Record validator using KOHA's MARC framework and other

This application uses a CSV export of Koha MARC framework to check if :

• Unmapped fields & subfields are not present
• Mandatory fields & subfields are present
• Non repeatable fields & subfields are not repeated
• Authorized values limited subfields have legal values

Uses pymarc 5.2.0 (originally developped with pymarc 4.2.2., the update to for compatibility was minimalist).

Setting up

You need to export from Koha :

• The default MARC framework as a CSV file
• All authorised values (or only those who could be used) adding branches and itemtypes to them, the order of column must be : category, authorised_value, lib, lib_opac

/* Example, remove the WHERE if you want to export every authorised value */
/* /!\ COLUMNS MUST FOLLOW THIS ORDER : category, authorised_value, lib, lib_opac */
SELECT category, authorised_value, lib, lib_opac
FROM authorised_values

UNION ALL 
SELECT "branches" AS category, branchcode AS "authorised_value", branchname as "lib", branchname as "lib_opac"
FROM branches

UNION ALL 
SELECT "itemtypes" AS category, itemtype AS "authorised_value", description as "lib", description as "lib_opac"
FROM itemtypes

Then you need to set up the following environment variables :

• RECORDS_FILE : full path to the file contining all the records to analyse
• ERRORS_FILE: full path to the file with errors (will be created / rewrite existing one)
• KOHA_MARC_FRAMEWORK_FILE : full path to the MARC framework file
• KOHA_AUTH_VAL_FILE : full path to the authorised values export file
• CONTROL_VALUES_FILE : full path to the control values XML file

A default control values XML is provided (do note that it's ArchiRès' one), the root must be fields, then it should always follow these conditions :

• fields root :
  • Does not use any attributes
  • Contains field nodes
• field nodes :
  • Must have an attribute tag (use 000 for the record label / leader)
  • Contains subfield nodes
• subfield nodes :
  • Must have an attribute code (use @ for controlfields and the record label / leader)
  • Can have attribute startPosition and / or endPosition :
    • Position starts at 0
    • Must contain only numbers (not even spaces)
    • Using only a start position means taking only the character at this position
    • Using only end position does nothing
    • Using both position takes all characters included in this interval (0 - 2 will be characters 0, 1 and 2)
  • Contains value nodes
• value nodes :
  • Must have an attribute value
  • Can have an attribute name (if name is not used, value will be used instead as a name)
  • Have no child

<fields>
    <field tag="181">
        <subfield code="c" startPosition="0" endPosition="2">
            <value value="nda" name="A"/>
            <value value="ndb" name="B"/>
        </subfield>
        <subfield code="b" startPosition="0">
            <value value="f" name="Fake Appollo"/>
        </subfield>
        <subfield code="a">
            <value value="lovecolor" name="Love Colored Master Spark"/>
        </subfield>
    </field>
</fields>