Filtering by multiple values on a field with lucene local indexes.

Hey Michael,

A few steps to achieve this (based on your example):

• In "employmentType" index field you should store keys (non-display names) joined by " " (space), i.e.: "full_time" and "part_time" in your case
• Then in Search fields menu in your Page Type settings for field "employmentType" you should tick Searchable and Tokenized tickboxes
• Rebuild your search index
• After these steps you should be able to perform this Lucene search: +employmentType:(full_time) +employmentType:(part_time)

Key thing is to store no-space keys joined by space. When you select "tokenize" setting then the system will break this field by space automatically. From my experience, I store GUIDs instead of keys there.

Dmitry, is this possible without doing a custom indexer class? Or no? Those values do actually get displayed on the front-end, so I can't alter them in the custom table to not have the spaces. I am able to achieve the search using a single value filter with the space by escaping the space and dash IE:

+employmentType:Full\ \-\ Time

Any documentation on how to acheive the below? I am a bit unclear on how to do this:

"Key thing is to store no-space keys joined by space. When you select "tokenize" setting then the system will break this field by space automatically. From my experience, I store GUIDs instead of keys there."

I already have the fields I need multiple values set as "Searchable" and "Tokenized". Currently using the "Whitespace Analyzer" as my analyzer.

Just for reference, this is a custom table, not a page type. I doubt that changes anything, but food for thought.

Okay, I actually just kind of tried a "hack" and made a second field in my custom table, and when importing this data, I'm removing all spaces and dashes and setting them to underscores. That gives me data like "full_time" and "part_time".

The new fields is set as "Searchable" and "Tokenized" in the custom table "Search Fields" area. Still not able to query off multiple as you put in your answer. Any idea why that might be? An example below of a real query I'm trying with a similar field:

+fld_jp_position_category_code:(allied_health) +fld_jp_position_category_code:(food_services)

This query is not working in Luke where i am testing everything. Should i try something other than the Whitespace analyzer?

I do see in the "Explain structure" tab of Luke that both of those clauses are "MUST" types and I'm looking for more of a "food_services" OR "allied_health" type of query. Bit confused.

Making some headway. Went even further and removed the underscores from the imported data as well in the second fields. So now allied_health is just alliedhealth and so on.

This syntax now works in Luke, but it's not working in Kentico. Any suggestions would be helpful:

+fld_jp_position_category_code:(nursing OR professional OR alliedhealth)

Dmitry,

I ended up getting it to work. I think after I figured out that final syntax i posted above, i needed to refresh my indexes and everything started working.

Have a functioning filter now! Thanks so much!