Merge tag '2023-12-11_withdrawn_journals_without_None' into develop

Update to withdrawn journals script
DOAJ · Dec 11, 2023 · 6de7cb8 · 6de7cb8
2 parents c76aab2 + cd0168f
commit 6de7cb8
Showing 1 changed file with 13 additions and 15 deletions.
diff --git a/portality/scripts/231127_withdrawn_journals.py b/portality/scripts/231127_withdrawn_journals.py
@@ -1,22 +1,21 @@
 from portality.models import Journal
 import csv
 
-
 NOT_IN_DOAJ = {
-    "query" : {
-        "bool" : {
-            "must" : [
-                {"term" : {"admin.in_doaj" : False}}
+    "query": {
+        "bool": {
+            "must": [
+                {"term": {"admin.in_doaj": False}}
             ]
         }
     }
 }
 
 IN_DOAJ = {
-    "query" : {
-        "bool" : {
-            "must" : [
-                {"term" : {"admin.in_doaj" : True}}
+    "query": {
+        "bool": {
+            "must": [
+                {"term": {"admin.in_doaj": True}}
             ]
         }
     }
@@ -25,6 +24,7 @@
 if __name__ == "__main__":
 
     import argparse
+
     parser = argparse.ArgumentParser()
     parser.add_argument("-o", "--out", help="output file path")
     args = parser.parse_args()
@@ -37,7 +37,7 @@
     with open(args.out, "w", encoding="utf-8") as f:
         writer = csv.writer(f)
         writer.writerow(["ID",
-                         "Journal Name",
+                         "Journal Title",
                          "Journal URL",
                          "E-ISSN",
                          "P-ISSN"
@@ -49,19 +49,17 @@
             in_doaj_issns.add(bibjson.get_one_identifier(bibjson.E_ISSN))
             in_doaj_issns.add(bibjson.get_one_identifier(bibjson.P_ISSN))
 
+        # Remove the value None so we don't match on all records
+        in_doaj_issns.discard(None)
 
         for journal in Journal.iterate(q=NOT_IN_DOAJ, keepalive='5m', wrap=True):
             bibjson = journal.bibjson()
             eissn = bibjson.get_one_identifier(bibjson.E_ISSN)
             pissn = bibjson.get_one_identifier(bibjson.P_ISSN)
-            if (eissn not in in_doaj_issns and pissn not in in_doaj_issns):
+            if eissn not in in_doaj_issns and pissn not in in_doaj_issns:
                 writer.writerow([journal.id,
                                  bibjson.title,
                                  bibjson.get_single_url(urltype="homepage"),
                                  eissn,
                                  pissn,
                                  ])
-
-
-
-