Handling of duplicates
This page informs about existing and planned features that help with detection & cleaning of duplicates:
refbase offers you to flag records as being a "duplicate" or "original" (master) record. If records have been flagged as duplicate (or original) entries, a small icon is displayed next to the particular records in list view to inform the user about it.
Currently, flagging of records has to be done manually and there is no web interface to do so. This will change, but in the meantime, admin users can flag records by entering SQL queries similar to the ones below into the form provided by sql_search.php.
In order to flag a record (say, with serial number '123') as an original record, use:
UPDATE refs SET orig_record = -123 WHERE serial = 123
In order to flag records with serials '456', '7890' and '13579' as duplicates and link them with their original record (in this example, the record with serial '123'):
UPDATE refs SET orig_record = 123 WHERE serial RLIKE "^(456|7890|13579)$"
Features in development
We have started to develop a simple duplicate detection mechanism. You can try this new feature online at our refbase beta database
To test this feature:
- log in to the database
- perform a search, e.g. http://beta.refbase.net/show.php?title=ice
- click on the "dups" link that's displayed in the status message
- hit the "Find Duplicates" button in the displayed form
This will find many (if not all) duplicate records for your current query. You can play with the form settings to tweak the results list.
The feature is currently only available in our refbase development version (within the bleeding-edge branch of our SVN source code repository at SourceForge):
Future versions should auto-detect duplicates, link duplicate entries with their original ones and allow users to migrate their user-specific data from a duplicate record to the original. It would be also nice if duplicate records could be hidden from the interface. See here for more info about the planned duplicate features.