Regexp and using the matched value returned/found

Ruby
1

Hi

I’m trying to do a regex search and using what I find a key to a Hash to
dig up some formatting info.

But what’s happening is the literal \1 or \0 (or \1 \0) is not being
resolved to what’s being found.

Any help would be awesome.

Ex: (Doesn’t work)

regex looks like this: @swi_start = /SWI((dm)|(ph)|(ty)|(st))st/

// executed in my code:
line.gsub!( swi_start, self.colorize_token( ‘\1’ ) )

also not working: line.gsub!( swi_start, self.colorize_token( ‘\1’

) )

def colorize_token( event )
puts “Event: " << event
return “<font color=”#{@@color_list[ event ]}”>"
end

Output:
Event: \1

Ex: (This worked!!!)

line.gsub!( swi_start, “#{@red_start}\1#{@red_end}”) then

Output:
dm

Lastly please notes I’ve tried using " instead of ’ in my method call
without any change.

2

Hi,

2009/4/13 Jean N. [email protected]:

Ex: (Doesn’t work)
 end

Lastly please notes I’ve tried using " instead of ’ in my method call
without any change.

You should use block format in that case
Try
line.gsub!(swi_start){colorize_token($1)}
or
line.gsub!(swi_start){|x| colorize_token(x)}

Regards,

Park H.

3

Excellent all

Thank you for the info. Using the block based notation worked much
better. (Also, Robert thank you for the additional info, very
informative).

My next question is:

Can someone direct me to a link to reading materials (my Google skills
seem to stink as I am not finding what I expect) on the pattern matching
variables I can use like $1 mentioned above as I am now trying to match
the WHOLE pattern and not just the “dm,ph,ty,st” portion.

So when I use:

line.gsub!(swi_start){colorize_token($1)}

The $1 only returns the aforementioned 2 char strings, I’d like to get
the whole thing “SWIdmst” or “SWItynd” etc etc.

Thanks.

4

On Mon, Apr 13, 2009 at 8:19 PM, Jean N. [email protected]
wrote:

[…]
Can someone direct me to a link to reading materials (my Google skills
seem to stink as I am not finding what I expect) on the pattern matching
variables I can use like $1 mentioned above as I am now trying to match
the WHOLE pattern and not just the “dm,ph,ty,st” portion.

The keywords seem to be ‘ruby pre-defined variables’…
First two hits:
http://is.gd/saBm
http://is.gd/saDy

[…]
The $1 only returns the aforementioned 2 char strings, I’d like to get
the whole thing “SWIdmst” or “SWItynd” etc etc.

$&

solidarity,
lasitha

5

On 13.04.2009 05:56, Jean N. wrote:

I’m trying to do a regex search and using what I find a key to a Hash to
dig up some formatting info.

But what’s happening is the literal \1 or \0 (or \1 \0) is not being
resolved to what’s being found.

Heesob gave you the solution already but I’d like to add a bit of
explanation why this is the solution.

Any help would be awesome.

Ex: (Doesn’t work)

regex looks like this: @swi_start = /SWI((dm)|(ph)|(ty)|(st))st/

The capturing groups are not needed. This is sufficient:

/SWI(dm|ph|ty|st)st/

// executed in my code:
line.gsub!( swi_start, self.colorize_token( ‘\1’ ) )

You need to be aware that all method arguments are evaluated before
the method is actually invoked. Since you need an argument for
#colorize which depends on the current match this can never work
because the #colorize is finished before #gsub! even starts matching.

Ex: (This worked!!!)

line.gsub!( swi_start, “#{@red_start}\1#{@red_end}”) then

No, this does not do the same as you attempted to do with the code above
because in this case the replacement cannot depend on the match. In
other words, you always insert the same color.

Output:
dm

Lastly please notes I’ve tried using " instead of ’ in my method call
without any change.

This is irrelevant in this case as the quote type does not affect the
time of evaluation (see above).

Kind regards

robert