Expression match references

Referencing matches in expressions is done using $. In addition to this a match group number may be added to just retrieve part of the expression match. The applicability and effect that this has depends on the type of expression used.

References to <split> Match Groups

In the following example a line matched by a parent <split> expression is referenced by a child <data> element.

<split delimiter="\n" >
  <data name="line" value="$"/>
</split>

A <split> element matches content up to and including the specified delimiter, so the above reference would output the entire line plus the delimiter. However there are various match groups that can be used by child <group> and <data> elements to reference sections of the matched content.

To illustrate the content provided by each match group, take the following example:

"This is some text\, that we wish to match", "This is the next text"

And the following <split> element:

<split delimiter="," escape="\">

The match groups are as follows:

• $ or $0: The entire content that is matched including the specified delimiter at the end

"This is some text\, that we wish to match",

• $1: The content up to the specified delimiter at the end

"This is some text\, that we wish to match"

• $2: The content up to the specified delimiter at the end and filtered to remove escape characters (more expensive than $1)

"This is some text, that we wish to match"

In addition to this behaviour match groups 1 and 2 will omit outermost whitespace and container characters if specified, e.g. with the following content:

"  This is some text\, that we wish to match  "  , "This is the next text"

And the following <split> element:

<split delimiter="," escape="\" containerStart="&quot" containerEnd="&quot">

The match groups are as follows:

• $ or $0: The entire content that is matched including the specified delimiter at the end

" This is some text\, that we wish to match " ,

• $1: The content up to the specified delimiter at the end and strips outer containers.

This is some text\, that we wish to match

• $2: The content up to the specified delimiter at the end and strips outer containers and filtered to remove escape characters (more computationally expensive than $1)

This is some text, that we wish to match

References to Match Groups

Like the <split> element various match groups can be referenced in a <regex> expression to retrieve portions of matched content. This content can be used as values for <group> and <data> elements.

Given the following input:

ip=1.1.1.1 user=user1

And the following <regex> element:

<regex pattern="ip=([^ ]+) user=([^ ]+)">

The match groups are as follows:

• $ or $0: The entire content that is matched by the expression

ip=1.1.1.1 user=user1

• $1: The content of the first match group

1.1.1.1

• $2: The content of the second match group

user1

Match group numbers in regular expressions are determined by the order that their open bracket appears in the expression.

References to <any> Match Groups

The <any> element does not have any match groups and always returns the entire content that was passed to it when referenced with $.