Deleting movie duplicates

[AimHere]

Hi,

How exactly does the "Delete movie duplicates" feature in "Optimize Database" work? What criteria does it use to determine whether a movie is a duplicate?

I ask because I did a test run on a copy of my database, which has some 1100 movies in it. After optimizing the database with "Delete movie duplicates" checked, it claimed to remove some 400 movies... apparently including at least 100 which were NOT actually duplicates of anything! These movies disappeared entirely from my main list. (Good thing I don't do experiments like this on my primary database!)

Aimhere

[nostra]

It checks if title is equal to title or original title of the other movie and year is same or one of the years is absent.

[AimHere]

And then it deletes the oldest entry[ies] in the database, leaving only the newest one? (I seem to recall this from another post.)

I did a little cleanup work on some fields in my database, going to try the "delete movie duplicates" function again and see what happens...

Aimhere

[nostra]

And then it deletes the oldest entry[ies] in the database, leaving only the newest one? (I seem to recall this from another post.)

It is not really determined which copy will be deleted, it will be just the first copy of the movie found by SQL query.

[AimHere]

And then it deletes the oldest entry[ies] in the database, leaving only the newest one? (I seem to recall this from another post.)

It is not really determined which copy will be deleted, it will be just the first copy of the movie found by SQL query.

Hmm, that seems rather arbitrary. It would be nice to be able to simply scan for duplicates without actually deleting them, and highlighting or grouping them in some way so they stand out, then I would be able to choose which one(s) to delete. I'd rather have PVD delete an "empty" Movie record (one with just the title and year) than a fully-populated one that I spent some considerable time filling in and checking!!!

Aimhere

[nostra]

I'll think about it

[AimHere]

One possible approach would be to have PVD delete the duplicate that has the least number of fields containing data. Movie records added when importing filmographies are likely only going to have a couple of fields populated (Title, original title, year, actors) while the ones *I* added (through "New Movie Master" or file scans) are gonna have the works... posters, screenshots, media information, genre, descriptions, etc. etc.

Aimhere

[nostra]

If I do so, the process of removing duplicates will become way too complex and slow, especially when invisible movies are to be checked, so I do not think it is a good idea.

Maybe I will implement the duplicate highlighting functionality.

[CAD]

i know pvd isnt a catalog program, but most people do use it as one.
Deleting duplicate functions need to be looked at with care as (especially now) people have the same movie in different formats
blueray/dvd

personally i would have it so it highlights duplicates and let the user make the choice what to do with them
Perhaps set the bookmarked flag ??

[nostra]

personally i would have it so it highlights duplicates and let the user make the choice what to do with them
Perhaps set the bookmarked flag ??

Yes, most probably it will work like this in future