Search Syntax

This search is based on the popular Apache Lucene search engine (possibly implemented on a Solr server) where the indexing of words in names and description:

• ignores case sensitivity. Therefore, search for Customer will also return customer, and CUSTOMER.
• ignores any space or special characters such as _ : . [ ] ( ) which are treated as word breakers by the search tokenizer. Therefore, search for "Customer ID" will also return: 2023_CUSTOMER_ID, Customer:ID, Customer[id].

When using the search in the UI:

• Before you type, a simple click will list the history of your previous searches.
• As you type, search is automatically activated offering you choices that you can select in the suggested list in matching priority order. Hit enter only to go to the advanced search window.

Basic Search

You may search for very precise results by specifying a search using the built-in query language (similar to popular web based search engines) as follows:

One can, of course, append these together into a search string. E.g.:

When searching for words with spaces or punctuation characters, such exact words must be quoted (e.g. “Product SKU”). By default, the search is only based on the name.

Case sensitivity varies by object. The search is case-sensitive search for objects, like Labels, and custom attributes and case-insensitive for text, like name and description, by default.

For system objects, like Labels, in general the UI is defined to allow you to pick the object (not the text of its name). You can define two Labels, web and Web, and search for any or all of them. When you need to search for both web and Web labels you should use the Equal ANY option.

The characters "_" (underscore), ":" (colon), "'" (single quote) and "." (dot) as well as white space characters are treated as word breakers by the search tokenizer. Therefore the search text "_end" will be tokenized as "pll" and thus will match both “example_end” and “append”. Furthermore, any search text that is not quoted by the double quote character (") will be surrounded by the wildcard character to match more objects. If you do not want wildcard characters to be added, quote the search text with the double quote character, i.e. make the search text "_ end ". You will find objects with only "_ end " not "append".

Semantic Search

In addition, the semantic search offers the use of the following key words in your search:

• Object Types: term, dataset, attribute, entity, table, column, file, field, database, model, data model, and report. Such key words can be used in singular or plural (e.g. terms).
• Model Categories: glossary, glossaries, database, databases, db, modeling, data modeling, bi, reporting, tableau, bo, cognos, and obi.

Note that the above key words do not need to be quoted when space is used (e.g. data model).

Then such key words can be used in a user friendly convenient syntax with additional qualifier key words as follows:

• customer table: returns all database tables having customer in the name.
• datasets with ssn : returns all tables (or files) with a column (or fields) having ssn in the name.
• tables with sensitive label: returns all tables with a label applied to them having sensitive in the name.
• customer in tableau: returns any tableau BI reports having customer in the name of any objects.

The search uses semantic search rules so that you may specify fairly sophisticated queries within a simple syntax. This syntax is documented below.

Semantic Search BNF Syntax

Example