How to extract file name from path's for both windows and nix format?

[att35]

Hi,

We have a data source containing File Path's from both Windows and Linux formats.  Applying regex separately works but how can I extract all file names regardless of format in a single search?

Following works for all Windows path's but for Linux, entire path gets extracted as file_name

| rex field=file_path "(?P<file_name>[^\\\]+)$"

 whereas this one works for Linux but does nothing on the Windows path.

 | rex field=file_path ".*\/(?<file_name>.*)$"

Is there a way to use both in same search? or a new regex that would work on either of the two formats?

[woodcock]

| rex field=file_path "([^\r\n\s\\\/]+[\\\/])+(?<file_name>.*)$"

[isoutamo]

Hi

Is this what you are looking for?

| makeresults 
| eval file_path="/as/bsb/asda/file.txt|c:\asda\asdadaa\aa\file.txt"
| eval file_path = split(file_path,"|")
| mvexpand file_path
``` Previous prepare sample data ```
| rex field=file_path "(?P<file_name>[^\\/\\\]+)$"

r. Ismo

[rut]

 | rex field=file_path "(?<file_name>[^\/\\\]+\..+)$"

Negation for / and \, match everything that has at least one dot. Tested on:

C:\test\test.tst
/test/test/test.tst
/test/test/test.tst.tst

[richgalloway]

This regex works in my limited testing.

| rex field=file_path "(?<file_name>[^\/\\]+)$"