Using KohaILS ? Better update it from time to time or be left behind!

Earlier in the day, young Sandeep pinged me with a question:

Can I add additional, separate columns of information to this table on the screen?

The screen he referred to was this one:

18685478_1887705531470539_114367369_n

As regular readers of this blog are know well by now, for any question about Koha, the very first piece of useful information is what is the version of Koha in this case?. Sandeep replied back saying it was “3.20.02”! I was like “Hmmmm! That’s a bit old…”

With two major versions of Koha coming out every May and November respectively, every major Koha version usually have a supported life cycle of about 18 months. The 3.20 series was launched in May 2015. Thus Sandeep’s Koha instance at nearly 24 months old, was already out of support lifecycle for nearly 6 months.

I pointed Sandeep to bug number 10855“Custom fields for subscriptions”. This feature to add custom fields to serial subscriptions became a part of Koha with the release of version 3.22 on Nov 26, 2015 (See https://koha-community.org/koha-3-22-released/ and look for “Custom fields for subscriptions (bug 10855)” under Serials).

Had his Koha been updated at anytime during the last 1 year, he would have found that the featured he wanted already built-in.

Harnessing Koha’s ExtendedPatronAttributes (aka patron custom fields)

Custom fields that make it possible to fit KohaILS to any type of library user category.

During from my frequent interactions with Indian Koha users over the last several years, one thing stands out rather starkly. There is surprisingly little understanding or use of the capabilities offered via Extended Patron Attributes. Since Koha ILS was written originally for public libraries (in fact public library consortia), its default patron data structures too are somewhat aligned with public libraries’ patron information capture.

When put to use in academic libraries e.g. K12 schools or colleges, the default patron information schema does not address the obvious need to capture data points like – programme enrolled, registration / roll numbers / class / sections details and in case of K12 schools – parent / guardian name and contact information etc. Many Koha users from the Indian sub-continent either (a) work their way around these issues without fully addressing their needs OR (b) often resort to direct editing of the koha.borrowers table schema, adding new custom fields and modifying the template files (.tmpl or .tt files) along with the Perl scripts that drive these templates.

The second approach is particularly problematic, as it means the modifications are usually hardcoded into scripts and templates. This is problematic is because doing so prevents the user from the benefit of seamless upgrades to their Koha. It means they have to do the whole “rinse-and-repeat” cycle everytime they upgrade their Koha, if they want their changes to persist. This makes Koha upgrades a costly, laborious, time-consuming process and error prone process.

Enter ExtendedPatronAttributes

According to the Koha manual –

“Patron attributes can be used to define custom fields to associate with your patron records. In order to enable the use of custom fields you need to set the ExtendedPatronAttributes system preference.”

About 10 years back, as Koha was finally moving out of version 2.x and moving into Koha 3.x, several new and exciting features were coming on to their own in Koha. One of this was the ability to easily handle custom fields for our patron records aka patron attributes. Around May 2008, my friend Galen Charlton, then with LibLime, wrote a (rather) large chunk of the code that brought in the support for custom fields to patron records. The ExtendedPatronAttributes system preference was the outcome and custom fields were here to stay.

Flash forward to present day

Recently we have been working with the Senior section library of Don Bosco School in Calcutta. Being a school library they needed to additionally support the following fields in their patron record (a) Class (b) Section (c) Roll No. (d) Parent / Guardian information i.e. name, contact info, relationship with student. Of course, we used defined these custom fields as patron attributes that were applicable only for their student patron category. In this blogpost, I will try to share just how easy and painless this is.

Step #1 : Enable ExtendedPatronAttributes syspref

This *is* the very first step and unless you enable this syspref, no matter what you do to define your custom fields (aka patron attributes), nothing would be visible.

Step #2 : Defining the Patron Attributes

In our case, we had two sets of information to be handled i.e. the student’s (a) academic details and (b) parent / guardian information. So we defined ST_CLASS, ST_SECTION and ST_ROLLNO patron attributes for the first part and ST_PNAME, ST_PRELATN and ST_PCONT as attributes for the second part.

2017-05-16_03-44-10

To make things easier for the library personnel (and reduce human error in the process), we pre-defined the values available to ST_CLASS, ST_SECTION and ST_PRELATN as authorised values.

2017-05-15_19-34-06

We also defined the PA_CLASS authorised value with the following two values : (a) ADETAILS (for academic details) and (b) GDETAILS (for guardian information).

2017-05-16_03-50-36

Step #3 : PA_CLASS – the secret sauce!

The Koha manual barely touches on this little nugget otherwise known as the PA_CLASS authorised value category. Before we proceed, just remember that in Koha Authorised value and Authority value mean two completely different things. Here we are talking about the former. If you have ever done any programming or done anything online, you must have used a drop-down / picklist to select a value. Well in Koha, authorised values are simply another name for picklists. These are defined as a category with the options added it. To the end-user, authorised values are presented as select drop-down HTML form element.

Now coming to PA_CLASS (or Patron Attribute Class), the first thing to remember is that by default Koha does not even define the PA_CLASS authorised value category. Rather it is left to the user to define it. There is a reason why it is not defined by default, not every library is going to used ExtendedPatronAttribute and unless you use it, PA_CLASS has no use. In our case, we have defined it and allotted it two values – ADETAILS and GDETAILS for now.

By allotting our custom fields (patron attributes) to either of these two PA_CLASS values allows us to logically group our two sets of information. By default, once done it looks like this:

2017-05-16_01-49-10

Order! Order! Order!

While chaos defines us, in life it is order that we look for, especially when we are attempting to present or capture information. So while the “Additional attributes and identifiers” grouping at the end of the page is adequate it is not ordered in the most logic flow with respect to the overall patron information form. Lucky for us, Koha has supported JQuery for donkey’s years. And JQuery provides us with a very nifty method .detach(). As the documentation says:

The .detach() method is the same as .remove(), except that .detach() keeps all jQuery data associated with the removed elements. This method is useful when removed elements are to be reinserted into the DOM at a later time.

blog-epa

By detaching the PA_CLASS grouped fieldsets from the DOM and by storing these into variables one at a time, we are now ready to insert them exactly where we want these fieldsets to be placed in the overall form. And that is exactly what we did.

epa-full

Conclusion

This way we did not touch the templates or the underlying business logic of KohaILS. All our definitions are stored in the database. As a result, we can now perform IN-SITU upgrades of Koha without worrying if we will break our forms if we upgraded.

Koha and the “magic” of XSLT : displaying new MARC fields on the OPAC.

A short tutorial on using XSLT to make MARC fields in your data visible on the OPAC.

Earlier today, Suresh Kumar Tejomurtula, a member of the venerable LIS-Forum mailing list being run out of IISc Bangalore, posted a question on that list:

Under language codes of 008/35-37 and also under 041$a I added language code. for eg: tel for Telugu language. But I do not see any difference to the view of the record, except that the marc tags contains that values. Adding these fields data will enable library team to understand the language of the material. My question is, How will the users of the OPAC, who do not know about marc language codes will understand the item language [from just the 3-letter code sequence].

Short answer : Using XSLT

Much of Koha’s superpowers on the OPAC (as also on the staff client) side comes from its judicious use of XSLT. When we search for documents in Koha, the result that is returned from the database by way of the various perl modules that perform much of Koha’s internal plumbing, comes in as an XML (eXtensible Markup Language) document. More precisely it is returned as a MARCXML record. Readers of this blog who are familiar with the MarcEdit software may have often encountered a MARCXML record. Those who are not so familiar may well like to read up a bit from here before proceeding with this post.

So what is XSLT?

Wikipedia gives the definition of XSLT as –

“XSLT (Extensible Stylesheet Language Transformations) is a language for transforming XML documents into other XML documents, or other formats such as HTML for web pages, plain text or XSL Formatting Objects, which may subsequently be converted to other formats, such as PDF, PostScript and PNG.”

Where to start?

At the OPAC level, the XSLT magic is primarily driven by two files i.e. (a) MARC21slim2OPACResults.xsl and (b) MARC21slim2OPACDetail.xsl. Koha’s default settings of how we see the search results on the OPAC and the document specific details in the Normal view, are defined in these two files.

N.B. Directly editing these two files is strictly not advised unless you are a XSLT guru.

Lucky for us that Koha’s system preferences provide options to override the defaults by creating new XSLT files and telling Koha to use the new ones instead. The screenshot below shows the default setting.

2017-05-04_03-05-15

The two files are available inside a folder named xslt under the locale of the selected theme you are using e.g. /usr/share/koha/opac/htdocs/opac-tmpl/bootstrap/en/xslt/ on a package-based installation. By default most of us in India would be using the English locale i.e. “en”. The default Koha theme is “bootstrap”. Thus if you are using a custom theme and/or different locale look for the files under that.

Step #1 : Copying the XSLT files

We will be fetching the 3-letter language code from MARC 008/35-37 of each marc record. Since our Koha instance here is named as “demo”, we’ll make a copy of the file MARC21slim2OPACResults.xsl as MARC21slim2OPACResults-demo.xsl and the MARC21slim2OPACDetail.xsl as MARC21slim2OPACDetail-demo.xsl.

N.B. You can give any name to the copies you make, but it is suggested that the less adventurous you are with the names, easier it will be for you to trace your steps back if you make a mistake while editing them. And trust me you will make mistakes, at least initially.

Step #2 : Editing MARC21slim2OPACResults-demo.xsl

The MARC21slim2OPACResults.xsl is what drives the display of results of a search on the Koha OPAC. And it does not show the value of 008/35-37 i.e. MARC language code, while displaying the search result. To add this facility, we need to do the following to our copy i.e. MARC21slim2OPACResults-demo.xsl. Around line no. 70 or there about, there will be this line <xsl:variable name="controlField008-30-31" select="substring($controlField008,31,2)"/>. Add the following line after it – <xsl:variable name="controlField008_35-37" select="substring($controlField008,36,3)"/>.

Explanation: we’re defining a new variable named controlField008_35-37 and in it we are storing the 3-digit value found at 008/35-37 from another variable – controlField008.

The second step in editing this file is to add the code to check if (a) 008/35-37 actually exists (e.g. if you have a blank 008 field, 008/35-37 won’t exist) and (b) it is set as something other than “und(i.e. undefined). The following code does that first and proceeds to match the value found in 008/35-37 against 23 different languages. These languages are English and the 22 official languages of India as per the Eighth Schedule of the Indian Constitution. You can add or remove any language from this list as you require. However, when you do, please make sure that the codes you setup in this list here are coming from the Code Sequence document of the MARC Code list for languages as published by the US Library of Congress.

This means that you must also be using these same exact codes in your MARC records in Koha.

2017-05-04_05-40-55

Where we place this code will depend on where we want this language to listed in the results. In our case, we decided to add it between Material type and Format. And thus in our case, the code was added at line no 637. If you too want language to be displayed between the Material Type and Format, then look out for this line <xsl:if test="string-length(normalize-space($physicalDescription))"> and add this block before.

Saved the file and moved on to the next section.

Step #3 : Editing MARC21slim2OPACDetail-demo.xsl

The process is similar to what we just did above. Open the file MARC21slim2OPACDetail-demo.xsl for editing and go to line no. 51 (or there about) and look for this line <xsl:variable name="controlField008" select="marc:controlfield[@tag=008]"/>. Add the following line immediately after that: <xsl:variable name="controlField008_35-37" select="substring($controlField008,36,3)"/>.

2017-05-04_06-02-17

Since we wanted the “Language of document” to come after Material type, we added the code immediate after the node <xsl:if test="$DisplayOPACiconsXSLT!='0'"> is closed but before <!--Series: Alternate Graphic Representation (MARC 880) -->; in our case, this worked out to be around line no. 195. Finally, we saved and closed the file.

Getting Koha to use the new XSLT files

The Koha system preference OPACXSLTResultsDisplay was changed from its original setting (i.e. “default”) and set to the path of our new file MARC21slim2OPACResults-demo.xsl i.e. /usr/share/koha/opac/htdocs/opac-tmpl/bootstrap/en/xslt/MARC21slim2OPACResults-demo.xsl. Likewise the other system preference OPACXSLTDetailsDisplay was changed to /usr/share/koha/opac/htdocs/opac-tmpl/bootstrap/en/xslt/MARC21slim2OPACDetail-demo.xsl. It was time to test our XSLT modifications.

And it works!

Since pictures are said to be worth a 1000 words, we will let before and after screen grabs do the talking here. Also if anyone wants to see directly how it actually looks after applying the changed XSLT, visit the URL http://demo-opac.l2c2academy.co.in/cgi-bin/koha/opac-detail.pl?biblionumber=18

Before : Using the default MARC21slim2OPACResults.xsl

chrome_2017-05-04_07-05-24

After : Using the new MARC21slim2OPACResults-demo.xsl

chrome_2017-05-04_07-17-09

Before : Using the default MARC21slim2OPACDetail.xsl

chrome_2017-05-04_07-48-28

After : Using the new MARC21slim2OPACDetail-demo.xsl

chrome_2017-05-04_07-49-09