Forum: Ruby best way to extract data from a string?

Hi
I have a string
str = "file name_Width34.56_Height45.97_.jpg"

what is the best way to extract the the data from the string?

file name => "file name.jpg"
width => "34.56"
hieght => "45.97"

2012/10/14 Erez Ben shoham <lists@ruby-forum.com>

> --
> Posted via http://www.ruby-forum.com/.
>
> string = "file name_Width34.56_Height45.97_.jpg"
string =~ /(.+)_Width(\d+.\d+)_Height(\d+.\d+)_.jpg/
puts $1, $2, $3

http://rubular.com is your friend, try it.

Hi,

the best way would be to use a proper data format. But I assume the 
strings are given and cannot be changed?

Jan E. wrote in post #1079799:
> Hi,
>
> the best way would be to use a proper data format. But I assume the
> strings are given and cannot be changed?

sorry for my English
the purpose of this is
there are photoshop .psd files localy on a mac.
and we need to upload small jpgs versions of them to a web app with the 
original width and height data in the fastest hassle free way.
the upload process  changes the size of the images.
so we created a photoshop script that automatically creates the small 
version jpgs and inserts the original width height data to the name of 
the file.
then the user drag and drop the small jpgs to the browser and then we 
upload the images and update the data accordingly.

Use regular expressions

*@ImranNazirMir* Follow Me on Pinterest
<http://pinterest.com/imrannazirmir/> View Imran Nazir's profile on
LinkedIn <http://uk.linkedin.com/pub/imran-nazir/2/402/a57>

str = 'file name_Width34.56_Height45.97_.jpg'

name, width, height, ext = str.split('_')
name << ext

width = width[5..-1]
height = height[6..-1]

puts name, width, height

--output:--
file name.jpg
34.56
45.97


> so we created a photoshop script that automatically creates the small
> version jpgs and inserts the original width height data to the name of
> the file.

Whoever wrote that script was not thinking ahead.

7stud -- wrote in post #1079843:
> Whoever wrote that script was not thinking ahead.

Very true. Anything (even a crappy CSV) would have been better than 
mangling all data in the file name. What if the original filename 
already includes an underscore (which is not that unlikely)? Good luck 
parsing it.