Update RecordPubli to add protection in _process_authors.

modules/invenio_tools/recordpubli.py

@@ -154,8 +154,14 @@ class RecordPubli(Record):
             df = df.drop(columns.difference(["a", "e", "u"]), axis="columns")
 
             # add columns first_name, last_name and fmt_name
-            # warning: in some case split create more than 2 columns
-            df[columns4names] = df.a.str.split(u",", expand=True)[[0, 1]]
+
+            # protection -- split create 1, 2 and more than 2 columns
+            # former append when the author name is 'ATLAS collaboration'
+            df1 = df.a.str.split(u",", expand=True)
+            if df1.shape[1] < 2:
+                continue
+
+            df[columns4names] = df1[[0, 1]]
             df["fmt_name"] = df.a
 
             df.first_name = df.first_name.str.strip()
@@ -226,7 +232,12 @@ class RecordPubli(Record):
             d700 = d100
 
         else:
-            d100 = d700 = DataFrame({"a": [""], "u": [""]})
+            d100 = d700 = DataFrame({
+                "a": [""],
+                "first_name": [""],
+                "fmt_name": [""],
+                "last_name": [""],
+                "u": [""]})
 
         # ....................................................................
         #

tests/invenio_tools/Record/test_com_cds2241325.py

+# -*- coding: utf-8 -*-
+"""TALK
+
+    http://cds.cern.ch/record/2241325
+
+    Search for ttH production (H-> bb) at the ATLAS experiment
+    ATLAS Collaboration
+    Epiphany 2017
+
+No corrections are applied to the record.
+Allow to test the brute force decoding with its mistakes.
+
+Note:
+    * First author not defined and equal to ATLAS Collaboration
+    * No author defined
+    * No conference defined
+"""
+import pytest
+
+
+from invenio_tools import load_record, RecordConf
+
+
+@pytest.fixture(scope="module")
+def record():
+    return load_record('cds.cern.ch', 2241325)
+
+
+def test_affiliations(record):
+    assert record.is_affiliations() == False
+    assert record.is_affiliation_for_all() == False
+
+
+def test_authors(record):
+    assert record.authors() == ""
+
+
+def test_first_author(record):
+    assert record.first_author() == ""
+
+
+def test_first_author_institutes(record):
+    assert record.first_author_institutes() == ""
+

tests/invenio_tools/Record/test_com_cds2260981.py

+# -*- coding: utf-8 -*-
+"""TALK
+
+    http://cds.cern.ch/record/2260981
+
+    Measurement of WW/WZ production in semileptonic decay channels
+    and search for anomalous gauge couplings with the ATLAS detector at 8 TeV
+    ATLAS Collaboration
+    Spalla, Margherita
+    Epiphany 2017
+
+No corrections are applied to the record.
+Allow to test the brute force decoding with its mistakes.
+
+Note:
+    * First author not defined and equal to ATLAS Collaboration
+    * One author defined
+    * No conference defined
+"""
+import pytest
+
+
+from invenio_tools import load_record, RecordConf
+
+
+@pytest.fixture(scope="module")
+def record():
+    return load_record('cds.cern.ch', 2260981)
+
+
+def test_affiliations(record):
+    assert record.is_affiliations() == True
+    assert record.is_affiliation_for_all() == True
+
+
+def test_authors(record):
+    assert record.authors() == u"Spalla, Margherita"
+
+
+def test_first_author(record):
+    assert record.first_author() == u"Spalla, Margherita"
+
+
+def test_first_author_institutes(record):
+    assert record.first_author_institutes() == "INFN Pisa and Universita' di Pisa, Dipartimento di Fisica"