Partial GSUB match / replacement

ruby-talk
1

I need to match a somewhat complicated match using gsub, then modify
part of it, leaving the rest intact

for example, replacing
"url(http://www.google.com)"
with
"url(HTTP://WWW.GOOGLE.COM)"

this is the closest answer I can come up with
gsub /(url\(['"]?)([^\)'"]+)(['"]?\))/, "\\1#{"\\2".capitalize}\\3"
or
gsub /(url\(['"]?)([^\)'"]+)(['"]?\))/, "\\1\\2.capitalize\\3"
but neither of these solutions work

It doen't have to be with gsub, I just thought this would be the most
straightforward way

ideas?

thanks

···

--
Posted via http://www.ruby-forum.com/.

2

I need to match a somewhat complicated match using gsub, then modify
part of it, leaving the rest intact

for example, replacing
"url(http://www.google.com)"
with
"url(HTTP://WWW.GOOGLE.COM)"

this is the closest answer I can come up with
gsub /(url\(['"]?)([^\)'"]+)(['"]?\))/, "\\1#{"\\2".capitalize}\\3"
or
gsub /(url\(['"]?)([^\)'"]+)(['"]?\))/, "\\1\\2.capitalize\\3"
but neither of these solutions work

It doen't have to be with gsub, I just thought this would be the most
straightforward way

ideas?

You can use a block with gsub, where the back-references are
accessible via the $1, $2, etc:

"url(http://www.google.com)".gsub(/(url\(['"]?)([^\)'"]+)(['"]?\))/) {"#{$1}#{$2.upcase}#{$3}"}

=> "url(HTTP://WWW.GOOGLE.COM)"

Note that I replaced #capitalize with #upcase to upcase all characters
not just the first one.

HTH,
Ammar

···

On Sun, Nov 21, 2010 at 1:04 AM, Shea Barton <shea@sysach.com> wrote:

3

You can use following regexp:
string.gsub!(/(url\(['"]?)(.+?)(["']?\))/) { "#{$1}#{$2.upcase}#{$3}" }

It is similar to yours with two differences:
1) I'm using (.+?)
question mark here means that + is no longer greedy, it will match only
to the point where next element in regexp is detected: and in your case
that is either ", ' or )

2) You cannot use functions on references like "\\2.capitalize" or
"\\2".capitalize. It is just enterpreted as string. If you want to get
matches as actual string representations, you should use a block.

[http://ruby-doc.org/core/classes/String.html#M000817]

hope this helps :slight_smile:

···

--
Posted via http://www.ruby-forum.com/.

4

"url(http://www.google.com)".sub( /\(.*?\)/ ){|s| s.upcase}
    ==>"url(HTTP://WWW.GOOGLE.COM)"

···

On Nov 20, 5:04 pm, Shea Barton <s...@sysach.com> wrote:

I need to match a somewhat complicated match using gsub, then modify
part of it, leaving the rest intact

for example, replacing
"url(http://www.google.com)"
with
"url(HTTP://WWW.GOOGLE.COM)"

this is the closest answer I can come up with
gsub /(url\(['"]?)([^\)'"]+)(['"]?\))/, "\\1#{"\\2".capitalize}\\3"
or
gsub /(url\(['"]?)([^\)'"]+)(['"]?\))/, "\\1\\2.capitalize\\3"
but neither of these solutions work

It doen't have to be with gsub, I just thought this would be the most
straightforward way

ideas?

thanks

--
Posted viahttp://www.ruby-forum.com/.

5

Thanks a bunch

Although the last example by w_a_x_man looks efficient and simpler, I
needed a very specific match, as it actually matching
(http://www.google.com) and not even considering the url, then just
trying to capitalize the parentheses.

I went with
"url(http://www.google.com)".sub /(url\(['"]?)([^\)'"]+)(['"]?\))/,
"#{$1}#{$2.upcase}#{$3}"

···

--
Posted via http://www.ruby-forum.com/.

6

Actually I'm having a bit more trouble.

When I use css.gsub /(url\(['"]?)([^\)'"]+)(['"]?\))/,
"#{$1}#{$2.upcase}#{$3}"

on a block of css, what happens is, every instance of url(...), gets
replaced with the upcase of the last instance of it.

so
body {
background: url(/images/site_background.png)
}
p {
background: url(/images/image.png);
margin-left: 10px;
}
=>
body {
background: url(/IMAGES/IMAGE.PNG)
}
p {
background: url(/IMAGES/IMAGE.PNG);
margin-left: 10px;
}

···

--
Posted via http://www.ruby-forum.com/.

7

My bad. had to put the replacement string in a block
When I use css.gsub(/(url\(['"]?)([^\)'"]+)(['"]?\))/)
{"#{$1}#{$2.upcase}#{$3}"}

···

--
Posted via http://www.ruby-forum.com/.