CODEWITHSUNDEEP
javascriptregex
user773737
user773737

Reputation:

trouble using string.replace with regex

Given something a regex like this:

http://rubular.com/r/ai1LFT5jvK

I want to use string.replace to replace "subdir" with a string of my choosing.

Doing myStr.replace(/^.*\/\/.*\.net\/.*\/(.*)\/.*\z/,otherStr)

only returns the same string, as shown here: http://jsfiddle.net/nLmbV/

If you view the Rublar, it appears to capture what I want it to capture, but on the Fiddle, it doesn't replace it.

I'd like to know why this happens, and what I'm doing wrong. A correct regex or a correct implementation of the replace call would be nice, but most of all, I want to understand what I'm doing wrong so that I can avoid it in the future.

EDIT

I've updated the fiddle to change my regex from:

/^.*\/\/.*\.net\/.*\/(.*)\/.*\z/

to

 /^.*\/\/.*\.net\/.*\/(.*)\/.*$/

And according to the fiddle, it just returns hello instead of https://xxxxxxxxxxx.cloudfront.net/dir/hello/Slide1_v2.PNG

Upvotes: 0

Views: 56

Answers (2)

Jon Cooke
Jon Cooke

Reputation: 92

Before I get marked down, I know the question asks about regexp. The reason for this answer URLs are nearly impossible to process reliably with a regexp without writing fiendishly complex regexps. It can be done, but it makes your head hurt! If you are doing this in a browser, you can use an A tag in your script to make things much simpler. The A tag knows how to parse them into pieces, and it lets you modify the pieces independently, so you only need to deal with the pathname:

//make a temporary a tag
var a = document.createElement('a');
//set the href property to the url you want to process
a.href = "scheme://host.domain/path/to/the/file?querystring"
//grab the path part of the url, and chop up into an array of directories
var dirs = a.pathname.split('/');
//set 2nd dir name - array is ['','path','to','file']
dirs[2]='hello';
//put the path back together
a.pathname = dirs.join('/');

a.href now contains the URL you want. More lines, but also more hair left when you come back to change the code later.

Upvotes: 1

TimWolla
TimWolla

Reputation: 32711

It's that little \z in your regex.

You probably forgot to replace it with a $ sign. JavaScript uses ^ and $ as anchors, while Ruby uses \A and \z.

To answer your edit:

The match is always replaced as a whole. You'll want to group both the left side and the right side of the to-be-replaced part and reinsert it in the replacement:

url.replace(/^(.*\/\/.*\.net\/.*\/).*(\/.*)$/,"$1hello$2")

Upvotes: 2

Related Questions