[R] regular expression change in R version 2.3.0?

[rgentlem]

Hi,
  This has been reverted in R-devel, so you should get the old behavior
in it. Gabor Grothendieck posted a work around for R 2.3.x.

best wishes
   Robert


Thomas Girke wrote:
> The interpretation of regular expressions with repetition
> quantifiers in the 'gregexpr' function seems to have changed 
> between R Version 2.2.0 and 2.3.0. The 'gsub' function, however, 
> gives the same results in R Versions 2.2.0 and 2.3.0. Below is 
> an example that demonstrates the version differences of the 
> 'gregexpr' function. I am not sure whether this new behavior 
> is an intended change or represents a bug. Personally, I liked
> the old behavior of this function more useful, since it is 
> consistent with the Perl regular expressions.
> 
> Here are my questions:
> (1) Is there a possibility to obtain from 'gregexpr' 
> the old output of R version 2.2.0 when using regular 
> expressions with repetition quantifiers.
> 
> (2) How can one be informed about regular expression changes 
> and the associated functions in new versions of R? 
> 
> 
> Here is the example code to demonstrate the version difference 
> of the 'gregexpr' function between versions 2.3.0 and 2.2.0:
> 
> # Example string
> x <- "xaaaaxaaaax"
> 
> # gregexpr in Version 2.2.0  (2005-10-06 r35749)
> gregexpr("[a]{1,}", as.character(x), perl=T)
> [[1]]
> [1] 2 7
> attr(,"match.length")
> [1] 4 4
> 
> # gregexpr in Version 2.3.0 (2006-04-24)
> gregexpr("[a]{1,}", as.character(x), perl=T)
> [[1]]
> [1]  2  3  4  5  7  8  9 10
> attr(,"match.length")
> [1] 4 3 2 1 4 3 2 1
> 
> # gsub gives expected output in Versions 2.2.0 & 2.3.0
> gsub("[a]{1,}", "_", as.character(x), perl=T)
> [1] "x_x_x"
> 
> 
> Thanks in advance for your help.
> 
> Thomas
> 

-- 
Robert Gentleman, PhD
Program in Computational Biology
Division of Public Health Sciences
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N, M2-B876
PO Box 19024
Seattle, Washington 98109-1024
206-667-7700
rgentlem at fhcrc.org