Parse a fixed width-encoded field into known columns. It can parse values of the form:

logscale
value 1 value 2 value 3, widths [8,8,8]
value 1value 2value 3, widths [7,7,7]

For a log line like this:

logscale
2017-02-22T13:14:01.917+0000 [main thread] INFO statsModule got result="117success 27% 3.14"

Using parseFixedWidth(result, columns=[count, status, completion, precision, sourcetask], widths=[3,9,4,10,10) will add these fields:

logscale
count: 117
status: success
completion: 27%
precision: 3.14

Sourcetask will not get assigned a value, as there were too few columns in the input for that. Values are trimmed after they have been extracted, i.e. success will become success from the above example.

Use the (unnamed) field parameter to specify which field should be parsed. Specify @rawstring to parse the raw string

ParameterTypeRequiredDefaultDescription
columns[string]true Names of columns to extract from field.
fieldstringfalse@rawstringField that holds the input in fixed width form. [a]
trimbooleanfalsetrueRemove leading and trailing white-space from fields after extracting.
widths[number]true Widths of columns.

[a] If an argument name is not given, field is the default argument.

Fixed width parse the result field from a log line:

accesslog
statsModule got result="117success 27% 3.14

This will add the following fields to the event:

  • count=117

  • status=success

  • completion=27%

  • precision=3.14

sourcetask will not get set as the input is too short.

logscale
parseFixedWidth(result, columns=[count, status, completion, precision, sourcetask], widths=[3,9,4,10,10])