Docs Expert Search Guide

Last modified 19 Feb 2026 01:17 +01:00

This page describes advanced search features of an Expert Search mode of documentation search. Expert Search is intended for users who are familiar with the documentation structure and want to quickly find relevant information. It by-passes the usual search multipliers and other scoring mechanisms and instead relies on manual selection on filter, search fields and other parameters.

Basic concepts

Expert Search can be accessed by clicking on the Expert Search toggle in the search modal or by typing qs: in the search input field.

It uses query_string query type of OpenSearch, with the backup option of simple_query_string in case of errors. This means that the search query syntax is the same as in the OpenSearch query string queries. For more information on the syntax, see the OpenSearch documentation.

Expert search enables features like:

  • Boolean operators in search queries (AND, OR, NOT)

  • Field-specific searches (e.g. title:, content:, tags:)

  • Grouping of search terms using parentheses

  • Wildcard searches using * and ?

  • Fuzzy searches using ~

  • Regular expression searches using /yourregexpattern/

  • And many more features of the OpenSearch query string syntax.

Default fuziness in query_string queries is 2, which means that it will match terms that are within an edit distance of 2 from the search term. This allows for some flexibility in search queries, but it can also lead to unexpected results if the search term is misspelled or if there are many similar terms in the index.
Regex patterns are applied only to the tokenized form of the content, which means that exact phrase matching is possible only if you specify the fields as xxx.keyword (e.g. title.keyword), but in this case, the search will be case-sensitive and will not allow for partial matches, also, text field does not have a keyword subfield, so multi-word regex patterns cannot be applied to the text field.

Difference between Expert Search and query_string OpenSearch query type

Expert Search mode offers a more user-friendly interface for constructing complex search queries, while still allowing users to take advantage of the powerful features of the OpenSearch query string syntax.

Differences that enable this behavior include:

  • The main difference is the presence of default search fields and their weak boosting, which allows users to get relevant results even if they are not familiar with the exact syntax of the query string queries. The fields and the boosting can be easily adjusted in the query.

  • Expert Search provides a backup mechanism to ensure that users get results even if their query contains syntax errors, which is not the case when using query_string directly.

  • Collapse clause used to filter out duplicate documents from different branches of the documentation. It does this by collapsing results that scored worse than the best scoring document, there is no multiplier for documents from branches that are not up-to-date.

Setting specific fields in the query will disable the default search fields and their boosting.

Backup query type

When using query_string, if the query contains syntax errors, the search will fail and return no results. If this happens, the search will display an warning with the error message and will automatically retry the search using simple_query_string query type. simple_query_string is a more forgiving query type that ignores invalid parts of the query and tries to execute the rest. It does not support all features of query_string, but it allows users to get results even if their query is not perfectly formatted.

Escaping special characters

In query_string queries, certain characters have special meanings (e.g. +, -, &&, ||, !, (, ), {, }, [, ], ^, ", ~, *, ?, :, \ and /). If you want to search for these characters as literals, you need to escape them with a backslash (\). Since the backslash itself is a special character, you need to escape it as well if you want to search for a literal backslash. For example, to search for nice: query!, you would need to enter nice\\: query\\! in the search input.

< and > characters are not escapable in query_string queries.

Example queries

In the following table, you can find some example queries that demonstrate the use of different features of the Expert Search mode.

Query Description

title:authentication AND keywords:oauth

This query searches for documents that have the word "authentication" in the title and the word "oauth" in the keywords.

text:(authentication OR authorization) AND NOT deprecated:true

This query searches for documents that have either "authentication" or "authorization" in the content

text:"midPoint is awesome"~5

This query searches for documents that have the exact phrase "midPoint is awesome" in the content, allowing for a maximum edit distance of 5.

/au[a-z]h/

This query uses regular expression to search for documents that have a word that starts with "au", followed by any lowercase letter, and ends with "h" in the content (e.g. "auth").

Fields

There are two types of fields that can be used in Expert Search queries: search fields and filter fields. In the default configuration, the filter fields are not used and the search query is executed against the following fields:

  • title (boosted by 2)

  • second_titles (boosted by 1.5)

  • third_titles (boosted by 1.2)

  • fourth_titles (boosted by 1.1)

  • fifth_titles (boosted by 1.0)

  • keywords (boosted by 2)

  • search-alias (boosted by 2.5)

  • text (no boost)

Here is the table of all available fields:

{}

Field name Description Type Default Default boost

title

The main title of the document

string

Yes

2.0

second_titles

The second level titles of the document

Array[string]

Yes

1.5

third_titles

The third level titles of the document

Array[string]

Yes

1.2

fourth_titles

The fourth level titles of the document

Array[string]

Yes

1.1

fifth_titles

The fifth level titles of the document

Array[string]

Yes

1.0

keywords

Keywords associated with the document

Array[string]

Yes

2.0

search-alias

Search alias of the document

string

Yes

2.5

text

The full text content of the document

string

Yes

1.0

type

The type/category of the document (e.g. Guide, Book, Reference)

string

No

X

branch

The branch slug of the document (e.g. before-4.8), in case of not branched document, the value is notBranched

string

No

X

midpointBranch

The name of the midpoint branch of the document in github (e.g. docs/before-4.8). Be careful, this field is not existant in documents that are not branched

string

No

X

url

The URL of the document

string

No

X

lastModificationDate

The date of the last modification of the document

date

No

X

preview

A short preview of the document content, it may not exist in some documents

string

No

X

description

A short description of the document, it may not exist in some documents

string

No

X

upkeep-status

The upkeep status of the document, possible options are: green, orange, yellow, red, this field may not exist in some documents

string

No

X

obsolete

Whether the content of the document is obsolete or not, may not exist

boolean

No

X

deprecated

Whether the content of the document is deprecated or not, may not exist

boolean

No

X

experimental

Whether the content of the document is experimental or not, may not exist

boolean

No

X

planned

Whether the content of the document is planned or not, may not exist

boolean

No

X

outdated

Whether the content of the document is outdated or not, may not exist

boolean

No

X

connector-name

The name of the connector associated with the document, does not exist in most of the documents

string

No

X

connector-protocol

The protocol of the connector associated with the document, does not exist in most of the documents

string

No

X

midpoint-feature

The name of the MidPoint feature associated with the document, does not exist in most of the documents

string

No

X

since

The version of MidPoint since which the content of the document is relevant, does not exist in most of the documents

string

No

X

upkeep-note

A note about the upkeep status of the document, does not exist in most of the documents

string

No

X
Was this page helpful?
YES NO
Thanks for your feedback