Editing GEDCOM Files: Don't Do It!

[Jeff89]

I asked a question in "Using Reunion 10" about why the Notes in my family files in Reunion 10 appeared like this:

Be fore s pending the las t thirt y five years o f his l ife in norther n In diana

and then discovered that that is how they appear in the GEDCOM file. That question was taken down in the Using Reunion 10 forum, apparently because it does not have anything to do with Reunion 10. Fair enough.

Yesterday, I wasted an entire day's worth of work by trying to edit the GEDCOM file to correct those mistakes and ended up unintentionally corrupting the file in such a way that when I imported the file to Reunion 10, the links between people were one-way - parents were visible, but children were not.

So here's the lesson learned: be exceedingly careful when you edit a GEDCOM file. You may end up ruining it. Instead, I would recommend that you copy and paste the Note field in Reunion to a word processor, correct the problems, save the changes and then copy and paste the notes back into Reunion. It is a fool's errand to try to edit a GEDCOM file.

Do note that because Reunion is not a word processor that it has its limitations in displaying text. I would recommend not making your lines longer than 55 picas in your word processor. If you make your lines longer and then copy and paste that file into Reunion, your text may not look right when you generate a report from Reunion. I haven't tried importing Notes to Reunion; that is something that might be worthwhile considering.

Jeff

[WilliamTaber]

Re: Editing GEDCOM Files: Don't Do It!

I too have seen the broken words and sentences upon exporting to a GEDCOM and importing into World Connect on Rootsweb. Now, I never see GEDCOMs with broken words like this:

Be fore s pending the las t thirt y five years o f his l ife in norther n In diana

For GEDCOM Export options, I now always choose Destination GEDCOM 5.5, Character Set ANSEL, Line Length 255. Ever since adopting the use of these option choices, I have never encountered the broken word problem in GEDCOMs I've exported. You might try that.

[donworth]

Re: Editing GEDCOM Files: Don't Do It!

Jeff,

I'm so sorry you had such a bad experience with editing your GED file!

I do have to say that I have not had as bad an experience, and I have edited GEDs a lot of times with success - notwithstanding a few missteps along the way. I suppose, over time, I figured out how the GED was structured enough that I could fix inadvertent mis-edits. And, since most of the mass editing I did was either with a global find and replace in a text editor, a change in the way I exported it from the original database, or by using an AppleScript or BASIC program I wrote, I was always able to go back a step if necessary and start over without losing too much time. I also only ever imported GEDs into a new, empty tree - never merging them into an existing tree that I cared about. That way if what I got was junky, I could just delete the tree and start over.

If folks do that they can probably safely play around with editing GEDs and learn as they go.

That said, the example line you gave looks pretty bad - I'm not sure the little program I wrote to fix inserted spaces would have worked since it wasn't as predictable where the gaps landed in your example as it was in mine.

Don

[Jeff89]

Re: Editing GEDCOM Files: Don't Do It!

For GEDCOM Export options, I now always choose Destination GEDCOM 5.5, Character Set ANSEL, Line Length 255. Ever since adopting the use of these option choices, I have never encountered the broken word problem in GEDCOMs I've exported. You might try that.

This particular GEDCOM file is one that I obtained 6 or 7 years ago and never did anything with because I had no time. Now that I'm retired, though ... If I ever get this one cleaned up and corrected, I'd be happy to export it using the options you specified. It's an astonishing mess, but one that affords wonderful learning opportunities. Thanks!

[Jeff89]

Re: Editing GEDCOM Files: Don't Do It!

Jeff,

I do have to say that I have not had as bad an experience, and I have edited GEDs a lot of times with success - notwithstanding a few missteps along the way. ... And, since most of the mass editing I did was either with a global find and replace in a text editor, ... I also only ever imported GEDs into a new, empty tree - never merging them into an existing tree that I cared about. That way if what I got was junky, I could just delete the tree and start over.
Don

Well, I used the global find and replace feature of Mariner Write and I ended up with the GEDCOM tag INID being changed to IndianaID! The CHAR tag changed to CHArkansas and so on. I unchecked the "Ignore Case" thinking that would help but it did no good, either - Mariner just changed every instance of ND to North Dakota! Maybe Mariner Write's editing function is god-awful and I should have used a dedicated editor program instead. I don't know. The bottom line is that the family file that I created with that GEDCOM had only one-way parent/child links. I thought that by using a leading space and a trailing space after IN in the "Find" box, for example, that I would prevent those kinds of errors, but it obviously didn't work.

I definitely agree with you about not importing into an existing family file - that would be a real tragedy!

Like I wrote - it's been a great educational experience!

Jeff