Advance/Query Builder Search

1 User Stories
2 Classes
- 2.1 FilterSearch
- 2.2 QueryGroup
- 2.3 QueryRule
- 2.4 QueryOperator
- 2.5 QueryField
- 2.6 SortRule
3 How to operate
- 3.1 Fields and Operators
- 3.2 Grouping

To let end users create complex search queries Asta7 offers a query builder or advanced search. Instead of a fixed advanced search users can create searches/filters of their own using a simple UI.

User Stories

https://stiftelsenasta.atlassian.net/wiki/spaces/AO/pages/1634598936

Classes

FilterSearch

This is the root class that saves queries against a user. It is also possible to make a filter global.

QueryGroup

This is used to group multiple queries. Based on the group condition it can either be grouped as an AND query or an OR query. A query group can contain multiple query groups or query rules.

QueryRule

This is used to link the query field, query operator, and user-provided value. It is also possible to negate the query using the negate property of this class.

QueryOperator

This is what generates individual ES queries like query_string, term, range, etc. It uses the query field and user-provided value and converts them to the appropriate ES query. For example, the ContainsQueryOperator of TextQueryField generates an ES query_string query with wildcards at both ends.

QueryField

This class holds information about the actual ES field (name, label, nested path, etc.).

SortRule

This is used for generating sorting rules with field names and order. There can be multiple sorting rules.

How to operate

For every query in the filter, the user has to select a field first. Similar to the basic mode tree select a dropdown of all the available fields will be available for every rule to select the field. Then based on the selected field a list of operators will appear, and based on the selected operator an appropriate value field will be shown.

Fields and Operators

Operators are dependent on fields and field types. Below is the list of all the current operators available:

Every operator can be negated with the not operator. All the text queries are case insensitive.

Field Type	Operator	Description

Field Type	Operator	Description
any	exists	Has value.
varchar/text/uuid	equals	Has value that equals the provided value ignoring the case.
varchar/text	matches	Has any token/word matching the provided value.
varchar	contains	Has value containing the provided value.
	starts with	Has value that starts with the user provided value.
	ends with	Has value that ends with the user provided value.
	greate than	Has value that is greater than the provided value alphanumerically and ignoring the case.
	greate than equals	Has value that is greater than or equal to the provided value alphanumerically and ignoring the case.
	less than	Has value that is less than the provided value alphanumerically and ignoring the case.
	less than equals	Has value that is less than or equals to the provided value alphanumerically and ignoring the case.
	between	Has value that is greater than or less than the provided value alphanumerically and ignoring the case.
integer/float/serial/date/timestamp	equals	Has value that equals the provided value. (Including time in case of timestamp)
	greate than	Has value that is greater than the provided value.
	greate than equals	Has value that is greater than or equal to the provided value.
	less than	Has value that is less than the provided value.
	less than equals	Has value that is less than or equals to the provided value.
	between	Has value that is greater than or less than the provided value.
boolean	yes	Has TRUE value.
	no	Has FALSE value.
codeTableRef fields	in	Has value that equals to any of the provided set of values.
Digital file content field (a text field)	matches	Has any token/word matching the provided value.

Grouping

Rules can be divided into any number of groups. It is also possible to nest groups in any way needed. For each group, the group condition (AND/OR) can also be selected.

When generating the ES query all the rules in a group will be combined in the ES bool query. Furthermore, all the rules with the same or matching nested paths will be grouped in the same ES nested query. This makes sure that multiple rules can be added for a nested level or its sub-levels.

For example, If the group contains the following rules

Nested Path	Field	Operator	Value

Nested Path	Field	Operator	Value
serie	serie.identifikator	equals	A
serie	serie.navn	starts with	Oslo
serie.stykke	serie.stykke.startdato	greater than	1920
serie.stykke	serie.stykke.slutdato	less than	1950
serie.stykke.mappe	serie.stykke.mappe.navn	matches	Test

An ES query like the following would be generated

{
  "bool": {
    "must": [{
      "nested": {
        "path": "serie",
        "query": {
          "bool": {
            "must": [{
              "term": {
                "serie.identifikator": "A"
              }
            }, {
              "wildcard": {
                "serie.navn": "Oslo*"
              }
            }, {
              "nested": {
                "path": "serie.stykke",
                "query": {
                  "bool": {
                    "must": [{
                      "range": {
                        "serie.stykke.startdato": {
                          "gt": "1920"
                        }
                      }
                    }, {
                      "range": {
                        "serie.stykke.slutdato": {
                          "lt": "1950"
                        }
                      }
                    }, {
                      "nested": {
                        "path": "serie.stykke.mappe",
                        "query": {
                          "match": {
                            "serie.stykke.mappe.navn": {
                              "query": "Test"
                            }
                          }
                        }
                      }
                    }]
                  }
                }
              }
            }]
          }
        }
      }
    }]  
  }
}