[R] regular expression in gsub() for strings with leading backslash

[Mike Miller]

On Fri, 29 Apr 2011, Duncan Murdoch wrote:

> On 29/04/2011 7:41 PM, Miao wrote:
>
>> Can anyone help on gsub() in R?  I have a string like something below, and
>> wanted to delete all the strings with leading backslash, including 
>> "\xa0On",
>> "\023, "\xab", and many others.   How should I write a regular expression
>> pattern in gsub()?  I don't care how many characters following backslash.
>
>
> If those are R strings, none of them contain a backslash.  In R, a backslash 
> would always be printed as \\.
>
> \x is the introduction to a hexadecimal encoding for a character; the next 
> two characters show the hex digits.  So your first string contains a single 
> character \xa0, the third one contains \xab, and so on.
>
> The \023 is an octal encoding for a single character.


If we were dealing with a leading backslash, I guess this would do it:

gsub("^\\\\.*", "", txt)

R would display a double backslash, but I believe that represents a single 
backslash.  So if the string were saved using write.table, say, only a 
single backslash would be stored.

> a <- "\\This is a string."
> a
[1] "\\This is a string."
> gsub("^\\\\", "", a)
[1] "This is a string."
> a
[1] "\\This is a string."
> gsub("^\\\\.*", "", a)
[1] ""
> gsub("^\\\\.*", "", c(a,"Another string","\\more"))
[1] ""               "Another string" ""
> write.table(a, file="a.txt", quote=F, row.names=F, col.names=F)

$ cat a.txt
\This is a string.

Apparently this is not what the OP really wanted.  The OP probably wanted 
to remove characters that were not from the regular ASCII set.


Mike

--
Michael B. Miller, Ph.D.
Minnesota Center for Twin and Family Research
Department of Psychology
University of Minnesota