Index Key Extraction (Authorities)

When a user browses the A-Z Index or clicks on the "Dictionary" icon, the system retrieves keys from the Inverted File. Often, these keys are normalized (converted to uppercase and stripped of accents) by the FST to facilitate searching.

The Problem:

• Stored in Index: GARCIA MARQUEZ, GABRIEL
• Desired Display: García Márquez, Gabriel

The Index Extraction module (autoridades.php) allows you to define a special PFT format (autoridades_opac.pft) that restores the original formatting of terms by extracting them directly from the record at runtime.

Script: autoridades.php File Created: bases/[db]/pfts/autoridades_opac.pft

1. How It Works

The mechanism relies on the Field ID defined in your database's FST (Field Selection Table).

1. Retrieval: The OPAC searches the Inverted File and finds a term.
2. Identification: It identifies the record (MFN) and the specific line in the FST (ID) that generated that term.
3. Extraction: It executes the autoridades_opac.pft script. This script uses the special command select e3 to switch logic based on the FST ID.
4. Display: It prints the formatted field (e.g., preserving mixed case and punctuation) instead of the raw index key.

2. Configuration Interface

The interface provides a text editor to modify the PFT and a helper section to view your current FST.

The FST Helper

At the bottom of the screen, the system lists your FST entries.

• Column 1 (ID): The numeric identifier (e.g., 100, 245). You need this number for the case statement.
• Column 2 (Technique): Indexing technique (usually 0).
• Column 3 (Format): The extraction format used to build the key.

3. The PFT Structure (select e3)

The file autoridades_opac.pft must use a specific switch structure. Do not include line breaks (/) inside the format logic, as this breaks the index list display.

Syntax:

select e3
    case [FST_ID_1]: [Format_to_extract_field]
    case [FST_ID_2]: [Format_to_extract_field]
    ...
endsel

Example Scenario (MARC21)

Suppose your FST looks like this:

100 0 "AU_"v100^a
245 0 "TI_"v245^a
700 0 "AU_"v700^a

• Note that both Tag 100 (Main Author) and Tag 700 (Added Entry) generate the prefix AU_.
• Note that Tag 100 has ID 100 and Tag 700 has ID 700.

Your autoridades_opac.pft should look like this:

select e3
    case 100: v100^a, " "v100^d
    case 700: v700^a, " "v700^d
    case 245: v245^a, " "v245^b
endsel

Result: When the user browses the "Authors" index:

1. The system finds AU_SHAKESPEARE, WILLIAM.
2. It sees it came from ID 100.
3. It runs case 100.
4. It displays: "Shakespeare, William".

4. Advanced Formatting

You can use standard ISIS formatting language to handle complex fields, subfields, and punctuation.

Example from autoridades_opac.pft:

select e3
    /* Personal Author */
    case 100: v100^a,v100^b,v100^c,v100^d ' ||',v4999[1],'$$$'s0

    /* Title (Handling 'Article' logic with mhl/mpl) */
    case 245: mhl,v245^a,mpl," "v245^b

    /* Subject with subfields separated by hyphens */
    case 650: v650^a, | - |v650^x, | - |v650^z
endsel

Performance

This PFT is executed for every single line shown in the dictionary browser. Keep the logic simple and efficient to ensure the list loads quickly.

Common Mistake

If you see keys in uppercase in the browser even after configuring this, check:

1. Did you save the file?
2. Does the ID in the case statement match the ID in the FST exactly?
3. Are you using e3 (Term ID) in the select command?