Migration from ISIS to Koha

From Koha Wiki
Jump to navigation Jump to search

This page is a translated version of Migration_from_ISIS_to_Koha_(Spanish) translated using google translate ...

This method has been tested with some success and leverages the versatility of ISIS formatting language, to generate a text file with the structure of fields and values ​​them at a "MARC Annotated" which is basically a file text that can be opened with MARCEdit, for final conversion to MARC. Finally, the script is used to load bulkmarcimport.pl records in Koha.

Obviously, this method involves a prior analysis of the database and correspondences between its component fields and MARC fields, along with a series of twists and turns (trial and error) to have a MARC file quality of at least acceptable. This is essential if one considers that the script used to populate bulkmarcimport.pl base Koha external data, but not to update them if they are present. In other words, it's best to ensure that the information contained in the MARC file is as correct as possible, before final import.


Check and check again partnerships KOHA MARC fields and fields in the Koha Management

PFT File Creation

Create a PFT file that displays the contents of the ISIS database in a manner similar to the following:

     =LDR 00000nam\\2200000\a\4500
     =020 \\$a9505044631
     =092 \\$a636(038) C152
     =100 \\$aTweeten, Luther G.$d1919-1999
     =245 10$aGlosario inglés castellano de términos usados en producción
     =650 04$aDiccionario Técnico
     =650 04$aInglés
     =650 04$aEspañol
     =650 04$aSilvicultura
     =650 04$aProducción forestal
     =650 04$aAlnus
     =650 04$aProducción Animal
     =653 \\$aAliso
     =942 \\$cBK
     =952 \\$pOAG13849$bOAG$dOAG

Obviously, our records CDS / ISIS have no "leader" (header) and lack some information that is necessary in KOHA, for example, the type of item (itemtype) or library code (branchcode).

In this example, Koha is set Branch Agricultural Sciences Library with the code 'OAG' and we have defined that the bar code (barcode) of each book you add the code library (branchcode), to differentiate it from the barcodes (barcodes) from other libraries.

The above output can be obtained with a PFT file like the following:

     '=LDR 00000nam\\\\2200000\\a\\4500'/
     (|=020 \\\\$a|v010/)/
     (|=100 \\\\$a|,v28^a/)/
     |=110 \\\\$a|,v29^p,if p(v29^s) then |$b|v29^s fi/
     |=111 \\\\$a|,v40^n/
     |=245 \\\\$a|v24^t,if p(v24^s) then | : $b|v24^s fi/
     |=591 \\\\$a|,v59/
     (|=650 \\\\$a|,v65/)/
     (|=653 \\\\$a|,v62/)/
     '=942 \\\\$cBK'/
     (|=952 \\\\$pOAG|v77|$aOAG$bOAG|/)
     |=999 \\\\$a|,v999/

We examine each of the lines of the previous format, to explain its function, so that it is more understandable and to facilitate their adaptation to other settings.

 '= LDR 00000nam \ \ \ \ 2200000 \ \ to \ \ 4500' / 

Prints a pseudo-leader. The leader of this example is the same in all registers. You may also be edited to build one from the information in each record.

From here, the idea is to be printed on each line the following:

'= <Nro MARC> field \ \ \ \ [identifier subfield] ISIS> field <Content'

For example, the following line ...

 (| = 020 \ \ \ \ $ a | V010 /) / 

prints ...

'= 20 \ \ \ \ $ a (Contents of field 10 at the base ISIS)'

Note: have to take into account those that are repeatable. In the case above, field 10 in ISIS is repeatable, so that the entire line is placed between brackets and so prints a '= $ 20 to' for each occurrence of field 10.

The next line prints the content of our call number that is in the field 75 of ISIS in the MARC 092

 | = 092 \ \ \ \ | V075 / 

This line prints the content of our field author, who is in the field ISIS 28, at 100 MARC

 (| = 100 \ \ \ \ $ a |, v28 ^ a /) / 

The other lines do similar work for the rest of the fields.

|=110 \\\\$a|,v29^p,if p(v29^s) then |$b|v29^s fi/
|=111 \\\\$a|,v40^n/
|=245 \\\\$a|v24^t,if p(v24^s) then | : $b|v24^s fi/
|=591 \\\\$a|,v59/
(|=650 \\\\$a|,v65/)/
(|=653 \\\\$a|,v62/)/

From here on begins Basically, for the import to work correctly ...

Koha, by default, expects to find at the time of import the type of item (itemtype) in field 942 subfield c of the base that is being imported or, in the case of activation of the system preference 'item-level_itypes' in the 952 subfield y.

'= 942 \ \ \ \ $ CBK' /

In addition, we also expect to find

  • 952 field subfield p ​​→ bar code item
  • in the 952 subfield a → the code of the library owns the book (homebranch)
  • in the 952 subfield b → the code of the library where the book is located (holdingbranch)
  • 952 field subfield and → the type of item (itemtype) the item

 (| = 952 \ \ \ \ $ Poag | v77 | $ aOAG $ Boag $ bk | /)

In this case, based on ISIS the item inventory number in the field is 77, so that the previous line produces an output simlar to this:

'= 952 \ \ \ \ $ $ Boag aOAG pOAG12345 $ $ bk'


Print the ISIS database using the PFT file mentioned in the previous point in a text document.

You may ustilizar WinISIS for this task, because to print a range of records, or the results of a search (for example, all the records we have present an inventory number, and print only those records, to not import analytics journals ).

Within WinISIS print options, define what you want to print (range of records or a search result), we changed obviously the default printer option "text file (Windows ANSI)" (there is an option ASCII DOS, but that brings problems with accents and other special characters) and then change the default file our PFT PFT file created in step 2.
By giving the print command, WinISIS present a dialog box to select the name and location of the generated text file. To make the process easier, it is best to give it a name and add the extension '. Mrk' (eg oag.mrk)

Edit text file in MARCEdit

Once the core dump, you can edit the file. Mrk created with the tool WinISIS MARCEdit. Basically, what will we do with this tool, first, find and replace all symbols' ^ 'which delimit subfields in ISIS by `$' that define subfields in MARC.

After the replacement, it only remains to use the function 'Compiling MARC file' or 'create MARC file' (the name and location of this function has changed between versions MARCEdit) which basically transforms this text file. Mrk in a file binary. mrc (following the example, a file would oag.mrc)

Import records to Koha using bulkmarcimport.pl.

The syntax for this would be basically:

       . / bulkmarcimport.pl-file / home / marc_files / oag.mrc

- Hilongo 16:42, 1 November 2010 (UTC)