Migrate CheckAndFix: methods required by the harvester Article.

languages/fr-fr.py

@@ -538,7 +538,7 @@
 'Reject conference dates is not well formed': 'Rejecté les dates de la conférence dates sont mal formatté',
 'Reject countries is not defined': "Rejeté le pays n'est pas définie",
 'Reject editor is not well formed': "Rejeté l'éditeur est mal formatté",
-'Reject incomplete paper reference': 'Rejeté la référence du papier est incomplète',
+'Reject incomplete paper reference. Check ': 'Rejeté la référence du papier est incomplète. Vérifiez ',
 'Reject invalid country': 'Rejeté pays inconnu',
 'Reject invalid publisher': "Rejeté l'abbréviatioin de la revue n'est pas valide",
 'Reject no %s authors': "Rejeté pas d'autheur(s) du %s",

modules/invenio_tools/recordpubli.py

 """ invenio_tools.recordpubli
 
 """
+import numpy as np
 import re
 
 
@@ -527,22 +528,27 @@ class RecordPubli(Record):
         return True
 
     def is_published(self):
-        """``True`` is the record is published and contains a complet set
-        of publication infromation (title, volume, year and pagination).
+        """``True`` is the record is published and contains a full set
+        of publication information (title, volume, year and pagination).
 
         Returns:
             bool:
 
         """
-        df = self[u"publication_info"]
-
-        query = \
-            (df.title.str.len() > 0) \
-            & (df.volume.str.len() > 0) \
-            & (df.year.str.len() > 0) \
-            & (df.pagination.str.len() > 0)
+        # NOTE
+        # * df.columns are title, volume, year and pagination
+        # * df can contains one or more rows due to erratum.
+        # * assume that the first row is the oldest one and corresponds tp
+        #   the first publication
+        # * the row contains empty string when the record is not published.
+        # * iloc[0] returns a serie where the index are the column's name
+        #
+        columns = (self[u"publication_info"].iloc[0]
+                   .replace("", np.nan)
+                   .dropna()
+                   .index)
 
-        return len(df[query]) > 0
+        return len(columns.intersection(PAPER_REFERENCE_KEYS)) == 4
 
     def is_with_erratum(self):
         """``True`` when the record contains erratum data.
@@ -774,7 +780,7 @@ class RecordPubli(Record):
         Returns:
             unicode:
                 * format are"YYYY-MM", "YYYY-MM-DD", "DD MMM YYYY", *etc.*
-                * Empty sring when not defined.
+                * Empty string when not defined.
 
         """
         return self._get(u"prepublication", u"date")

tests/basis/test_04_RecordPubli.py

@@ -160,6 +160,13 @@ def test_collaboration(reccds):
 def test_is_published(reccds):
     assert reccds.is_published()
 
+    # Paper is published but there are error in the paper reference
+    # Correct reference is Eur. Phys. J. C 75 (2015) 158
+    # But volume is not defined and pagination is wrong (75)
+
+    record = load_record("cds.cern.ch", 1753190)
+    assert not record.is_published()
+
 
 def test_is_with_erratum(reccds):
     assert not reccds.is_with_erratum()

tests/basis/test_05_RecordConf.py

@@ -80,22 +80,3 @@ def test_conference_url(reccds, recins):
 def test_conference_year(reccds, recins):
     assert reccds.conference_year() == "2010"
     assert recins.conference_year() == "2010"
-
-
-# def test_reference_conference_id(reccds):
-#     from pprint import pprint
-#     pprint(reccds["meeting_name"])
-#     pprint(recins["meeting_name"])
-#     assert reccds.reference_conference_id() == "1181092"
-#
-#
-# def test_reference_conference_key(reccds):
-#     assert reccds.reference_conference_key() == "rome20101206"
-#
-#
-# def test_reference_conference_proceeding(reccds):
-#     assert reccds.reference_conference_proceeding() == "1313736"
-#
-#
-# def test_reference_conference_talk(reccds):
-#     assert reccds.reference_conference_talk() == "1313736"

tests/basis/test_10_CheckAndFix_article.py

+# -*- coding: utf-8 -*-
+"""test_10_CheckAndFix_article
+
+    * Test CheckAndFix methods for article:
+        - format_editor
+        - publisher
+        - paper_reference
+        - submitted
+        - year
+        - format_author
+        - get_my_authors
+
+    * Same article in cds.cern.ch and inspirehep.net
+      Phys. Rev. D 95 (2017) 052005
+
+"""
+import pytest
+from harvest_tools.checkandfix import CheckAndFix
+from harvest_tools.exception import ToolException
+from invenio_tools import load_record
+
+
+@pytest.fixture(scope="module")
+def reccds():
+    return load_record("cds.cern.ch", 2242641)
+
+
+@pytest.fixture(scope="module")
+def recins():
+    return load_record("inspirehep.net", 1509922)
+
+
+@pytest.fixture(scope="module")
+def svc():
+    return CheckAndFix()
+
+
+def test_format_editor(svc, reccds, recins):
+
+    # cds
+    assert reccds.paper_editor() == "Phys. Rev. D"
+    assert reccds.paper_volume() == "95"
+
+    svc.format_editor(reccds)
+
+    assert reccds.paper_editor() == "Phys. Rev. D"
+    assert reccds.paper_volume() == "95"
+
+    # inspire
+    assert recins.paper_editor() == "Phys.Rev."
+    assert recins.paper_volume() == "D95"
+
+    svc.format_editor(recins)
+
+    assert recins.paper_editor() == "Phys. Rev. D"
+    assert recins.paper_volume() == "95"
+
+
+def test_publisher(svc, reccds):
+    assert svc.publisher(reccds) is None
+
+
+def test_paper_reference(svc, reccds):
+
+    # check recovery procedure using DOI
+    # remove the publisher and volume information
+    paper_ref = reccds.paper_reference()
+
+    reccds["publication_info"].loc[0, ["title", "volume"]] = ["", ""]
+    svc.paper_reference(reccds)
+
+    assert reccds.paper_reference() == paper_ref
+
+    # Paper is published but there are error in the paper reference
+    # Correct reference is Eur. Phys. J. C 75 (2015) 158
+    # But volume is not defined and pagination is wrong (75)
+    # It is not possible to recover it from the doi data.
+    record = load_record("cds.cern.ch", 1753190)
+    with pytest.raises(ToolException):
+        svc.paper_reference(record)
+
+
+def test_submitted(svc, reccds, recins):
+
+    assert reccds.submitted() == "19 Jan 2017"
+    assert recins.submitted() == "2017-01-19"
+
+    svc.submitted(reccds)
+    reccds.submitted()
+
+    assert reccds.submitted() == "2017-01-19"
+
+    # test the case 19 01 2017
+    reccds[u"prepublication"][u"date"] = "19 01 2017"
+    svc.submitted(reccds)
+    assert reccds.submitted() == "2017-01-19"
+
+    # test the case 2017
+    reccds[u"prepublication"][u"date"] = "2017"
+    svc.submitted(reccds)
+    assert reccds.submitted() == "2017-01"
+
+
+def test_format_authors(svc, reccds):
+
+    authors = reccds.authors_as_list()
+
+    assert len(authors) == reccds["number_of_authors"]
+    assert authors[0] == u"Aaij, Roel"
+    assert authors[1] == u"Adeva, Bernardo"
+    assert authors[344] == u"Koopman, Rose"
+    assert authors[-1] == u"Zucchelli, Stefano"
+
+    svc.format_authors(reccds, fmt="F. Last")
+    authors = reccds.authors_as_list()
+
+    assert authors[0] == u"R. Aaij"
+    assert authors[1] == u"B. Adeva"
+    assert authors[344] == u"R. Koopman"
+    assert authors[-1] == u"S. Zucchelli"
+
+
+def test_get_my_authors(svc, reccds):
+
+    svc.format_authors(reccds, fmt="F. Last")
+    assert svc.get_my_authors(reccds, sep=u"|", sort=True) is None
+
+    my_authors = reccds.my_authors.encode("utf-8")
+
+    assert my_authors == "J. Arnau Romeu|E. Aslanides|J. Cogan|" \
+                         "K. De Bruyn|R. Le Gac|O. Leroy|" \
+                         "G. Mancinelli|M. Martin|A. Mordà|" \
+                         "J. Serrano|A. Tayduganov|A. Tsaregorodtsev"