Ok guys,

If you wanted to change the following paragraph in VIM using only commands. How would you replace the uppercase letters of every word except for those starting a sentence to lowercase?


The vi editor has a command, tilde (~),
that changes lowercase letters to
uppercase and vice versa.
Unfortunately the ~ command does not
work with a Unit of Measure or
a Repeat Factor, so you have to change
the case one character at a time.

:-D and, no I don't know how, or else I wouldn't be asking

Good luck, and thank you

I would move the curser over the character in question...then press [r] then type the replacement. Repeat for all characters that need to be changed. So in your case...place the curser over the "U" in "Unit of Measure", then press 'r' (literally: the 'r' key) then 'u' and presto chango, it's a lowercase 'u'. Repeat.

I haven't built it, but a look behind regular expression should be able to handle this. You just need to look for a character not after a period, and lower case it. The result would be something like this:

Normal Regex to match with a look behind:
/(?<=\.)(\w)/

VIM command to lowercase all uppercase letters
:%s/\([A-Z]\)/\l\1/g

Now, this will convert all upper-case to lower-case. You need to figure out now how to incorporate look behind matching in a VIM regex. Then, you can look for all uppercase letters not after a period and lower case them. Now, you have to also be careful of the first character and those on a new line. so you don't want to convert "^[A-Z]". This means you have to look at treating the block as new-lines as well.

As I said, this is a start, as it is a complex regex, so take this and see if you can finish it up.

You'll want to use something like sed or perl to do this.

This link will give you some useful input.

http://www.netsys.com/sunmgr/1997-10/msg00027.html

Looks like Andrew's got the better option.

Thanks.

Thanks for your help guys, I'll try these out and let you know my results.

I'm trying to improve upon my Vim skills and was randomly googling for "vim regex challenge" when I came upon this thread, and decided to pick up the challenge.

I think I've come up with something that finds an uppercase alphabetical character that is not immediately preceded by the beginning of the file, the beginning of a line or a ., ! or ? followed by a space or tab and substitutes it with itself in lowercase.

If you accept every prompt, the following text:

The vi editor has a command, tilde (~),
that changes lowercase letters to
uppercase and vice versa.
Unfortunately the ~ command does not
work with a Unit of Measure or
a Repeat Factor, so you have to change
the case one character at a time.
Hello. I am Joe.
HOLA! I am José! Why are you looking at me weird? Is something WRONG?

becomes:

The vi editor has a command, tilde (~),
that changes lowercase letters to
uppercase and vice versa.
Unfortunately the ~ command does not
work with a unit of measure or
a repeat factor, so you have to change
the case one character at a time.
Hello. I am joe.
Hola! I am josé! Why are you looking at me weird? Is something wrong?


And it goes:

%s/\%^\@<!\_^\@<!\([.!?]\s\)\@<!\(\u\)/\l\2/gc


I'm sure it's very verbose - it could probably be shorter with magic or very magic. Possibly, the three negative lookbehinds could be reduced to one somehow, or substituted with something more elegant.

This is my first post on this ancient-by-internet-standards thread on a forum I've never seen before, but I just couldn't leave it hanging after constructing my own attempt