* migrate from Legacy - keep ids

[teresa]

I would like to migrate a GEDCOM files from Legacy, but keep the same ids as I had in Legacy. In Legacy I don't fill gaps (because I export to TNG and reusing ids could cause errors / confusion)
When I move the GEDCOM to Family Historian I find that most people have the same ids, but some are changed because the gaps are filled. How can I stop this happening?

Thank you.

[tatewise]

It was my understanding that FH does NOT renumber Individual Records unless one or more of the imported numbers are very large (more than about 10 digits).
But if just one Record number is large, then they all get renumbered.
Unfortunately, this feature is not documented.

Could you check that NONE of your Legacy Individual Record numbers are large.

Maybe it is possible to renumber them in Legacy and TNG, so they import unchanged into FH.

Alternatively, it would be possible to edit the GEDCOM file to adjust the large numbers.

[teresa]

Thanks for reply.
The largest RIN in my Gedcom from Legacy is 13892, so nowhere near 10 digits. And they are not all getting renumbered, but most are as as soon as FH encounters a gap it fills it with the next record (and I have a gap at i148).
As I said, I wish to keep my Legacy numbers because of my website; if I renumbered in TNG it would not only be enormous hassle but also inconvenience anyone with bookmarks to people as the url includes the id, so I don't consider renumbering in Legacy an option for me.
I guess if I didn't have gaps it would be fine, and maybe FH keeping numbers when there are no gaps is why you thought it didn't renumber.

What I could do is fill the gaps with private dummy people but I thought there should be a way of importing to FH without that hassle. It seems from what I've read that if Legacy held the nos. in a non-standard place then FH has a way of copying them, so the same numbers can be kept, so I thought it should be able to cope with them coming across in the standard place.

[tatewise]

I am convinced FH does NOT renumber and does retain gaps, but only if the RIN is entirely numeric and less than 10 digits.

Could you please post a snippet or two from the GEDCOM file showing how Legacy formats the RIN.
The parts I am interested in look like this:

Code: Select all

0 @I12345@ INDI
1 NAME John /Smith/
1 SEX M

where 1234 is the RIN.

The gap at RIN = 148 and RIN = 13892 would be good choices.

[teresa]

Here is a chunk from my GEDCOM file as exported from Legacy - id 146 and 148 (I was slightly wrong in what I said before, the gap is 147, so the first renumbered is 148):

0 @I146@ INDI
1 NAME Charles Middleton /Inman/
2 GIVN Charles Middleton
2 SURN Inman
1 SEX M
1 BIRT
2 DATE 1 Mar 1869
2 PLAC Harrogate, Yorkshire
1 DEAT
2 DATE 27 Oct 1940
2 PLAC Knaresborough Reg Dist, Yorkshire
2 NOTE @DI146@
1 OCCU Cabinet maker, 1891.
1 EDUC scholar (1881)
1 _UID ECE5711D350D42EF9F9E9347588AA9188C5F
1 CHAN
2 DATE 15 Dec 2011
3 TIME 12:48
1 FAMS @F86@
1 FAMS @F1387@
1 FAMC @F70@
1 NOTE @NI146@
0 @NI146@ NOTE
1 CONC Living with grandparents etc  in 1871;
1 CONT with grandfather (Philip Inman) in 1881 & 1891. 
1 CONT Found family in 1901 living at High Bondend - transcribed on ancestry as Suman
1 CONC ! Photo with recast bells in 1925(?) (as he was parish clerk)
0 @DI146@ NOTE
1 CONC ref: q4 1940 Knaresborough RD 9a 321aged 71
0 @I148@ INDI
1 NAME George /Shaw/
2 GIVN George
2 SURN Shaw
1 SEX M
1 BIRT
2 DATE Abt 1854
2 PLAC Masham, Yorkshire
2 NOTE @BI148@
1 DEAT
2 DATE Bet 1901 and 1911
2 PLAC (probably), Knaresborough, Yorkshire
2 NOTE @DI148@
1 OCCU Bricklayer
2 DATE 1891-1901
2 PLAC Knaresborough, Yorkshire
1 _UID 326577BF756C4E59AB6C342E35DCB4BE5185
1 CHAN
2 DATE 17 May 2011
3 TIME 09:06
1 FAMS @F87@
1 NOTE @NI148@
0 @NI148@ NOTE
1 CONC In 1881 George was a local bailiff staying at the White Horse, in Knaresboroug
1 CONC h High St. - although his wife and baby son were living in Knaresborough.
0 @BI148@ NOTE
1 CONC Aged 46 at time of 1901 census
0 @DI148@ NOTE
1 CONC Hannah was shown as a widow in 1911 census, but so many George Shaws it's a bi
1 CONC t difficult to determine which is death of this one.
1 CONT Most likely seems one whose death was recorded in Knaresborough RD in Q1 1903
1 CONC , age given as 49 (9a 71).

I opened this in Family Historian, went to menu under 'file' and clicked save.
Here is a chunk from the same file after saving in FH:

0 @I146@ INDI
1 NAME Charles Middleton /Inman/
2 GIVN Charles Middleton
2 SURN Inman
1 SEX M
1 BIRT
2 DATE 1 MAR 1869
2 PLAC Harrogate, Yorkshire
1 DEAT
2 DATE 27 OCT 1940
2 PLAC Knaresborough Reg Dist, Yorkshire
2 NOTE @N236@
1 OCCU Cabinet maker, 1891.
1 EDUC scholar (1881)
1 FAMC @F70@
1 FAMS @F86@
1 FAMS @F1105@
1 NOTE @N235@
1 _UID ECE5711D350D42EF9F9E9347588AA9188C5F
1 CHAN
2 DATE 15 DEC 2011
3 TIME 12:48:00
0 @I147@ INDI
1 NAME George /Shaw/
2 GIVN George
2 SURN Shaw
1 SEX M
1 BIRT
2 DATE ABT 1854
2 PLAC Masham, Yorkshire
2 NOTE @N238@
1 DEAT
2 DATE BET 1901 AND 1911
2 PLAC (probably), Knaresborough, Yorkshire
2 NOTE @N239@
1 OCCU Bricklayer
2 DATE BET 1891 AND 1901
2 PLAC Knaresborough, Yorkshire
1 FAMS @F87@
1 NOTE @N237@
1 _UID 326577BF756C4E59AB6C342E35DCB4BE5185
1 CHAN
2 DATE 17 MAY 2011
3 TIME 09:06:00
0 @I148@ INDI
1 NAME Emma ("Em") /Shaw/
2 GIVN Emma ("Em")
2 NICK Em
2 SURN Shaw
...

I have tried opening the Legacy GEDCOM direct, and opening Family Historian, then starting new project and importing the GEDCOM - same result each time.

Last person / highest id in GEDCOM - as from Legacy:
0 @I13892@ INDI
1 NAME Ann //
2 GIVN Ann
1 SEX F
1 BIRT
2 DATE Abt 1818
2 PLAC Buckfastleigh, Devon
2 NOTE @BI13892@
1 DEAT Y
1 _UID CF286D3E635A4EFB9E025D0B37119509966A
1 CHAN
2 DATE 23 Sep 2013
3 TIME 10:08
1 FAMS @F5957@
0 @BI13892@ NOTE
1 CONC from the 1861 census where shown as aged 42

Thanks for your interest,
Teresa

[Jane]

If you open the file in something like PSPAD and do a list search for 0 @I do any of the ID's have alphabetic characters in them other than I or very long numbers?

As Mike says FH only renumbers when it finds a reference number it can't process, so we need to find the lndividual which is causing the renumber to be triggered

[tatewise]

For example, notice how all the Note Records have been completely renumbered (NOTE @N236@, etc) because in Legacy they start with BI and DI and NI rather than N.
For the number to be retained in FH, Note Record Ids must all start with N and be numeric thereafter.
Similarly, Individual Record Ids must all start with I and be numeric thereafter.

This appears to be a common convention, but not required by the GEDCOM specification, which only requires all the Ids to be unique, and does not require them to be retained when transferred from system to system.

[Jane]

I have done some experiments and if FH renumbers any record types it renumbers all of them.

Have you purchased FH yet? If so I could do a plugin to copy the Individual IDs to a custom ID and once FH has imported you can use the Work with Record Identifers to copy the the ID's back to the main ones.

If you don't have a license yet, if you send me the file, I will do the plugin and send the file back to you.

[teresa]

Thank you, yes I have purchased FH. The plugin you suggest would be great.

Teresa

[Jane]

Can you give this one a try, it will copy the numeric part of the ID into the Automated Record ID field.

Once you have run it and opened the updated version in to FH, then use the Tools>Work with Record Identifiers to copy the numbers back into the Record Id fields

Legacy - Copy Record IDs to Automated IDs.fh_lua

Copy IDs in gedcom to Automated Field

(1.54 KiB) Downloaded 345 times

[tatewise]

Jane forgot to mention that you can perform that trick with all 8 Record Types if you wish.

But if any Record Type, such as the Note Records, uses the same numerical part more than once, then the request to copy its Automated Record Id will be ignored.