[Isis-users] Othority data base

Egbert De Smet egbert.desmet at uantwerpen.be
Wed Jul 31 15:13:55 CEST 2019


then most probably the script already puts the PFT in a loop, creating a 'nested repeatable format' which results in the known 'semantic error' of wxis (CISIS).

You could test your records in the terminal with mx to check whether the ID and OCC are stored correctly, e.g. if your database is named 'test' and you have a FST 'test.fst' you could use the following command :

mx test fst=@

without anything more it would present each record one by one just on the screen with first the listing of the field-contents, followed by the FST-output, which looks like e.g. :

mfn=     1
  1  "1"
  1  "2"
  1  "V1_1^m1^o1^c1^l1"
  1  "V1_2^m1^o2^c1^l1"
 11  "V11_1^m1^o1^c1^l1"
 11  "V11_2^m1^o2^c1^l1"

This is a single record with 2 occurrences of v1 with resp. values 1 and 2, but indexed with two different ID's (1 and 11 and accordingly different prefixes). The interesting part is ^m1^o2^c1^l1 which indicates in ^o the occurrence counter, in this case 2.

So now please check if your author-names have different occurrence counters reflecting their position in v100 and v400. If they are correct then it might be a bug indeed of the '$$$' extract-as technique with the autoridades.pft, because then ABCD should be able to correctly extract the same occurrence of v100 as the one read (listed as) of v400.

Egbert de Smet
Universiteit Antwerpen

From: Edwin Hübner <edwin.hubner at gmail.com>
Sent: Wednesday, July 31, 2019 3:01 PM
To: Egbert De Smet
Cc: Soraya Ferraz Lima; isis-users
Subject: Re: [Isis-users] Othority data base

Dear Egbert,
Yes, I had already tested this alternative  before, using in the autoridades.pft
case 400: (v400*4/)'$$$'(v100/)  or
case 400: (v400*4/)'$$$'v100
but unfortunately presents the following error:
WXIS|fatal error|unavoidable|fmt_gener/semantic_error|

Edwin Hübner

Em qua, 31 de jul de 2019 às 09:19, Egbert De Smet <egbert.desmet at uantwerpen.be<mailto:egbert.desmet at uantwerpen.be>> escreveu:


OK, then your 'occurrence counter' must be correctly stored in your IF, so that doesn't explain the problem.

Maybe you could try - as I don't have easily good testing material at hand here - to use the brackets to make the field treated as a repeatable field :

case 400: (v400*4/)'$$$'(v100/)

so as to force both the 'list as' and 'extract as' values to come as separate entries.

I don't see any examples of such repeated groups in the series of 'autoridades.pft' in the ABCD-demo databases, but in the ABCDWIKI pages on this issue (http://abcdwiki.net/wiki/es/index.php?title=Control_de_terminolog%C3%ADa:_Listas_de_autoridades) I can see that it simply refers to 'the format to be used to transfer the field in the record' (el formato a utilizar para transferir el campo al registro (extraer como), so in principle a repeatable group should be allowed and ensures that there is an occurrence counter being followed as that seems to be the problem.

I have no idea whether it would make a meaningful difference but it's worth a try. Otherwise Guilda Ascencio should be called to the rescue to check her implementation in the PHP-scripts (or the IsisScript).

By the way, the FST examples of the WIKI-page use the method  '0' for the indexing but that should not create any difference in the resulting strings :

400  0 mpu,(|AU_|v400^*,|, |V400^b,'%'/)/

Egbert de Smet
Universiteit Antwerpen

From: Edwin Hübner <edwin.hubner at gmail.com<mailto:edwin.hubner at gmail.com>>
Sent: Wednesday, July 31, 2019 1:08 PM
To: Soraya Ferraz Lima
Cc: Egbert De Smet; isis-users
Subject: Re: [Isis-users] Othority data base

Hi Egbert,
My fst for fields 100 and 400 is as follows:
100 5 mpu,if p(v100^a) then '/AU_/',v100^a,| |v100^b,| |v100^c,| |v100^d,| - |v100^x,'%',fi,
400 5 mpu,if p(v400^a) then '/AU_/',(v400^a,| |v400^b,| |v400^c,| |v400^d,| - |v400^x|%|/),fi,

Remember that when I click on a name that is the first occurrence of field 400, it assigns the name of field 100 (as expected), but if the name that is clicked is another occurrence, it assigns field 400 instead of 100

Edwin Hübner

Em qua, 31 de jul de 2019 às 07:12, Soraya Ferraz Lima <soyferrazlima at gmail.com<mailto:soyferrazlima at gmail.com>> escreveu:
Hi Edwin!

Here we use like this

Case 400: v400^a, '$$$', v700

Since the first column is for the field list and after the $$$ is how it will be stored.

Fst 400 0 |AU_|v400^a

At least that is the way I understand

Em terça-feira, 30 de julho de 2019, Edwin Hübner <edwin.hubner at gmail.com<mailto:edwin.hubner at gmail.com>> escreveu:
Thank you Egbert for your interest in helping. Sorry I didn't explain it more clearly because it was clear to me.
  Please see further explanation in the attached document.
Edwin Hübner

Em ter, 30 de jul de 2019 às 03:45, Egbert De Smet <egbert.desmet at uantwerpen.be<mailto:egbert.desmet at uantwerpen.be>> escreveu:


sorry but your problem and request is not that clear to me. Probably that is also why no reply came...

Can you please explain the exact situation : are you clicking on an occurrence in a PFT (a hyperlink ?), in a picklist of a worksheet or what exactly do you want to get ?

But before everything : please note that v100 * 4 should actually be v100*4 (no spaces in between). Perhaps that causes problems.

Egbert de Smet
Universiteit Antwerpen

From: isis-users <isis-users-bounces+egbert.desmet=ua.ac.be at iccisis.org<mailto:ua.ac.be at iccisis.org>> on behalf of Edwin Hübner <edwin.hubner at gmail.com<mailto:edwin.hubner at gmail.com>>
Sent: Monday, July 29, 2019 1:19 PM
To: isis-users
Subject: [Isis-users] Othority data base

I have asked for this problem before, but didn't receive help. So I'm asking again:

I have a MARC bibliographic database and a authority database for the author names normalization. The problem is: when field 400 - See tracing, for example, has more than one occurrence and is clicked on a name other than the first occurrence, ABCD does not assign the name of field 100, as expected, but the name of field 400 it self. I wonder if anyone has had this problem and how did solve it?

My autoridades.pft is as follows:
select e3
case 100: v100 * 4 '$$$' v100
case 110: v110 * 4 '$$$' v110
case 111: v111 * 4 '$$$' v111
case 400: v400 * 4 '$$$' v100
case 410: v410 * 4 '$$$' v110
case 411: v411 * 4 '$$$' v111
Thanks for any help.
Edwin Hübner
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.iccisis.org/pipermail/isis-users/attachments/20190731/52789aa8/attachment.html>

More information about the isis-users mailing list