Best Method to Combine Merge GEDCOM files?

[Doug120745]

I have collected numerous GEDCOM files (50-60) from various family members and others. I'm now retired and now want to merge/extract these files into one entity. There are obviously duplicates, some with more information than others. I am trying to decide what methods to use. I've thought of trying a GEDCOM to SQL parser but defining the schema and all of the tables and dependencies seems overwhelming. I've also tried using manual duplicate/view/edit functions in some of the available genealogy apps, but that seems tedious and non-productive.

Has anyone faced this same situation?

Can TNG do this somehow? 

If I import all 50-60 files into TNG will I end up with a mess?

And how did you go about solving it?

[theKiwi]

TNG can do this, but it would be very nerve wracking and to my mind quite clumsy.

You'd be better to do this in a desktop software that has a robust Match and Merge feature, but even then, it's still somewhat nervewracking - you pretty much need to know the answer to "what should be left over" before you start, so you can know you've got there when you're done.

Roger

[edglimited]

4 hours ago, Doug120745 said:

If I import all 50-60 files into TNG will I end up with a mess?

Hate to say it, but you already have one, even before you begin. What Roger has suggested, would be a way of doing it. How successful that end result would be, would be up to you. I attempted something along the same lines some time ago, so as bat shit crazy as it was at the time, it might work for you.

You need to put all your Gedcoms in order of size, largest kb, to smallest kb. Then work out, what is in that largest file. Is it actual data on the people you want in your final tree, or is it sources and notes, that make up the kb's. Then you need to look at that file in a program like TNG. Does everything look right, names, places, dates of birth, dates of death, children, dates of birth..........you can see where this is going.!!!!!

If you're happy with what you have, then use that Gedcom as the basis for your final tree and slowly merge each Gedcom one by one. Again......"Does everything look right, names, places, dates of birth, dates of death, children, dates of birth"..........now you can see where this is going.!!!!! 🙃

If you want a program, that will do the merge but will take you some time (outside of TNG, hit me up)..Assuming your not in the rubber house or laughing continuously.

 

Cheers Drew

 

[Doug120745]

Thanks Roger. I think this is going to be a long manually performed task. Unless I can find someone who has written a gedcom-to-SQL parser program. However, I think that TNG has the pieces internally--TNG imports gedcom and stores it in SQL. And you can export a gedcom file from the MySQL database. What is unknown to me is how TNG would handle duplicate gedcom imports from different gedcom files. I guess I can experiment by setting up a throw-away test copy instance site and try it out. If it doesn't work out, I can trash it and start over with a fresh copy.

Thanks again.

[Doug120745]

Drew, I see that this is not going to be an easy task. The problem appears to be that the gedcom files are not really what one would define as a true database. Rather it is a list of gedcom field statements not necessarily in the same order. I see some Google hits on "gedcom parsers" and will take a look at them. ...still thinking about this...

 

[Chris Lloyd]

Have a look at mudcreek software.

 

[Doug120745]

Thanks, Chris

I downloaded the trial of GENViewer2 and installed it, but could not get it to run. I looked at the task manager window and it was loaded there and occupying 14%  of the CPU time. It was spinning its wheels evidently looking for or waiting for some system event. I'm running the latest version of Windows 10. Too bad as GENViewer looks to be quite useful.

[tngrlkrz]

3 hours ago, Doug120745 said:

GENViewer2

I still have GenViewer 1.23 circa 2007, and it still runs on my Windows 10 system.  I could PM it to you, if you think it would help.  Being 15 years old, not sure if useful though for your task.