Merge tag '3725_improve_script' into develop

Split results into multiple files
DOAJ · Nov 6, 2023 · 225665c · 225665c
2 parents 6c85bda + 4fdec0b
commit 225665c
Showing 1 changed file with 19 additions and 4 deletions.
diff --git a/portality/scripts/230609_find_articles_with_invalid_issns.py b/portality/scripts/230609_find_articles_with_invalid_issns.py
@@ -22,9 +22,19 @@
     parser.add_argument("-o", "--out", help="output file path", required=True)
     args = parser.parse_args()
 
-    with open(args.out, "w", encoding="utf-8") as f:
-        writer = csv.writer(f)
-        writer.writerow(["ID", "PISSN", "EISSN", "Journals found with article's PISSN", "In doaj?", "Journals found with article's EISSN", "In doaj?", "Error"])
+
+    with open(args.out+"notfound.csv", "w", encoding="utf-8") as f_notfound, open(args.out+"-identical.csv", "w", encoding="utf-8") as f_identical, open(args.out+"-others.csv", "w", encoding="utf-8") as f_others:
+        writer_notfound = csv.writer(f_notfound)
+        writer_notfound.writerow(["ID", "PISSN", "EISSN", "Journals found with article's PISSN", "In doaj?",
+                         "Journals found with article's EISSN", "In doaj?", "Error"])
+
+        writer_identical = csv.writer(f_identical)
+        writer_identical.writerow(["ID", "PISSN", "EISSN", "Journals found with article's PISSN", "In doaj?",
+                         "Journals found with article's EISSN", "In doaj?", "Error"])
+
+        writer_others = csv.writer(f_others)
+        writer_others.writerow(["ID", "PISSN", "EISSN", "Journals found with article's PISSN", "In doaj?",
+                         "Journals found with article's EISSN", "In doaj?", "Error"])
 
         for a in models.Article.iterate(q=IN_DOAJ, page_size=100, keepalive='5m'):
             article = models.Article(_source=a)
@@ -54,4 +64,9 @@
                             j_e_in_doaj.append(jobj.is_in_doaj())
                         else:
                             j_e_in_doaj.append("n/a")
-                writer.writerow([id, pissn, eissn, j_p, j_p_in_doaj, j_e, j_e_in_doaj, str(e)])
+                if (str(e) == "The Print and Online ISSNs supplied are identical. If you supply 2 ISSNs they must be different."):
+                    writer_identical.writerow([id, pissn, eissn, j_p, j_p_in_doaj, j_e, j_e_in_doaj, "Identical ISSNs"])
+                elif (str(e) == "ISSNs provided don't match any journal."):
+                    writer_notfound.writerow([id, pissn, eissn, j_p, j_p_in_doaj, j_e, j_e_in_doaj, "No matching journal found."])
+                else:
+                    writer_others.writerow([id, pissn, eissn, j_p, j_p_in_doaj, j_e, j_e_in_doaj, str(e)])