From eeb456c16d016d8523023f787597efae7a6317b9 Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@archive.org>
Date: Mon, 26 Jul 2021 15:11:44 -0700
Subject: better parsing of year as integer in refs pipeline

---
 fatcat_scholar/schema.py | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

(limited to 'fatcat_scholar/schema.py')

diff --git a/fatcat_scholar/schema.py b/fatcat_scholar/schema.py
index bc6b016..0fcf56e 100644
--- a/fatcat_scholar/schema.py
+++ b/fatcat_scholar/schema.py
@@ -301,9 +301,12 @@ class RefTarget(BaseModel):
 
 
 def clean_small_int(raw: Optional[str]) -> Optional[int]:
-    if not raw or not raw.isdigit():
+    if not raw or not raw.strip().isdigit():
+        return None
+    try:
+        val = int(raw.strip())
+    except ValueError:
         return None
-    val = int(raw)
     if abs(val) > 30000:
         return None
     return val
@@ -318,6 +321,7 @@ def test_clean_small_int() -> None:
     assert clean_small_int("1200003") == None
     assert clean_small_int("-123") == None
     assert clean_small_int("48844") == None
+    assert clean_small_int("1990²") == None
 
 
 def doi_split_prefix(doi: str) -> str:
-- 
cgit v1.2.3