Regex for ^ (the caret symbol)?

On 13-01-21 3:20 PM, Jeff Newmiller wrote:

Apparently Extended RegExp syntax eliminated the
"^-is-an-ordinary-character-except-for-two-uses" meaning that I am
familiar with from the Basic RegExp usage, since GNU grep with the -e
option also refuses to match the carat unless it is escaped. The TRE
library treats BRE as obsolete, so we only get ERE and Perl regexes in
R. So I guess it isn't a bug, but is rather a "feature".

I re-read the ?regex help page, and I think it does actually say this,
so we don't even have a documentation error as I thought before.  When
it is saying that ^ is a plain character except when it comes first, it
is talking about first within a character class, e.g. [a^] meaning "a"
or "^" as opposed to [^a] meaning "not a".

Duncan Murdoch


---------------------------------------------------------------------------

Jeff Newmiller                        The     .....       .....  Go
Live...
DCN:<jdnewmil at dcn.davis.ca.us>        Basics: ##.#.       ##.#.  Live
Go...
                                       Live:   OO#.. Dead: OO#..  Playing
Research Engineer (Solar/Batteries            O.O#.       #.O#.  with
/Software/Embedded Controllers)               .OO#.       .OO#.
rocks...1k
---------------------------------------------------------------------------

Sent from my phone. Please excuse my brevity.

Duncan Murdoch <murdoch.duncan at gmail.com> wrote:

On 13-01-21 1:05 PM, Jeff Newmiller wrote:

So what is the special behavior of the ^ symbol when not at the

beginning of the string that occurs when it is not escaped?

I think it retains its meaning as an assertion that it occurs at the
beginning of the line, and so a pattern like "a^b" could never match
anything.  It's not very useful in this context, but I expect it's
easier to implement in the case of complicated patterns, where some
paths through the pattern put it at the beginning and others don't,
e.g.

(a|)^b

has two possible patterns:  a^b and ^b.

Duncan Murdoch

---------------------------------------------------------------------------

Jeff Newmiller                        The     .....       .....  Go

Live...

DCN:<jdnewmil at dcn.davis.ca.us>        Basics: ##.#.       ##.#.  Live

Go...

                                        Live:   OO#.. Dead: OO#..

Playing

Research Engineer (Solar/Batteries            O.O#.       #.O#.  with
/Software/Embedded Controllers)               .OO#.       .OO#.

rocks...1k

---------------------------------------------------------------------------

Sent from my phone. Please excuse my brevity.

Duncan Murdoch <murdoch.duncan at gmail.com> wrote:

On 13-01-21 11:48 AM, Jeff Newmiller wrote:

I am not sure I understand what worked perfectly, since it is my

understanding that ^ is only special at the beginning of the regex

(to

anchor the pattern at the beginning of the target string) or as the
first character of a character set (to indicate exclusion of the

listed

characters). In any other position the caret should behave like an
ordinary character. That is, your original pattern should have

worked

as-is. This is supported by the help page documentation for regex in
the paragraph below the definition of [:xdigit:]. I think this is a

bug

in R.

It's a documentation error rather than a bug.  The ^ character is
special anywhere in the extended RE syntax defined by the TRE

library

or the Perl-compatible library that we use.  This is inconsistent

with

the POSIX standard, which might be what you were thinking of.

Duncan Murdoch

---------------------------------------------------------------------------

Jeff Newmiller                        The     .....       .....  Go

Live...

DCN:<jdnewmil at dcn.davis.ca.us>        Basics: ##.#.       ##.#.

Live

Go...

                                         Live:   OO#.. Dead: OO#..

Playing

Research Engineer (Solar/Batteries            O.O#.       #.O#.

with

/Software/Embedded Controllers)               .OO#.       .OO#.

rocks...1k

---------------------------------------------------------------------------

Sent from my phone. Please excuse my brevity.

mtb954 at gmail.com wrote:

Hi Tsjerk, many thanks...that worked perfectly!

Mark Na



On Mon, Jan 21, 2013 at 9:36 AM, Tsjerk Wassenaar

<tsjerkw at gmail.com>

wrote:

Oh, I'm jetlagged. ^ is a control character for 'start of

string'.

In

the

context of a character set it means negation: [^a-z].

Ciao,

Tsjerk


On Mon, Jan 21, 2013 at 4:33 PM, Tsjerk Wassenaar

<tsjerkw at gmail.com>wrote:

Hi Mark Na,

Try:

grepl("latitude\\^2",temp)

^ is a control character for negation, so you have to escape it.

Cheers,

Tsjerk


On Mon, Jan 21, 2013 at 4:26 PM, <mtb954 at gmail.com> wrote:

Hello R-helpers,

I am trying to search for string that includes the caret

symbol,

using

the
following code:

grepl("latitude^2",temp)


And R doesn't like that. It gives me:

temp<-c("latitude^2","latitude and

latitude^2","longitude^2","longitude

and longitude^2")

temp

[1] "latitude^2"                "latitude and latitude^2"

"longitude^2"

               "longitude and longitude^2"

grepl("latitude^2",temp)

[1] FALSE FALSE FALSE FALSE


I think this must a regex problem, but I can't find out to

specify

the

caret using regex.

I would appreciate any help you could provide.

Many thanks,

Mark Na

           [[alternative HTML version deleted]]

code.

--
Tsjerk A. Wassenaar, Ph.D.

post-doctoral researcher
Biocomputing Group
Department of Biological Sciences
2500 University Drive NW
Calgary, AB T2N 1N4
Canada

--
Tsjerk A. Wassenaar, Ph.D.

post-doctoral researcher
Biocomputing Group
Department of Biological Sciences
2500 University Drive NW
Calgary, AB T2N 1N4
Canada

    [[alternative HTML version deleted]]