that ncbi_disease is correctly partitioned despite the extra _
+ lang, package, processor = prepare_resources.split_model_name('en_ncbi_disease_nertagger.pt')
+ assert lang == 'en'
+ assert package == 'ncbi_disease'
+ assert processor == 'ner'
+
+ # assert that processors with _ in them are also okay
+ lang, package, processor = prepare_resources.split_model_name('en_pubmed_forward_charlm.pt')
+ assert lang == 'en'
+ assert package == 'pubmed'
+ assert processor == 'forward_charlm'
+
+