Okay. Here's what happened.

I transferred the text file "bases" by sftp to my spare computer here. It's a 20 MB text file, with a weird file exptension .gw, that IS a text file (file manager says so) and that my program (Geneweb) understands. This file transfer took under 2 seconds for the 20 MB. No complaints.

I tried opening it... Abiword would do that --it tried --but at 85% completed "building" the file (which Abiword said it was doing), and since it didn't seem to be getting past that 85% and by then, both cores of the processor in that computer were up to about 83 degrees and the mobo was also getting hot. So I aborted Abiword, assuming that it might be getting stuck wherever the first (only??) glich is. BTW I think this glitch happened when the landlady's handman was up here in the kitchen fixing the string-pull on a light in the kitchen without turning off the power, and shorted the hot wire to the negative. I ended up having to buy a new router AND modem because of that.

Anyway, I also assume that command line functions such as cp will be equally easy to make happen.

So now I think I need some way of seeing the text in this file, so that I can find where the glitched-up text is, that is hanging up the whole operation. Somehow or other, I still need a way to see the text, and/or to write some script that can find the glitches in it. I know the name of one of the families I was working on at the time, and know that that surname DOES hang and not present all of its descendants, so I could start there.

Thanks in advance for the help.

Liz J

On 8 October 2014 14:20, E Johnson <iris.gates@gmail.com> wrote:

Tim,

I'll try the cp; don't remember if I tried that or not. Will take a little while. I'm going to ssh in from the other box, so I can leave this one less-occupied while I attempt it. Will report back.

Thanks,
Liz
On 8 October 2014 14:08, Tim Keller <turbofx@gmail.com> wrote:
What do you mean you can't copy the original file? if you do a "cp file1 file2" do you get an error?

Barring that, doing a quick search of gedcom parsers in python shows there's a whole pile of them. Without too much trouble we could likely put together a program that would take this original file and at least validate it if not go thorough it to the point where the parser barfs and say "Record X" is the problem.
On Wed, Oct 8, 2014 at 1:28 PM, E Johnson <iris.gates@gmail.com> wrote:
Thanks, Keith.

The program doesn't crash at all, and I don't think it is broken. It just can't do certain things that it used to do. This program is called Geneweb, and it's free, GNU, open-source, etc. http://opensource.geneanet.org/projects/geneweb. I am running version 5.01 which is not the latest, but I'm keeping it because in versions starting from 6, they changed something, and a mod that I installed (that I didn't write) that allows a display of three generations before and after the individual won't work any more.
The data in the text file (called the base), containing various types of information about the individual, also contains some numbers and so forth, linking that person to his/her parents, children, spouse, etc. The program will access this file display an individual and parents, children, etc, if known in a web page that is nicely formatted; that's fine. What happens since May 2012 is that from the top page there is a choice of displaying all surnames (alphabetical), or a box in which one can input a surname, which then will result in a display of all persons with that name. That won't work now, and neither will the all-surnames display work. One can also choose to display persons sorted on title or estate, which also doesn't work.

The problem really is in the text file (the base). I have several different databases here, and the others, one of which contains 50,000 individuals, perform perfectly. But in my big database, it hangs, seeming to get lost, when asked to display surnames or estates.
The data in the text file is supposed to look like this, more or less:
0 @I1@ INDI
1 NAME Bob /Cox/
1 SEX M
1 FAMS @F1@
1 CHAN
2 DATE 11 FEB 2006
0 @I2@ INDI
1 NAME Joann /Para/
1 SEX F
1 FAMS @F1@
1 CHAN
2 DATE 11 FEB 2006
0 @I3@ INDI
1 NAME Bobby Jo /Cox/
1 SEX M
1 FAMC @F1@
1 CHAN
2 DATE 11 FEB 2006
0 @F1@ FAM
1 HUSB @I1@
1 WIFE @I2@
1 MARR
1 CHIL @I3@
0 TRLR
I also can't create a backup or copy to anywhere, which is crucial, and I can't move the base file, either to another location on the hard drive or to another device. I also can't get it to re-sort, which would be nice, because that cleans up "ghosts" or leftover info on individuals that don't "exist" any more because they have been merged with another one who turned out to be the same person. This problem first cropped up when I had either 86,000 individuals or 82,000 individuals, 2 1/2 years ago.

If I ever use the wrong terminology, I aploogize, but I'm not an IT specialist. I am a Linux User located in Worcester. Last time I wrote a program, it was in Basic. I should have said, I guess, is that there must be some bad characters in this text file "base", which makes it impossible for the program that uses it to perform certain functions, such as Sort, or to display certain data, such as all persons of the surname {Whatever}.

Liz

I didn't know what could be meant by "broken text file", now I see
that you mean "broken file processing program". Do you have
the source for the geneology program? If not, insert rant here.

Does the program crash as soon as you import the file, or
only when you do something particular? My first impulse
would be to put printf commands into the input procedure.

Do you have a spec for what the data is supposed to look like,
or is this one of those take-the-money-and-(run|hide|
sell updates)
programs, whose authors think that's super double secret eye pee?

> The text file has been broken since May 2012. This is a genealogical
> database, that runs in a program that accesses the file, which consists of
> names, dates, places, etc, plus notes. I think the bogus character might be
> an at sign @ but if so, it won't help to search for it, because GEDCOMS
> contain several @ for each individual. My database contains 115,000
> individuals. I think I know what family (surname) I was working on when it
> broke, though, and it still hangs there. I guess I could grep the file for
> that surname and get it to output a couple of lines after it finds each
> instance.
_______________________________________________
Wlug mailing list
Wlug@mail.wlug.org
http://mail.wlug.org/mailman/listinfo/wlug
--
I am leery of the allegiances of any politician who refers to their constituents as "consumers".
_______________________________________________
Wlug mailing list
Wlug@mail.wlug.org
http://mail.wlug.org/mailman/listinfo/wlug