Question

In the following content samples, I wrapped the lines to make it easier to read on Stackoverflow (so you don’t have to scroll to the right to look at the examples).

Content A:

"Lorem Ipsum\r\n
[img]http://example.org/first.jpg[/img]\r\n
[img]http://example.org/second.jpg[/img]\r\n
more lorem ipsum ..."

Content B:

"Lorem Ipsum\r\n
[img caption="Sample caption"]http://example.org/third.jpg[/img]
[img]http://example.org/fourth.jpg[/img]"

Content C:

"Lorem Ipsum [img]http://example.org/fifth.jpg[/img]\r\n
more lorem ipsum\r\n\r\n
[img caption="Some other caption"]http://example.org[/img]"

What I’ve tried:

content.match(/\[img\]([^<>]*)\[\/img\]/imu)
return example: "[img]...[/img]\r\n[img]...[/img]
content.scan(/\[img\]([^<>]*)\[\/img\]/imu)
return example: "...[/img]\r\n[img]..."

What I would like to accomplish when running the scan/match/regex solution over the above 3 Content Examples is to get every occurence of [img]...[/img] and [img caption="?"]...[/img] and put it in an array for later use.

Array
  1 : A : [img]http://example.org/first.jpg[/img]
  2 : A : [img]http://example.org/second.jpg[/img]
  3 : B : [img caption="Sample caption"]http://example.org/third.jpg[/img]
  4 : B : [img]http://example.org/fourth.jpg[/img]
  5 : C : [img]http://example.org/fifth.jpg[/img]
  6 : C : [img caption="Some other caption"]http://example.org[/img]

It would also be helpful to limit the “stripped content” to only where there is an open and closign tag, meaning when there is a [img] / [img caption="?"] and a missing [/img] afterwards, to ignore it.

I’ve read the http://www.ruby-doc.org/core-1.9.3/String.html up and down but can’t find anything that seem to work for this.

Update:

So I figured that this:

\[img([^<>]*)\]([^<>]*)\[\/img\]

will find either:

[img]something[/img]

and:

[img caption="something"]something[/img]

Now I just need to know how to catch every occurence inside the different contents. I can always just get it from the first to the last [img][/img] tags, so when there is other Lorem Ipsum in between it will get grabbed too.

Answer 1

You can use /\[img(?:\s+caption=".+")?\].+?\[\/img\]/ to scan the documents:

regex = /\[img(?:\s+caption=".+")?\].+?\[\/img\]/

text = <<EOT
Lorem Ipsum
[img]http://example.org/first.jpg[/img]
[img]http://example.org/second.jpg[/img]
more lorem ipsum ...

Content B:

Lorem Ipsum
[img caption="Sample caption"]http://example.org/third.jpg[/img]
[img]http://example.org/fourth.jpg[/img]

Content C:

Lorem Ipsum [img]http://example.org/fifth.jpg[/img]
more lorem ipsum

[img caption="Some other caption"]http://example.org[/img]
EOT

array = text.scan(regex)
puts array

Which generates:

[img]http://example.org/first.jpg[/img]
[img]http://example.org/second.jpg[/img]
[img caption="Sample caption"]http://example.org/third.jpg[/img]
[img]http://example.org/fourth.jpg[/img]
[img]http://example.org/fifth.jpg[/img]
[img caption="Some other caption"]http://example.org[/img]

If you want to ignore the tags and only grab the content, change the regexp to:

regex = /\[img(?:\s+caption=".+")?\](.+?)\[\/img\]/

Running again with that change returns:

http://example.org/first.jpg
http://example.org/second.jpg
http://example.org/third.jpg
http://example.org/fourth.jpg
http://example.org/fifth.jpg
http://example.org

(Rubular proof)

If you need to look for different tags, you can generate an “OR” list easily:

Regexp.union(%w[foo img bar])
=> /foo|img|bar/

If you need to make sure that “magic” characters are escaped beforehand:

Regexp.union(%w[foo img bar].map{ |s| Regexp.escape(s) })