Improved algorithm to find cppm authors.

modules/harvest_tools.py

@@ -71,34 +71,35 @@ def cppm_authors(record, db=None, id_project=None, id_team=None):
     @type id_team: int or None
     @param id_team:
     
-    @rtype: unicode or None
-    @return: author names separated by ', ' or UNKNOWN or none 
+    @rtype: str
+    @return: author names separated by comma or an empty string 
+    when there is no authors
     
     """
     # find authors of the institute signing the record
     rep = record.find_authors_by_institute(current.app.reg_institute, 
                                            family_name_fr)
+    
     if rep:
         return rep
     
-    # nothing found, check that the institute list is defined
-    # if institutes are not defined, use the rescue list store in the database
-    if not record.is_institute_defined():
-
-        if db == None or id_team == None:
-            return UNKNOWN
-
-        row = db.cppm_authors(year=record.year(), 
-                              id_projects=id_project, 
-                              id_teams=id_team)
-        if not row:
-            return None
+    # nothing found, use the rescue list store in the database
+    #
+    # NOTE: the following code cover all the cases: 
+    # institutes not defined or defined only for a fraction of the authors
+    #
+    if db == None or id_team == None:
+        return ''
+
+    row = db.cppm_authors(year=record.year(), 
+                          id_projects=id_project, 
+                          id_teams=id_team)
+    if not row:
+        return None
 
-        reference = row['authors'].split(', ')
-        return fix_cppm_authors(record, reference)
+    reference = row['authors'].split(', ')
+    return fix_cppm_authors(record, reference)
         
-    return None
-
 
 def family_name_fr(x):
     """Extract the family name when the full name is encoded as C{J. Doe}.
@@ -143,14 +144,14 @@ def fix_cppm_authors(record, reference):
     @type reference: list
     @param reference: list of author names
     
-    @rtype: unicode
+    @rtype: str
     @return: author names separated by comma or an empty string 
     when there is no authors
     
     """
     s1 = set(record.authors_as_list())
     s2 = set(reference)
-    
+
     # sort according to family name
     li = list(s1.intersection(s2))
     li.sort(key=family_name_fr)
@@ -1125,22 +1126,6 @@ class Articles(PublicationsTool):
                                      volume=volume,
                                      year=year)
 
-        # In April 2014 the pages field of some inspirehep record was modified
-        # the following code handle this case
-        if not id:
-            id = get_id(db.publications, id_projects=self.harvester.id_projects,
-                                         id_publishers=id_publisher,
-                                         id_teams=self.harvester.id_teams,
-                                         title=title,
-                                         volume=volume,
-                                         year=year)
-            
-            if id:
-                if self.selector.mode != DRY_RUN:
-                    db.publications[id] = dict(pages=pages)
-                    
-                self.logs[-1].modify(MSG_FIX_PAGE)
-                
         # fix orign field
         if id and not db.publications[id].origin:
             if self.selector.mode != DRY_RUN: