Check entries with the same origin field value.

modules/check_tools.py

@@ -86,6 +86,12 @@ def check_publication(row):
         text = T('Report numbers contains "Note :" or ";"')
         li.append(text)
         
+    # duplicate by origin
+    ids = duplicate_origin(row.publications)
+    if len(ids) > 1:
+        text = T("Entries with duplicate origin [%s]") % ', '.join(ids)
+        li.append(text)
+    
     # specific fields for article
     if row.categories.usual == 'article':
         
@@ -153,7 +159,7 @@ def check_publication(row):
     return (li, ids)
 
 
-def _extend_ids(db, query, ids):
+def extend_ids(db, query, ids):
     """helper functions
     
     @type db: gluon.dal.DAL
@@ -210,16 +216,16 @@ def duplicate_article(publication):
     query = ((qmain) & (db.publications.title == publication['title']))
     query = ((query) & (db.publications.volume == publication['volume']))
     query = ((query) & (db.publications.pages == publication['pages']))
-    _extend_ids(db, query, ids)
+    extend_ids(db, query, ids)
 
     # publisher, volume and pages
     query = ((qmain) & (db.publications.volume == publication['volume']))
     query = ((query) & (db.publications.pages == publication['pages']))
-    _extend_ids(db, query, ids)
+    extend_ids(db, query, ids)
     
     # publisher and title
     query = ((qmain) & (db.publications.title == publication['title']))
-    _extend_ids(db, query, ids)
+    extend_ids(db, query, ids)
 
     return ids
 
@@ -259,17 +265,44 @@ def duplicate_conference(publication):
     query = ((qmain) & (db.publications.conference_title == publication['conference_title']))
     query = ((query) & (db.publications.conference_dates == publication['conference_dates']))
     query = ((query) & (db.publications.conference_town == publication['conference_town']))
-    _extend_ids(db, query, ids)        
+    extend_ids(db, query, ids)        
 
     # title, conference date and conference town
     query = ((query) & (db.publications.conference_dates == publication['conference_dates']))
     query = ((query) & (db.publications.conference_town == publication['conference_town']))
-    _extend_ids(db, query, ids)        
+    extend_ids(db, query, ids)        
     
     # title, conference title and conference town
     query = ((qmain) & (db.publications.conference_title == publication['conference_title']))
     query = ((query) & (db.publications.conference_town == publication['conference_town']))
-    _extend_ids(db, query, ids)        
+    extend_ids(db, query, ids)        
+    
+    return ids
+
+
+def duplicate_origin(publication):
+    """Look for publications with the same value in the origin field.
+
+    @type publication: dict or gluon.storage.Storage
+    @param publication: contains the publication fields and theirs values
+    
+    @rtype: list
+    @return: list of ids corresponding to duplicate entries
+    
+    """
+    ids = []
+    db = current.globalenv['db']
+
+    # protection against empty origin field
+    if not publication['origin']:
+        return ids
+
+    # look for publication with the same origin field
+    query = db.publications.origin == publication['origin']
+    set = db(query)
+    if set.count():
+        for row in set.select():
+            ids.append(str(row.id))
     
     return ids
 
@@ -301,7 +334,7 @@ def duplicate_report(publication):
     if 'id' in publication:
         qmain = ((qmain) & (db.publications.id != publication['id']))
     
-    _extend_ids(db, qmain, ids)        
+    extend_ids(db, qmain, ids)        
             
     return ids
     
\ No newline at end of file

static/CHANGELOG

@@ -4,6 +4,7 @@ HEAD
   - Bugs fixed.
   - Add and deploy the methods Record.host, oai and oai_url.
   - Add the application property reg_institute.
+  - Check entries with the same value in the origin field.
 
 0.8.5 (May 2014)
   - Migrate to plugin_dbui 0.6.1.1 and web2py 2.9.5