Mike Griffin

I was messing around with some text manipulation recently. You know the type of thing, copy a table from a webpage and transform it to a CSV file or to JSON, nothing too difficult, but it can be time consuming.

To ease the tedium and save my fingers from some typing, I wanted to do a major search and replace in vim. I like to think I’m pretty competent in vim, and with regexes, so I thought this would be straightforward. In the end it was, but it took some time to find out how to do non greedy searches.

First things first, always use \v at the start of your regex, this allows you to use special characters without having to escape them. It’s the difference between /a\+b\+c\+ and /\ca+b+c+. It might not seem much there, but wait till you see the monstrosity I came up with at the end of this article!

In vim, the + character is used to match more than one of the previous character, for exacmple /\vg+ would match gggggg. This is all well and good until you want to match all up to a space. Take this string as an example:

Mike GRIFFIN        Liffey Valley AC    M    M35    1090

Using a regex of /\v.+ will match

Mike GRIFFIN        Liffey Valley AC    M    M35    

which is way more than I wanted.

Enter the non greedy multiplier, {-}.

Using a regex of /\v.{-} on the string above gives

Mike 

which is better.

That set me on my way to be able to transform lines like this:

16     Mike GRIFFIN        Liffey Valley AC    M    M35    1090    (17:17)    17:20    

into:

  - place: 16
    name: Mike GRIFFIN
    club: Liffey Valley AC
    category: M35
    time: 17m20s

The final regex was this maserpiece, or monstrosity, but it got the job done:

:%s/\v^(\d+)\s+(.{-})  +(.{-})  +\S  +(\S{2,3})  +\d+  +\S+  +(\d+):(\d+)/  - place: \1\r    name: \2\r    club: \3\r    category: \4\r    time: \5m\6s/