GEDCOM problems
Steve Hayes
compare notes, and encountered unexpected problems.
He sent me a GEDCOM file, and I tried to import it into PAF 4, but failed.
I imported it into Legacy 5, and then re-exported it as a GEDCOM file, which
PAF 4 then imported with no problems.
I also sent him a GEDCOM file exported by PAF 4, which his program could not
handle.
I imported that file into Legacy with no problems and reexported it, and have
sent him the reexported version, in the hope that he will be able to import
it.
I've also asked him what program he uses -- does anyone know of any progrtams
that have problems in exchanging data with PAF?
Here are some of the errors reported:
GEDCOM Import Log - 12 Nov 2009
File: J:\John Sandercock-Mary Riddle.GED
Created by:
Submitted by: <UNKNOWN>
Individuals: 0
Marriages: 0
Sources: 0
Repositories: 0
Errors: 3702
The following lines in the GEDCOM file could not be processed due to
unrecognizable tags or invalid data:
ERROR 1: John Sandercock-Mary Riddle.GED, line 1: Bad level number.
ERROR 2: Level number doesn't increment by one. Skipped lines 2 through 2.
ERROR 3: Level number doesn't increment by one. Skipped lines 3 through 3.
ERROR 4: Level number doesn't increment by one. Skipped lines 4 through 4.
ERROR 5: Level number doesn't increment by one. Skipped lines 5 through 5.
ERROR 6: Level number doesn't increment by one. Skipped lines 6 through 6.
ERROR 7: Level number doesn't increment by one. Skipped lines 7 through 7.
ERROR 8: Level number doesn't increment by one. Skipped lines 8 through 8.
ERROR 9: Level number doesn't increment by one. Skipped lines 9 through 9.
ERROR 10: Level number doesn't increment by one. Skipped lines 10 through 10.
ERROR 11: Level number doesn't increment by one. Skipped lines 11 through 11.
ERROR 12: Level number doesn't increment by one. Skipped lines 12 through 12.
ERROR 13: Level number doesn't increment by one. Skipped lines 13 through 13.
ERROR 14: Level number doesn't increment by one. Skipped lines 14 through 14.
ERROR 15: Level number doesn't increment by one. Skipped lines 15 through 15.
ERROR 16: Level number doesn't increment by one. Skipped lines 16 through 16.
ERROR 17: Level number doesn't increment by one. Skipped lines 17 through 17.
ERROR 18: Level number doesn't increment by one. Skipped lines 18 through 18.
ERROR 19: Level number doesn't increment by one. Skipped lines 18 through 19.
ERROR 20: John Sandercock-Mary Riddle.GED, line 20: Ignoring unexpected tag
'NAME'.
1 NAME David Frank Coppin
ERROR 21: John Sandercock-Mary Riddle.GED, line 21: Ignoring unexpected tag
'ADDR'.
1 ADDR 1665 E 1185 N
Logan
Utah 84341
ERROR 22: John Sandercock-Mary Riddle.GED, line 24: Ignoring unexpected tag
'CTRY'.
2 CTRY USA
ERROR 23: John Sandercock-Mary Riddle.GED, line 25: Ignoring unexpected tag
'PHON'.
1 PHON 435-753-5855
ERROR 24: John Sandercock-Mary Riddle.GED, line 26: Ignoring unexpected tag
'EMAIL'.
1 EMAIL cop...@comcast.net
ERROR 25: John Sandercock-Mary Riddle.GED, line 28: Ignoring unexpected tag
'NAME'.
1 NAME Mary /Riddle/
ERROR 26: John Sandercock-Mary Riddle.GED, line 29: Ignoring unexpected tag
'SURN'.
2 SURN Riddle
ERROR 27: John Sandercock-Mary Riddle.GED, line 30: Ignoring unexpected tag
'GIVN'.
2 GIVN Mary
ERROR 28: John Sandercock-Mary Riddle.GED, line 31: Ignoring unexpected tag
'_MARNM'.
2 _MARNM Sandercock
(I won't give the full list, but would be interested to know if anyone can
spot what is causing the errors)
--
Steve Hayes from Tshwane, South Africa
Web: http://hayesfam.bravehost.com/stevesig.htm
Blog: http://methodius.blogspot.com
E-mail - see web page, or parse: shayes at dunelm full stop org full stop uk
Peter J Seymour
> I've just been trying to exchange GEDCOM files with a correspondent, so we can
> compare notes, and encountered unexpected problems.
>
> He sent me a GEDCOM file, and I tried to import it into PAF 4, but failed.
>
> I imported it into Legacy 5, and then re-exported it as a GEDCOM file, which
> PAF 4 then imported with no problems.
>
> I also sent him a GEDCOM file exported by PAF 4, which his program could not
> handle.
>
> I imported that file into Legacy with no problems and reexported it, and have
> sent him the reexported version, in the hope that he will be able to import
> it.
>
> I've also asked him what program he uses -- does anyone know of any progrtams
> that have problems in exchanging data with PAF?
>
> Here are some of the errors reported:
>
It's a long time since I fiddled with PAF and without seeing the actual
file I am only guessing, but possibilities for the problems might be:
- Unexpected binary values at start of file
- Leading blanks or zeros on records
- Unexpected end-of-line style
- Wrapped or broken lines including embedded end-of-line sequences
- Some of the tags are actually considered invalid by the reading program
That doesn't actually answer your question but it may help.
Regards
Peter
Joe Makowiec
> I've also asked him what program he uses
Open the GEDCOM file in a text editor - notepad will do if you're on
Windows. Generally, genealogy programs will indicate near the top of a
GEDCOM which program generated the file.
--
Joe Makowiec
http://makowiec.org/
Email: http://makowiec.org/contact/?Joe
Usenet Improvement Project: http://twovoyagers.com/improve-usenet.org/
Joe Makowiec
> (I won't give the full list, but would be interested to know if
> anyone can spot what is causing the errors)
And as a followup to my previous post... Can you post the first 30 lines
or so of the file? It might help somebody spot an error.
Gene Y.
>
> (I won't give the full list, but would be interested to know if anyone can
> spot what is causing the errors)
>
>
First guess is an improper gedcom format. Simple solution is for the
other person to download the free version of legacy and exchange the
file as a legacy file. Smallest loss of data that way and it doesn't
cost anything.
--
Gene Y.
http://myyoungs.atspace.com/index.htm
Researching Young, Zies, Harer & Cox with
Legacy Family Tree
Denis Beauregard
<haye...@hotmail.com> �crivait dans soc.genealogy.computing:
>GEDCOM Import Log - 12 Nov 2009
>
> File: J:\John Sandercock-Mary Riddle.GED
> Created by:
> Submitted by: <UNKNOWN>
> Individuals: 0
> Marriages: 0
> Sources: 0
> Repositories: 0
> Errors: 3702
>
An idea : the GEDCOM file is from a Mac file and the line
separator is not compatible. CR-LF in DOS/Windows, CR
in Unix/Linux, LF in Mac. Cure : open with a text editor
and save it again in text format.
Denis
--
Denis Beauregard - g�n�alogiste �m�rite (FQSG)
Les Fran�ais d'Am�rique du Nord - www.francogene.com/genealogie--quebec/
French in North America before 1722 - www.francogene.com/quebec--genealogy/
Sur c�d�rom � 1770 - On CD-ROM to 1770
Steve W. Jackson
Denis Beauregard <denis.b-at-f...@fr.invalid> wrote:
> Le Thu, 12 Nov 2009 09:41:15 +0200, Steve Hayes
> <haye...@hotmail.com> �crivait dans soc.genealogy.computing:
>
> >GEDCOM Import Log - 12 Nov 2009
> >
> > File: J:\John Sandercock-Mary Riddle.GED
> > Created by:
> > Submitted by: <UNKNOWN>
> > Individuals: 0
> > Marriages: 0
> > Sources: 0
> > Repositories: 0
> > Errors: 3702
> >
>
>
> An idea : the GEDCOM file is from a Mac file and the line
> separator is not compatible. CR-LF in DOS/Windows, CR
> in Unix/Linux, LF in Mac. Cure : open with a text editor
> and save it again in text format.
>
>
> Denis
Actually, those line ending settings aren't right. LF is what's been
used in the Unix world for quite some time. CR is used in older Mac
apps, but newer ones use the Unix style since Mac OS X is based on some
variant of BSD.
While the line ending character is a possible source of trouble, there's
also the matter of which character set is used. Most programs that can
read or write GEDCOM files can be told what to use. I use Reunion for
Macintosh, for example, and it allows me to choose the character set if
I create a GEDCOM. I can choose ANSEL, Mac or Windows, maybe others I'm
not recalling just now. And I can also tell it what to expect in the
GEDCOM it reads.
--
Steve W. Jackson
Montgomery, Alabama
Denis Beauregard
<stevew...@knology.net> wrote in soc.genealogy.computing:
>In article <4f4of5h0golqisv3f...@4ax.com>,
> Denis Beauregard <denis.b-at-f...@fr.invalid> wrote:
>
>> Le Thu, 12 Nov 2009 09:41:15 +0200, Steve Hayes
>> <haye...@hotmail.com> �crivait dans soc.genealogy.computing:
>>
>> >GEDCOM Import Log - 12 Nov 2009
>> >
>> > File: J:\John Sandercock-Mary Riddle.GED
>> > Created by:
>> > Submitted by: <UNKNOWN>
>> > Individuals: 0
>> > Marriages: 0
>> > Sources: 0
>> > Repositories: 0
>> > Errors: 3702
>> >
>>
>>
>> An idea : the GEDCOM file is from a Mac file and the line
>> separator is not compatible. CR-LF in DOS/Windows, CR
>> in Unix/Linux, LF in Mac. Cure : open with a text editor
>> and save it again in text format.
>>
>>
>> Denis
>
>Actually, those line ending settings aren't right. LF is what's been
>used in the Unix world for quite some time. CR is used in older Mac
>apps, but newer ones use the Unix style since Mac OS X is based on some
>variant of BSD.
You're right, I inverted old Mac and Unix. Otherwise, that problem
is quite possible. Sometimes, I download a file from a Linux server
and open it in a DOS or Windows text editor and all lines are merged
in one line. But considering there are 3702 errors, I see now that
the problem of Steve can't be there.
>While the line ending character is a possible source of trouble, there's
>also the matter of which character set is used. Most programs that can
>read or write GEDCOM files can be told what to use. I use Reunion for
>Macintosh, for example, and it allows me to choose the character set if
>I create a GEDCOM. I can choose ANSEL, Mac or Windows, maybe others I'm
>not recalling just now. And I can also tell it what to expect in the
>GEDCOM it reads.
Wrong character set means the accented letters or diacritics are not
converted correctly. That can't lead to no line understood at all.
Seeing the first lines of the GEDCOM file is probably the best mean to
get a realistic diagnostic...
singhals
> (I won't give the full list, but would be interested to know if anyone can
> spot what is causing the errors)
Steve --
it looks like a GED4 export.
For format, here's the submitter/compiler part (since that
was the part you gave the line-by-line for) from a recent
GED5.5 format:
0 @S1@ SUBM
1 NAME Pxxx Rxxx Bxxx
1 ADDR xxxx Pacific
2 CONC Alameda, CA 90000
2 CONC (e-mail addy here)
1 PHON 555-222-9999
1 COMM Generated for Cheryl for the Cresap Society
Further in, your GED should resemble this:
0 @I51282@ INDI
1 NAME Isaac /Kuykendall/
2 SURN Kuykendall
2 GIVN Isaac
1 SEX M
1 BIRT
2 DATE 30 Aug 1839
2 PLAC Cumberland, Alleghany, Maryland
1 DEAT
2 DATE 29 Nov 1909
2 PLAC Springfield, , W. Virginia
1 BURI
2 PLAC Old Burying Grounds, Springfield, , W. Virginia
1 FAMS @F33490@
1 FAMC @F33395@
1 NOTE
2 CONC
2 CONC
2 CONC
2 CONC
2 CONC
2 CONC
1 CHAN
2 DATE 30 Jul 2008
3 TIME 17:26:31
BY CONTRAST: Here's the same bits exported to GED4x
0 @SUB1@ SUBM
1 NAME me2 myself
1 ADDR line1
2 CONT line3
1 CTRY usa
and
0 @I36@ INDI
1 NAME James /Largent/
1 SEX M
1 BIRT
2 DATE 1 SEP 1753
2 PLAC Forks of Capon,Hampshire,VA
1 DEAT
2 DATE 1813
1 FAMS @F18@
1 FAMS @F19@
1 FAMC @F1@
1 NOTE
2 CONT
1 NOTE
2 CONT
2 CONT
0 @I37@ INDI
Then again, I suppose, it could be the cursed GED5.2, in
which case, taking through Legacy may be your only choice. (g)
Good luck.
Cheryl
Patrick Texier
> It's a long time since I fiddled with PAF and without seeing the actual
> file I am only guessing, but possibilities for the problems might be:
> - Unexpected binary values at start of file
Yes, an UTF-8 file with a fucking BOM (Byte-Order-Marker).
With a BOM, the file is no more ASCII-compatible with three binaries
digit before 0 HEAD.
--
Patrick Texier
vim:syntax=mail:ai:ts=4:et:tw=72
Steve Hayes
<mako...@invalid.invalid> wrote:
>On 12 Nov 2009 in soc.genealogy.computing, Steve Hayes wrote:
>
>> I've also asked him what program he uses
>
>Open the GEDCOM file in a text editor - notepad will do if you're on
>Windows. Generally, genealogy programs will indicate near the top of a
>GEDCOM which program generated the file.
Found from the person who sent it that he is using PAF 5.5, but surely they
should be compatible?
Here's the first bit:
n++0 HEAD
1 SOUR PAF
2 NAME Personal Ancestral File
2 VERS 5.2.18.0
2 CORP The Church of Jesus Christ of Latter-day Saints
3 ADDR 50 East North Temple Street
4 CONT Salt Lake City, UT 84150
4 CONT USA
1 DEST Other
1 DATE 11 Nov 2009
2 TIME 00:20:44
1 FILE John Sandercock-Mary Riddle.ged
1 GEDC
2 VERS 5.5
2 FORM LINEAGE-LINKED
1 CHAR UTF-8
1 LANG English
1 SUBM @SUB1@
0 @SUB1@ SUBM
And what is that strange stuff in the first line? What put that in it?
And why was PAF 4 fazed by it, but Legacy apparently wasn't?
Patrick Texier
> n++0 HEAD
> And what is that strange stuff in the first line? What put that in it?
A UTF-8 BOM (xEF xBB xBF) written probabely by a Windows text editor.
Use an editor to remove it. It's easy with NotePad++
<http://notepad-plus.sourceforge.net>.
Peter J Seymour
to be a byte order mark (unnecessary for UTF-8 in this case as the
content style is explicitly labelled). I seem to recall that some
versions of PAF will put it there and it may have messed up PAF 4's
processing of the header record.
That doesn't explain the later errors though.
Peter
Denis Beauregard
>Found from the person who sent it that he is using PAF 5.5, but surely they
>should be compatible?
>
>Here's the first bit:
>
>n++0 HEAD
>1 SOUR PAF
>2 NAME Personal Ancestral File
>
>And what is that strange stuff in the first line? What put that in it?
Probably some glitch during the transmission of the file.
>And why was PAF 4 fazed by it, but Legacy apparently wasn't?
Because PAF 4 is a very old software with no mean to read a file if
there is a glitch in the 1st line (to explain why PAF read no line at
all), and because Legacy is much more recent (I presume you read it
with some Legacy 6.x release) and either detected the first line was
corrupted and useless and discarded it or detected n++ was useless and
read the remaining part of the file correctly.
In other words, I presume more recent softwares will take into account
the fact that a GEDCOM file may be corrupted.
Steve Hayes
wrote:
>Le Mon, 16 Nov 2009 18:27:15 +0200, Steve Hayes a �crit�:
>
>> n++0 HEAD
>> And what is that strange stuff in the first line? What put that in it?
>
>A UTF-8 BOM (xEF xBB xBF) written probabely by a Windows text editor.
UTF-8 BOM?
What is it? Why would it be put there?
Steve Hayes
wrote:
Iy makes sense to me why the others were messed up -- if it is expecting the
first line to be level 0, and it doesn't find it, then all subsequent lines
will be the wrong level number.
Steve Hayes
wrote:
>On Thu, 12 Nov 2009 09:07:54 +0000, Peter J Seymour wrote:
>
>> It's a long time since I fiddled with PAF and without seeing the actual
>> file I am only guessing, but possibilities for the problems might be:
>
>> - Unexpected binary values at start of file
>
>Yes, an UTF-8 file with a fucking BOM (Byte-Order-Marker).
>
>With a BOM, the file is no more ASCII-compatible with three binaries
>digit before 0 HEAD.
Ah, so that is what a BOM is -- but what puts it there?
Joe Makowiec
> On Mon, 16 Nov 2009 17:36:31 +0100, Patrick Texier
> <p.te...@alussinan.org> wrote:
>
>>Le Mon, 16 Nov 2009 18:27:15 +0200, Steve Hayes a �crit�:
>>
>>> n++0 HEAD
>>> And what is that strange stuff in the first line? What put that in
>>> it?
>>
>>A UTF-8 BOM (xEF xBB xBF) written probabely by a Windows text
>>editor.
>
> UTF-8 BOM?
>
> What is it? Why would it be put there?
http://unicode.org/faq/utf_bom.html#BOM
It's put there by whatever generates the file.
singhals
> Le Mon, 16 Nov 2009 18:27:15 +0200, Steve Hayes
> <haye...@hotmail.com> �crivait dans soc.genealogy.computing:
>
>
>>Found from the person who sent it that he is using PAF 5.5, but surely they
>>should be compatible?
>>
>>Here's the first bit:
>>
>>n++0 HEAD
>>1 SOUR PAF
>>2 NAME Personal Ancestral File
>>
>>And what is that strange stuff in the first line? What put that in it?
>
>
> Probably some glitch during the transmission of the file.
>
>
>>And why was PAF 4 fazed by it, but Legacy apparently wasn't?
>
>
> Because PAF 4 is a very old software with no mean to read a file if
> there is a glitch in the 1st line (to explain why PAF read no line at
> all), and because Legacy is much more recent (I presume you read it
> with some Legacy 6.x release) and either detected the first line was
> corrupted and useless and discarded it or detected n++ was useless and
> read the remaining part of the file correctly.
>
> In other words, I presume more recent softwares will take into account
> the fact that a GEDCOM file may be corrupted.
>
>
> Denis
>
Those three chr$ at the beginning the file are put there for
a reason "known but to God" by PAF5's export routine. PAF4
doesn't use 'em, doesn't know what to do with 'em and
chokes. A dyscompatibility known since week 2 of PAF5.
Odd that the rest of the GED didn't seem to match PAF5
output, though.
The work-around is to open *all* GEDs in notepad or equiv,
delete that line, and resave.
Cheryl
Nigel Bufton
"singhals" <sing...@erols.com> wrote in message
news:mailman.220.12584...@rootsweb.com...
If you open a UTF-8 file in Notepad, it will recognize the BOM header and
not show it. If your save as UTF-8 encoding, Notepad will add the BOM. You
would need to be sure that you change the encoding to ANSI when you save the
file if you want Notepad not to add the BOM.
Nigel
singhals
You could be right. I've never had a problem using my
method, though.
Cheryl
Patrick Texier
> If you open a UTF-8 file in Notepad, it will recognize the BOM header and
> not show it. If your save as UTF-8 encoding, Notepad will add the BOM. You
> would need to be sure that you change the encoding to ANSI when you save the
> file if you want Notepad not to add the BOM.
Notepad is the worst text editor never written.
With NotePad++, <http://notepad-plus.sourceforge.net> you can write
ANSI, UTF-8 (bomb or without BOM), UCS-2 (Big or Little Endian).
singhals
> Le Wed, 18 Nov 2009 09:04:01 -0000, Nigel Bufton a �crit :
>
>
>>If you open a UTF-8 file in Notepad, it will recognize the BOM header and
>>not show it. If your save as UTF-8 encoding, Notepad will add the BOM. You
>>would need to be sure that you change the encoding to ANSI when you save the
>>file if you want Notepad not to add the BOM.
>
>
> Notepad is the worst text editor never written.
You've clearly never used edlin. Or you've suffered
merciful amnesia. (g)
Cheryl