RegEx with Capturing Groups Search Field Expression Plugin
The RegEx with capturing groups plugin runs a regular expression on the message body of a Log Event and returns only the text inside the capturing group you choose, not the whole match. Use it to pick a value out of its surroundings, such as 054 from CityId="054" or the 4th field of a delimited flat file record.
- ✅ Returns one, several, or all capturing groups of each match
- ✅ Global and Multi line options for record-by-record flat files
- ✅ Works on any text format in the message body

Animated walkthrough: the pattern [\w.]+@([\w.-]+) with group 1 on a logged Order-to-Cash order confirmation returns northwind.example.com, the group, not the whole address orders@northwind.example.com.
When to use it
Four RegEx plugins cover the combinations of where the text is and what part of the match you want. For every plugin, see Choose a plugin.
| Your data | You want | Use |
|---|---|---|
| Message body | The whole match | RegEx |
| Message body | Only the part inside a capturing group ( ) |
RegEx with capturing groups (this page) |
| One context value (you know the key) | The whole match | RegEx on Message Context |
| One context value (you know the key) | Only the part inside a capturing group | RegEx on Message Context with capturing groups |
| One context value (you know the key) | The value as it is, no pattern | Message Context Key |
| A group you need to transform or combine | Nested functions | Formula with regexGroup |
Expression syntax
Select RegEx with capturing groups in the Type list. The plugin shows these fields:

The RegEx with capturing groups plugin and its siblings in the Type list.
| Field | What goes in it |
|---|---|
| Expression | The regular expression, with at least one capturing group ( ). The help text under the field ends with "The pattern ignores case and white space: write a space as \s or [ ] and # as \# (an unescaped # starts a comment)." |
| RegEx groups | The group or groups to return: a single number (1), a list (1,3,4,5,6), a range (1,3-6), or group names (value1,value2, with no space after the comma). Group 0 is the whole match, and named groups are numbered after the unnamed ones. Left empty, the plugin returns the whole match. The help text under the field: "The groups to return, comma-separated: group numbers (1,3,4 or a range 1,3-6; 0 is the whole match, named groups are numbered after the unnamed ones) and/or group names (value1,value2). Leave empty to return the whole match." |
| Global | Checked: return every match. Unchecked: stop after the first match. |
| Multi line | Checked: ^ and $ match the start and end of each line, not only of the whole body. |
- Groups are numbered from 1 in the order their opening parentheses appear.
(?: )is a non-capturing group and gets no number. - A repeated group keeps the value of its last repetition.
^(?:(.*?)\|){4}therefore returns the 4th pipe-delimited field of a line, counting from 1. - Matching ignores case: the pattern
cityIdmatchesCityIdin Example 1. - Spaces in the pattern are ignored, and
#starts a comment:a bmatchesab, nota b. Write a literal space as\sor[ ], and a literal#as\#. - Each value is trimmed of leading and trailing white space, and an empty group is dropped.
- A pattern may run for 30 seconds per event at most.
- The Global checkbox above belongs to the plugin. The dialog has a second Global under Apply to, which applies the expression to all Message Types (see Add or manage Search Field).
- The plugin reads the whole message body; for large XML or JSON payloads, prefer XPath or JSON Path.
- The Test Expression tab has its own Global and Multi line checkboxes; tick them there before you press Test.

The plugin's fields with their help texts.

The plugin's Global checkbox, ticked on the Test Expression tab.
Configure and test the expression as described in Add or manage Search Field and Test Your Expression. When nothing matches, Test Expression reports Processed with warning: The Expression yields no result:

Test result when the pattern matches nothing in the message body.
Examples
Example 1: Extract city IDs from an order message
Input
<ns0:Orders xmlns:ns0="Common.Schemas/Nodinite/1.0">
<Order>
<Id>101</Id>
<Amount>1337</Amount>
<City CityId="054">Karlstad</City>
</Order>
<Order>
<Id>102</Id>
<Amount>10</Amount>
<City CityId="08">Stockholm</City>
</Order>
</ns0:Orders>
Expression
cityId="([0-9]+)"
- RegEx groups:
1 - Global: checked
- Multi line: unchecked
Result
054
08
The same pattern in the RegEx plugin returns CityId="054" and CityId="08".

Test Expression: group 1 of each match, two unique values. The first City line is blurred in the screenshot; it is the same as in the input above.
Example 2: Extract the 4th field from each pipe-delimited order record
Input
ORD|101|ExampleCompany1|456|Company Name ACME |Dieselgate valley 1|123 45|Flameburg|
ORD|102|ExampleCompany2|789|Company Name MECA |Radiator Spring 420|543 21|Apocalypsenburg|
Expression
^(?:(.*?)\|){4}
- RegEx groups:
1 - Global: checked
- Multi line: checked
Result
456
789
The group repeats four times per line and keeps the last value: ORD, 101, ExampleCompany1, then 456. Multi line makes ^ match at the start of each record, and Global returns both records.

Test Expression: the 4th field of each flat file record.
Example 3: Extract a field from the line that starts with a tag
Input
ST|123
KV|Z|AN|1337|ABC123
ID|1912121212|
Expression
^KV(?:(.*?)\|){4}
- RegEx groups:
1 - Global: checked
- Multi line: checked
Result
1337
After KV, the first repetition captures the empty text before the first |, then Z, AN, and 1337. Counting the tag KV as field 1, {4} returns the 4th field. Only one line starts with KV, so Global makes no difference here.

Test Expression: the 4th field of the line tagged KV.
Example 4: Extract the 3rd field from a semicolon-delimited header line
Input
HDR;ORDER;42
KV;Z;AN;1337;ABC123
ID;1912121212;
Expression
^HDR(?:(.*?)[;\r\n]){3}
- RegEx groups:
1 - Global: unchecked
- Multi line: unchecked
Result
42
Each repetition ends at a ; or a line break, so the last field of the line is also captured. Counting the tag HDR as field 1, {3} returns the 3rd field: the empty text before the first ;, then ORDER, then 42. Without Multi line, ^ matches only at the start of the body, where the HDR line is.

Test Expression: the 3rd field of the HDR line, with Global and Multi line unchecked.