From 3260b880c0003b7d55bd026b9f1d259256648b0e Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@archive.org>
Date: Wed, 12 Aug 2020 12:27:34 -0700
Subject: transform: more string cleaning

---
 tests/test_scrub.py | 20 +++++++++++++++++++-
 1 file changed, 19 insertions(+), 1 deletion(-)

(limited to 'tests')

diff --git a/tests/test_scrub.py b/tests/test_scrub.py
index 6541e13..b142c10 100644
--- a/tests/test_scrub.py
+++ b/tests/test_scrub.py
@@ -1,4 +1,4 @@
-from fatcat_scholar.schema import scrub_text
+from fatcat_scholar.schema import scrub_text, clean_str
 
 
 def test_scrub() -> None:
@@ -12,3 +12,21 @@ def test_scrub() -> None:
 
     for raw, fixed in vectors:
         assert fixed == scrub_text(raw)
+
+
+def test_clean_str() -> None:
+    vectors = [
+        (
+            "Di� Hekimli�i Fak�ltesi ��rencilerinde Temporomandibular Eklem Rahats�zl�klar�n�n ve A��z Sa�l��� Al��kanl�klar�n�n De�erlendirilmesi",
+            "Di� Hekimli�i Fak�ltesi ��rencilerinde Temporomandibular Eklem Rahats�zl�klar�n�n ve A��z Sa�l��� Al��kanl�klar�n�n De�erlendirilmesi",
+        ),
+        ("<jats:p>blah thing", "blah thing"),
+        ("title with <i>italics</i>", "title with italics"),
+        ("title with <sup>partial super", "title with partial super"),
+        ("", None),
+        ("&NA", None),
+        (None, None),
+    ]
+
+    for raw, fixed in vectors:
+        assert fixed == clean_str(raw)
-- 
cgit v1.2.3