Developer forum

Forum » PIM » Another day, another question...

Another day, another question...

Stephen Anthony Jackson
Stephen Anthony Jackson
Reply

Does anyone know why some of my text fields in the PIM are having their values lower-cased in the search filters, when the values in the PIM and the product list on the same page are all upper case? 

I cannot find a setting that dictates case on the filters

Cheers

//Steve


Replies

 
Adrian Ursu Dynamicweb Employee
Adrian Ursu
Reply

Hi Stephen,

There is a setting of the index builder to analyze all product fields. by default, it is set to true. THis means that all values are analyzed and lowercased.
You have 2 options:
1. Change the setting to false
2. Create non-analyzed fields based on your regular fields and use those for filtering. Your facets will now reflect the correct casing of the values.

For searching it is better to keep them analyzed.

Adrian

 
Stephen Anthony Jackson
Stephen Anthony Jackson
Reply

Default = true? Ouch. I would argue that analyzing should be something you enable when needed. Considering the majority of use cases will involve filtering on specific values (also values you dont want splitting up as analysis does), then it makes no sense that it is enabled by default?  Only free text search fields need to be analyzed.

Thanks for the tip though!

 
Stephen Anthony Jackson
Stephen Anthony Jackson
Reply

Hi Adrian

These are product category fields though.

And in the index builder, none of them are set to be analyzed.

But still they are lower cased

 
Adrian Ursu Dynamicweb Employee
Adrian Ursu
Reply

Hi Stephen,

In this case, it could be another cause for your case.
But I know too little about your setup to advise properly.

Adrian

 
Stephen Anthony Jackson
Stephen Anthony Jackson
Reply

Apologies. I was looking in the wrong place. Tired. Yes in the index schema extender, where the defaul definitions are, its set to be analyzed.

How can I change the default behaviour so NO fields are analyzed?

 

//Steve

 
Adrian Ursu Dynamicweb Employee
Adrian Ursu
Reply

Hi Stephen,

Go to your index, click on each index builder (if you have multiple builders) and look for this setting:

It needs to be set to "True" instead of "False".

Adrian

 
Nuno Aguiar Dynamicweb Employee
Nuno Aguiar
Reply

Hi Stephen,

 

Beware on changing this default because you may be introducing a lot of UX issues. The reason why it's analyzed by default is because users will search for your product data not necessarily in the same casing of the data, so this ensures that if they search SKU "ABC123" or "abc213" they all get the same result.

 

I know I'm using the SKU (aka Product number), which is a field you'll liked handle differently, but consider free text search on descriptions and names and such.

 

From my personal experience I rather have the default be set to "True", and handle the exceptions (creating non-analyzed fields) as needed.

 

Best Regards,

Nuno Aguiar

 
Stephen Anthony Jackson
Stephen Anthony Jackson
Reply

Hi Nuno. 

In this user case, there is no search feature on default fields, or any other fields for that matter. The product catalogue is small with 4 well defined categories which are presented in seperate areas and linked from links at top of page. The lists themselves are then filtered purely by facets. So there is currently no need for any freetext searching, and this the default lower-casing is just annoying. But thanks for the tip! It makes sense ofc when it comes to search matching. If we do have a case where we need free-text search on a large text field, I will just activate analyze for that field only. Personally, I would prefer that to the default behaviour to be honest. It's not a big deal to remember to activate analyze, is it? Or is the problem that you cant change the extender fields individually?

 
Nuno Aguiar Dynamicweb Employee
Nuno Aguiar
Reply

Hi Stephen,

 

  • It's not a big deal to remember to activate analyze, is it?
    No, it is not.
     
  • Or is the problem that you cant change the extender fields individually?
    That's not a problem either

 

My callout comes from my personal experience based on the most likely scenario. One can say that over the years I've had lesser issues (during project development and post go-live/support desk) by having the default behavior to Analyze all fields and then handle the necessary ones as exceptions, versus the opposite.

Obviously that's my/our experience based on the customers/projects we faced, hence our preferred/recommended path (and exceptions do apply). 

 

Best Regards,

Nuno Aguiar

 
Stephen Anthony Jackson
Stephen Anthony Jackson
Reply
  • Or is the problem that you cant change the extender fields individually?
    That's not a problem either

But I cannot change the extender fields. I click them and nothing opens. 

Is there somewhere to change the indexing settings in the ecommerce field type settings, cos I cant find that either?

Best Regards

//Steve

 
Nuno Aguiar Dynamicweb Employee
Nuno Aguiar
Reply

Hi Stephen,

 

I'm starting to wonder I may have mis-interpreted your "extender fields" question. Can you rephrase and/or point me to what you mean?

 

And what it would mean in the concept of them needing to be Analyzed vs Non-Analyzed?

 

//Nuno

 
Stephen Anthony Jackson
Stephen Anthony Jackson
Reply

No you understood it fine :)

 

You must be logged in to post in the forum