Parsing question

All,

A quick question hopefully.... I have data such as the following: yyyy-mm-dd hh:mm:ss - <some message>yy-mm-dd hh:mm:ss - <some message> (and this can be several timestamps long)

What I need to do is to split the first three dates and messages into fields and drop remaining data from field.

Using Text to Columns tool the separator is a date so can't use that. Using regex I'm not sure how I tokenize the data into fields. Maybe there's a different tool to be used. Any help appreciated.

Thanks,

Jeff

Parse

Expression

Data Investigation

Accepted answers

apathetichell

Sorry - saw that part later - try this with another regex tokenize/split to rows on the original data- -\<(.*)\> then you can union in via position your two tables.

If that's not working can you post a data sample?

All comments

apathetichell

regex tokenize. If it's exactly 3 and always 3 you can keep them as columns, otherwise add a recordid beforehand and split to rows...

your formula should be something like:

(\d{4}-\d{2}-\d{2}\s\d{2}\:\d{2}\:\d{2})

For your tokenizing. If that doesn't work then post some sample data and I can fix it... After converting you can add a date/time tool or formula and use the todate or datetimeparse functions (and if you've split to rows it's only one column to match for date/time)

jeffv

Thanks for response.... it won't always be three sometimes 1, sometimes multiple, but I only need to keep the first three entries.

Probably need some clarification though... using the formula you suggest with tokenize I only see dates in the three fields it creates. What I'd like to see id the date and the message. Putting ".*" at the end doesn't quite get it either as then it gets everything. (\d{4}-\d{2}-\d{2}\s\d{2}\:\d{2}\:\d{2}.*)

Thoughts?

Thanks,

Jeff

Quick Links

This months top contributors

AkimasaKajitani 387

mceleavey 383

mbarone 337

Hollingsworth 335

LanisC 335